Method and device for drafting, revising, automatic numbering and figure numbering, and medium

Abstract

Disclosed in the present application are a method and device for drafting, revising, automatic numbering and figure numbering. The method comprises: receiving a first instruction for patent-related text, the first instruction comprising a revision end instruction; in response to the first instruction, determining pre-revision content in the patent-related text; using the pre-revision content to perform matching on the remaining parts or selected parts of the patent-related text; and on the basis of post-revision content, revising all or part of the target content obtained by means of matching, the post-revision content being content obtained on the basis of the pre-revision content. In the present application, it is not necessary to manually search for, locate and revise the remaining corresponding content one by one after patent-related text is revised, and the remaining corresponding content can be automatically matched and located and revised on the basis of pre-revision content, thereby achieving convenience and rapidness and reducing the risk of errors during the process.

Description

Methods, apparatuses, and media for authoring, revising, auto-labeling, and figure labeling Cross-reference to Related Applications This application claims priority to Chinese Patent Application No. 2024110017271, filed on July 24, 2025, Chinese Patent Application No. 2024110300400, filed on July 29, 2024, Chinese Patent Application No. 2024113469463, filed on September 25, 2024, Chinese Patent Application No. 2024117520261, filed on November 30, 2024, and Chinese Patent Application No. 2024119355485, filed on December 24, 2024, the entire contents of which are incorporated herein by reference. TECHNICAL FIELD The present application belongs to the technical field of data processing, and particularly relates to a method and device for authoring, revising, auto-labeling, and figure labeling. BACKGROUND Patent-related texts need to follow certain formal requirements, and there are several sets of corresponding content in different positions, and each set of corresponding content needs to be consistent. For example, the technical content in the claims needs to be consistent with the technical content in the specification. However, if one of the corresponding contents in a group is revised during the writing process or after the writing is completed, the remaining other corresponding contents in the group also need to be revised synchronously. The existing revision method generally manually finds and revises each revision, which is tedious and prone to errors. SUMMARY Embodiments of the present application provide a method and device for authoring, revising, auto-labeling, and figure labeling to solve the technical problem of tedious and error-prone file editing. To solve the above technical problem, an aspect of the present application provides a revision assistance method for patent-related texts, which includes: receiving a first instruction for the patent-related texts, the first instruction including a revision end instruction; determining pre-revision content in the patent-related texts in response to the first instruction; matching the remaining part or selected part of the patent-related texts with the pre-revision content; and revising all target content or part of the target content obtained by matching based on post-revision content, the post-revision content being content obtained based on the pre-revision content. To solve the above technical problem, an aspect of the present application provides an authoring assistance method for patent-related texts, which includes: reading a data structure, forming a corresponding claim according to a first preset rule for a set of elements corresponding to each claim and a set of relationships, the data structure defining a number of elements and their relationships of a protected subject matter, and defining a set of elements and a set of relationships constituting each claim. To solve the above technical problems, an aspect of the present application provides a patent-related text writing assistance method, which comprises: a client sending a data structure to a server, the data structure defining a plurality of elements of a protected subject and their relationships, and defining a set of the elements and a set of the relationships constituting each claim; and the client receiving a claim from the server, the claim being formed by the server according to a first preset rule based on the set of the elements and the set of the relationships corresponding to each claim. To solve the above technical problems, an aspect of the present application provides a patent-related text writing assistance method, which comprises: a client receiving a first instruction for editing a data structure; the client sending the first instruction to a server to make the server edit a data structure, the edited data structure defining a plurality of elements of a protected subject and their relationships, and defining a set of the elements and a set of the relationships constituting each claim; and the client receiving a claim or a processing completion notification from the server, the claim being formed by the server according to a first preset rule based on the set of the elements and the set of the relationships corresponding to each claim, and the processing completion notification including a processing result of forming the claim. To solve the above technical problems, an aspect of the present application provides a computer device, which comprises: a processor; and a memory connected to the processor and used for storing program instructions, the program instructions being executable by the processor to implement the above-mentioned patent-related text writing assistance method. To solve the above technical problems, an aspect of the present application provides a computer-readable storage medium, which is used for storing program instructions, the program instructions being executable to implement the above-mentioned patent-related text writing assistance method. To solve the above technical problems, an aspect of the present application provides a writing assistance method, which comprises: reading a file name of a patent-related text, the file name comprising a file identifier, the file identifier being associated with writing reference information of the patent-related text; and obtaining the writing reference information associated with the file identifier; wherein the obtained writing reference information is displayed simultaneously with the content of the patent-related text. To solve the above technical problems, an aspect of the present application provides a text writing assistance method, which comprises: a server receiving a file name sent by a client, the file name comprising a file identifier, the file identifier being associated with writing reference information of a text; the server obtaining the writing reference information associated with the file identifier; and the server sending the writing reference information to the client to make the client display the obtained writing reference information simultaneously with the content of the text. To solve the above technical problems, an aspect of the present application provides a writing assistance method for a text, which comprises: inputting content in a text editing application by using an input method; obtaining the content by using a plug-in of the text editing application, and searching for a matching paragraph or a matching sentence in or out of the text by taking at least part of the characters of the content as a keyword; and displaying the matching paragraph or the matching sentence as writing reference information on the same screen as the content. To solve the above technical problems, an aspect of the present application provides a writing assistance method for a text, which comprises: receiving first content input in a text, or receiving first content input in an input method interface; searching for a matching paragraph or a matching sentence in the text or a designated document by taking the first content as a keyword, wherein the matching paragraph or the matching sentence comprises the first content and second content, or the matching paragraph or the matching sentence comprises the second content and related content of the first content, and the matching paragraph or the matching sentence is obtained by using a delimiting character in the text or the designated document, wherein the delimiting character is a general character adjacent to the front or the back of the matching paragraph or the matching sentence; and displaying the matching paragraph or the matching sentence as writing reference information or recommended information on the same screen as the text. To solve the above technical problems, an aspect of the present application provides a computer device, which comprises: a processor; and a memory connected to the processor and used for storing program instructions, wherein the program instructions can be executed by the processor to implement the writing assistance method provided above. To solve the above technical problems, an aspect of the present application provides a computer readable storage medium, which is used for storing program instructions, wherein the program instructions can be executed to implement the writing assistance method provided above. To solve the above technical problems, an aspect of the present application provides an automatic labeling method for a patent-related text, which comprises: determining a claim or a specification range in the patent-related text; delimiting structural names in the claim or the specification range; hierarchically dividing at least two of the delimiting structural names; and labeling the at least two of the delimiting structural names according to the hierarchical division. To solve the above technical problems, an aspect of the present application provides an automatic labeling method for a patent-related text, which comprises: determining a claim or a specification range in the patent-related text; delimiting structural names in the claim or the specification range; and labeling at least two of the delimiting structural names in a random labeling manner or in a pre-defined labeling manner. To solve the above technical problems, an aspect of the present application provides an automatic modification method of patent-related text, which comprises: receiving an artificial revision end instruction for the patent-related text; in response to the artificial revision end instruction, or in response to a preset condition being met, determining pre-revision content in the patent-related text, and defining a structural name in a claim scope or a specification; in response to the pre-revision content matching the structural name, matching the structural name or the pre-revision content with a remaining part or a selected part of the patent-related text; and revising all target content or part of the target content obtained by matching based on post-revision content. To solve the above technical problems, an aspect of the present application provides a computer device, which comprises: a processor; a memory connected to the processor and used for storing program instructions, the program instructions being executable by the processor to implement the automatic numbering method of patent-related text or the automatic modification method of patent-related text provided above. To solve the above technical problems, an aspect of the present application provides a numbering method of a drawing, which comprises: displaying a drawing, and displaying at least one of a name set and a number set, a name in the name set corresponding to a number in the number set; receiving a drag start or end instruction of dragging the name itself or a first anchor point thereof, or the number itself or a second anchor point thereof to a first position in or outside the drawing; in response to the drag start or end instruction, generating a line, and dragging or copying the name or the number to a second position in or outside the drawing, one end of the line extending to the first position, the other end of the line extending to one side of the dragged or copied name or one side of the dragged or copied number; or the line being located below the name or the number. To solve the above technical problems, an aspect of the present application provides a numbering method of a drawing, which comprises: displaying a drawing, and displaying at least one of a name set and a number set, a name in the name set corresponding to a number in the number set; receiving a first instruction of selecting or clicking the name or the number, and receiving a second instruction of clicking a first position in or outside the drawing; in response to the first clicking and the second instruction, generating a line, and moving or copying the name or the number to a second position in or outside the drawing, one end of the line extending to the first position, the other end of the line extending to one side of the moved or copied name or one side of the moved or copied number; or the line being located below the name or the number. To solve the above technical problems, one aspect of the present application provides a method for labeling a drawing, the method comprising: displaying a drawing; receiving a first click on a first position in or outside the drawing; in response to the first click, generating a line extending to the first position at one end and extending to a name or a label at the other end, or generating the name or the label at a second position below the line. To solve the above technical problems, one aspect of the present application provides a computer device, comprising: a processor; a memory connected to the processor and configured to store program instructions executable by the processor to implement the above-provided method for labeling a drawing, or implement the above-provided method for labeling a drawing. BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is a flowchart of an embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 2 is an interface diagram of a first interface of an embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 3 is a flowchart of another embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 4 is an interface diagram of a second interface of an embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 5 is a structural diagram of an embodiment of the computer device according to the present application; FIG. 6 is a structural diagram of an embodiment of the computer-readable storage medium according to the present application; FIG. 7 is a structural diagram of a client and a server in an embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 8 is a structural diagram of a client and a server in another embodiment of the method for assisting revision of patent-related texts according to the present application; FIG. 9 is an interface diagram of an embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 10 is a data structure diagram of an embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 11 is a data structure diagram of another embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 12 is a diagram of an example product of an embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 13 is an interface diagram of a first interface of an embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 14 is an interface diagram of a second interface of an embodiment of the method for assisting drafting of patent-related texts according to the present application; FIG. 15 is a structural diagram of an embodiment of the computer device according to the present application; FIG. 16 is a structural diagram of an embodiment of the computer readable storage medium provided in the present application; FIG. 17 is a structural diagram of a client and a server in another embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 18 is an interface diagram of an embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 19 is an interface diagram of another embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 20 is an interface diagram of an embodiment of the method for assisting in writing in the present application; FIG. 21 is an interface diagram of another embodiment of the method for assisting in writing in the present application; FIG. 22 is a data structure diagram of an embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 23 is a data structure diagram of another embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 24 is a diagram of an example product of an embodiment of the method for assisting in writing a patent-related text in the present application; FIG. 25 is a structural diagram of an embodiment of the computer device provided in the present application; FIG. 26 is a structural diagram of an embodiment of the computer readable storage medium provided in the present application; FIG. 27 is a flow diagram of an embodiment of the method for automatically labeling a patent-related text in the present application; FIG. 28 is an interface diagram of a first interface of an embodiment of the method for automatically modifying a patent-related text in the present application; FIG. 29 is an interface diagram of a second interface of an embodiment of the method for automatically modifying a patent-related text in the present application; FIG. 30 is a structural diagram of an embodiment of the computer device provided in the present application; FIG. 31 is a structural diagram of an embodiment of the computer readable storage medium provided in the present application; FIG. 32 is a structural diagram of a client and a server in an embodiment of the method for automatically modifying a patent-related text in the present application; FIG. 33 is a flow diagram of an embodiment of the method for labeling a figure in the present application; FIG. 34 is an effect diagram of moving a mouse to a label in an embodiment of the method for labeling a figure in the present application; FIG. 35 is an effect diagram of moving a name and a label to a first position in the embodiment shown in FIG. 34; FIG. 36 is an effect diagram of labeling a wheel in the embodiment shown in FIG. 34; FIG. 37 is an effect diagram of completing labeling a car in the embodiment shown in FIG. 34; FIG. 38 is an effect diagram of removing a name and leaving only a label in the embodiment shown in FIG. 34; FIG. 39 is a schematic diagram of an interface of another embodiment of the method of labeling of the drawings of the present application; FIG. 40 is a schematic diagram of an interface of still another embodiment of the method of labeling of the drawings of the present application; FIG. 41 is a schematic diagram of the effect of moving a mouse to a housing of still another embodiment of the method of labeling of the drawings of the present application; FIG. 42 is a schematic diagram of the effect of displaying a name and label of the embodiment shown in FIG. 41; FIG. 43 is a schematic diagram of the effect of labeling a housing of the embodiment shown in FIG. 41; FIG. 44 is a schematic diagram of the structure of an embodiment of a computer device of the present application; FIG. 45 is a schematic diagram of the structure of an embodiment of a computer readable storage medium of the present application; FIG. 46 is a schematic diagram of the structure of a client and server in an embodiment of the method of labeling of the drawings of the present application. DETAILED DESCRIPTION The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those of ordinary skill in the art without any creative work fall within the scope of protection of the present application. Reference to "an embodiment" or "the embodiment" in this text means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor does it necessarily exclude other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments. Currently, part of the work of patent workers is to write patent documents to form patent-related texts, and then submit the patent application through official channels. Patent applications generally include invention applications, utility model applications or design applications. Among them, the patent-related texts involved in invention applications and utility model applications generally include abstracts, claims, descriptions, etc. Due to the requirements of the patent law, part of the content in the abstract, the claims and the description is substantially the same, and part of the content in different claims or different positions of the same claim in the claims may also be substantially the same, and part of the content in different positions of the description may also be substantially the same. Therefore, if a sentence or a few words in the patent-related texts are modified, other places of the patent-related texts may also need to be manually found out and modified consistently. The current word editing tool provides a "find" and "replace" function. During the revision process, the word editing software interface needs to be operated, the "find" and "replace" window needs to be opened, the content to be found and replaced needs to be input in the input box of the window, and then "find" or "replace all" needs to be clicked. The above "find" and "replace" functions provide a faster way than manually finding one by one, but as the patent work develops towards higher quality requirements, the work of applicants and patent agencies becomes more standardized, and the patent-related texts may need to be repeatedly modified and improved several times before being submitted as a patent application through official channels. Each modification and improvement may require several or more places to be found and replaced, and the word editing software interface needs to be frequently operated, the "find" and "replace" window needs to be opened, the content to be found and replaced needs to be input in the input box of the window, and then "find" or "replace all" needs to be clicked, so the work efficiency is still relatively low. Moreover, the current "find" and "replace" functions of the word editing software can only match 100% exactly, such as replacing "abc", which can only find other content with "abc" string, but cannot find content with "a11bc" with labels, or "a said bc" content, and vice versa, resulting in a lot of matching blind spots, which need to be manually found out and modified consistently, and the work efficiency cannot be further improved. The patent-related text revision assistance method, device and storage medium provided by the present application will be described in detail below in combination with embodiments, which can partially or completely solve the above technical problems. Please refer to FIG. 1, which is a flowchart of an embodiment of the patent-related text revision assistance method of the present application. The patent-related text revision assistance method comprises: S11: receiving a first instruction for the patent-related text, the first instruction comprising a revision end instruction. The revision end instruction refers to an instruction reflecting completion of at least one revision action operation. The revision end instruction can be actively triggered or passively issued during the operation of the revision action. Optionally, receiving the first instruction for the patent-related text comprises: in response to at least one of clicking an operation item in the interface, pressing a preset key, or inputting a preset instruction, triggering a revision end instruction. Specifically, at least one of actively clicking the operation item, pressing the preset key, or inputting the preset instruction can represent that the current revision action has been completed and a revision end instruction is desired to be triggered. The operation item can be a button in the interface. For example, a "run" button on a software interaction interface. During operation, the operation item can be interacted with through common operation methods such as clicking and sliding. Pressing the preset key can be pressing a certain physical key on the keyboard or other input. For example, the "F1" key on the keyboard can be set as the preset key. The preset key can also be a combination key, for example, it can be set to trigger a revision end instruction when Ctrl+F1 on the keyboard is pressed at the same time. Inputting the preset instruction can be inputting a specific string of characters to trigger the revision end instruction. The input of the preset instruction can be input in a specific preset area. Only input in the preset area can trigger the revision end instruction. The input of the preset instruction can also detect the input content in real time. As long as the condition is met, the revision end instruction is triggered. For example, it is set to trigger a revision end instruction as long as a specific format command symbol such as "run modification follow" is detected. Through the above setting, the content to be revised can be adjusted repeatedly several times first, and a revision end instruction is actively triggered only once after, for example, the revision is completed or after a certain stage. The content before revision can be processed in batches in one operation, which is more efficient. In addition, the revision end instruction is actively triggered, which can clearly reflect the content and number of revisions, and is not easy to have unexpected operations beyond the expected range, and is stable to use. Optionally, receiving the first instruction for the patent-related text comprises: in response to a preset change in the position of the cursor, triggering a revision end instruction. Specifically, when the position of the cursor changes, such as the cursor position changes from the current line to another line, it means that an operation has been performed. Each operation is detected and triggers a revision end instruction, that is, the revision end instruction is passively triggered in real time during operation. Through the above setting, when the revision is completed, the revision end instruction has been triggered in real time, and there is no need to perform an operation of triggering the revision end instruction after each revision is completed. The embodiment has fewer operation steps and is convenient to use. The preset change can be moving to the next character position or the previous character position, or jumping to other character positions. Alternatively, the preset change is detecting a line change of the cursor. Thus, on the basis of real-time triggering, a certain number of redundant triggering times can be reduced, and processing speed is taken into account. S12: In response to the first instruction, determining the pre-revision content in the patent-related text. Specifically, the pre-revision content refers to content before a revision occurs. The pre-revision content should include at least one character or format symbol. The pre-revision content can include changed content formed by operations such as addition, deletion, and change. For example, when "abcd" in the patent-related text is revised to "abzz", the pre-revision content at least includes "cd". Alternatively, when "abcd" in the patent-related text is revised to "ab", the pre-revision content at least includes "cd". For another example, when "ab" in the patent-related text is revised to "abcd", the pre-revision content at least includes "ab". When the pre-revision content is not detected in response to the first instruction, an end instruction can be triggered, and subsequent steps are no longer continued. Alternatively, determining the pre-revision content in the patent-related text specifically includes: Determining the third content that has been revised in the patent-related text, such as a background record of the last or several times of the third content that has been revised, or determining the third content with a revision format in the patent-related text. The revision format can be a format reflecting a revision operation in other software. For example, the revision format can be a format generated by the "revision" function in a word processing software. Determining the third content can be taking several continuous characters with the revision format as one third content. For example, the patent-related text includes "abcdefg", wherein "ab" and "efg" have revision formats, "ab" is taken as one third content, and "efg" is taken as another third content. Based on the third content, the fourth content of the minimum matching unit is determined, the fourth content has a length greater than or equal to the length of the third content, and the fourth content is taken as the pre-revision content. Determining the fourth content can be done by directly using each part containing the third content as the fourth content. For example, when the third content is "ab", the fourth content is also directly determined to be "ab". Alternatively, the fourth content can be determined by finding a complete word or phrase corresponding to the position of the third content. When determining the fourth content, the location of the third content can be detected, and based on the position of the third content, the text can be expanded in both directions according to semantics until a complete word or phrase is obtained as the fourth content. For example, patent-related text contains the semantically complete "abcdefg", where "de" is determined as the third content. In this case, the text can be expanded forward and backward based on "de" until the semantically complete "abcdefg" is obtained as the fourth content. When several third contents are adjacent, causing the complete words or phrases obtained by expanding from these third contents to overlap, the results of these expanded words or phrases can be merged to form only one fourth content. For example, patent-related text contains the semantically complete "abcdefg", where "bc" and "de" are identified as third content. "abcdefg" can then be obtained by expanding "bc" and "de" into the surrounding text. If the results of expanding "bc" and "de" into the surrounding text are found to be identical, then two separate fourth content entries for "abcdefg" are not created; only one fourth content entry for "abcdefg" is formed. Understandably, one purpose of determining the fourth content is to make the matching more accurate. Therefore, the specific strategy for obtaining the fourth content based on the third content can be set according to the actual situation. In addition, when the revision type is "change," the determination of the fourth content should be based on the text before the revision, and the revised content can be ignored. With the above settings, the location of revisions can be directly indicated using existing software's revision formats and background records. Furthermore, determining the fourth content based on the third content and using the fourth content as the original content increases the uniqueness of the original content within the patent-related text, reduces noise during matching, and ensures accurate positioning. In one specific embodiment, determining the fourth content of the minimum matching unit based on the third content includes: Starting from the third content, find the first first separator that appears in the preceding and following text. Using the first first separator that appears in the preceding and following text as the ending point, take the content between the ending point of the preceding and following text as the fourth content. The first separator can be a comma, period, semicolon, space, paragraph break, or other special symbol. For example, if the patent-related text contains "abc,defgh,ijk.", where "fg" is the third content, then the text is expanded forward and backward until the first separator comma is encountered as the endpoint, thus determining the fourth content as "defgh". Thus, the text containing the third content is cut according to the first first delimiter appearing in the preceding context and the following context, so that a sentence segment containing certain unique features can be easily obtained, so that when the fourth content is taken as the minimum matching unit, false matching noise is not easily generated. In another specific embodiment, determining the fourth content of the minimum matching unit comprises: The fourth content of the minimum matching unit is determined by taking the third content as the center and taking the preset length of the preceding context and the following context as the boundary. The preset length can be a value preset in advance, for example, it can be 3, 4, 5, etc. The preset length can be set to be adjustable or a fixed value. The fourth content is obtained by concatenating the characters of the preceding context, the third content, and the characters of the following context in turn. In addition, when the length of the preceding context and the following context of the paragraph in which the third content is located is less than the preset length, the maximum preceding context and the following context that can be cut off of the paragraph in which the third content is located can be taken as the cut-off point to determine the fourth content of the minimum matching unit. For example, when the third content is located at the end of the paragraph, the preceding context is cut according to the preset length, and the following context is cut according to the actual remaining character length of the paragraph. Through the above setting, the fourth content of the minimum matching unit can be obtained simply and quickly, the function is simple to implement, and the processing speed is fast. Optionally, the revised content in the patent-related text is determined, specifically comprising: The content in the selected range in the patent-related text is taken as the revised content. When the selected range of content contains content that is changed by revision, the selected range of content as a whole is the revised content and is taken as the minimum matching unit. When the selected range of content does not contain content that is changed by revision, i.e., the selected range of content is all original text, an end instruction can be triggered, and subsequent steps are no longer continued. Through the above setting, the revised content in the patent-related text does not need to rely on the revision format of other software, and the application range is wider. In another other embodiment, the revised content in the patent-related text is determined, specifically comprising: The third content determined after comparing the current version of the patent-related text and the historical version of the patent-related text is taken as the revised content. The historical version of the patent-related text can be another old version file, an old version of the patent-related text located at another text position in the same file, etc. By comparing the historical version of the patent-related text with the current patent-related text, the part of the current patent-related text that is changed relative to the historical version of the patent-related text can be obtained. The changed part can be taken as the revised content. Through the above setting, it is determined that the pre-revision content in the patent-related text does not need to rely on the revision format of other software, and does not need to be selected. The changed part in the patent-related text can be obtained at one time in batches, especially suitable for the case where the changed part is not marked and only the historical version content is left. It is convenient to use and widely applicable. S13: matching the remaining part or the selected part of the patent-related text with the pre-revision content. Specifically, the remaining part or the selected part of the patent-related text is matched with the pre-revision content, so that the position of the content in the remaining part or the selected part of the patent-related text which needs to be modified synchronously corresponding to the pre-revision content can be obtained. For example, the pre-revision content is "abc", then all "abc" in the remaining part or the selected part of the patent-related text is matched. The matching can be performed in the entire remaining part of the patent-related text, that is, full-text matching. The matching can also be performed only in the selected part, and the other part is not processed. The specific implementation of the matching can be any character matching technology that can locate the position, which will not be described here. Optionally, the matching the remaining part or the selected part of the patent-related text with the pre-revision content comprises: In response to not matching the target content based on the pre-revision content, a second delimiter in the pre-revision content is searched. Specifically, in the process of matching the remaining part or the selected part of the patent-related text with the pre-revision content, the case that no corresponding result is matched may occur. In this embodiment, it is considered as an abnormal case that the pre-revision content in the patent-related text has corresponding content in the remaining part of the patent-related text as a high probability case. The case that no corresponding result is matched may be caused by character change, long sentence splitting, grammar change, etc. Therefore, in this embodiment, a second delimiter in the pre-revision content is searched as a judgment basis for further processing. The second delimiter can be a comma, a semicolon, a period, a space, etc. In response to the absence of the second delimiter, a conjunction in the pre-revision content is replaced with a subject of a previous sentence, or the subject of the previous sentence is inserted at the beginning of the pre-revision content. When the second separator is not found, it means that the probability of not matching the corresponding result due to long sentence splitting is low, and the probability of not matching the corresponding result due to syntax change should be considered first. For example, a sentence in the pre-revision content replaces the subject of the sentence with a conjunction word due to writing habits when writing, and the subject is not replaced with a conjunction word when writing the remaining part of the patent-related text. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the conjunction word in the pre-revision content with the subject of the previous sentence, so that the matching result is not easily missed. For example, the first sentence of the pre-revision content is "a includes b.", the pre-revision content is "also includes c.", and in the remaining part of the patent-related text, it is recorded as "a includes b" and "a includes c". At this time, it may not be possible to get the matching result directly using "also includes c.". After replacing "also" with the subject "a" of the previous sentence, the correct matching can be achieved. In response to the presence of the second separator, the pre-revision content is split into a preceding part and a plurality of subsequent parts based on the second separator, and the subject of the preceding part is replaced with a keyword in each subsequent part or inserted at the beginning of each subsequent part. When the second separator is found, it means that the probability of not matching the corresponding result due to long sentence splitting is high. For example, two short sentences in the pre-revision content are connected by a colon, "and" and the like, and when writing the remaining part of the patent-related text, they are split into two short sentences, and the colon, "and" and the like in the subsequent short sentence are replaced with the subject. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the keyword in each subsequent part with the subject of the preceding part, so that the matching result is not easily missed. For example, the pre-revision content is "a includes b, and includes cd.", and in the remaining part of the patent-related text, it is recorded as "a includes b. a includes cd.". At this time, due to long sentence splitting, it may not be possible to get the matching result. After replacing the "and" in the subsequent part with the subject "a" of the preceding part, the correct matching can be achieved. The keyword can be set according to the actual situation, which can be a common word or symbol used to separate long sentences, which is not limited here. The position of the pre-revision content and the remaining part of the patent-related text can be set according to the actual situation. For example, in a specific embodiment, the pre-revision content is located in the claims of the patent-related text. The remaining part or selected part of the patent-related text is at least partially located in the specification of the patent-related text. For another specific embodiment, the pre-revision content is located in the specification of the patent-related text, and the remaining part or selected part of the patent-related text is at least partially located in the claims of the patent-related text. S14: revising all target content or part of the target content obtained by matching based on the post-revision content, the post-revision content being the content obtained based on the pre-revision content. In particular, the revised content can be new content changed from the pre-revision content, or content after deletion of the pre-revision content, or new content added on the basis of the pre-revision content. The revised content can be directly changed from the pre-revision content. The revised content can also be new content added at a position based on the pre-revision content, or a blank generated after deletion at a position based on the pre-revision content. For example, when "abcd" in the patent-related text is revised to "abzz" due to a change operation, the pre-revision content at least includes "cd". The revised content can be "zz". When "abcd" in the patent-related text is revised to "ab" due to a deletion operation, the pre-revision content can be "abcd". The revised content can be "ab". When "ab" in the patent-related text is revised to "abcd" due to an addition operation, the pre-revision content at least includes "ab". The revised content can be "abcd". For another example, when "abcd" in the patent-related text is revised to "ab" due to a deletion operation, the pre-revision content can be "cd". The revised content can be empty. When "ab" in the patent-related text is revised to "abcd" due to an addition operation, the pre-revision content at least includes "ab". The revised content can be "cd". All target content or part of the target content will be revised to be consistent with the revised content. All matched target content can be uniformly revised, or only part of the matched target content can be revised. When the pre-revision content is determined by the revision format, the target content can also be revised by the revision format during revision, so as to facilitate subsequent observation of the position of the target content to which the revision is applied. Optionally, a positioning mark is formed at the revised target content. The positioning mark can be used to quickly locate the revised target content. The positioning mark can be a comment in a word processing software, etc. Optionally, before or simultaneously with matching the pre-revision content to the remaining part or selected part of the patent-related text, the following is included: The pre-revision content, the remaining part, and the selected part that are matched are subjected to de-noising processing. The de-noising processing includes at least one or all of the following: removing punctuation, removing reference numerals, removing definite articles, and removing demonstrative pronouns. The pre-revision content can interfere with the matching due to the difference in the form of the text from the target content. For example, when the pre-revision content is located in a claim and its content is "the a includes b", the "the" in the pre-revision content needs to be removed, and when the remaining content of the patent-related text contains a reference numeral, for example, its content is "a11 includes b12", the reference numeral in the remaining content of the patent-related text also needs to be removed, and then the pre-revision content is matched with the remaining content of the patent-related text. By the above setting, the remaining part or selected part of the patent related text is matched with the pre-revision content, and is not easily disturbed by noise, and the matching is accurate. Optionally, revising all target contents or part of the target contents obtained by matching based on the revised content comprises: The first content obtained by the de-noising processing of the revised content replaces the second content in the target content, and the remaining part of the target content except the second content is not changed. In the embodiment, when there is noise in the revised content, the first content obtained by the de-noising processing can be different from the revised content. When there is no noise in the revised content, the first content can also be the same as the revised content, that is, the revised content does not need to be modified to obtain the first content directly. The second content is the content in the target content that needs to be modified based on the change of the pre-revision content. When revising the target content, only the part of the target content corresponding to the change of the pre-revision content is changed, and the whole target content is not directly changed. For example, in the patent related text, "the a includes b" is revised to "the c includes b", and the target content is "a11 includes b12". At this time, the revised content is de-noised to obtain "c", and the second content is "a11", the revised content is replaced by the second content to obtain "c includes b12". In this way, the remaining content in the target content is not easily changed to cause errors, and the revision is accurate. Optionally, after the first content obtained by the de-noising processing of the revised content replaces the second content in the target content, and the remaining part of the target content except the second content is not changed, it further comprises: In response to the revised content having a corresponding figure number, adding a figure number to the revised content. For example, after obtaining "c includes b12" in the above embodiment, the corresponding figure number "13" of "c" is also added to obtain "c13 includes b12". In this way, the revised content can automatically complete the figure mark, and is not easily omitted to cause formal errors. When the revised content does not have a corresponding figure number or does not need to add a figure number, the above process can not be performed. To more clearly illustrate the workflow of the revision assisting method of the patent-related text, the following is an example of a change revision. Please refer to FIG. 2, which is an interface diagram of the first interface of an embodiment of the revision assisting method of the patent-related text. Optionally, the revision assisting method of the patent-related text can have a first interface 100. The patent-related text can be edited in the first interface 100. The patent-related text includes at least a first part 10 and a second part 20. As shown in FIG. 2, the content of the original text "ABCD" can be edited in the first part 10, and then the operation item 30 "Run" button is clicked. At this time, "CD" in the first part 10 has a revision format, "CD" is determined as the third content 11, and "ZZ" is the revised content 13. Next, based on the third content 11, the context is extended forward and backward. The extension process is based on the pre-revision text, and the revised content 13 "ZZ" is skipped until the first comma position is reached as the cutoff, and the fourth content 12 "ABCD" is obtained as the pre-revision content. Then, based on the fourth content 12, the target content 22 "ABCD" in the second part 20 is matched and located to the second content 21 "CD" in the second part 20, and the revised content 13 is replaced with the second content 21, i.e., the content synchronization revision is completed. After the content synchronization revision, the revised content 13 replaced in the second part 20 can also be labeled. In the above matching and replacing process, a de-noising process is performed, which includes at least one of removing punctuation, removing labels, removing definite articles, and removing demonstrative pronouns. As can be seen from the above embodiments, the present application can not need to manually input the keywords for searching in the "Find" or "Replace" interface each time, i.e., the pre-revision content (i.e., the keywords for searching in the above) in the patent-related text can be determined by the computer, and subsequent matching and replacing can be performed, which greatly improves the revision efficiency of the patent-related text and reduces the error probability. The patent-related text processed by the embodiments of the present application is relatively neat and accurate, reduces reading difficulty, and is convenient for subsequent review and reading. Moreover, since the embodiments of the present application can perform de-noising processing on the pre-revision content, the remaining part, and the selected part for matching, the "Find" or "Replace" has a generalization function, which can more comprehensively and accurately find the matching target, and greatly reduces the problems of missing changes due to manual one-by-one searching and not finding. Secondly, since the embodiments of the present application replace the second content in the target content with the first content of the revised content, the remaining content in the target content is not changed, so that only the part that needs to be modified in the target content will be affected by the change, and the remaining content will not be processed, which is not easy to cause unexpected changes and form errors, and the revision is accurate. Further, the third content changed can be extended to the fourth content according to the context before and after the third content in the matching process. Compared with the third content which can be only a few words, the fourth content can contain more unique features, so that the matching is more accurate and less likely to form noise interference. When determining the fourth content, the text can be cut off by the first delimiter as the cut-off point, or the text can be directly cut off according to the preset length. The fourth content formed by cutting off the text can have good uniqueness. In addition, the pre-revision content in the embodiments of the present application can be determined by revision format, selected range in the patent-related text, or comparison between the current version of the patent-related text and the historical version of the patent-related text. The determination of the pre-revision content is various and can be freely selected according to different occasions. When the pre-revision content is determined by the revision format, the positioning is accurate, and the revision format is easy to observe, so that the content to be revised can be intuitively reflected. When the pre-revision content is determined by the selected range, the part of the text to be processed can be conveniently adjusted by selecting different ranges, and the effective range of the revision is controllable. When the pre-revision content is determined by comparing the current version of the patent-related text with the historical version of the patent-related text, only one historical version of the patent-related text needs to be stored in advance, or even automatically stored by the software of the embodiments of the present application. Editing under the revision format is not necessary, and repeated operations for selecting the selected range are not necessary. Therefore, time and labor are saved, and the operation is simple. This is especially suitable for occasions where the revision format is not opened or forgotten to be opened during the revision of the patent-related text. When matching the target, the embodiments of the present application further identify whether the pre-revision content contains the second delimiter when the matching fails. When the second delimiter is not found, it means that the probability of not matching the corresponding result due to long sentence splitting is low, and the probability of not matching the corresponding result due to syntax change should be considered first. At this time, the embodiments of the present application replace the conjunction in the pre-revision content with the main sentence of the previous sentence and then perform matching, so that the above situation is included in the matching range. When the second delimiter is found, it means that the probability of not matching the corresponding result due to long sentence splitting is high, and the embodiments of the present application replace the key word in each subsequent part with the subject of the previous part, so that the above situation is included in the matching range. It can be seen that the embodiments of the present application can also adapt to situations such as character change, long sentence splitting, and syntax change during matching, and are less likely to miss the matching result. The above matching and revision process in the embodiments of the present application can be actively triggered by at least one of responding to the operation item in the interface, pressing the preset key, or inputting the preset instruction, or passively triggered according to the revision operation in response to the preset change of the cursor position. The setting method is flexible, and the revision operation can be well reacted. Therefore, the use is convenient. Referring to FIG. 3, FIG. 3 is a flow diagram of another embodiment of the revision assisting method of the patent-related text. Optionally, the revision assisting method comprises: S21: generating a history record, the history record comprising at least part of information related to the revision of the matched target content or part of the target content based on the revised content. The history record comprises information related to the corresponding revision operation. For example, the history record can comprise the time of the revision operation, the content summary of the revision, the effective range of the revision, etc. S22: restoring the target content to the state before the revision corresponding to the history record in response to clicking the history record. When the history record is clicked, the content changed by the corresponding revision operation will be restored to the content before the revision operation. The way of restoring the target content to the state before the revision corresponding to the history record can be recording the old version of the text content, and covering the current text content with the old version of the text content when the click occurs. It can also be storing the old version of the file, and covering the current file with the old version of the file when the click occurs. Referring to FIG. 4, FIG. 4 is an interface diagram of a second interface of an embodiment of the revision assisting method of the patent-related text. Optionally, the revision assisting method of the patent-related text comprises a second interface 200, and the history record is displayed on the second interface 200. The history record can record the revision number, the revision position, the revision content, and the operation time, etc. The revision position can be the paragraph number of the position where the revision occurs in the patent-related text. The revision content can record the summary of the content where the revision operation occurs. The operation time can be the time when the revision operation is triggered. The second interface 200 can be clicked, as shown in FIG. 4, when the history record with the number 2 in the second interface 200 is clicked, the patent-related text will be rolled back to the state before the operation of deleting aa. Referring to FIG. 5, FIG. 5 is a structural diagram of an embodiment of the computer device 50. Optionally, the computer device 50 comprises a memory 51 and a processor 52 coupled with each other, and the processor 52 is configured to execute program instructions stored in the memory 51 to implement the steps of any of the above-mentioned embodiments of the revision assisting method of the patent-related text. In a specific implementation scenario, the computer device 50 can include but is not limited to a microcomputer, a server, and in addition, the computer device 50 can also include a notebook computer, a tablet computer, and other mobile devices, which are not limited here. Specifically, the processor 52 is configured to control itself and the memory 51 to implement the steps of any of the above patent-related text revision assisting method embodiments. The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 can be an integrated circuit chip having a processing capability. The processor 52 can also be a general purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general purpose processor can be a microprocessor or the processor can also be any conventional processor. In addition, the processor 52 can be implemented by an integrated circuit chip together. Referring to FIG. 6, FIG. 6 is a structural diagram of an embodiment of the computer readable storage medium provided in the present application. The computer readable storage medium 60 of the embodiment of the present application stores program instructions 61, which, when executed, implement the method provided by any embodiment of the patent-related text revision assisting method of the present application and any non-conflicting combination. The program instructions 61 can form a program file and be stored in the computer readable storage medium 60 in the form of a software product, so that a computer device (which can be a computer, a server, or a network device, etc.) executes all or part of the steps of the method of each embodiment of the present application. The aforementioned computer readable storage medium 60 includes a U disk, a mobile hard disk, a ROM (Read-Only Memory), a RAM (Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes, or a terminal device such as a computer, a server, a mobile phone, a tablet, etc. Referring to FIG. 7, FIG. 7 is a structural diagram of a client and a server in an embodiment of the patent-related text revision assisting method. The present application provides a patent-related text revision assisting method, which includes that the client 80 receives a first instruction for a patent-related text. The first instruction includes a revision end instruction. The client 80 sends the first instruction to the server 70, so that the server 70 determines the pre-revision content in the patent-related text, and matches the remaining part or the selected part of the patent-related text with the pre-revision content. All target content or part of the target content obtained by the matching is revised based on the post-revision content, and the post-revision content is obtained based on the pre-revision content. The client 80 receives the revision result and displays the patent-related text that has been revised in the interface. In particular, the revision assisting method of patent-related text can also be performed by the server 70. In this embodiment, the client 80 is mainly used to provide an interactive window. The patent-related text can be located on the server 70 and displayed by the client 80. That is, the client 80 can not store the patent-related text and only serve as a display function, and the processing of the patent-related text is performed by the server 70. For example, the patent-related text can be an online cloud operation client 80 interacting with the server 70. After receiving the first instruction from the client 80, the server 70 performs the steps of the revision assisting method of patent-related text in this application, completes the revision of the patent-related text, and sends the result to the client 80 for display. For example, the client 80 can only be responsible for providing an interactive window. The patent-related text is directly stored on the server 70. After receiving the first instruction, the client 80 sends the first instruction to the server 70. The server 70 processes and completes the revision of the patent-related text. Then the client 80 displays the revision result on the server 70 through the window interface of the client software. Through the above setting, the client does not need to perform complex operations, which is conducive to reducing the load of the client computer device. At the same time, the operation process is performed on the server, and the data security is better. Please refer to FIG. 8, which is a structural schematic diagram of a client and a server in another embodiment of the revision assisting method of patent-related text. Optionally, the present application provides a revision assisting method of patent-related text, which comprises: The client 80 receives the first instruction for the patent-related text, and the first instruction includes a revision end instruction. The client 80 sends the first instruction and the patent-related text to the server 70, so that the server 70 determines the pre-revision content in the patent-related text and matches the remaining part or selected part of the patent-related text with the pre-revision content. Based on the pre-revision content, the revision position and revision manner of all target contents obtained by matching are determined. The client 80 receives the revision position and revision manner, and revises the patent-related text based on the revision position and revision manner. Specifically, in the present embodiment, some steps of the revision assisting method for patent-related text are performed by the server 70, and other steps are performed by the client 80. For example, the server 70 can be configured to perform the complex key steps, and the client 80 can be configured to perform the other simple steps. The patent-related text can be stored in the client 80 and not stored in the server 70. After receiving the first instruction, the client 80 sends the first instruction and the patent-related text to the server 70. After receiving the first instruction and the patent-related text, the server 70 processes the patent-related text according to some steps of the revision assisting method for patent-related text, and obtains the revision position and the revision manner corresponding to each pre-revision content. The server 70 returns the revision position and the revision manner to the client 80, and the client 80 performs the revision on the patent-related text according to the revision position and the revision manner. For example, the client 80 can send the patent-related text to the server 70, and the server 70 is responsible for performing the steps of the revision assisting method for patent-related text related to matching positioning and calculating revision manner. The server 70 obtains, for example, that the revision position is "the 20th character of the first paragraph of the specification", and the revision manner is "change the original character to abc". Then the server 70 returns the revision position and the revision manner to the client 80, and the client 80 does not need to be responsible for matching positioning and calculating revision manner, and directly performs the revision on the patent-related text according to the revision position and the revision manner.

[0010] For example, the client 80 can send the patent-related text to the server 70, and the server 70 is responsible for performing the steps of the revision assisting method for patent-related text related to matching positioning and calculating revision manner. The server 70 obtains, for example, that the revision position is "the 20th character of the first paragraph of the specification", and the revision manner is "change the original character to abc". Then the server 70 returns the revision position and the revision manner to the client 80, and the client 80 does not need to be responsible for matching positioning and calculating revision manner, and directly performs the revision on the patent-related text according to the revision position and the revision manner. Through the above arrangement, the key steps can be placed in the server for execution, and the simple steps can be performed by the client, so that the load pressure of both the server and the client can be considered, and the data security is good. Optionally, the present application provides a revision assisting method for text, which comprises: receiving a first instruction for the text, the first instruction comprising a revision end instruction; determining pre-revision content in the text in response to the first instruction; matching the pre-revision content with the remaining part or the selected part of the text; and revising all target content or part of the target content obtained by the matching based on post-revision content, the post-revision content being content obtained based on the pre-revision content. The text in the present example can be an article, a report, a thesis, a novel, or a scientific and technological document, etc. Optionally, the revision assisting method for patent-related text of the present application is executed by program instructions, the program instructions are installed in the existing word editing software in the form of plug-ins, and the program instructions act on the patent-related text opened by the word editing software to realize the above-mentioned revision assisting method for patent-related text. Optionally, the revision assisting method for patent-related text of the present application is executed by program instructions, the program instructions are independently run as computer software and at least contain a word editing function. The program instructions act on the patent-related text in the computer software to realize the above-mentioned revision assisting method for patent-related text. Optionally, the revision assisting method for patent-related texts is implemented by program instructions, and the program instructions include at least two parts, one of which is located in a local computer device, and the other of which is located in a server. At least one step in the above-mentioned revision assisting method for patent-related texts is executed by the server and can return a result, and the local computer executes other steps by communicating with the server and obtaining the result, so as to implement the above-mentioned revision assisting method for patent-related texts. Optionally, the revision assisting method for patent-related texts is implemented by program instructions, and the program instructions are all located in a local computer, and the above-mentioned revision assisting method for patent-related texts is implemented by executing the program instructions by the local computer. Optionally, the revision assisting method for patent-related texts is all executed on a server, and the execution result is displayed by a local computer, that is, a client. At present, part of the work of patent workers is to write patent files to form patent-related texts, and then to submit the patent application through official channels. The patent application generally includes an invention application, a utility model application or a design application. Among them, the patent-related texts involved in the invention application and the utility model application generally include an abstract, a claim, a specification, etc. Among them, the writing of the claim requires a high requirement, not only to meet the hard requirements of laws and regulations, but also to meet the requirements of the patent layout of the applicant and the further requirements of high-quality patent applications. In terms of technology, law, profession and logic, it is as perfect as possible to ensure that the patent application can be authorized while achieving the purpose of protecting the invention to the greatest extent. Therefore, at present, some inexperienced patent agents find it difficult to start writing, and they have no idea how to write in the face of complex technical materials. The claims written may not meet the requirements of the above-mentioned relevant regulations, or the quality of the written claims is low, and it takes a long time. In an embodiment of the present application, a revision assisting method for patent-related texts is proposed, which realizes human-computer interaction by software and can perform related operations on the software interface to assist the agent to write high-quality and efficient patent-related texts that meet the requirements. The method comprises: S100: editing a data structure; The data structure defines a plurality of elements of the protected subject matter and the relationship thereof, and defines a set of elements and a set of relationships constituting each claim. For example, the elements of the product claim generally include the elements, components or structures of the product, and the components or structures are generally composed of elements, etc. The relationships of the product claim generally include position relationship, connection relationship, action relationship, signal relationship or electromagnetic field relationship, etc. For example, the elements of the method claim generally include steps of the method, and the relationships of the method claim generally include the sequence of the steps. Optionally, as shown in FIG. 9, to facilitate the editing of the data structure, a first area 1000 for editing the data structure and a second area for displaying the formed claims or the contents of the technical disclosure or the prior art or the existing technology can be displayed on the software interface. The second area displays the contents of the technical disclosure, which allows the user to view the relevant technical contents of the technical disclosure, the prior art, etc. while conceiving the elements of the claims and their relationships, facilitating the editing of the set of elements, the set of relationships, and the definition of the elements and their relationships involved in each claim in the first area 1000. The second area displays the contents of the prior art, which facilitates the comparison of the invention and the prior art and better understanding of the invention. Optionally, the second area can switch between displaying the claims, the technical disclosure, and the prior art, or display at least two of the above-mentioned three in parallel. If the claims have not been formed, only the technical disclosure and the prior art can be displayed. Optionally, the second area can be hidden in the interface, such as being hidden and then being triggered to be displayed in the interface by moving the mouse to the side bar, while the first area 1000 is reduced in size and the visible information in the first area 1000 is adaptively adjusted in size and position to be adaptively displayed in the first area 1000. Optionally, the technical disclosure and the prior art of the second area can also be set to be hidden and invisible, and the corresponding technical disclosure or prior art can be triggered to be displayed in the interface or the display area can be enlarged or reduced by moving the mouse to the corresponding position of the side bar or the top or bottom area. Of course, the size of the first area 1000 can also be adjusted or even hidden. Optionally, as shown in FIG. 10, a tree diagram or a flowchart of the chain structure can be edited in the first area, and the information and their relationships in the edited chain structure diagram form the data structure 3000 of the present application. Optionally, the data structure 3000 in the first area is displayed in the form of a tree diagram or a flowchart. Taking the tree diagram as an example, the tree diagram has a root node 3100 and branch nodes 3200, the root node 3100 is used to edit and display the subject matter of the claims for protection, such as the largest product name involved in the invention. For example, the invention is made for a mobile phone, and the root node 3100 can be edited as a mobile device or a smart terminal. The first-level branch nodes can be the second level of the root node 3100, and the second layer elements such as a circuit board, a camera assembly, and a communication assembly can be edited. The second-level branch nodes are the third level of the root node 3100, and the third layer elements such as the radio frequency circuit and the antenna in the communication assembly can be edited, and so on, until all the elements involved in the invention are edited, and a complete structure decomposition is formed in the form of a tree diagram. The root node 3100 and the branch nodes are collectively referred to as nodes herein. The process of completing the tree diagram is also the process of repeatedly considering, repeatedly modifying, and gradually improving the claim architecture and layout by the agent in a visual manner. Compared with directly drafting the claims in the text editing software, the embodiment of the application makes it easier to start drafting, clearer, more comprehensive in thinking, and less error and omission. Optionally, in the process of editing the data structure, the software receives a first instruction of text editing, node adding, node deleting, or changing the position of a selected node in the tree diagram or flowchart; that is, the first instruction of text editing, node adding, or changing the position of the selected node can be directly generated by right-clicking the first area or clicking the corresponding operation button of the interface. The software responds to the first instruction to perform text editing, node adding, node deleting, or changing the position of the selected node in the tree diagram or flowchart. For example, right-clicking the first area and selecting the option “node adding” causes the software to add a node at the next level of the node at the position where the right-click is located or the node pointed to by the right-click or at the previous level. Alternatively, dragging the third-level node a to the area where the second-level node is located generates the first instruction, and the software responds to the first instruction to move the third-level node a to the position of the second-level node, so that the node a becomes a second-level node. The original second-level node can be stored at the same level as the node a or can be swapped with the node a. Meanwhile, the software can automatically adjust the other nodes of the tree diagram or the shape or size of the tree diagram to adapt to the change of the node a. Optionally, after the editing of the lowest-level node is completed, the software can automatically add lower-level nodes to be edited on the basis of the lowest-level node to facilitate possible node addition, without the need for manual node addition. When reading the data structure subsequently, the nodes with unedited text are automatically ignored. Optionally, after the text editing, node adding, or changing the position of the selected node in the tree diagram or flowchart, the software performs the following operations: The software receives a second instruction of defining the relationship of the selected node in the tree diagram. After the elements of the claim are edited in the tree diagram, the relationship of the elements needs to be defined, and therefore the software receives the corresponding second instruction. Referring to FIG. 11, the software responds to the second instruction to draw a line connecting the selected nodes in the first area, one end of the line being connected to the selected node such as A11 and the other end being a text input box, and the cursor being located in the text input box. If the number of selected nodes is greater than or equal to two, the software draws one or several lines connecting all the selected nodes in the first area, each end of the several lines being connected to the corresponding selected node such as A12, the other end being a text input box, and the cursor being located in the text input box. If the drawn line connects two or more nodes, a text box is displayed at the middle of the line or at the adjacent end of several lines, and the cursor is located in the text input box. For example, in FIG. 11, one end of the line is connected to B2 and the other end is connected to A12, and a text box is displayed at the middle of the line or an area in which text can be edited is formed. Optionally, if the first area cannot accommodate the tree diagram, the content is automatically reduced in size or part of the content is automatically hidden, or the range of the first area is automatically expanded. The function of drawing a line is to visually see which element or which elements have a relationship, and the line defines the link of the element and its relationship and the relationship link between different elements, which form part of the data structure. The appearance of the text box and the cursor in the text box simplify the operation, and the cursor does not need to be moved and the editing state does not need to be entered. After the line is drawn, the editing state is entered by default, and text can be directly input in the text box. Optionally, the number of text input boxes is one or two. For the case of one, the text input box is divided into a relationship input area and a technical problem input area. For the case of two, one text input box is used to input the relationship and the other text box is used to input the technical problem. For example, the keyboard "shift" or other control keys can select any two or more nodes, and the software automatically draws a line to connect the two nodes to the corresponding text boxes, and the cursor is automatically placed in any position in the text box, or clicking one of the text boxes can input the relationship description of the selected nodes, and clicking the other text box can input the technical problem to be solved. Optionally, for the tree diagram, several technical problems and relationships can be edited, and the order of each technical problem can be set, or "first technical problem", "second technical problem", etc. can be directly input in different text boxes to represent different claims, so that the software can recognize them. Each technical problem corresponds to a claim, and after the order of the technical problem is defined, the software can sort the subsequent formed claims according to the order of the technical problem. The data structure of the embodiments of the present application is not limited to the above form, and can also include a graph and elements and relationships on the graph. For example, referring to FIG. 12, a user can directly draw a line on the graph, the starting point of the line is the position of the element corresponding to the product in the graph, and the end point is outside the range of the product. The user can edit the name of the element at the end point. In order to reflect the hierarchical relationship of the elements, the names of multiple parallel elements can also be used as the starting point, and the line can be continued to point to the name of the same upper element. In addition, the line can be continued to edit the relationship of the elements, and finally the data structure is formed. S200: reading the data structure; After the data structure is edited or obtained, the system generates a third instruction by clicking the "OK" button or pressing a function key, etc. The software reads the data structure in response to the third instruction, or directly reads the data structure after obtaining the data structure, and parses the data structure to obtain a set of elements and a set of relationships constituting each claim. If the data structure contains information of a technical problem, the technical problem, a set of elements corresponding to the technical problem, a set of relationships, and the sorting information of the technical problem are obtained. Optionally, after parsing the data structure, a protected subject in the data structure is determined according to a second preset rule, and a first claim is determined under the determined protected subject, and a set of elements and a set of relationships associated with the first claim are read. For example, the second preset rule includes a first preference for finding a technical problem ranked first, and the first technical problem and a set of elements and a set of relationships associated with the first technical problem are read in the parsed data structure. The technical problem ranked first means that it is determined as a protected subject when the data structure is edited, and therefore finding the technical problem ranked first means that the first claim is determined under the determined protected subject. Then, the second preset rule is followed to read the technical problems ranked second, third, etc. and a set of elements and a set of relationships associated with the technical problems in sequence. Optionally, there can be two or more protected subjects in the data structure, and each protected subject can be marked with an independent claim, such as being marked with different letters, the serial numbers of the first set of claims are named in the form of Am, where m is an Arabic numeral. The serial numbers of the second set of claims are named in the form of Bn, where n is obtained by adding one to the Arabic numeral of the last claim in the first set of claims. The two sets of claims are distinguished by the first letters A and B. The serial numbers of other sets of claims are named in the same way. The order of the independent claims can be determined according to the order of the first letters, and the independent claim following the first letter is placed after the independent claim corresponding to the set of claims preceding the first letter. In this way, the software can conveniently and accurately sort the claims and determine which is an independent claim and which is a dependent claim subordinate to which independent claim. If a combination of letters is used to mark the claims, the sorting of the technical problems can be omitted. Of course, the naming method of the serial numbers is not limited to this, as long as the software can recognize the predetermined rule. S300: Forming a corresponding claim according to a first preset rule for a set of elements and a set of relationships corresponding to each claim; The first preset rule is a rule of forming a claim meeting at least part of the formal requirements by using the set of elements and the set of relations. For example, the format, the order of sentences, the punctuation marks and the like required by the claim. Optionally, the order of appearance of the elements in the set of elements associated with the first claim in the first claim is determined by using the set of relations associated with the first claim. For example, refer to Fig. 12 and the data structure in the following table: Table 1: Set of elements Table 2: Set of relations Table 1 and Table 2 are optional data structures in the embodiments of the present application. In Table 1, the largest protection subject is the advertisement module in column a, the second level elements are the drawing board, the frame, the pin, the bolt, the elastic member and the safety rope in column b, the third level elements are the elements in column c, and the fourth level elements are the elements in column d. Among them, the strip board and the decoration in column c belong to the drawing board in column b, and the relations of other elements are similar. When reading the data structure in the form of a table, the software will parse that the strip board and the decoration belong to the drawing board according to the area relations of the table and the area positions of the elements in the table, and that the first mounting hole belongs to the strip board. The mounting plate belongs to the frame, the plate hole and the first rope hole belong to the mounting plate, and so on. The set of relations in Table 2 reflects the different relations of the elements in Table 1. Table 1 and Table 2 are associated by the addresses such as B6 in column b and the like. In Table 2, the first technical problem, i.e., the technical problem corresponding to the first claim, is "easy to disassemble and assemble", and the set of relations is "B6 is connected with b1 at one end and passes through d3 at the other end, b8 is inserted into d7 and is located on the side of c3 opposite to b1". According to the order of appearance of the elements in the first claim, the template corresponding to the first claim is used to form the organization structure of the elements and the organization structure of the relations of the first claim. The first claim is an independent claim, and therefore its template can be defined as follows: The above italicized and underlined serial number "1" and "advertising module" are variables, and the rest are template constants. The serial number is given according to a rule, such as "1" for the first claim, "2" for the second claim, and so on. The "advertising module" is the protected subject matter, which can be obtained from the set of elements in Table 1, with the leftmost element in the table, or the root element in the chain structure, as one of the protected subject matters. Of course, elements in the middle area of the chain structure can also be used as protected subject matters, which can be defined and set according to the situation. The template ends with a period, and only one period is allowed for one claim. The period is placed at the end of the template. Between "characterized in that" and "comprising", and between "comprising" and ".", there is a fillable space for the remaining technical features. Optionally, the set of elements includes a set of product elements, the set of relationships includes a set of element relationships, the first claim is an independent claim, and the next line of the independent claim and before the period is defined as filling in the upper necessary elements (elements) of the lowermost necessary elements (elements). Optionally, in Table 2 "set of relationships", the necessary elements and their relationships represent the lowermost necessary elements and their relationships, and if necessary, the upper necessary elements and their relationships. It should be noted that whether the "necessary elements and their relationships" are necessary depends on the technical problem to be solved, so the "necessary elements", "necessary elements" and "necessary relationships" referred to in this text all refer to "elements", "elements" and "relationships" respectively. Different technical problems have different "necessary elements", "necessary elements" and "necessary relationships". First, determine the lowermost necessary elements, and then use the set of relationships associated with the first claim to determine the order of elements in the set of elements associated with the first claim in the first claim. The method can include: Using the set of relationships associated with the first claim, all elements in the first claim are constructed into a subject-predicate ordering link, and the elements in the subject-predicate ordering link are sorted in reverse order of their sequence. For example, all elements and their relationships in the first claim are "B6 is connected to b1 at one end, passes through d3 at the other end, b8 is inserted into d7 and located on the side of c3 opposite b1". Optionally, to ensure that the constructed subject-predicate ordering link meets the requirements, the software automatically checks whether a closed loop is achieved in each claim. The closed loop condition includes whether all elements are associated with at least one other element within a claim or between different claims, and whether the subject-predicate ordering link has only one, otherwise a prompt of "cannot form a closed loop" is given. The software further analyzes all elements and their relationships in the first claim and obtains two links "b8-d7(b1), b6-b1(d3)". It is analyzed that the two links constitute a substantial single link through b1, which meets the requirements. The lowermost necessary elements of the single link or the two links are b1, d7 and d3. It should be noted that the lowermost necessary element referred to here refers to the lowermost element among the elements involved in the current claim. The lowermost necessary elements are sorted. Since b1 is the object of the two links, it is ranked first. d7 and d3 are in a parallel relationship in the link, so the order of appearance of the necessary relationship is used for sorting. d3 is ranked after b1, and d7 is ranked after d3. All upper necessary elements of the lowermost necessary elements are determined. The upper necessary element of b1 is b1 itself. The upper necessary elements of d7 are c7 and b6. The upper necessary elements of d3 are c3 and b3. The principle of determination is to obtain from the upper node for the chain structure. Then, the software determines the sorting of all upper necessary elements: b1, b3, b6 and b8. Continue to fill in the necessary elements and their necessary relationships determined above in the independent claim writing template. Optionally, according to the sorting of the lowermost necessary elements, list the inclusion relationships of the upper necessary elements and the corresponding necessary relationships in sections. Optionally, in the inclusion relationship, "have" or "form" is used for holes, cavities, recesses, gaps, spaces, channels, airways, liquid channels and openings. The default is "include". According to the above method, the first claim is as follows: The italicized and underlined contents above are variables, and the rest are template constants. No "said" needs to be added in front of the first-appearing element. The "said" needs to be added in front of the non-first-appearing element. It can be understood that the first preset rule is not unique and can be set differently as needed. For example, it is not necessary to consider whether a closed loop is achieved in each claim. The condition of the closed loop does not necessarily include only one claim, and whether all elements are associated with at least another element between different claims, etc. For example, a smart large model can be introduced to input elements and their relationships to obtain a claim. Optionally, the second claim is continued to be completed subsequently. Specifically, the set of elements and the set of relationships associated with the second claim are read under the determined protected subject matter to determine the second claim. Specifically, scan table one and table two to obtain the second technical problem; the determined protected subject is the element within the chain structure node. Obtain the necessary elements associated with the second technical problem and their relationship "B10 elastically abuts between c3 and b8"; The second claim and its subsequent claims need to be determined as independent claims or dependent claims before being constructed, because the templates of the two are not the same. The way to determine whether it is an independent claim or a dependent claim has been described before. According to the first preset rule, the set of elements corresponding to each claim and the set of relationships are formed into a corresponding claim, which includes: According to the element sorting method of the first claim, determine the order of the elements appearing in the second claim, apply the template corresponding to the second claim to form the organization structure of the elements and the organization structure of the relationships of the second claim, which specifically includes: i. Fit into the writing template of the dependent claim, and give the serial number according to the rule: ii. The next line under the title of the dependent claim, if there is no new lower level necessary element, directly fill in the lowest level necessary element and its necessary relationship, if there is a new lower level necessary element, fill in the upper level necessary element of the lower level necessary element, and according to the same processing process of the independent claim, we get: The above italicized and underlined content is a variable, and the rest is a template constant. The dependent claim is different from the independent claim, and the reference relationship needs to be defined. Optionally, the method for determining the reference relationship between claims includes: Find another claim that is closest to the elements and relationships associated with the current claim and has the same protected subject, and determine the reference serial number of the current claim as the serial number of the found another claim. For example, the current claim, i.e. claim 2, contains c3 and b8, and c3 and b8 appear in claim 1, not in other claims (i.e. the set of elements and relationships corresponding to the technical problem, defined by the technical problem before the claim is formed, not mentioned later). Therefore, the elements and relationships of claim 2 are closest to claim 1, and the same protected subject, so claim 2 needs to reference claim 1. Optionally, in addition to the above embodiments, the way to determine whether an independent claim or a dependent claim can also include: i. Get all elements from the set of elements and the set of relationships associated with the current claim; ii. judging whether a new element appears in the current claim, while the relationship involves elements already appearing in other claims and the protected subject matter is the same, if both are, defining as a dependent claim, if not, not a dependent claim of the same protected subject matter. In the same protected subject matter, all elements are newly appearing, determining as an independent claim. For example, the elements and relationship of the current claim, claim 2, include "B10, elastically abutting between the c3 and the b8", a new element "B10" appears, while the relationship "elastically abutting between the c3 and the b8" involves elements "c3 and b8" already appearing in claim 1, and the protected subject matter is the same, defining as a dependent claim. After claim 2 is completed, for the remaining claims, continue reading the data structure, including: traversing the data structure, determining the remaining protected subject matters in the data structure in turn according to a second preset rule, forming all claims under each protected subject matter; The second preset rule includes sorting the protected subject matters from small to large or from large to small according to the containing relationship between the protected subject matters. For example, the protected subject matter "advertising module" contains "frame", and the sorting from large to small is that a set of claims of the protected subject matter "advertising module" is constructed first, and then a set of claims of the protected subject matter "frame" is constructed. For method claims, optionally, the set of elements includes a set of method steps, and the set of relationships includes a set of time before and after relationships between the steps, and the set of elements associated with the first claim is determined according to the set of relationships associated with the first claim, including: The time before and after relationships of all steps in the first claim are determined according to the set of relationships associated with the first claim. As described above, the elements of the method claims generally include steps of the method, and the relationships of the method claims generally include the sequence of the steps. The sequence of the steps can be directly obtained by reading the data structure representing the arrangement of the steps, that is, which step is in front and which step is behind, or determining that the b step is between the a step and the b step, and the like. For the case of sorting technical problems by serial numbers in the data structure, determining the first claim under the determined protected subject matter includes: Searching for the claim with the first serial number under the determined protected subject matter as the first claim; or For each element, search the set of elements, the set of relations of all claims under the determined protected subject matter, obtain all elements following the expression forms of "include", "contain", "have", "form", "consist of", "consist essentially of", or "provided with" vocabulary, all obtained elements are defined as the set of first appearing elements, the remaining elements are defined as the set of later appearing elements, and all elements in the set of elements, the set of relations of all claims are checked, the set of elements, the set of relations of claims that only appear the first appearing element set are taken as the first claim, wherein "*" refers to a word or symbol, formula. For the relationship description of the element set, the relationship set, such as the expression forms of "include", "contain", "have", "form", "consist of", "consist essentially of", or "provided with", it means that the claim adds new elements, and the first claim is usually an independent claim, so it must contain new elements, and generally only appears the first appearing element, and there is no element of another claim in the same set of claims, so such claim is determined as an independent claim. For some environmental elements (environmental features), they will not be introduced in the form of "include", "contain", "have", "form", or "provided with". Therefore, by appearing such environmental features in all claims but not being introduced in this way, the logic can be recognized as not being an element of another claim in the same set of claims, to determine the independent claim. For the relationship description of the element set, the relationship set in which the elements exist, determine the first claim under the determined protected subject matter, including: Exclude the first appearing element corresponding to the first claim, continue to process the set of elements, the set of relations of the remaining claims in the same way, and determine the set of elements, the set of relations of claims that only appear the first appearing element in the set of elements, the set of relations of the remaining claims as the second claim, and determine that the second claim refers to the first claim. Wherein, the set of elements of claims that only appear the first appearing element indicates that the current claim adds new elements on the basis of the first claim, and the current claim is a further limitation of the elements appearing in the first claim, so the current claim is set to refer to the first claim; or In the set of elements, the set of relations of the remaining claims, if the set of elements, the set of relations of the claim of the first occurrence does not occur again, it indicates that no new element is added, and the current claim is only a limitation of the function or more relations of the cited claim, at this time, the set of elements, the set of relations of the claim with the same elements as the cited claim is taken as the second claim or the claim arranged in sequence, and the cited claim in front is cited. Optionally, when the dependent claim is specifically constituted, the reading data structure comprises: determining the second claim under the determined protected subject, and reading the set of steps, the set of relations associated with the second claim; The set of elements, the set of relations corresponding to each claim is formed according to the first preset rule to form the corresponding claim, which comprises: According to the step sorting method of the first claim, the order of the steps appearing in the second claim is determined, the template corresponding to the second claim is applied to form the organization structure of the steps of the second claim, all the determined protected subjects are traversed, and all the claims under all the determined protected subjects are formed. In this way, all the claims under all the determined protected subjects are formed by traversing the determined protected subjects. Optionally, after all the dependent claims are completed, it is checked whether there is a remaining protected subject, and if there is, the previous logic is continued to process until the writing of all the claims is completed. Different from the prior art, the embodiment of the application mainly defines the data structure, so that the claims can be formed according to the first preset rule by using a computer, the quality of the claims can be improved, the claims themselves can be more in line with relevant regulations, and the time for writing the patent-related text can be greatly reduced and the work efficiency can be improved. Specifically: 1) The user can be guided to gradually sort, arrange and refine the unclear and incomplete understanding and recognition in the mind into a data structure with clear structure and rigorous logic by means of the constitution and formation method of the data structure, so as to reduce the difficulty of technical understanding and the difficulty of claim writing; 2) As long as the data structure is edited according to the rules, the first preset rule can be used to automatically construct part or the whole set of claims in line with relevant regulations and user requirements, so that the efficiency of claim writing is greatly improved; 3) After the claims are constructed, the user can modify the claims in the software by comparing the data structure and the claims; if the data structure is updated, the software can automatically construct the claims based on the updated data structure again, so that the modification efficiency is greatly improved. Optionally, after forming the corresponding claims, the other parts of the specification can be completed. Optionally, the elements in the specification can be further labeled, and the method comprises: i. Using the relationship between elements, generate a label strategy that matches the relationship; Optionally, if the relationship between different elements is an upper and lower layer containing relationship, the generated label of the corresponding upper layer element is less than the label of the corresponding lower layer element, and if the relationship between different elements is a same layer parallel relationship, the generated labels of the corresponding different elements belong to the same order of magnitude. For example, A includes B, a label strategy can be given that A and B belong to the same label range, such as giving A a label of 100, and B a label of 110, so that label 110 and label 100 belong to the number range of 100 and above, and a certain number below, and the label of B is greater than the label of A. ii. Labeling some or all elements of the patent-related text according to the label strategy. Search the full text or selected part of the specification, for example, search for A, find A and replace it with A100, or directly insert the 100 label after A. The above-mentioned "search" and "replace" functions provide a faster way than searching one by one for the user, but as the patent work develops towards higher quality requirements, the work of the applicant and the patent agency becomes more standardized, and the patent-related text may need to be modified and improved repeatedly before it is submitted as a patent application through official channels. Each modification and improvement may require several or even more search and replace operations, and the user needs to frequently operate the text editing software interface, open the "search" and "replace" windows, enter the content to be searched and replaced in the input field of the window, and then click "search" or "replace all". The work efficiency is still relatively low. Moreover, the current "search" and "replace" functions of the text editing software can only match 100% exactly, such as replacing "abc", which can only find other content with "abc" string, but cannot find content with label such as "a11bc", and cannot find content such as "abc", and vice versa. There are many matching blind spots that need to be found out and modified consistently by the user, which cannot further improve the work efficiency. Optionally, the search and replace method comprises: S11: receiving a first instruction for the patent-related text, the first instruction comprising a manual revision end instruction. The artificial revision end instruction refers to an instruction reflecting that the user completes at least one revision operation. The artificial revision end instruction can be actively issued by the user or passively triggered in the process of the user's operation, such as after inserting the 100 label after A, moving the cursor to the next line, and passively triggering the artificial revision end instruction. In an embodiment, receiving the first instruction for the patent-related text comprises triggering the artificial revision end instruction in response to at least one of the user clicking an operation item in the user interface, pressing a preset key, or inputting a preset instruction. Specifically, the user actively clicks at least one of the operation item, presses the preset key, or inputs the preset instruction, which means that the user confirms that the current revision operation has been completed and wants to trigger the artificial revision end instruction. The operation item can be a button in the user interface, such as the "run" button on the software interaction interface. The user can interact with the operation item through common operation methods such as clicking and sliding. Pressing the preset key can be a certain physical key position on the keyboard or other input. For example, the "F1" key on the keyboard can be set as the preset key. The preset key can also be a combination key, such as setting the artificial revision end instruction to be triggered when Ctrl+F1 on the keyboard is pressed at the same time. Inputting the preset instruction can be that the user inputs a specific string of characters to trigger the artificial revision end instruction. The input preset instruction can be input in a specific preset area, and only input in the preset area to trigger the artificial revision end instruction. The input preset instruction can also detect the input content of the user in any position in real time, and trigger the artificial revision end instruction as long as the condition is met. For example, it is set to trigger the artificial revision end instruction as long as a specific format of command symbol such as "run modification follow" is detected. Through the above setting, the user can first repeatedly adjust the content to be revised, and actively trigger the artificial revision end instruction only after the artificial revision is completed or after a certain stage. The content before revision can be processed in batches in one operation, which is more efficient. In addition, the artificial revision end instruction is actively issued by the user, and the user can clearly master the content and number of revisions, so that unexpected operations beyond the user's expectations are less likely to occur, and the use is stable. In another embodiment, receiving the first instruction for the patent-related text comprises triggering the artificial revision end instruction in response to a preset change in the user's cursor position. Specifically, when the user changes the position of the text input cursor through the keyboard or mouse, such as moving the cursor position from the current line to another line, it means that the user has performed an operation. Each operation of the user is detected and triggers the artificial revision end instruction, that is, the artificial revision end instruction is passively triggered in real time when the user operates. Through the above setting, the user revises the completion to have triggered the artificial revision end instruction in real time, and the user does not need to additionally trigger the artificial revision end instruction operation after each revision is completed. The user operation steps are less, and the use is convenient. The preset change can be moving to the next character position or the previous character position, or jumping to other character positions. In an embodiment, the preset change is detecting that the user cursor occurs line feed. Thus, on the basis of real-time triggering, a certain number of redundant triggering times can be reduced, and processing speed is considered. S12: In response to the first instruction, determining the pre-revision content in the patent-related text. Specifically, the pre-revision content refers to the content before the change caused by the user operation. The pre-revision content should at least include one character or format symbol. The pre-revision content can contain changed content formed by adding, deleting, changing, etc. For example, when the user revises “abcd” in the patent-related text to “abzz”, the pre-revision content at least includes “cd”. Or, the user revises “abcd” in the patent-related text to “ab”, and the pre-revision content at least includes “cd”. For another example, the user revises “ab” in the patent-related text to “abcd”, and the pre-revision content at least includes “ab”. When responding to the first instruction, and no pre-revision content is detected, the end instruction can be triggered, and the subsequent steps are no longer continued. In an embodiment, determining the pre-revision content in the patent-related text specifically includes: Determining the third content that has been revised in the patent-related text, such as the background record of the last or several times of the third content that has been revised, or determining the third content with a revision format in the patent-related text. The revision format can be a format reflecting the revision operation in other software. For example, the revision format can be a format generated by the “revision” function in a word processing software. Determining the third content can be to take a number of continuous characters with a revision format as a third content. For example, the patent-related text contains “abcdefg”, wherein “ab” and “efg” have revision formats, “ab” is taken as a third content, and “efg” is taken as another third content. Based on the third content, the fourth content of the minimum matching unit is determined, the fourth content length is greater than or equal to the third content length, and the fourth content is taken as the pre-revision content. The fourth content can be determined directly as the third content. For example, when the third content is "ab", the fourth content is also determined directly as "ab". The fourth content can also be determined by searching for a complete word or sentence segment corresponding to the position of the third content. When determining the fourth content, the position of the third content can be detected, and the text before and after the position of the third content can be expanded based on the position of the third content and according to semantics until a complete word or sentence segment is obtained as the fourth content. For example, the patent-related text contains "abcdefg" with complete semantics, and "de" is determined as the third content. At this time, the text before and after "de" can be expanded until "abcdefg" with complete semantics is obtained as the fourth content. When several third contents are adjacent, and the complete word or sentence segment obtained by expanding the several third contents overlaps, the results of the word or sentence segment obtained by expanding the several third contents can be combined to form only one fourth content. For example, the patent-related text contains "abcdefg" with complete semantics, and "bc" and "de" are determined as the third content, respectively. Then, "abcdefg" can be obtained by expanding the text before and after "bc", and "abcdefg" can be obtained by expanding the text before and after "de". At this time, it is detected that the results of expanding the text before and after "bc" and "de" are the same, and therefore, only one fourth content "abcdefg" is formed instead of two fourth contents "abcdefg". It can be understood that one of the purposes of determining the fourth content is to make the matching more accurate, and therefore, the specific strategy of expanding the fourth content based on the third content can be set according to actual conditions. In addition, when the revision type is change, the fourth content should be determined based on the text before the revision, and the content after the revision can be ignored. Through the above settings, the user can directly indicate the position of the revision through the revision format, background recording, and other methods of existing other software. In addition, the fourth content is determined based on the third content, and the fourth content is used as the content before the revision, which can make the content before the revision more unique in the patent-related text, and the matching is less likely to form noise and the positioning is accurate. In a specific embodiment, determining the fourth content of the minimum matching unit based on the third content includes: The first delimiter appearing in the text before and after the third content is searched as a starting point, and the first delimiter appearing in the text before and after the third content is searched as an end point. The content between the end points of the text before and after the third content is used as the fourth content. The first separator can be a comma, a period, a semicolon, a space, a paragraph separator, or other special symbols, etc. For example, the patent-related text contains "abc, defgh, ijk.", where "fg" is the third content, and the fourth content is determined as "defgh" by extending the context before and after until the first separator comma is reached as the cutoff point. In this way, the text containing the third content is cut off according to the first first separator appearing in the preceding context and the following context, and the sentence segment containing certain unique features can be easily obtained, so that when the fourth content is taken as the minimum matching unit, it is not easy to produce false matching noise. In another specific embodiment, determining the fourth content of the minimum matching unit comprises: Taking the third content as the center and the preset length of the preceding context and the following context as the boundary, the fourth content of the minimum matching unit is determined as the preset length of the preceding context and the third content and the preset length of the following context. The preset length can be a value preset in advance, for example, it can be 3, 4, 5, etc. The preset length can be set by the user or can be a fixed value. The fourth content is obtained by concatenating the characters of the preset length of the preceding context, the third content, and the characters of the preset length of the following context in turn. In addition, when the length of the preceding context and the following context of the paragraph in which the third content is located is less than the preset length, the maximum preceding context and the following context that can be cut off in the paragraph in which the third content is located are determined as the fourth content of the minimum matching unit. For example, when the third content is located at the end of the paragraph, the preceding context is cut off according to the preset length, and the following context is cut off according to the actual remaining character length of the paragraph. Through the above setting, the fourth content of the minimum matching unit can be obtained simply and quickly, the function is simple to implement, and the processing speed is fast. In another embodiment, determining the pre-revision content in the patent-related text comprises: Taking the content selected by the user in the patent-related text as the pre-revision content. When the selected range of content contains content that has been changed by user operation, the selected range of content as a whole is the pre-revision content and is taken as the minimum matching unit. When the selected range of content does not contain content that has been changed by user operation, i.e. the selected range of content is all original text, an end instruction can be triggered, and the subsequent steps are no longer continued. Through the above setting, the determination of the pre-revision content in the patent-related text does not need to rely on the revision format of other software, and the application range is wider. In another other embodiment, determining the pre-revision content in the patent-related text comprises: Taking the third content determined after comparing the current version of the patent-related text and the historical version of the patent-related text as the pre-revision content. The historical version of the patent-related text can be another old version of the file, an old version of the patent-related text in other text positions in the same file, etc. By comparing the historical version of the patent-related text with the current patent-related text, the part of the current patent-related text that has changed relative to the historical version of the patent-related text can be obtained. The changed part can be used as the pre-revision content. Through the above setting, the pre-revision content in the patent-related text does not need to rely on the revision format of other software, and does not need to be selected and operated. All changed parts in the patent-related text can be obtained at one time, especially suitable for the case where the changed part is not marked and only the historical version content is left. It is convenient to use and has a wide range of applications. S13: Match the remaining part or the selected part of the patent-related text with the pre-revision content. Specifically, matching the remaining part or the selected part of the patent-related text with the pre-revision content can obtain the position of the content in the remaining part or the selected part of the patent-related text that needs to be modified synchronously corresponding to the pre-revision content. For example, the pre-revision content is “abc”, and all “abc” in the remaining part or the selected part of the patent-related text is matched. The matching can be performed in the entire remaining part of the patent-related text, that is, full-text matching. The matching can also be performed only in the selected part, and the other part is not processed. The specific implementation of the matching can be any character matching technology that can locate the position, which will not be described here. In an embodiment, matching the remaining part or the selected part of the patent-related text with the pre-revision content comprises: In response to not matching the target content based on the pre-revision content, a second delimiter in the pre-revision content is found. Specifically, in the process of matching the remaining part or the selected part of the patent-related text with the pre-revision content, the case of not matching the corresponding result can occur. In this embodiment, the pre-revision content in the patent-related text will have corresponding content in the remaining part of the patent-related text as a high probability case. The case of not matching the corresponding result is first identified as an abnormal case. The case of not matching the corresponding result can be caused by character changes, long sentence splitting, syntax changes, etc. Therefore, in this embodiment, a second delimiter in the pre-revision content is first found as a judgment basis for further processing. The second delimiter can be a comma, a semicolon, a period, a space, etc. In response to the absence of the second delimiter, a conjunction in the pre-revision content is replaced with the subject of the previous sentence, or the subject of the previous sentence is inserted at the beginning of the pre-revision content. When the second separator is not found, it means that the probability of not matching the corresponding result due to long sentence splitting is low, and the probability of not matching the corresponding result due to syntax change should be considered first. For example, a sentence in the pre-revision content replaces the subject of the sentence with a conjunction word due to writing habits when writing, and the subject is not replaced with a conjunction word when writing the remaining part of the patent-related text. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the conjunction word in the pre-revision content with the subject of the previous sentence, so that the matching result is not easily missed. For example, the first sentence of the pre-revision content is "a includes b.", the pre-revision content is "also includes c.", and in the remaining part of the patent-related text, it is recorded as "a includes b" and "a includes c". At this time, it may not be possible to get the matching result directly using "also includes c.". After replacing "also" with the subject "a" of the previous sentence, the correct matching can be achieved. In response to the presence of the second separator, the pre-revision content is split into a preceding part and a plurality of subsequent parts based on the second separator, and the subject of the preceding part is replaced with a keyword in each subsequent part or inserted at the beginning of each subsequent part. When the second separator is found, it means that the probability of not matching the corresponding result due to long sentence splitting is high. For example, two short sentences in the pre-revision content are connected by a colon, "and" and the like, and when writing the remaining part of the patent-related text, they are split into two short sentences, and the colon, "and" and the like in the subsequent short sentence are replaced with the subject. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the keyword in each subsequent part with the subject of the preceding part, so that the matching result is not easily missed. For example, the pre-revision content is "a includes b, and includes cd.", and in the remaining part of the patent-related text, it is recorded as "a includes b. a includes cd.". At this time, due to long sentence splitting, it may not be possible to get the matching result. After replacing the "and" in the subsequent part with the subject "a" of the preceding part, the correct matching can be achieved. The keyword can be set according to the actual situation, which can be a common word or symbol used to separate long sentences, which is not limited here. The position of the pre-revision content and the remaining part of the patent-related text can be set according to the actual situation. For example, in a specific embodiment, the pre-revision content is located in the claims of the patent-related text. The remaining part or selected part of the patent-related text is at least partially located in the specification of the patent-related text. In another specific embodiment, the pre-revision content is located in the specification of the patent-related text, and the remaining part or selected part of the patent-related text is at least partially located in the claims of the patent-related text. S14: revising all target content or part of the target content obtained by matching based on the post-revision content, the post-revision content being the content obtained based on the pre-revision content. Specifically, all target contents or part of the target contents will be revised to be consistent with the pre-revision contents. All matched target contents can be uniformly revised, or only part of the matched target contents can be revised. When the pre-revision contents are determined by revision formats, the target contents can also be revised by the revision formats when the target contents are revised, so as to facilitate the user to observe the position of the target contents to which the revision is applied. In an embodiment, a positioning mark is formed at the revised target content. The user can quickly locate to the revised target content through the positioning mark. The positioning mark can be a comment in the word editing software and the like. In an embodiment, the following is included before or simultaneously with the matching of the remaining part or the selected part of the patent-related text with the pre-revision contents: The pre-revision contents, the remaining part, and the selected part that are matched are subjected to de-noising processing. The de-noising processing includes at least one or all of the following: removing punctuation marks, removing reference signs, removing definite articles, and removing demonstrative pronouns. The pre-revision contents can interfere with the matching due to the difference in the form of the words from the target contents. For example, when the pre-revision contents are located in the claims and the contents are "a includes b", the " " in the pre-revision contents needs to be removed, and when the remaining contents in the patent-related text contain reference signs, for example, the contents are "a11 includes b12", the reference signs in the remaining contents in the patent-related text also need to be removed, and then the pre-revision contents are matched with the remaining contents in the patent-related text. Through the above setting, the matching is not easily disturbed by noise when the pre-revision contents are matched with the remaining part or the selected part of the patent-related text, and the matching is accurate. Further, in an embodiment, all target contents or part of the target contents obtained by matching are revised based on the post-revision contents, and the post-revision contents are contents obtained based on the pre-revision contents, including: The post-revision contents are determined based on the pre-revision contents. The post-revision contents are parts of the patent-related text that change based on the pre-revision contents. When the pre-revision contents are formed by the user changing, the post-revision contents can be new contents after the change. For example, when the user revises "abcd" in the patent-related text to "abzz", it can be determined that the post-revision contents are "zz". When the pre-revision contents are formed by the user deleting contents, the post-revision contents can be empty. For example, the user revises "abcd" in the patent-related text to "ab", and it is determined that the post-revision contents are empty. When the pre-revision contents are formed by the user adding contents, the post-revision contents can be newly added contents. For example, the user revises "ab" in the patent-related text to "abcd", and it can be determined that the post-revision contents are "cd". The first content after being denoised is substituted for the second content in the target content, and the remaining content in the target content other than the second content is not changed. In this embodiment, the second content is the content in the target content that needs to be modified. When the target content is revised, only the part of the target content that is changed before the revision is changed, and the entire target content is not directly changed. For example, in the patent-related text, "a includes b" is revised to "c includes b", and the target content is "a11 includes b12". At this time, the revised content is denoised to obtain "c", and it is confirmed that the second content is "a11". The revised content is replaced by the second content to obtain "c includes b12". In this way, the remaining content in the target content is less likely to be changed unexpectedly and cause errors, and the revision is accurate. In one embodiment, after the first content after being denoised is substituted for the second content in the target content, and the remaining content in the target content other than the second content is not changed, it further includes: Adding a figure reference to the revised content. For example, after obtaining "c includes b12" in the above embodiment, the corresponding figure reference "13" is added to "c" to obtain "c13 includes b12". In this way, the revised content can be automatically completed with figure references, and it is less likely to be omitted and cause formal errors. To more clearly illustrate the workflow of the patent-related text writing assistance method, the following examples are given. Please refer to FIG. 13, which is an interface diagram of the first interface of an embodiment of the patent-related text writing assistance method. In one embodiment, the patent-related text writing assistance method can have a first interface 100. The user can edit the patent-related text in the first interface 100. The patent-related text includes at least a first part 10 and a second part 20. As shown in FIG. 13, the user can edit the original text "ABCD" in the first part 10, and then click the operation item 30 "Run" button. At this time, "CD" in the first part 10 has a revision format, "CD" is determined as the third content 11, and "ZZ" is the revised content 13. Then, based on the third content 11, the context is expanded forward and backward. The expansion process is based on the text before the revision, and the revised content 13 "ZZ" is skipped until the first comma position is reached as the cutoff, and the fourth content 12 "ABCD" is obtained as the content before the revision. Then, based on the fourth content 12, the target content 22 "ABCD" in the second part 20 is matched and located to the second content 21 "CD" in the second part 20, and the revised content 13 is replaced by the second content 21, i.e. the content synchronization revision is completed. After the content synchronization revision, the revised content 13 replaced in the second part 10 can also be labeled. The denoising processing includes at least one of removing punctuation, removing label, removing definite article, and removing demonstrative pronoun. As can be seen from the above embodiments, the present application can not require the user to input the keyword for searching each time in the "search" or "replace" interface, i.e., the computer can determine the pre-revision content (i.e., the keyword for searching in the above) in the patent-related text and perform subsequent matching and replacing, which greatly improves the revision efficiency of the patent-related text and reduces the probability of human error. Moreover, since the pre-revision content, the remaining part, and the selected part for matching are subjected to denoising processing in the embodiments of the present application, the "search" or "replace" has a generalization function, which can more comprehensively and accurately find the matching target, and greatly reduces the problems of troublesome manual searching and missing revision due to failure to find the matching target. Secondly, since the first content of the target content is replaced by the second content subjected to denoising processing in the embodiments of the present application, the remaining content of the target content except the second content is not changed, so that only the part that needs to be modified in the target content is affected, and the remaining content is not processed, which is not easy to cause unexpected changes and form formal errors, and the revision is accurate. Further, the third content that is changed can be extended to the fourth content by the context in the embodiments of the present application when matching, compared to the third content that can be only scattered words, the fourth content can contain more unique features, so that the matching is more accurate and is not easy to form noise interference. When determining the fourth content, the text can be cut off by the first delimiter as the stop point, or the text can be directly cut off by setting a preset length, so that the way of constructing the fourth content can be flexibly selected, and the fourth content formed by cutting off the text can have good uniqueness. In addition, the pre-revision content in the embodiments of the present application can be determined by revision format, or can be determined by the user selecting the range in the patent-related text, or can be determined by comparing the current version of the patent-related text and the historical version of the patent-related text, so that the way of determining the pre-revision content is various and can be freely selected to adapt to different occasions. When the pre-revision content is determined by the revision format, not only is the positioning accurate, but the revision format is easy to observe, and the user can intuitively understand the content that will be revised. When the pre-revision content is determined by the selected range, the user can easily decide which part of the text to process according to their own needs, and the user has stronger control over the revision. When the pre-revision content is determined by comparing the current version of the patent-related text and the historical version of the patent-related text, the user only needs to store a historical version of the patent-related text in advance, or even automatically store a historical version of the patent-related text by the software of the embodiment of the present application. It can not be necessary to edit under the revision format, and it is not necessary to repeatedly select the selected range, which saves time and effort and is simple to operate, and is especially suitable for occasions where the revision format is not opened or forgotten to open during the revision of the patent-related text. When matching the target, the embodiment of the present application also identifies whether the pre-revision content contains the second separator when the matching fails. When the second separator is not found, it means that the probability of not matching the corresponding result due to long sentence splitting is low, and the probability of not matching the corresponding result due to syntax changes should be considered first. At this time, the embodiment of the present application will replace the conjunction in the pre-revision content with the main sentence of the previous sentence and then perform matching, and the above situation is included in the matching range. When the second separator is found, it means that the probability of not matching the corresponding result due to long sentence splitting is high, and the embodiment of the present application will replace the key word in each subsequent part with the subject of the previous part, and the above situation is included in the matching range. It can be seen that the embodiment of the present application can also adapt to situations such as character changes, long sentence splitting, and syntax changes during matching, and is not easy to miss the matching result. The above matching and revision process in the embodiment of the present application can be triggered by the user actively by responding to at least one of the user clicking the operation item in the user interface, pressing the preset key, or inputting the preset instruction, or being triggered passively according to the user's operation in response to the user's cursor position changing by a preset, and the setting method is flexible. It can well respond to the user's revision operation and is convenient to use. In an embodiment, all target contents or part of the target contents matched are revised based on the revised content, and the revised content is the content obtained based on the pre-revision content, and the revised content includes: S21: Generate a history record, the history record includes at least part of the information related to the revision of all target contents or part of the target contents matched based on the revised content, and the revised content is the content obtained based on the pre-revision content. The history record contains information related to the corresponding revision operation. For example, the history record can include the time of the revision operation, the content summary of the revision, the effective range of the revision, etc. S22: In response to the user clicking the history record, restoring the target content to the state before the revision corresponding to the history record. When the user clicks the history record, the content changed by the corresponding revision operation is restored to the content before the revision operation. The way to restore the target content to the state before the revision corresponding to the history record can be to record the old version of the text content, and when the user clicks, the old version of the text content is overlaid on the current text content. It can also be to store the old version of the file, and when the user clicks, the old version of the file is overlaid on the current file. Referring to FIG. 14, FIG. 14 is an interface diagram of a second interface of an embodiment of the patent-related text writing assistance method. In an embodiment, the patent-related text writing assistance method includes a second interface 200, and the history record is displayed on the second interface 200. The history record can record the revision number, the revision position, the revision content, and the operation time, etc. The revision position can be the paragraph number of the position where the revision occurs in the patent-related text. The revision content can record the summary of the content of the revision operation. The operation time can be the time when the revision operation is triggered. The second interface 200 can be clicked, as shown in FIG. 14, when the user clicks the history record with the number "2" in the second interface 200, the patent-related text will be rolled back to the state before the operation of "deleting aa". Referring to FIG. 15, FIG. 15 is a structural diagram of an embodiment of the computer device 50. In an embodiment, the computer device 50 includes a memory 51 and a processor 52 coupled with each other. The processor 52 is configured to execute program instructions stored in the memory 51 to implement the steps of any of the above patent-related text writing assistance method embodiments. In a specific implementation scenario, the computer device 50 can include but is not limited to a microcomputer, a server, and in addition, the computer device 50 can also include a notebook computer, a tablet computer, and other mobile devices, which are not limited here. Specifically, the processor 52 is configured to control itself and the memory 51 to implement the steps of any of the above patent-related text writing assistance method embodiments. The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 can be an integrated circuit chip with processing capability. The processor 52 can also be a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. In addition, the processor 52 can be implemented by an integrated circuit chip. Please refer to Fig. 16, which is a structural schematic diagram of an embodiment of the computer readable storage medium provided in the present application. The computer readable storage medium 60 of the embodiment of the present application stores program instructions 61, which, when executed, implement the method provided by any embodiment of the writing assistance method for the patent-related text and any non-conflicting combination. The program instructions 61 can form a program file and be stored in the computer readable storage medium 60 in the form of a software product, so that a computer device (which can be a personal computer, a server, or a network device, etc.) executes all or part of the steps of the method of each embodiment of the present application. The aforementioned computer readable storage medium 60 includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes, or a terminal device such as a computer, a server, a mobile phone, or a tablet. Please refer to Fig. 17, which is a structural schematic diagram of the client and the server in an embodiment of the writing assistance method for the patent-related text. The present application provides a writing assistance method for the patent-related text, which includes that the client 60 sends a data structure to the server 70, the data structure defines a plurality of elements of a protected subject and their relationships, and defines a set of elements and a set of relationships constituting each claim; the method further includes receiving a claim from the server, the claim is formed by the server 70 according to a first preset rule based on the set of elements and the set of relationships corresponding to each claim. Alternatively, The client 60 receives a first instruction for editing the data structure; The client 60 sends the first instruction to the server 70, so that the server 70 edits the data structure, and the edited data structure defines a plurality of elements of a protected subject and their relationships, and defines a set of elements and a set of relationships constituting each claim; The client 60 receives a claim or a processing completion notification from the server 70, the claim is formed by the server based on the set of elements and the set of relationships corresponding to each claim according to the first preset rule, and the processing completion notification includes the processing result of forming the claim. For example, the client 60 can be responsible for user interaction only. The data structure is directly stored on the server 70. After the user operates on the client 60, the client 60 generates a first instruction and sends it to the server 70. The server 70 processes the data and completes the editing. Then, the client 60 displays the editing result on the server 70 through the window interface of the client software. Through the above setting, the client does not need to perform complex operation, which is beneficial to reduce the load of the user computer device, and meanwhile, the operation process is performed on the server, and data security is better. Through the above setting, the key step can be placed on the server for execution, and the simple step is executed by the client, so that the load pressure of the server and the client can be considered, and data security is good. In an embodiment, the application provides a text writing assisting method, which comprises: receiving a first instruction for a text, the first instruction comprising an artificial revision ending instruction; determining pre-revision content in the text in response to the first instruction; matching the pre-revision content with a remaining part or a selected part of the text; and revising all target content or part of the target content obtained through the matching based on post-revision content, the post-revision content being content obtained based on the pre-revision content. The text in the embodiment can be an article, a report, a thesis, a novel or a scientific and technological document, etc. In an embodiment, the application provides a text writing assisting method, which is executed through program instructions, the program instructions being installed in an existing word processing software in the form of a plug-in, and acting on a patent-related text opened by the word processing software to realize the above-mentioned patent-related text writing assisting method. In an embodiment, the application provides a text writing assisting method, which is executed through program instructions, the program instructions being installed in an existing word processing software in the form of a plug-in, and acting on a patent-related text opened by the word processing software to realize the above-mentioned patent-related text writing assisting method. In an embodiment, the application provides a text writing assisting method, which is executed through program instructions, the program instructions being installed in an existing word processing software in the form of a plug-in, and acting on a patent-related text opened by the word processing software to realize the above-mentioned patent-related text writing assisting method. In an embodiment, the application provides a text writing assisting method, which is executed through program instructions, the program instructions being installed in an existing word processing software in the form of a plug-in, and acting on a patent-related text opened by the word processing software to realize the above-mentioned patent-related text writing assisting method. In an embodiment, the application provides a text writing assisting method, which is executed through program instructions, the program instructions being installed in an existing word processing software in the form of a plug-in, and acting on a patent-related text opened by the word processing software to realize the above-mentioned patent-related text writing assisting method. Currently, a part of the work of a patent agent is to write a patent document, form a patent related text, and then submit the patent application through an official channel. The patent application generally includes an invention application, a utility model application or a design application. Among them, the patent related text involved in the invention application and the utility model application generally includes an abstract, a claim, a specification, etc. In a patent agency, it often needs to proxy several or even hundreds of patent applications of different applicants. Different applicants have different requirements for patent applications, and patent agents need to distinguish different requirements of different applicants for differentiated patent writing. However, in many cases, the agent may forget the related requirements of the applicant or confuse the different requirements of different applicants, resulting in low efficiency of writing work or problems in the quality of the completed patent application file that do not meet the requirements of the applicant. On the other hand, for patent agents, it often takes several years of training to gain some work experience to meet the growing requirements of applicants. However, all patent agents need to start from zero experience, although the files completed by these junior agents will generally be audited by internal teachers, but before the audit, the junior agents lack experience and knowledge, and have no idea how to proceed, even if they have written a part of the content, but they are not sure if it is correct, and then modify it repeatedly, which is very inefficient. In addition to the patent agency industry, generally speaking, for enterprises with office needs, text editing work is often related to data and information within the enterprise or related enterprises. When data and information within the enterprise or related enterprises are needed, it is often necessary to manually search for files from related databases, which is very inconvenient. To solve the above technical problems, in an embodiment of the present application, a writing assistance method is proposed, which uses the existing file naming rules with identifiers of modern enterprises to automatically obtain relevant reference information to assist agents to write high-quality and efficient text that meets the requirements. The method comprises: Step 100: reading a file name, the file name comprising a file identifier, the file identifier being associated with writing reference information of a patent related text; Referring to FIG. 18, in the embodiment, the user opens a text application program or a third-party program and enters a text editing interface 100, and the system automatically extracts the file name 300 of the patent related text. Alternatively, when the user inputs characters using an input method in the text editing interface, the system automatically extracts the file name 300 of the patent related text. Of course, the timing of the system automatically extracting the file name of the patent related text is not limited to the above two cases. The system referred to in the embodiment of the present application is a program, application, plug-in or tool, etc. that implements the writing assistance method of the embodiment of the present application. Step 200: obtaining the writing reference information associated with the file identifier; The client can send the file name extracted by the system to the server, and the server looks up the writing reference information corresponding to the file identifier in the association between the file identifier and the writing reference information, and returns it to the client. Optionally, the association between the file identifier and the writing reference information, and the set of writing reference information are stored in the server in advance. The writing reference information can exist in the form of a file, that is, one file identifier is associated with one or more files. The writing reference information can also exist in the form of storing multiple information units in a file, and one file identifier is associated with one or more information units. The writing reference information can be laws and regulations, process control, writing requirements, writing skills, writing experience or forbidden words, etc. Table three: one of the association between the file identifier and the writing reference information Table three is an example of the association between the file identifier and the writing reference information. For convenience of understanding, the file identifier and the writing reference are directly taken as the two objects in the mapping relationship. In actual use, one side in the association relationship is the client number or the application type, and the other side is the writing reference information. In table three, "CN" in the file identifier "CNXYP202412345" represents China, "XY" represents the client number, "P" represents a patent, "2024" represents the year, the first digit "1" following "2024" represents the application type, which is an invention application, and the remaining digit sequence represents the number of the application. The server reads the file identifier in the file name to obtain the client name XY, and then obtains the specific requirements in the writing reference information "file a" and "file b" associated with the file identifier through the association relationship. In addition, the priority in the association relationship represents the priority level of the corresponding writing reference information in subsequent display or other operations. The higher the priority, the higher the priority of the writing reference information displayed, for example, when the screen display area is not enough, the writing reference information with higher priority is displayed first. Table four: another example of the association between the file identifier and the writing reference information Table four is another example of the association between the file identifier and the writing reference information. The writing reference information of this association is divided into two categories, one is basic information, which can be referred to by all application files, and the other is classification information, which is only associated with specific file identifiers, and other file identifiers are not associated with the system. For example, the file identifier CNXYP202418888 indicates that the corresponding application file is an invention application file, so the identifier is associated with information unit 1 of file a, but not associated with information unit 2. The file identifier CNXZP20242000 indicates that the corresponding application file is a utility model application file, so it is associated with information unit 1 and information unit 2 of file a. Wherein, after the writing reference information is returned to the client, the client displays the writing reference information 201 obtained while displaying the content of the text. The display mode of the writing reference information 201 includes at least the following two options: 1. The text and the writing reference information 201 are displayed on the same screen without affecting each other. For example, the text and the writing reference information 201 are displayed side by side, or the text is displayed as the main content and the writing reference information is displayed in the form of a sidebar 200. If the writing reference information displayed in the sidebar 200 is too much to be reasonably displayed on the screen, the writing reference information 201 with high priority can be highlighted, or only part of the writing reference information 201 with high priority is displayed, or only part of the writing reference information 201 is displayed. When the user clicks on a part of the displayed writing reference information 201, the system automatically expands and displays the non-displayed part of the writing reference information 201, replacing or covering the display content in other areas of the screen. 2. While inputting content in the text, the writing reference information is displayed in the form of content recommendation near the current cursor. Once the user selects a writing reference information, the writing reference information is filled in behind the current cursor as part of the text content. If the writing reference information is not selected, the writing reference information changes position following the cursor movement, providing reference to the user without blocking the user's view and affecting the user's input. In addition, the writing reference information is associated with not only the file identification but also the input content in the text, such as inputting "claim", the system believes that the user is writing a claim, and therefore recommends writing-related laws and regulations, writing experience, writing requirements, writing skills, etc. Optionally, at least part of the embodiments of the present application are implemented by using plug-ins of text editing application programs, such as MICROSOFT WORD application program or WPS application program. These text editing application programs are relatively closed, and generally only plug-ins can be used to operate the text editing application programs themselves or the text opened by them. For example, reading the file name of the text, reading the content input into the text by using the input method, and enabling the text to display the writing reference information in the split screen or window, etc. can be realized by using plug-ins supported by the text editing application programs to realize these functions. Optionally, obtaining the writing reference information associated with the file identification comprises: Step i: obtaining the client information associated with the file identification; The server pre-selects the definition rules of the file identification and the association relationship between the file identification and the client information. The client information is parsed from the file identification by the definition rules of the file identification, and the writing reference information is obtained according to the association relationship. Step ii: obtaining the writing reference information associated with the client information, and the writing reference information corresponding to at least two client information is different. Optionally, the writing reference information corresponding to all the client information is different, but the writing reference information corresponding to part of the client information can be the same, and the specific case is defined specifically. Optionally, the writing reference information includes writing requirements, and the method further includes: i: automatically modifying the content of the patent-related text according to the writing requirements; or ii: marking the part not meeting the writing requirements in the content of the patent-related text or prompting in a pop-up window. The embodiment can realize higher degree of automation, can let the system as an auditor to audit the text, further, can directly replace the user to modify the text. Specifically, the writing reference information can be defined in advance, and the part easy to be recognized and understood by the computer is defined separately according to the predetermined rule, which is convenient for subsequent system reading and checking the text according to the writing reference information. If the place not meeting the writing reference information is found, it is modified according to the writing reference information. For example, the writing reference information requires that certain words cannot appear, and the system automatically replaces the related words in the text with new words meeting the requirements. For the part not convenient for automatic modification, the part not meeting the writing requirements in the content of the text or the pop-up window prompt can be marked. Optionally, in order to more accurately locate the writing reference information, the method further includes: obtaining the module position of the current cursor in the patent-related text, the module position including at least one of the abstract module position, the claim module position, the invention name position, the technical field module position, the background technology module position, the invention or utility model content module position, the drawing description module position, and the specific implementation manner module position; The invention content module can be further divided into a technical problem module (referred to as a technical problem module) required to be solved by the invention or utility model, a technical solution module, and a technical effect module. The method further includes: obtaining the writing reference information associated with the file identifier and the module position of the current cursor in the patent-related text. That is, when the patent-related text is a patent application document, the writing reference information corresponding to the current cursor (or mouse) position in the patent application document can be recommended according to the current cursor (or mouse) position in the patent application document. For example, if the current cursor is located in the position of the claims module, the writing reference information related to the writing of claims is recommended. The method of determining the position of the current cursor includes determining the position of the current cursor between the two module names in the patent application document, and the position of the module corresponding to the previous module name is regarded as the position of the current cursor. For example, if the current cursor is located between the technical field and the background technology, it is considered that the current cursor is located in the technical field module. If there is no module name after the current cursor, it is determined as the last module. Optionally, the writing requirement includes that the disabled words cannot be used, and the method of automatically modifying the content of the patent-related text according to the writing requirement after obtaining the writing reference information associated with the file identifier includes: i: searching the content of the patent-related text with the disabled words as keywords; For example, after the text is processed, the user closes the text, and the system intervenes at this time. That is, after receiving the instruction to close the text, the system automatically pauses the closing of the text, obtains the writing reference information with the file identifier of the text, identifies that the writing reference information includes the disabled words, and then searches the text with each disabled word; ii: marking the searched results in the patent-related text, or replacing the results in the patent-related text with the replacement words corresponding to the disabled words, and each disabled word is preconfigured to correspond to at least one replacement word. Optionally, in order to select the replacement word, the method of replacing the searched results with the replacement words corresponding to the disabled words includes: i: positioning the current display position of the patent-related text to the searched results, and displaying at least two selectable replacement words near the results; This function facilitates the user to select the correct replacement word to replace all the searched results. ii: receiving a first instruction of selecting one of the replacement words; For example, the user double-clicks one of the replacement words to generate the first instruction. The method of replacing the searched results with the replacement words corresponding to the disabled words includes: replacing the results with the selected replacement words in response to the first instruction. Of course, the method of automatically modifying the content of the patent-related text according to the writing requirement is not limited to modifying the disabled words, but also can be modifying the multiple references in the claims to single references, modifying the incorrect reference relationship, modifying the wrong characters, adding "the" in front of all elements appearing in the claims, modifying the incorrect numbers or serial numbers, etc. Optionally, in order to provide the user with information related to the current input during the drafting process, the obtaining of the drafting reference information associated with the file identification comprises, before the obtaining of the drafting reference information associated with the file identification: i: receiving the current input in the display interface of the patent-related text; For example, a technical term, a patent application number, or a legal term, or even a paragraph of text. ii: searching the document set corresponding to the current user group with at least part of the current input as the keyword, different user groups corresponding to different document sets; or searching the current text; The server has pre-stored the document set corresponding to the current user group, such as the relevant documents of the patent drafting specification, or the patent files partially or completely handled by the current patent agency, or the files required by the client. iii: displaying the matching paragraphs or matching sentences in the searched document set, the current text that match the keyword. For example, if the current input is "creativity", the relevant discussions on creativity in the Patent Law, the Detailed Rules for the Implementation of the Patent Law, and the Examination Guidelines can be searched and displayed on the left or right side of the screen. For example, if the current input is "electric vehicle", the patent files partially or completely handled by the current patent agency can be automatically searched, and the relevant patent files related to electric vehicles can be found and summarized and displayed on the left or right side of the screen. Optionally, the process of searching for matching paragraphs or matching sentences that match the keyword can be performed in different ways. The matching paragraph is relatively simple to obtain, that is, as long as a paragraph contains a keyword, the first carriage return symbol is identified from the keyword to the rear in the text, and at least the content between the keyword and the first carriage return symbol can be taken as the matching paragraph. As for the sentence in the matching sentence, it is also relatively easy to obtain, that is, as long as a sentence contains a keyword, the first punctuation symbol is identified from the keyword to the rear in the text, and at least the content between the keyword and the first punctuation symbol can be taken as the matching sentence. As for the word in the matching sentence, a word library can be obtained by segmenting the text, and the keyword can be matched in the word library. Alternatively, a simpler method is used, which does not require pre-configuration of the word library, but uses some writing features of the text to define some boundary symbols to "capture" the words. For example, nouns are a kind of words that need to be "captured" mainly, especially for the patent-related texts for protecting products, there are a large number of nouns, due to the particularity of the patent-related texts, there are a large number of "self-created" or "self-defined" nouns in the patent-related texts, these "self-created" or "self-defined" nouns do not exist in the general dictionary, nor in the professional dictionary, therefore, when writing the patent-related texts, if trying to use the input method on the market to recommend these nouns, the effect is often poor. The reason is that the input method on the market generally uses two methods to recommend words, one is to match the dictionary directly, the other is to count the word frequency, and recommend the nouns according to the height of the word frequency. The limitation of the dictionary has been described before, and the limitation of the statistical word frequency also exists, either the network statistics of all users or the local statistics of the text. The word frequency counted by the network does not match the nouns in the text, because the word frequency of the text in the network is negligible, and the word frequency of the text may not be input by the input method. The local statistics can only recommend the words that have been input in the text, but for the pinyin input method, if there are multiple words with the same first pinyin or the same full pinyin, accurate word recommendation cannot be obtained. More importantly, almost all input methods only produce corresponding characters or words according to the user's keystrokes, and do not produce characters or words other than keystrokes. For example, when the user types "zl" two first pinyins representing "patent", only "data", "instruction", "patent" and other recommendations will appear, and "patent law" or "patent-related texts" and other recommendations outside the actual input will not appear, which is difficult to effectively improve the text editing efficiency. The present inventors have realized, after research, that the reasons why the input method does not actually land in the actual input in addition to the association recommendation are: first, the thinking inertia of existing industry personnel, and second, it is difficult to achieve the optimal solution in terms of accurate recommendation, vocabulary delimitation, and cost payment. Regarding accurate recommendation, the current input method has relatively high accuracy in short vocabulary recommendation, but as the vocabulary length increases, the accuracy drops sharply to almost unusable, because there are too many words with full pinyin or the same first pinyin. When a single word, the recommendation accuracy cannot be 100% due to this reason. When multiple words form the association recommendation described above, through the combination probability analysis of probability theory, it can be concluded that even if the accuracy of a single word is relatively high, when multiple words are combined into long words, the recommendation accuracy probability of long words can become very low. Second, it is difficult to delimit a single short word, and it is even more difficult to delimit long words. Of course, natural language processing methods that are currently popular can be used for analysis, or AI models can be used for analysis. However, these tools are often high in cost or poor in real-time performance. The speed of inputting content using an input method is generally the speed of typing a keyboard, and the input method must keep up with the speed of typing the keyboard, which limits the application of high-processing-capacity word segmentation tools, forming a research bottleneck for long-word recommendation. Therefore, optionally, the present embodiment proposes a writing assistance method with high recommendation accuracy, reasonable vocabulary delimitation, low cost, and fast response, comprising: i. receiving first content input in a text, or receiving first content input in an input method interface; Optionally, the first content input in the text using the input method or the first content input in the input method interface using the input method can be received. Receiving the first content input in the text using the input method means that after the first content is input in the text using the input method, the first content is obtained using a third-party program such as a plug-in of a text application. Receiving the first content input in the input method interface using the input method means that when the input method is used, the first content input in the input method interface is obtained using the input method itself. For a pinyin input method, the first content appears in the candidate window of the input method interface as a candidate word for the user to select. ii. using the first content as a keyword to search for a matching paragraph or a matching sentence in a text or a specified document, the matching paragraph or the matching sentence including the first content and the second content, or the matching paragraph or the matching sentence including the second content and the related content of the first content, the matching paragraph or the matching sentence being obtained by delimiting the text or the specified document through a delimiting character, the delimiting character being a general character adjacent to the front or the rear of the matching paragraph or the matching sentence; There are at least two ways to achieve this step. First, using an input method, search for matching paragraphs or phrases in the text or a specified document using the first content as a keyword. The matching phrases or paragraphs are then displayed as candidate words in the input method interface for selection and insertion into the text. In other words, the input method completes the process of inputting the first content into the text, obtaining the first content (which can be omitted), searching for matching paragraphs or phrases, and displaying matching paragraphs or phrases for selection and insertion into the text. Referring to Figure 20, when inputting text 210 using the Pinyin input method 310 in text 110, typing the first two letters 320 of "ZL" in the input method interface will display the first content 330 "patent". Directly searching text 110 using "patent" as a keyword will yield matching phrases 340 such as "patent law" and "patent-related documents". Alternatively, a plugin for a text editing application can be used to search for matching paragraphs or phrases in the text or a specified document. The plugin's interface then displays these matching phrases or paragraphs as candidate words, ready to be selected and entered into the text. In other words, the input method can handle the initial input, while the plugin for the text editing application handles other tasks, such as obtaining the initial content, searching for matching paragraphs or phrases, and displaying them for selection and entry. Figure 21 illustrates the simultaneous use of the input method 500 and the plugin 600 of the text editing application to input text 230. When inputting text 400 using the Pinyin input method 500 in text 230, typing the first two letters "ZL" 510 in the input method interface will bring up multiple candidate words such as "patent", "data", and "self-awareness". If the candidate word "patent" is selected, "patent" will be entered as the first content 330 in text 230. At this time, the input method interface is closed, and the plugin of the text editing application searches for "patent" in text 2300. Matching phrases 610 such as "patent law" and "patent-related documents" are displayed in the suggestion window 600 for the user to select when inputting text 230. Matching paragraphs or phrases includes both the first and second content, meaning the search results will include longer words or sentences than the first content, rather than simply recommending words actually typed, as current input methods often do. Optionally, the first and second content can be adjacent, with the second content following the first. For example, if you actually type "patent," the second content could be "law," "related text," or "application documents," etc. Alternatively, the first content can follow the second. The matching paragraph or matching sentence is defined by a delimiter character in the text or the specified document, which means that the matching paragraph or matching sentence is located and delimited directly using the existing characters in the text or the specified document without natural language processing or AI model analysis. These existing characters can be named as delimiter characters because they can delimit and locate the words and phrases such as nouns in the text. In addition, these existing characters are usually used together with the matching paragraph or matching sentence, so they can also be general characters. For example, the delimiter characters include any one of the front general characters in a sentence in the patent-related text, such as “including”, “having”, “containing”, “provided with”, and also include any one of the back general characters in a sentence in the patent-related text, such as punctuation marks and symbols. For example, in the product claim, the open-ended limitation is usually used in the claim limitation structure, that is, the limited structure is introduced by “including”, “having”, “provided with” or “containing”, as shown in the following independent claim: 1. An advertising module, comprising: B1; B3, including c3, the c3 having d3; B6, including c7, the c7 having d7, the b6 being connected to the b1 at one end and passing through the d3 at the other end; B8, being interposed in the d7 and located on the side of the c3 opposite to the b1. In the above independent claim, B1, B3, B6 and B8 are all elements of the advertising module (the actual element names are replaced by English letters and numbers for convenience of description), and are also nouns. Because the delimiter character “including” is used to introduce the limited structure, and must be located before the limited B1, B3, B6 and B8, it belongs to the front general character. Similarly, “having”, “provided with” and the like also belong to the front general character. After these nouns B1, B3, B6 and B8 are introduced, punctuation marks or conjunctions such as “and”, “with” and “as well as” usually follow, so punctuation marks, conjunctions and symbols belong to the back general character. In addition, the front general character and the back general character both appear in the same sentence, such as the above independent claim which can be regarded as a sentence with only one period in the entire paragraph. Of course, in some scenarios, those ending with a comma, semicolon or colon are also regarded as a sentence. The main purpose of using the front general character and the back general character to locate the matching sentence is to determine the first occurrence of the word in the claim, because the first occurrence of the word such as the noun in the claim is more easily delimited than the second and subsequent occurrences of the word. For the scenario where the second content is located behind the first content, the first content and the back general character can be used to delimit the second content, that is, the content sandwiched between the first content and the back general character is read out as the second content for recommended display. iii. The matched paragraph or matched phrase is displayed on the screen as a writing reference or recommendation information. As mentioned above, after finding the second content, the second content can be displayed on the screen as a matched paragraph or matched phrase alone, or the first content and the second content can be combined into a matched paragraph or matched phrase and put into the suggestion window as a candidate content or candidate phrase. After the user selects the matched paragraph or matched phrase, the matched paragraph or matched phrase can be filled into the text. Obviously, because the high-cost tool is not used for vocabulary extraction, but the matched paragraph or matched phrase is delimited by using the content of the text itself, the long vocabulary recommendation is accurately and extremely low-cost provided, the real-time recommendation can be achieved, the input speed is not affected, and the text editing efficiency is greatly improved. Optionally, the number of the matched paragraphs or matched phrases is several, and the matched paragraphs or matched phrases are displayed, including: i. receiving a second instruction of selecting one of the matched paragraphs or matched phrases; After the user observes, the user feels that one of the matched paragraphs or matched phrases is more suitable, and then the user double-clicks the matched paragraph or matched phrase. ii. combining the current content with the selected matched paragraph or matched phrase in response to the second instruction. When the user selects a matched paragraph or matched phrase, the current content is combined with the selected matched paragraph or matched phrase, and there are two processing methods. First, the selected matched paragraph or matched phrase is directly put behind the current cursor of the text for the user to edit. Second, only the content behind the key word in the matched paragraph or matched phrase is filled behind the current cursor of the text, so that the filled content and the content of the text are combined together. By directly recommending the matched paragraph or matched phrase, the user does not need to retype too much content, and the input efficiency is greatly improved. Optionally, if the user feels that only part of the content in the matched paragraph or matched phrase is needed, the matched paragraph or matched phrase is displayed, including: i. receiving a third instruction of selecting part of the content of the matched paragraph or matched phrase; For example, the user selects part of the content of the matched paragraph or matched phrase by using a mouse. ii. combining the selected part of the content with the matched paragraph or matched phrase in response to the third instruction. For example, after the selection by the mouse is completed, the mouse is released, and the system automatically fills the selected content behind the current cursor of the text. Optionally, in order to better match, the searching of the document set corresponding to the current user group by taking at least part of the current content as a key word includes: search the document set corresponding to the current user group with the keyword obtained after the de-noising processing, the de-noising processing including removing "the", "this", "of", "above" or "below". Because the patent application file has special provisions for the text content, such as in the claim, the second occurrence of the name needs to be preceded by "the", therefore, when the system takes the name as a keyword, it may also take "the" as part of the keyword, but "the" may not exist in the document set corresponding to the current user group, resulting in a matching failure. Therefore, the de-noising processing before the search can make the search operation more accurate and smooth. Optionally, the number of the matching paragraphs or matching sentences is a plurality, and the displaying of the matching paragraphs or matching sentences comprises: The plurality of matching paragraphs or matching sentences are sorted according to the distance from the current cursor position, and the most proximate matching paragraph or matching sentence before the current cursor is placed at the top of the sorting and displayed in the candidate window below the current cursor. During the input of the text, the user usually focuses on describing the same thing in a paragraph or adjacent paragraphs, so the probability of the occurrence of the thing in the same content area is relatively large, and therefore, when the user inputs the corresponding text of the thing, the most proximate matching paragraph or matching sentence is preferably placed at the top of the sorting to facilitate the user to view and select. Optionally, the displaying of the matching paragraphs or matching sentences comprises: The most proximate matching paragraph or matching sentence placed at the top of the sorting and the remaining matching paragraphs or matching sentences are displayed differently, and the display is in the form of split screens or superimposed display in the window, respectively displaying the page of the text where the most proximate matching paragraph or matching sentence is located and the page where the current cursor is located. When the user views the matching paragraphs or matching sentences searched, in order to guide the user to view, the most proximate matching paragraph or matching sentence placed at the top of the sorting and the remaining matching paragraphs or matching sentences are displayed differently; at the same time, in order to enable the user to understand more about the origin of the matching paragraphs or matching sentences searched, the split screens are used, one screen interface displays the page where the current cursor is located, and the other screen interface displays the page of the text where the most proximate matching paragraph or matching sentence is located, facilitating the user to understand more information. Optionally, the most proximate matching paragraph or matching sentence placed at the top of the sorting in the plurality of matching paragraphs or matching sentences is set as a default selected item, and the displaying of the page where the current cursor is located and the page of the text where the most proximate matching paragraph or matching sentence is located comprises: receiving a fourth instruction to reselect the matched paragraph or matched sentence, displaying the reselected matched paragraph or matched sentence in a distinguished manner, and displaying a page where the reselected matched paragraph or matched sentence is located. If the user wants to know the content of the page where the other matched paragraph or matched sentence is located after the split screen view, the user can select the other matched paragraph or matched sentence, and after selection, the screen is refreshed to display the page where the reselected matched paragraph or matched sentence is located. Optionally, if the split screen is already in a split screen state before the split screen, the split screen content of the page where the matched paragraph or matched sentence is located replaces the original split screen content, and after the user fills the selected content after the current cursor, the system can return to the original state, i.e., a state without split screen. Optionally, to facilitate the user to better obtain relevant information, after opening the current file, the system obtains or automatically obtains links of other files in the same folder, and arranges these other division links on the top side or side of the current file. When the links are clicked, the split screen position displays the files of the links. Generally, related files are stored in the same folder, such as a related technical disclosure, a comparison file, and an archive of the same patent application. In the prior art, after the user opens the technical disclosure, the user needs to switch windows to open other files in the same folder. Because of window switching, the user needs to perform too many operations, and sometimes the switching is incorrect, resulting in back-and-forth switching and wasting time. In the embodiment of the present application, the links of other files in the same folder are automatically obtained and arranged on the top side or side of the current file. When the links are clicked, the split screen position displays the files of the links, without the need to switch windows back and forth, saving time and improving accuracy. Optionally, the input text is searched in the same folder, and if the technical disclosure in the same folder has the same text, the search results and the page where the search results are located are displayed in the right split screen, which is beneficial for a patent agent to refer to related information of the technical disclosure for writing. Optionally, after automatic split screen, the size of the split screen content is automatically adjusted to facilitate the user to view. Optionally, to reduce user operations as much as possible, after receiving the current content input in the display interface of the patent-related text, the following operations are performed: In response to whether the distance between the current cursor and the bottom of the current text is less than a threshold value, the current text is automatically scrolled backward, so that the distance between the current cursor and the bottom of the current text is not less than the threshold value. That is, since the current text will automatically scroll backward due to the current cursor position being too low, the current cursor follows the scrolling of the text and returns to the appropriate position after the text is scrolled, so that the user does not have to manually scroll the pages of the text with the mouse or other keys frequently, and user operations are minimized as much as possible. The embodiment is not limited to being applied to patent-related text, and can be applied to any text, and can be independent of any of the preceding embodiments. Optionally, the displaying of the searched document set, the matching paragraph or the matching sentence in the current text that matches the keyword comprises: Optionally, the displaying of the searched document set, the matching paragraph or the matching sentence in the current text that matches the keyword comprises: The existing input method application will pop up a candidate window when the user types, but before the user confirms, whether the typed content or the candidate content, they are all displayed in the popped-up candidate window, and only when the user selects the candidate content, the selected candidate content will be filled into the text input space behind the current cursor. In this process, the user's vision will switch between the left side of the current cursor and the candidate content in the candidate window. The embodiment of the present application pre-fills the matching paragraph or the matching sentence into the text input space behind the current cursor, so that the user's vision may not have to switch between the left side of the current cursor and the candidate content in the candidate window, and the user's vision only needs to focus on the left and right of the current cursor, thereby saving the user's energy. Optionally, in some embodiments of the present application, switching between the existing input method and the system used in the embodiments of the present application is required, which involves the analysis, use and problem avoidance of the focus in the user interface (UI). In user interface design, focus is a very important concept, which refers to the control or element that can currently receive user input. When a control has focus, the user can directly interact with it through the keyboard, such as inputting text, navigating options through arrow keys, etc. The concept of focus is not limited to keyboard input, but can be extended to touch screen click events or other forms of user input. It can be seen that in the case of using the existing input method and the system of the embodiment of the present application simultaneously, the capturing of the focus is a necessary operation. Not only the existing input method needs to capture the focus accurately and timely, but also the system of the embodiment of the present application needs to capture the focus accurately and timely, so as not to cause the input method to be unable to be called up in time to input content in the text input process, the suggestion window generated by the system of the embodiment of the present application to be unable to be called up, or the input method to be unable to be called up again after the suggestion window generated by the system of the embodiment of the present application appears. The suggestion window generated by the system of the embodiment of the present application is defined as being able to display matching paragraphs or matching sentences, or writing reference information and the like. The inventor of the present application found that when the input method is used to input content and the suggestion window is called up, and the input method is used to continue inputting content without paying attention to the suggestion window, the first character (hereinafter referred to as "key" or "key message") inputted is "swallowed", that is, the input method cannot be called up in time, so that after the input method is called up, only the second and third inputted characters are displayed in the input method window, and the first inputted character does not appear. In some fast input scenarios, the second and third characters can also be "swallowed". More specifically, the normal running scenario should be: When the suggestion window is called up, the focus needs to be on the suggestion window to facilitate the up and down switching of options, so when the keyboard is pressed at this time, the calling position of the input method will necessarily be on the suggestion window. Then if the input method is used to continue inputting content without paying attention to the suggestion window, the suggestion window needs to disappear, the focus needs to be handed over to the text application, and the input method needs to be displayed in the candidate window suspended state. However, due to the design of the computer operating system itself or the incoordination between the input method and the computer operating system, problem 1 occurs: the input method positions the candidate window coordinates through the placeholder box of the text (representing the focus), and saves the to-be-inserted position of the current cursor in the text. When the user selects the content of a certain option in the candidate window of the input method, such as pressing space to determine, the placeholder box disappears, and the selected option content of the candidate window is sent to the to-be-inserted position (the position behind the placeholder box) of the text. According to the embodiment of the present application, the option content sent to the to-be-inserted position of the text will call up the suggestion window, and the focus will be switched to the suggestion window (the placeholder box is initially bound to the suggestion window). When the input method is used to continue inputting content without paying attention to the suggestion window after the suggestion window is called up, that is, the keyboard is pressed, triggering the suggestion window to disappear. After the suggestion window disappears, the input method loses the binding object, so although the focus will return to the text application, the to-be-inserted position of the input method is lost, triggering the candidate window of the input method to close, that is, this key only triggers the window to close, and the activation of the input method is interrupted due to the window closing, causing "swallowed words". The above technical problem is difficult to find, mainly because the user uses different types of computers or different versions, different operating systems or different versions, different input methods or different versions, and different application environments in the computer. Some users may not have found the above technical problem, so the present inventors have also accidentally discovered the above technical problem in a large number of attempts and switching different computers for research and development, and have started to improve it. Alternatively, one solution provided by the embodiments of the present application is that the system of the embodiments of the present application records which key is pressed at the moment of pressing the keyboard at the same time or after the suggestion window is awakened, and then triggers the candidate window form to close. Then the previously recorded key is sent to the operating system again to trigger a key input event and activate the input method for the second time. In this way, there is no problem of swallowing words, and it seems that the candidate window of the input method has never disappeared. Of course, alternatively, the speed of user input can be used to determine which keys are pressed at the moment of pressing the keyboard, or no key is pressed, so that no matter how fast the user inputs, the input method can correctly capture the characters corresponding to the actual input keys of the user when the input method is activated for the second time. In order to ensure that the system of the embodiments of the present application records how many keys are pressed at the moment of pressing the keyboard at the same time or after the suggestion window is triggered to generate, all key information is not missed, the system starts timing when the suggestion window is triggered to generate, and records what key is pressed within a relatively long preset time, defined as the first key combination (including the case of only one key information), and then starts to record new keys in real time when the input method is activated for the second time, defined as the second key combination (including the case of only one key information), and splices the first key combination and the second key combination in the order of first and second, and removes the adjacent keys between the two key combinations, such as the first key combination is AB, and the second key combination is B, then the spliced AB is AB, and the pinyin or letter B is removed. Then, in this process, if the input method is self-closed due to the loss of the to-be-inserted position, and then the recorded keys are sent in immediately, it will cause the current cursor (placeholder) to look normal when typing, but at the moment when the user selects the content of the candidate window of the input method, such as pressing space to determine, the current cursor will jump to the "0" position at the beginning of the document. The present inventors speculate that the input method uses the to-be-inserted position created for the first time in the process of being awakened for the second time, but the to-be-inserted position has been lost at this time, that is, it becomes a parameter "null", which causes the input method to incorrectly pass null to the text application, so that the selection of the text application is suddenly jumped to the beginning of the document (0 position) due to the influence of null. There are many factors causing the above technical problems, and each system and application is independent and does not tell each other the core technology or technical details, and there is no suitable or realistic third party to effectively coordinate the interaction of both or more parties. Therefore, it is difficult to find technical problems, and it is also difficult to solve them after finding technical problems. After long-term thinking, the first solution proposed by the inventor is to refresh the cache information of the input method from the moment the input method is used to input content in the text editing application to the moment the suggestion window is invoked, so that the input method rebinds the required insertion position based on the current cursor position. At this time, the candidate window insertion position of the input method is correct, and pressing the space bar will not cause the current cursor (placeholder) in the text to jump to the beginning of the text. The second solution is to send the previously recorded keys to the operating system again to trigger a key entry event. The principle of this method is that if the loss of the insertion position causes problems, then do not let the input method record the insertion position at the beginning. At the moment of selecting the content of a candidate window generated by the input method or when the suggestion window is invoked, the input method is cut off or disabled. At this time, the user thinks he is typing in Chinese input method state, and pressing the a key actually inputs English a. Then the background closes the suggestion window and swallows the English a, switches the input state of the input method to Chinese, inputs an a in the Chinese state, invokes the Chinese a, and since the insertion position has not been lost since the beginning, the process of binding the insertion position does not occur. Therefore, null will not appear, and the cursor will not jump to the beginning of the text. The whole process is user-agnostic. The third solution is to close the input method after the first input using the input method before the input method obtains the key entry message again, so that the input method does not bind the focus on the suggestion window. That is, the input method only binds the focus before the suggestion window is invoked and after the second invocation. Therefore, the focus position bound by the input method can always be changed, so that the input method loses the binding object and causes the input position to be affected when the input method is invoked for the second time. Optionally, if the writing reference information is too much to be displayed completely on the screen, the display of the obtained writing reference information includes: scrolling and highlighting the writing reference information in the display interface of the patent-related text. Optionally, the method further comprises: i: receiving a sixth instruction to close the patent-related text; ii: in response to the sixth instruction, displaying the writing reference information in a pop-up window before closing the patent-related text. The pop-up window displays the drafting reference information before the patent-related text is closed. In case the user forgets, the user can be reminded so as to not miss some important information such as the requirement of a client before and after the patent-related text is closed. Optionally, the drafting reference information comprises a data structure, the data structure defining a plurality of elements of the protected subject matter and the relationship thereof, and defining a set of elements and a set of relationships constituting each claim. For example, the elements of a product claim generally comprise elements, components or structures of the product, the components or structures generally also being constituted by elements, and the relationships of the product claim generally comprise position relationship, connection relationship, action relationship, signal relationship, or electromagnetic field relationship. For another example, the elements of a method claim generally comprise steps of the method, and the relationships of the method claim generally comprise the sequence of the steps. Optionally, as shown in FIG. 19, to facilitate the editing of the data structure, a first area 1000 and a second area can be displayed on the software interface, the first area 1000 being used for editing the data structure, and the second area being used for waiting for display or displaying the formed claims, or displaying the content of the technical disclosure, or displaying the prior art. The second area displays the content of the technical disclosure, which can facilitate the conception of the elements and the relationships thereof of the claims while looking at the related technical content of the technical disclosure, the prior art, etc., and facilitate the editing of the set of elements, the set of relationships, and the definition of the elements and the relationships thereof involved in each claim in the first area 1000. The second area displays the content of the prior art, which facilitates the comparison of the differences between the invention and the prior art and better understanding of the invention. Optionally, the second area can switch to display the claims, the technical disclosure, and the prior art, or can display at least two of the above three in parallel. If the claims have not been formed, only the technical disclosure and the prior art can be displayed. Optionally, the second area can be hidden in the interface, for example, after being hidden, the mouse is moved to the side bar to trigger the second area to be displayed in the interface, and at the same time, the first area 1000 is reduced, and the visible information of the data structure in the first area 1000 is adaptively adjusted in size and position to be adaptively displayed in the first area 1000. Optionally, the technical disclosure and the prior art of the second area can also be set to be hidden and invisible, the mouse is moved to the corresponding position of the side bar or the top or bottom area to trigger the corresponding technical disclosure or prior art to be displayed in the interface, or to enlarge or reduce the display area. Of course, the size of the first area 1000 can also be adjusted or even hidden. Optionally, as shown in FIG. 22, the information in the chain structure diagram obtained by editing in the first area and the relationship thereof can be combined to form the data structure 3000 of the present application. Optionally, the data structure 3000 in the first area is displayed in the form of a tree diagram or a flowchart. Taking the tree diagram as an example, the tree diagram has a root node 3100 and branch nodes 3200, the root node 3100 is used to edit and display the subject matter protected by the claim, such as the largest product name involved in the invention. For example, the root node 3100 can be edited as a mobile device or a smart terminal if the invention is made for a mobile phone. The first-level branch nodes can be the second level of the root node 3100, and the second layer elements such as a circuit board, a camera assembly, and a communication assembly can be edited. The second-level branch nodes are the third level of the root node 3100, and the third layer elements such as a radio frequency circuit and an antenna in the communication assembly can be edited, and so on, until all elements involved in the invention are edited, and a complete structure decomposition is formed in the form of a tree diagram. The root node 3100 and the branch nodes are collectively referred to as nodes herein. The process of completing the tree diagram is also a process of repeatedly considering, repeatedly modifying, and gradually improving the claim architecture and layout in a visual manner by the agent. Compared with directly drafting the claim in a text editing software, the present application makes it easier to start drafting, more clear, more comprehensive in thinking, and less errors and omissions. Optionally, in the process of editing the data structure, the software receives a first instruction for text editing, node adding, node deleting, or changing the position of a selected node in the tree diagram or flowchart; that is, the first instruction for text editing, node adding, or changing the position of the selected node can be generated by directly right-clicking or clicking the corresponding operation button on the interface in the first area. The software responds to the first instruction to perform text editing, node adding, node deleting, or changing the position of the selected node in the tree diagram or flowchart. For example, right-clicking the first area and selecting the option "node adding", the software adds a node at the next level of the node where the right-click is located or at the previous level of the node where the right-click is directed. Alternatively, dragging the third-level node a to the area where the second-level node is located, the system generates the first instruction, and the software responds to the first instruction to move the third-level node a to the position of the second-level node, and the node a becomes a second-level node. The original second-level node can be stored at the same level as the node a or can be transposed with the node a. At the same time, the software can automatically adjust other nodes of the tree diagram or the shape or size of the tree diagram to adapt to the change of the node a. Optionally, after the editing of the lowest level node is completed, the software can automatically add lower level nodes to be edited on the basis of the lowest level node to facilitate possible node addition, without the need for manual node addition. When reading the data structure subsequently, the nodes without edited text are automatically ignored. The method of automatically adding lower nodes to be edited by the software based on the lowest nodes can include: I. The software automatically or in response to instructions reads the content of the lowest nodes, combines the preset prompt words, and sends them to the artificial intelligence large model through the API interface provided by the artificial intelligence large model, requiring the artificial intelligence large model to return recommended or reference information of lower nodes. The preset prompt words can include requiring the artificial intelligence large model to combine the lowest nodes, the technical disclosure, or combine the content of the technical disclosure and the tree diagram / flowchart and its generation rules to guess what the user wants to add to the lowest nodes. II. Display the recommended or reference information of the lower nodes returned by the artificial intelligence large model on the same screen without blocking the lowest nodes of the tree diagram / flowchart. III. If the user selects a recommended or reference information, generate nodes at the lower level of the tree diagram / flowchart lowest nodes, and the content of these nodes can be edited by the user. Optionally, after the selected node in the tree diagram or flowchart is edited, the node is added, or the node position is changed, it includes: The software receives a second instruction for defining the relationship of the selected node in the tree diagram. After the elements of the claim are edited in the tree diagram, the relationship of the elements needs to be defined, so the software receives the corresponding second instruction. Referring to FIG. 23, the software responds to the second instruction by drawing a line connecting the selected nodes in the first area. One end of the line connects the selected node such as A11, and the other end is a text input box, and the current cursor is in the text input box. If the number of selected nodes is greater than or equal to two, draw one or more lines connecting all selected nodes in the first area. Each end of the one or more lines connects the corresponding selected node such as A12, and the other end is a text input box, and the current cursor is in the text input box. If the drawn line connects two or more nodes, a text box is automatically displayed at the middle of the line or the adjacent end of the one or more lines, and the current cursor is in the text input box. For example, in FIG. 20, one end of the line connects B2, and the other end connects A12. A text box is displayed in the middle of the line or forms an area where text can be edited. Similarly, defining the relationship of the elements can also use the artificial intelligence large model, and the method includes: i、In response to selecting two or more nodes, read the contents of the selected nodes, combine the preset prompt words, and send them to the artificial intelligence large model through the API interface provided by the artificial intelligence large model, and ask the artificial intelligence large model to return the recommended or reference information of the relationship of the selected nodes. The preset prompt words can include asking the artificial intelligence large model to combine the selected nodes, the technical disclosure, or combine the contents of the technical disclosure and the tree diagram / flowchart and its generation rules, and guess what relationships the user wants to define. ii、Display the recommended or reference information of the relationship returned by the artificial intelligence large model on the screen, and do not block the tree diagram / flowchart; iii、If the user selects a recommended or reference information, generate a relationship near the selected node. The content of these relationships can be edited by the user. Similarly, if the user edits a node or a relationship, the original tree diagram / flowchart and the updated tree diagram / flowchart can be combined with the preset prompt words and sent to the artificial intelligence large model, and the artificial intelligence large model is required to return the modification recommendations of other nodes or relationships after editing the nodes or relationships. Of course, the artificial intelligence large model described above can be replaced by a rule base that defines the generation and modification rules of the tree diagram / flowchart and also defines the rules for patent drafting. Optionally, the formation of the tree diagram / flowchart is not limited to user editing. The technical disclosure and the preset prompt words can be sent to the artificial intelligence large model, and the artificial intelligence large model can generate one or several optional tree diagrams / flowcharts based on the technical disclosure and recommend them to the user. Optionally, if the screen of the first area cannot accommodate the tree diagram, the content can be automatically reduced or part of the content can be automatically hidden, or the range of the first area can be automatically expanded. The purpose of drawing lines is to visually see which elements or which elements have relationships between them. At the same time, lines define the links of elements and their relationships, and the relationship links between different elements. These link information constitutes part of the data structure. The appearance of the text box and the current cursor in the text box simplifies the operation, eliminating the need to move the current cursor and enter the editing state. After drawing a line, it can be automatically entered into the editing state, and the text can be directly input in the text box. Optionally, the number of text input boxes is one or two. For the case of one, the text input box is divided into a relationship input area and a technical problem input area. For the case of two, one text input box is used to input the relationship, and the other text box is used to input the technical problem. For example, the keyboard "shift" or other control keys can select any two or more nodes, the software automatically draws a line, and connects the two nodes to the corresponding text box. The current cursor automatically falls into any position of the text box, or clicking one of the text boxes can input the relationship description of the selected node, and clicking the other text box can input the technical problem to be solved. Alternatively, for tree graphs, several technical problems and relationships can be edited, and the order of each technical problem can be set, or "first technical problem", "second technical problem", etc. can be directly input in different text boxes to represent different claims, so that the software can be identified. Each technical problem corresponds to a claim, and after defining the order of the technical problem, the software can sort the subsequent formed claims according to the order of the technical problem. The data structure of the embodiments of the present application is not limited to the above form, and can also include graphs and elements and relationships on the graphs. For example, referring to FIG. 21, the user can directly draw a line on the graph, the starting point of the line is the position of the element corresponding to the product in the graph, and the end point is outside the range of the product. The user can edit the name of the element at the end point. In order to reflect the hierarchical relationship of the elements, the name of a plurality of parallel elements can also be used as the starting point, and the line can be continued to point to the name of the same upper element. In addition, the line can also be continued to edit the relationship of the elements, and finally form a data structure. The above data structure can be used as reference information for writing, and can be searched out when relevant content is input in the current text. For example, when the user edits the above data structure or a similar data structure in the current text, the system can automatically recommend it, and the user can directly use the recommended data structure, and modify and edit it on the basis of the recommended data structure. The use method of the above data structure is described below: S200: reading the data structure; After the data structure is edited or obtained, the "OK" button or the function key is clicked, and the system generates a third instruction. The software reads the data structure in response to the third instruction, or directly reads the data structure after obtaining the data structure, and analyzes the data structure to obtain the set of elements and the set of relationships constituting each claim. If the data structure contains information of the technical problem, the technical problem, the set of elements corresponding to the technical problem, the set of relationships, and the order information of the technical problem are obtained. Alternatively, after analyzing the data structure, a protected subject in the data structure is determined according to the second preset rule, a first claim is determined under the determined protected subject, and the set of elements and the set of relationships associated with the first claim are read. For example, the second preset rule includes that the first technical problem is preferred to be found, and then the first technical problem is found in the parsed data structure, and the set of elements and the set of relations associated with the first technical problem are read. The first technical problem means that the first technical problem is determined as a protected subject when the data structure is edited, and therefore the first technical problem is found, which means that the first claim is determined under the determined protected subject. Then, according to the second preset rule, the second, third, and the like technical problems and the set of elements and the set of relations associated with the technical problems are read in sequence. Optionally, there are two or more protected subjects in the data structure, and the independent claims corresponding to each protected subject can be marked, such as marked with different letters, the serial numbers of the first set of claims are named in the form of Am, where m is an Arabic numeral. The serial numbers of the second set of claims are named in the form of Bn, where n is an Arabic numeral obtained by adding one to the last claim of the first set of claims. The two sets of claims are distinguished by the first letters A and B. The serial numbers of other sets of claims are named in the same way. The order of the independent claims can be determined according to the order of the first letters. The independent claim after the first letter is arranged after the independent claim corresponding to the first set of claims before the first letter. In this way, the software can conveniently and accurately sort the claims, determine which is an independent claim, and determine which is a dependent claim subordinate to which independent claim. If a combination of letters is used to mark the claims, the order of the technical problems can be omitted. Of course, the naming method of the serial numbers is not limited to this, as long as the software can recognize the predetermined rules. S300: Forming a corresponding claim according to the set of elements and the set of relations corresponding to each claim according to the first preset rule; The first preset rule is a rule for constructing a claim that meets at least part of the formal requirements using the set of elements and the set of relations. For example, the format, sentence order, punctuation marks, and the like required by the claim. Optionally, the set of elements associated with the first claim is determined according to the set of relations associated with the first claim, and the order of the elements in the set of elements associated with the first claim in the first claim is determined. For example, refer to FIG. 24 and the following table of data structure: Table Five: Set of Elements Table Six: Set of Relations Table five and table six are optional data structures in the embodiments of the present application. In table five, the biggest protected subject is the advertisement module in column a, the second level element is the drawing board, frame, pin, bolt, elastic member and safety rope in column b. The third level element is each element in column c, and the last level element is each element in column d. Among them, the strip board and the decoration element in column c belong to the drawing board in column b, and other elements have similar relationships. When the software reads the data structure in the form of table, it will parse that the strip board and the decoration element belong to the drawing board according to the area relationship of the table and the area position of the element in the table, and the first mounting hole belongs to the strip board. The mounting plate belongs to the frame, and the plate hole and the first rope hole belong to the mounting plate, and so on. The relationship set of table four reflects the different relationships of each element in table three. Table three and table four are associated through the address of column a such as B6. In table six, the first technical problem corresponding to the first claim is "easy to disassemble", and the relationship set is "B6 connects b1 at one end, passes through d3 at the other end, b8 is inserted into d7, and is located on the side of c3 opposite to b1". According to the order of elements appearing in the first claim, the template corresponding to the first claim is used to form the organization structure of the elements and the organization structure of the relationships of the first claim. The first claim is an independent claim, so its template can be defined as follows: The above italicized and underlined serial numbers "1" and "advertisement module" are variables, and the rest are template constants. The serial number is given according to the rule, such as the first claim, output "1", the second claim, output "2", and so on. "Advertisement module" is the protected subject, which can be obtained from the element set of table two. The leftmost element in the table, or the root element in the chain structure, is one of the protected subjects. Of course, elements in the middle area of the chain structure can also be used as protected subjects, which can be defined and set according to the situation. The last of the template is a period, and only one period is allowed in a claim. The period is placed at the end of the template. Between "characterized in that" and "including", and between "including" and ".", is the fillable space of the remaining technical features. Optionally, the set of elements includes a set of product elements, the set of relations includes a set of element relations, the first claim is an independent claim, the next line of the independent claim and before the period is defined as filling in the upper necessary elements (elements) of the lowermost necessary elements (elements). Optionally, in Table Four "Set of Relations", the necessary elements and their relations embody the lowermost necessary elements and their relations, and the upper necessary elements and their relations if necessary. It should be noted that the "necessary elements and their relations" referred to here are not necessarily necessary, depending on the technical problem to be solved, so the "necessary elements", "necessary elements" and "necessary relations" referred to in this paper all refer to "elements", "elements" and "relations" respectively. Different technical problems, at least part of the "necessary elements", "necessary elements" and "necessary relations" are not the same. First, determine the lowermost necessary elements, and then determine the order of the elements in the set of elements associated with the first claim in the first claim using the set of relations associated with the first claim. The method can include: Using the set of relations associated with the first claim, all elements in the first claim are constructed into a subject-object ordering link, and the elements in the subject-object ordering link are sorted in reverse order of their sequence. For example, all elements and their relations in the first claim are "B6 is connected to bl at one end, passes through d3 at the other end, b8 is inserted into d7 and is located on the side of c3 opposite b1". Optionally, in order to ensure that the constructed subject-object ordering link meets the requirements, the software automatically checks whether a closed loop is achieved in each claim. The closed loop condition includes whether all elements are associated with at least another element within a claim or between different claims, and whether the subject-object ordering link is only one, otherwise a prompt that the closed loop cannot be achieved is given. The software (system) further analyzes all elements and their relations in the first claim and obtains two links "b8-d7(b1), b6-b1(d3)". Analysis shows that these two links constitute a substantial single link through b1, which meets the requirements. The lowermost necessary elements of the single link or two links are b1, d7 and d3. It should be noted that the lowermost necessary elements referred to here refer to the lowermost elements involved in the current claim. Sort the lowermost necessary elements. Since b1 is the object of the two links, it is placed first, and d7 and d3 are in a parallel relationship in the link, so the necessary relationship needs to be sorted using the order of appearance. d3 is placed after b1, and d7 is placed after d3. Determine all upper necessary elements of the lowermost necessary element, the upper necessary elements of b1 is b1 itself, the upper necessary elements of d7 is c7, b6, the upper necessary elements of d3 is c3, b3, the principle of determination is single chain upward acquisition, and branches are discarded. The upward acquisition refers to that for the chain structure, the node located at the upper level is acquired. Then, the software determines the order of all upper necessary elements: b1, b3, b6, b8. Continue to fill in the determined necessary elements and their necessary relationships in the writing template of the independent claim. Optionally, according to the order of the lowermost necessary element, the containing relationship of the upper necessary element is listed in sections, and the corresponding necessary relationship is listed. Optionally, in the containing relationship, “has” or “forms” is used for hole, cavity, recess, gap, space, channel, airway, liquid channel, opening, and “includes” is used by default. According to the above method, the first claim is as follows: The above italic and underlined content are variables, and the rest are template constants. The element appearing for the first time does not need to be added with “the” in front of it, and the element appearing for the first time needs to be added with “the” in front of it. It can be understood that the first preset rule is not unique, and different settings can be made according to needs. For example, it is not necessary to consider whether a closed loop is realized in each claim, the condition of the closed loop does not necessarily include only in one claim, whether all elements are associated with at least one other element between different claims, and the like. For example, by introducing a smart large model, the element and its relationship are input, and the claim is obtained. Optionally, after the claim is formed, if the user modifies the corresponding table, tree diagram or flowchart, the software can automatically send the table, tree diagram or flowchart before and after the modification, the claim, and the preset prompt word to the artificial intelligence large model, so that the artificial intelligence large model gives the corresponding updated claim based on the modified table, tree diagram or flowchart. The embodiment of the application also provides a text writing assistance method, comprising: i: The server receives the file name sent by the client, the file name comprising a file identifier, and the file identifier being associated with text writing reference information; ii: The server acquires the writing reference information associated with the file identifier; iii: The server sends the writing reference information to the client, so that the client displays the acquired writing reference information while displaying the content of the text. Referring to FIG. 25, FIG. 25 is a structural schematic diagram of an embodiment of the computer device 50 provided in the present application. In an embodiment, the computer device 50 comprises a memory 51 and a processor 52 coupled with each other. The processor 52 is configured to execute program instructions stored in the memory 51 to implement the steps of any of the above-described embodiments of the writing assistance method. In a specific implementation scenario, the computer device 50 can include, but is not limited to, a microcomputer, a server, and the like. In addition, the computer device 50 can also include a notebook computer, a tablet computer, and the like mobile devices, without limitation. Specifically, the processor 52 is configured to control itself and the memory 51 to implement the steps of any of the above-described embodiments of the writing assistance method. The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 can be an integrated circuit chip having a processing capability of signals. The processor 52 can also be a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like. In addition, the processor 52 can be jointly implemented by integrated circuit chips. Referring to FIG. 26, FIG. 26 is a structural schematic diagram of an embodiment of the computer readable storage medium provided in the present application. The computer readable storage medium 60 of the embodiment of the present application stores program instructions 61, which, when executed, implement the method provided by any embodiment of the writing assistance method of the present application and any non-conflicting combination. The program instructions 61 can form a program file and be stored in the computer readable storage medium 60 in the form of a software product, so that a computer device (which can be a personal computer, a server, or a network device, etc.) executes all or part of the steps of the method of each embodiment of the present application. The aforementioned computer readable storage medium 60 includes various media that can store program codes, such as a U disk, a mobile hard disk, a ROM (Read-Only Memory), a RAM (Random Access Memory), a magnetic disk or an optical disk, or a terminal device such as a computer, a server, a mobile phone, a tablet, etc. In an embodiment, the patent-related text writing assisting method is executed by program instructions, which are installed in the existing word editing software in the form of plug-ins, and act on the patent-related text opened by the word editing software to realize the patent-related text writing assisting method. In an embodiment, the patent-related text writing assisting method is executed by program instructions, which are installed in the existing word editing software in the form of plug-ins, and act on the patent-related text opened by the word editing software to realize the patent-related text writing assisting method. In an embodiment, the patent-related text writing assisting method is executed by program instructions, which are installed in the existing word editing software in the form of plug-ins, and act on the patent-related text opened by the word editing software to realize the patent-related text writing assisting method. In an embodiment, the patent-related text writing assisting method is executed by program instructions, which are installed in the existing word editing software in the form of plug-ins, and act on the patent-related text opened by the word editing software to realize the patent-related text writing assisting method. In an embodiment, the patent-related text writing assisting method is executed by program instructions, which are installed in the existing word editing software in the form of plug-ins, and act on the patent-related text opened by the word editing software to realize the patent-related text writing assisting method. It can be understood that the embodiments of the present application utilize the naming rules of the identification file adopted by the enterprise, and can automatically extract the file identification of the file name during text editing, and automatically obtain the related data in the enterprise or the related enterprise by using the file identification, and then provide relevant and strong information support for text editing, without relying on manual memory or searching, thereby improving work efficiency. When the embodiment of the application is applied to the patent writing assistance scenario, optionally, the patent agent can obtain information guidance like a senior teacher in real time without manual assistance during the writing process, and real-time help is provided for how to write, how to write, and how to modify; optionally, the patent agent can understand the requirements of the applicant, the law firm and other related parties in time without manual assistance during the writing process, so as to avoid complaints or criticism due to forgetting the requirements of the applicant or the law firm; optionally, during the writing process or at the end, the text obtained by the patent agent can also be modified or prompted according to the default or user-defined requirements without the patent agent's awareness, which is equivalent to providing the patent agent with an unwearied and stable writing assistant, assisting in checking and modifying; optionally, since the system can automatically search the current file or the server related text by taking at least part of the input content as a keyword, compared with the existing input method which can only provide public vocabulary or high-frequency vocabulary recommendation, the recommendation mechanism of the embodiment of the application is based on the principle of proximity or the principle of early, the principle of proximity is based on the fact that the user will input the same content in the same text area, and the principle of early is that in the claim, the name of the thing will be described first, and then other relationships will be described, so that the first occurrence position of the thing is obtained, the boundary of the name of the thing is accurately defined, and the search result with very high relevance and very high accuracy can be obtained, and the search result is not limited to vocabulary, but can also be a paragraph, a sentence or a data structure, which greatly improves the input efficiency. At present, part of the work of patent workers is to write patent files to form patent related texts, and then submit them as patent applications through official channels. Patent applications generally include invention applications, utility model applications or design applications. Among them, the patent related texts involved in invention applications and utility model applications generally include abstracts, claims, descriptions and the like. Due to the requirements of the patent law, part of the content in the abstract, the claim and the description is substantially the same, part of the content in different claims or different positions of the same claim in the claim may also be substantially the same, and part of the content in different positions of the description may also be substantially the same. Therefore, if a sentence or several words in the patent related text are modified, other places of the patent related text may also need to be found out by the user one by one for consistency modification. The current word processing tool provides "find" and "replace" functions. The user needs to operate the word processing software interface, open the "find" and "replace" window, input the content to be found and replaced in the input field of the window, and then click "find" or "replace all". The above-mentioned "search", "replace" function provides a more efficient way for users to find, but as the patent work towards higher quality requirements, the applicant, patent agency work more standardized, patent-related text as a patent application through official channels before submission, may need to be repeated several times to modify, improve. Each modification and improvement, there may be several or more places to find and replace the operation, the user needs to frequently operate the text editing software interface, open the "search", "replace" window, enter the content to be searched and replaced in the input field of the window, and then click "search" or "replace all", the work efficiency is still relatively low. And the current text editing software "search", "replace" function, can only be 100% match, such as replace "abc", can only find out other "abc" string content, and can not find out "a11bc" such content with label, also can not find out "a said bc" such content, and vice versa, resulting in a lot of matching blind area, the user needs to find out and make consistent modification, can not further improve the work efficiency. On the other hand, the corresponding structure name in the specification label is a time-consuming and careful process, especially in the case of complex product structure, the structure name is various, and the label also needs to follow the relationship between the structure name to distinguish. The existing label is generally found by manual one by one, which is tedious and prone to errors. The patent-related text automatic labeling method, device and storage medium provided by the present application are described in detail below in conjunction with embodiments, which can partially or completely solve the technical problems of low processing efficiency and accuracy of patent-related text. Please refer to FIG. 27, which is a flowchart of an embodiment of the patent-related text automatic labeling method of the present application. FIG. 27 only shows the operation process of the claim, and the patent-related text automatic labeling method comprises: S100: Determine the claim or specification range in the patent-related text; The purpose of determining the claim or specification range is to determine the structure name of each element of the product, so as to facilitate subsequent labeling of the structure name. In the computer "eye", which words in a sentence in a file are correct structure names, or whether there is a structure name in the sentence, is invisible or unknown. The computer "eye" of a file is only a collection of binary coded data, and cannot distinguish which is a structure name and which is not, nor can it distinguish the left and right boundaries of the structure name. Therefore, the claim or specification range in the patent-related text is first determined. The patent-related texts of an invention application or a utility model application generally include an abstract, a claim, a specification, an abstract drawing, and a specification drawing. If the claim or the specification is a separate document, the scope of the claim or the specification is the entire document. If the claim and at least one of the abstract, the specification, the abstract drawing, and the specification drawing constitute a document, or the specification and at least one of the abstract, the claim, the abstract drawing, and the specification drawing constitute a document, a means is needed to determine the scope of the claim or the specification. The following is described by taking the determination of the scope of the claim as an example: S110: searching for "claim" in the patent-related texts; The patent-related texts have a fixed format, and the module name is listed before each part. The "claim" is the module name, and the module name means that the claim will be found after the module name. S120: in response to finding the "claim", taking the found result as a first boundary, and searching for "technical field", "specification abstract", "abstract drawing", "specification drawing", "specification", or a document end symbol after the "claim"; If the patent-related texts have a next module after the "claim" module, the next module is one of "technical field", "specification abstract", "abstract drawing", "specification drawing", "specification", and if there is no next module, the end of the document is the document end symbol. S130: in response to finding "technical field", "specification abstract", "abstract drawing", "specification drawing", "specification", or a document end symbol, taking the found result as a second boundary, and taking the region between the first boundary and the second boundary as the scope of the claim. S200: defining the structural names in the scope of the claim or the specification; As described above, the computer cannot distinguish which are the structural names and which are not, and cannot distinguish the left and right boundaries of the structural names. Therefore, it is necessary to determine which are the correct structural names in the claim to facilitate subsequent labeling. There can be various methods for defining the structural names in the scope of the claim, such as: Using the first word leading to the structural name in the claim and the tail symbol after the structural name, searching for the first content between the first word and the tail symbol or the second content between adjacent first words in the scope of the claim, and taking the first content or the second content as the structural name, or further determining the structural name from the first content or the second content. The first word includes any one of "containing", "having", "provided with", and "including", and the tail symbol includes a punctuation mark. As mentioned above, the format of the claim is fixed, and the description of the structure name is also fixed, so the format of the structure name can be located by locating the first word and the end symbol. Here is an example of a claim: 1. An advertisement module, comprising: B1; B3, comprising c3, the c3 having d3; B6, comprising c7, the c7 having d7, the b6 connecting to the b1 at one end and passing through the d3 at the other end; B8, being inserted into the d7 and located on the side of the c3 opposite to the b1. In the above claim, the advertisement module comprises B1, B3, B6 and B8. Here, "comprising" is the first word, which leads to the following B1, B3, B6 and B8. Therefore, the first word is defined as a word that can lead to the structure name. After the structure name is led, a punctuation mark is generally followed, so the punctuation mark is defined as the end symbol. With the combination of the first word and the end symbol, the first content and the second content containing the structure name can be generally defined. Further, for B3 and the like, the structure name such as c3 is also included, and a punctuation mark is generally followed after c3, so the structure name such as c3 can be further determined. However, the first content and the second content may not only contain the structure name, but also contain some words such as adjectives, so these irrelevant words need to be filtered out. The method of filtering out these irrelevant words, that is, the method of further determining the structure name from the first content and the second content, includes: Finding the content after the last "of" and before the end symbol in the first content and the second content as the structure name. When the structure name has multiple adjectives, find the content after the last "of" and before the end symbol as the structure name. When there is only one adjective, the first "of" is also the last "of". The method of defining the structure name in the scope of the claim also includes: Determining at least two same words in the scope of the claim, and at least one of the same words has a definite article "the", and the other does not have a definite article "the", and taking the same word as the structure name. As mentioned above, the format of the claim is fixed, and the structure name appearing for the second time needs to be preceded by "the", so by using this fixed format, as long as two same words are found, and the first word does not have "the" or "this", and the second word has "the" or "this", the same word is taken as the noun. Optionally, the number of the same words determined in the scope of the claim is at least two, including: S210: determining a word with two words in the scope of the claim, and traversing all the same words with two words in the scope of the claim and the same as the word; The word with one word is not operable, so the number of words of the word is first determined to be two, two words of the word are read at the beginning of the claim, and are traversed in the scope of the claim; of course, two words of the word can also be randomly read in the claim, and some formatted words such as “according to”, “claim” and the like are filtered out; S220: in response to finding more than one same word, adding one character to the left or right of the word to the word as another word with three words, continuing to refer to the step of traversing all the same words with two words in the scope of the claim and the same as the word, and traversing and subsequent steps for another word, and continuing to loop; In the claim, the structure name generally appears at least twice, if more than one same word can be found, it is possible that the number of the same word is more, and then the range of the word is continuously expanded, such as finding other words to the left or right of the same word, if the word is found, the word is added to the same word, so that the number of the new same word becomes three, and then the step of traversing all the same words with two words in the scope of the claim and the same as the word is continued, and the traversal and subsequent steps are continued, as long as more than one same word can be found, the loop is continuously performed; S230: in response to being unable to find more than one same word, determining another word with two words in the scope of the claim, and continuing to refer to the step of traversing all the same words with two words in the scope of the claim and the same as the word, and traversing and subsequent steps for another word, and continuing to loop until the scope of the claim is traversed. If in the process of looping to find more than two same words, it is found that after increasing the number of words of the word, another same word cannot be found any more, it is indicated that the new word with increased number of words is not a correct structure name, and then the loop finding is terminated, and the word with the same word found last time is taken as the structure name. Re-read two words of the word in the claim, and find the same word of the new word again until the entire claim is read, and all the structure names are defined. S300: hierarchically classifying the at least two structure names defined; In order to have a label with a hierarchy, the different structure names can be hierarchically classified according to the relationship between the different structure names during or after the definition of the structure name, for example: The two structure names immediately before and after the first word are classified as upper and lower levels; For example, in the aforementioned claim example, the first word in the technical feature "B3, including c3," is "including", and the two structural names immediately before and after "including" are B3 and c3, so B3 is the upper level of c3, and c3 is the lower level of B3. S400: According to the hierarchical division, at least two structural names are labeled. That is, according to the label of the upper and lower levels, the two structural names of the upper and lower levels are labeled. For example, the label of B3 is 10, and the label of c3 is 11. According to the above method, all structural names are labeled. Optionally, the above steps S300 and S400 can also be replaced in the following way: At least two structural names defined are labeled in a random label manner or in a pre-defined label manner. That is, the structural names can not be divided into levels, but all structural names are labeled in a random label manner, or all structural names are labeled according to a pre-defined label manner. For example, according to the order of the appearance of the structural names, they are sequentially numbered in the order of Roman numerals 1, 2, 3, and so on. Optionally, if the user finds that some structural names cannot be recognized by the program, the user can manually supplement the label, so that after labeling at least two structural names, it includes: S410: Receive a first instruction for inserting a label in the patent-related text; The user selects a structural name in the patent-related text that is missing a label and inserts a label after it, thus triggering the first instruction. S420: In response to the first instruction, a label is inserted in the patent-related text; The user's action of supplementing the label is complete, and the label has been inserted after the structural name that is missing a label. S430: Taking the above one label as a third boundary, determining a structural name immediately before the third boundary; Although the user can clearly know which structural name is missing a label, the computer does not know, but it can use the above action of inserting a label to identify it, and then make the same word consistent in other places. Because the label is on the right side of the structural name, the above inserted label can be taken as the third boundary, that is, the right boundary, so that the computer obtains a structural name immediately before the third boundary. The left boundary of a general structural name can be "of" or "said", so by using the left boundary and the right boundary, the structural name that is missing a label is determined. S440: Search for the structural name that is missing a label in the patent-related text, and insert the above one label after the searched content; For example, in the specification or claims, find the structure name of the defined missing label, and insert the above-mentioned one label to the right of the structure name. Alternatively, if the user is not satisfied with the program automatic labeling, or the error labels a position that should not be labeled, after labeling at least two structure names, the method comprises: S440: receiving a second instruction to delete or replace a label; The user deletes or replaces a label, triggering the second instruction. S450: in response to the second instruction, deleting or replacing the label; The program responds to the second instruction and deletes or replaces the label. S460: determining a structure name immediately preceding the fourth boundary as a fourth boundary, and determining a structure name immediately preceding the fourth boundary as a fourth boundary; After deleting or replacing the label, the remaining same label also needs to be deleted or replaced, which can be directly searched in the entire text and then deleted or replaced. Of course, to be more accurate, the structure name immediately preceding the fourth boundary can be determined as a fourth boundary, and the determined structure name can be positioned. S470: searching for a structure name in the patent-related text, and deleting or replacing the label of the searched content. Search in the specified part of the patent-related text to locate all the same words. After locating the structure name, the label to the right of the structure name can be deleted or replaced. Or Alternatively, if it is necessary to rename the structure name, the method comprises: S480: receiving a third instruction to revise the patent-related text; The user modifies a structure name, triggering the generation of the third instruction. S490: in response to the third instruction, determining whether the revision is directed to a structure name or immediately adjacent to a structure name; The user makes modifications in the patent-related text, and the computer cannot directly know whether the modification is directed to a structure name or whether the structure name is renamed, so it is necessary to determine whether the revision is directed to a structure name or immediately adjacent to a structure name. The method for determining whether the revision is directed to a structure name can be to directly search for the revised content in all structure names, if a structure name is found, then search for the revised position content with the found structure name, and if a matching item is also found, it is determined that the modification is directed to the structure name. S491: In response to the revision being directed to one structural name, or immediately adjacent to one structural name, replace at least another one structural name in the patent-related text with the revised one structural name. After determining that the modification is directed to the structural name, perform a full-text search and replace. Optionally, the range of the at least two structural names marked can be limited to the specific embodiments of the specification, can include the claims and the specific embodiments of the specification, can be only the claims, or can even be a user-selected area or selected preface or postface. The structural name mark in the claims needs to be added in parentheses. Optionally, the mark in the claims needs to be added in parentheses. Optionally, the timing of the partial or full-text search and replace, modification, and how to automatically modify after the timing is reached can be as follows: S11: Receive an artificial revision end instruction directed to the patent-related text. The artificial revision end instruction refers to an instruction reflecting that the user has completed at least one revision operation. The artificial revision end instruction can be actively issued by the user, or can be passively issued in the user's operation process. In an embodiment, receiving an artificial revision end instruction directed to the patent-related text comprises: in response to at least one of the user clicking an operation item in the user interface, pressing a preset key, or inputting a preset instruction, triggering an artificial revision end instruction. Specifically, at least one of the user actively clicking an operation item, pressing a preset key, or inputting a preset instruction can represent that the user confirms that the current revision action has been completed and wants to trigger an artificial revision end instruction. The operation item can be a button in the user interface. For example, the "run" button on the software interaction interface. The user can interact with the operation item through common operation methods such as clicking and sliding. Pressing the preset key can be a certain physical key position on the keyboard or other input. For example, the "F1" key on the keyboard can be set as the preset key. The preset key can also be a combination key, for example, it can be set to trigger an artificial revision end instruction when Ctrl+F1 on the keyboard is pressed at the same time. Inputting a preset instruction can be that the user inputs a specific string of characters to trigger an artificial revision end instruction. The input of the preset instruction can be input in a specific preset area. Only input in the preset area triggers an artificial revision end instruction. The input of the preset instruction can also detect the input content of the user in real time. As long as the condition is met, an artificial revision end instruction is triggered. For example, it is set to trigger an artificial revision end instruction as long as a specific format of command symbol is detected, such as "run modification follow". Through the above setting, the user can first repeatedly adjust the content to be revised, and only actively trigger the manual revision end instruction after, for example, the manual revision is completed or after a certain stage. The content before revision can be processed in batches in one operation, which is more efficient. In addition, the manual revision end instruction is actively issued by the user, and the user can clearly master the revised content and times, and is less likely to have unexpected operations beyond the user's expectations, and is stable to use. In another embodiment, receiving the manual revision end instruction for the patent-related text comprises: in response to a preset change in the user's cursor position, triggering a manual revision end instruction. Specifically, when the user changes the position of the text input cursor through the keyboard or mouse, such as moving the cursor position from the current line to another line, it means that the user has performed an operation. Each operation of the user is detected and triggers a manual revision end instruction, that is, the manual revision end instruction is triggered passively in real time when the user operates. Through the above setting, the user revises the content and triggers the manual revision end instruction in real time after completing the revision. The user does not need to perform the operation of triggering the manual revision end instruction after each revision is completed, and the user operation steps are less, and the use is convenient. The preset change can be moving to the next character position or the previous character position, or jumping to other character positions. In an embodiment, the preset change is to detect that the user's cursor changes line. Thus, on the basis of real-time triggering, the number of redundant triggers can be reduced, and the processing speed is considered. S12: In response to the manual revision end instruction, or in response to satisfying a preset condition, determining the content before revision in the patent-related text. In response to satisfying the preset condition, the preset delay time length after receiving the manual revision end instruction has arrived, such as 1 minute after receiving the manual revision end instruction, which satisfies the preset condition. This can determine the content before revision in the patent-related text when the user is in a state of not editing the file or not operating the computer, without affecting the user's use experience. In particular, the pre-revision content refers to the content before the user operation occurs. The pre-revision content should include at least one character or format symbol. The pre-revision content can include the changed content formed by the operations of adding, deleting, changing, etc. For example, when the user revises "abcd" in the patent-related text to "abzz", the pre-revision content at least includes "cd". Or, the user revises "abcd" in the patent-related text to "ab", the pre-revision content at least includes "cd". For another example, the user revises "ab" in the patent-related text to "abcd", the pre-revision content at least includes "ab". Or, the user renames a structure name or re-labels the label after the structure name. When the end instruction is responded to and no pre-revision content is detected, the end instruction can be triggered and the subsequent steps are no longer continued. The way of determining the pre-revision content as a structure name has been described above, and thus will not be repeated here. In an embodiment, determining the pre-revision content in the patent-related text specifically includes: In addition, determining the third content that has been revised in the patent-related text, such as recording the last or several times of the third content that has been revised in the background, or Determining the third content with a revision format in the patent-related text. The revision format can be a format reflecting the revision operation in other software. For example, the revision format can be a format generated by the "revision" function in a word processing software. Determining the third content can be to take a number of continuous characters with a revision format as a third content. For example, the patent-related text includes "abcdefg", wherein "ab" and "efg" have a revision format, then "ab" is taken as a third content and "efg" is taken as another third content. Based on the third content, determining the fourth content of the minimum matching unit, the fourth content has a length greater than or equal to the length of the third content, and the fourth content is taken as the pre-revision content. Determining the fourth content can be done by directly using each part containing the third content as the fourth content. For example, when the third content is "ab", the fourth content is also directly determined to be "ab". Alternatively, the fourth content can be determined by finding a complete word or phrase corresponding to the position of the third content. When determining the fourth content, the location of the third content can be detected, and the text can be expanded in both directions based on the location of the third content until a complete word or phrase is obtained as the fourth content. For example, in patent-related text containing "abcdefg", where "de" is determined to be the third content, it can be expanded forward and backward based on "de" until "abcdefg" is obtained as the fourth content. When several third contents are adjacent, causing the complete words or phrases obtained by expanding from these third contents to overlap, the results of expanding from these third contents can be merged to form only one fourth content. For example, in patent-related text containing "abcdefg", where "bc" and "de" are determined to be the third content respectively, it can be expanded forward and backward based on "bc" to obtain "abcdefg", and based on "de" to obtain "abcdefg". If the results of expanding "bc" and "de" to the preceding and following text are the same, then two separate fourth contents of "abcdefg" will no longer be formed, but only one fourth contents of "abcdefg" will be formed. Understandably, one purpose of determining the fourth content is to make the matching more accurate. Therefore, the specific strategy for obtaining the fourth content based on the third content can be set according to the actual situation. In addition, when the revision type is "change," the determination of the fourth content should be based on the text before the revision, and the revised content can be ignored. With the above settings, users can directly indicate the location of revisions using existing software's revision formats and background records. Furthermore, determining the fourth content based on the third content and using the fourth content as the original content enhances the uniqueness of the original content within the patent-related text, reduces noise during matching, and ensures accurate positioning. In one specific embodiment, determining the fourth content of the minimum matching unit based on the third content includes: Starting from the third content, find the first first separator that appears in the preceding and following text. Using the first first separator that appears in the preceding and following text as the ending point, the content between the ending points of the preceding and following texts is taken as the fourth content. The first separator can be a comma, period, semicolon, space, paragraph break, or other special symbol. For example, if the patent-related text contains "abc,defgh,ijk.", where "fg" is the third content, then the text is expanded forward and backward until the first separator comma is encountered as the endpoint, thus determining the fourth content as "defgh". Thus, according to the first first separator appearing in the foregoing and the following, the text containing the third content is intercepted, and a sentence segment containing a unique feature can be easily obtained, so that when the fourth content is taken as a minimum matching unit, false matching noise is not easily generated. In another specific embodiment, determining the fourth content of the minimum matching unit comprises: The third content is taken as the center, and the preset length of the foregoing and the following is taken as the boundary, to determine the fourth content of the minimum matching unit as the preset length of the foregoing and the following content and the third content. The preset length can be a value preset in advance, for example, 3, 4, 5, etc. The preset length can be set by the user or can be a fixed value. The fourth content is obtained by concatenating the characters of the foregoing preset length, the third content, and the characters of the following preset length in sequence. In addition, when the length of the foregoing and the following of the paragraph in which the third content is located is less than the preset length, the maximum foregoing and following of the paragraph in which the third content is located that can be intercepted is taken as the cutoff point to determine the fourth content of the minimum matching unit. For example, when the third content is located at the end of the paragraph, the foregoing content is intercepted according to the preset length, and the following content is intercepted according to the actual remaining character length of the paragraph. Through the above setting, the fourth content of the minimum matching unit can be obtained simply and quickly, the function is simple to implement, and the processing speed is fast. In another embodiment, the revised content in the patent-related text is determined, specifically comprising: The content selected by the user in the patent-related text is taken as the revised content. When the selected content contains content that is changed by user operation, the selected content as a whole is the revised content and is taken as the minimum matching unit. When the selected content does not contain content that is changed by user operation, i.e., the selected content is all original text, an end instruction can be triggered, and the subsequent steps are no longer continued. Through the above setting, the revised content in the patent-related text does not need to rely on the revision format of other software, and the application range is wider. In another other embodiment, the revised content in the patent-related text is determined, specifically comprising: The third content determined after comparing the current version of the patent-related text and the historical version of the patent-related text is taken as the revised content. The historical version of the patent-related text can be another old version file, an old version of the patent-related text located at another text position in the same file, etc. By comparing the historical version of the patent-related text with the current patent-related text, the part of the current patent-related text that is changed relative to the historical version of the patent-related text can be obtained. The changed part can be taken as the revised content. In addition, the structure name in the scope of the claims or the specification can also be defined, and the method of defining the structure name can refer to the foregoing embodiments, which will not be described herein. By the foregoing method, one-key automatic labeling can be realized without manual labeling, without manual recognition of the structure name, and without manual arrangement of the structure name, thereby greatly improving the editing efficiency and reducing the error rate of manual labeling. In addition, the automatic recognition of the structure name can also be used to quickly and consistently label the missing label, and to quickly and consistently modify the label, so as to facilitate the response to various labeling scenarios. S13: Matching the remaining part or the selected part of the patent-related text with the structure name or the content before revision. Specifically, matching the remaining part or the selected part of the patent-related text with the content before revision can obtain the position of the content in the remaining part or the selected part of the patent-related text that needs to be modified synchronously. For example, if the content before revision is “abc”, all “abc” in the remaining part or the selected part of the patent-related text is matched. The matching can be performed in the entire remaining part of the patent-related text, that is, full-text matching. The matching can also be performed only in the selected part, and the other part is not processed. The specific implementation of the matching can be any character matching technology that can locate the position, which will not be described herein. In an embodiment, matching the remaining part or the selected part of the patent-related text with the content before revision comprises: In response to not matching the target content based on the content before revision, a second delimiter in the content before revision is found. Specifically, in the process of matching the remaining part or the selected part of the patent-related text with the content before revision, the case of not matching the corresponding result can occur. In this embodiment, the content before revision in the patent-related text has corresponding content in the remaining part of the patent-related text as a probable case. The case of not matching the corresponding result is first identified as an abnormal case. The case of not matching the corresponding result can be caused by character changes, long sentence splitting, grammar changes, and the like. Therefore, in this embodiment, a second delimiter in the content before revision is first found as a basis for further processing. The second delimiter can be a comma, a semicolon, a period, a space, or the like. In response to the absence of the second delimiter, a conjunction in the content before revision is replaced with the subject of the previous sentence, or the subject of the previous sentence is inserted at the beginning of the content before revision. When the second separator is not found, it means that the probability of not matching the corresponding result due to long sentence splitting is low, and the probability of not matching the corresponding result due to syntax change should be considered first. For example, a sentence in the pre-revision content replaces the subject of the sentence with a conjunction word due to writing habits when writing, and the subject is not replaced with a conjunction word when writing the remaining part of the patent-related text. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the conjunction word in the pre-revision content with the subject of the previous sentence, so that the matching result is not easily missed. For example, the first sentence of the pre-revision content is "a includes b.", the pre-revision content is "also includes c.", and in the remaining part of the patent-related text, it is recorded as "a includes b" and "a includes c". At this time, it may not be possible to get the matching result directly by using "also includes c.". After replacing "also" with the subject "a" of the previous sentence, the correct matching can be achieved. In response to the presence of the second separator, the pre-revision content is split into a preceding part and a plurality of subsequent parts based on the second separator, and the key word in each subsequent part is replaced with the subject of the preceding part, or the subject of the preceding part is inserted at the beginning of each subsequent part. When the second separator is found, it means that the probability of not matching the corresponding result due to long sentence splitting is high. For example, two short sentences in the pre-revision content are connected by a colon, "and" and the like, and when writing the remaining part of the patent-related text, they are split into two short sentences, and the colon, "and" and the like in the subsequent short sentence are replaced with the subject. At this time, although the meaning is consistent, the text is not consistent. However, the above situation can be included in the matching range by replacing the key word in each subsequent part with the subject of the preceding part, so that the matching result is not easily missed. For example, the pre-revision content is "a includes b, and includes cd.", and in the remaining part of the patent-related text, it is recorded as "a includes b. a includes cd.". At this time, it may not be possible to get the matching result due to long sentence splitting. After replacing the "and" in the subsequent part with the subject "a" of the preceding part, the correct matching can be achieved. The key word can be set according to the actual situation, which can be a common word or symbol used to separate long sentences, which is not limited here. The positions of the pre-revision content and the remaining part of the patent-related text can be set according to the actual situation. For example, in a specific embodiment, the pre-revision content is located in the claims of the patent-related text. The remaining part or selected part of the patent-related text is at least partially located in the description of the patent-related text. For another specific embodiment, the pre-revision content is located in the description of the patent-related text, and the remaining part or selected part of the patent-related text is at least partially located in the claims of the patent-related text. In addition, the trigger condition of matching the rest or selected part of the patent-related text with the structural name or the pre-revision content can include that the pre-revision content matches the structural name, that is, only when the pre-revision content matches the structural name, the rest or selected part of the patent-related text is matched with the structural name or the pre-revision content. S14: revising all the target content or part of the target content matched based on the post-revision content. Optionally, all the target content or part of the target content will be revised to be consistent with the post-revision content. All the matched target content can be revised uniformly, or only part of the matched target content can be revised. When the pre-revision content is determined by the revision format, the target content can also be revised by the revision format when being revised, so as to facilitate the user to observe the position of the target content to which the revision is applied. In an embodiment, a positioning mark is formed at the revised target content. The user can quickly locate to the revised target content through the positioning mark. The positioning mark can be a comment in the word editing software, etc. In an embodiment, before or at the same time of matching the rest or selected part of the patent-related text with the pre-revision content, the following steps are included: De-noising processing is performed on the pre-revision content, the rest, and the selected part for matching. The de-noising processing includes at least one or all of the following: removing punctuation, removing reference signs, removing definite articles, and removing demonstrative pronouns. The pre-revision content can interfere with the matching because of the different text forms from the target content. For example, when the pre-revision content is in the claim and its content is “the a includes b”, the “the” in the pre-revision content needs to be removed, and when the rest of the content in the patent-related text contains reference signs, for example, its content is “a11 includes b12”, the reference signs in the rest of the content in the patent-related text also need to be removed, and then the pre-revision content is matched with the rest of the content in the patent-related text. Through the above setting, when the rest or selected part of the patent-related text is matched with the pre-revision content, it is not easy to be disturbed by noise, and the matching is accurate. Optionally, in an embodiment, the post-revision content is determined based on the pre-revision content. The post-revision content is determined based on the pre-revision content. The revised content is the part of the patent-related text that has changed based on the pre-revision content. When the pre-revision content is changed by the user, the revised content can be the new content after the change. For example, when the user revises "abcd" in the patent-related text to "abzz", it can be determined that the revised content is "zz". When the pre-revision content is deleted by the user, the revised content can be empty. For example, the user revises "abcd" in the patent-related text to "ab", and it is determined that the revised content is empty. When the pre-revision content is added by the user, the revised content can be the newly added content. For example, the user revises "ab" in the patent-related text to "abcd", and it can be determined that the revised content is "cd". The first content after the revised content is denoised replaces the second content in the target content, without changing the remaining content in the target content except the second content. In this embodiment, the second content is the content in the target content that needs to be modified accordingly. When revising the target content, only the part of the target content that has changed corresponding to the pre-revision content will be changed, and the entire target content will not be changed directly. For example, the patent-related text is changed from "the a includes b" to "the c includes b", and the target content is "a11 includes b12". At this time, the revised content after denoising is "c", and the second content is "a11", and the revised content is replaced by the second content to obtain "c includes b12". In this way, the remaining content in the target content is less likely to be changed unexpectedly and cause errors, and the revision is accurate. In an embodiment, after the first content after the revised content is denoised replaces the second content in the target content, without changing the remaining content in the target content except the second content, it further includes: An additional figure number is added to the revised content. For example, after obtaining "c includes b12" in the above embodiment, an additional figure number "13" corresponding to "c" is added to obtain "c13 includes b12". In this way, the revised content can be automatically completed with the figure mark, and it is less likely to be omitted and cause formal errors. To more clearly illustrate the workflow of the method for automatically modifying patent-related text according to the present application, the following is an example of a change revision. Please refer to FIG. 28, which is an interface diagram of the first interface of an embodiment of the method for automatically modifying patent-related text according to the present application. In an embodiment, the method for modifying the mark of patent-related text can have a first interface 100. A user can edit the patent-related text in the first interface 100. The patent-related text includes at least a first part 10 and a second part 20. As shown in FIG. 28, the user can edit the content of the original text "ABCD" in the first part 10, and then click the operation item 30 "Run" button. At this time, "CD" in the first part 10 has a revision format, "CD" is determined as the third content 11, and "ZZ" is the revised content 13. Next, based on the third content 11, the context is extended forward and backward. The extension process is based on the pre-revision text, and the revised content 13 "ZZ" is skipped until the first comma position is reached as the cutoff, and the fourth content 12 "ABCD" is obtained as the pre-revision content. Then, based on the fourth content 12, the target content 22 "ABCD" in the second part 20 is matched and located to the second content 21 "CD" in the second part 20, and the revised content 13 is replaced with the second content 21, i.e., the content synchronization revision is completed. After the content synchronization revision, the revised content 13 replaced in the second part 10 can also be marked. The above matching and replacement process can perform a de-noising process, which includes at least one of removing punctuation marks, removing marks, removing definite articles, and removing demonstrative pronouns. As can be seen from the above embodiments, the present application can not require the user to input the keyword for searching in the "Find" or "Replace" interface each time, i.e., the computer can determine the pre-revision content (i.e., the keyword for searching in the above) in the patent-related text and perform subsequent matching and replacement, which greatly improves the revision efficiency of the patent-related text and reduces the probability of human error. Moreover, since the embodiments of the present application can perform de-noising processing on the pre-revision content, the remaining part, and the selected part for matching, the "Find" or "Replace" has a generalization function, which can more comprehensively and accurately find the matching target, greatly reducing the trouble of finding one by one by human and the missing revision problem due to the inability to find. Secondly, since the embodiments of the present application replace the second content in the target content with the first content of the revised content, the remaining content in the target content is not changed, so that only the part that needs to be modified in the target content will be affected by the change, and the remaining content will not be processed, which is not easy to cause unexpected changes and form errors, and the revision is accurate. Further, the third content changed can be extended to the fourth content by the context in advance when matching in the embodiment of the application. Compared with the third content which can be only scattered words, the fourth content can contain more unique features, so that the matching is more accurate and is not easy to form noise interference when matching. When determining the fourth content, the text can be intercepted by the first delimiter as the cut-off point, or the text can be directly set to a preset length, and the way of constructing the fourth content can be flexibly selected. The fourth content formed by intercepting the text can have good uniqueness. In addition, the pre-revision content in the embodiment of the application can be determined by the revision format, or by the user selecting the range in the patent-related text, or by comparing the current version of the patent-related text and the historical version of the patent-related text. The way of determining the pre-revision content is various, and can be freely selected to adapt to different occasions. When the pre-revision content is determined by the revision format, the positioning is accurate, and the revision format is easy to observe, so that the user can intuitively master the content to be revised. When the pre-revision content is determined by the selected range, the user can easily decide which part of the text to process according to his own needs, and the user has stronger control over the revision. When the pre-revision content is determined by comparing the current version of the patent-related text and the historical version of the patent-related text, the user only needs to store a historical version of the patent-related text in advance, or even automatically store a historical version of the patent-related text by the software of the embodiment of the application. It can not be necessary to edit under the revision format, and it is not necessary to repeatedly select the selected range, which saves time and effort, and the operation is simple, especially suitable for occasions where the revision format is not opened or forgotten to be opened during the revision of the patent-related text. When matching the target, the embodiment of the present application further identifies whether the pre-revision content contains the second separator when the matching fails. When the second separator is not found, it means that the probability of the failure to match the corresponding result due to the long sentence splitting is low, and the failure to match the corresponding result due to the syntax change should be considered first. At this time, the embodiment of the present application replaces the conjunction in the pre-revision content with the main sentence of the previous sentence and then performs the matching, and the above case is included in the matching range. When the second separator is found, it means that the probability of the failure to match the corresponding result due to the long sentence splitting is high, and the embodiment of the present application replaces the key word in each subsequent part with the subject of the previous part, and the above case is included in the matching range. It can be seen that the embodiment of the present application can also adapt to the situations of character change, long sentence splitting, syntax change and the like during matching, and is not easy to miss the matching result. The above matching and revision process in the embodiment of the present application can be triggered by the user actively by at least one of responding to the user clicking the operation item in the user interface, pressing the preset key or inputting the preset instruction, or triggered passively according to the user operation in response to the preset change of the cursor position of the user, and the setting mode is flexible, and the revision operation of the user can be well reacted, and the use is convenient. Optionally, the all target content or part of the target content matched is revised based on the revised content, and the revised content is the content obtained based on the pre-revision content and includes: S21: generating a history record, the history record including at least part of the information related to the revision of the all target content or part of the target content matched based on the revised content. The history record contains information related to the corresponding revision operation. For example, the history record can include the time of the revision operation, the content summary of the revision, the effective range of the revision, and the like. S22: in response to the user clicking the history record, restoring the target content to the state before the revision corresponding to the history record. When the user clicks the history record, the content changed by the corresponding revision operation is restored to the content before the revision operation. The way of restoring the target content to the state before the revision corresponding to the history record can be recording the old version of the text content, and when the user clicks, the old version of the text content covers the current text content. It can also be to store the old version of the file, and when the user clicks, the old version of the file covers the current file. Referring to FIG. 29, FIG. 29 is an interface diagram of a second interface of an embodiment of the automatic modification method of the patent-related text. In an embodiment, the automatic modification method of the patent-related text comprises a second interface 200, and a history record is displayed on the second interface 200. The history record can record a revision number, a revision position, revision content, and operation time, etc. The revision position can be the paragraph number of the position in the patent-related text where the revision occurs. The revision content can record an abstract of the content of the revision operation. The operation time can be the time when the revision operation is triggered. The second interface 200 can be clicked, as shown in FIG. 29. When the user clicks the history record with the number "2" in the second interface 200, the patent-related text will be rolled back to the state before the operation of "deleting aa". Referring to FIG. 30, FIG. 30 is a structural diagram of an embodiment of the computer device 50. In an embodiment, the computer device 50 comprises a memory 51 and a processor 52 coupled with each other. The processor 52 is configured to execute program instructions stored in the memory 51 to implement the steps of any of the above-mentioned embodiments of the automatic modification method of the patent-related text. In a specific implementation scenario, the computer device 50 can include but is not limited to a microcomputer, a server, and in addition, the computer device 50 can also include a notebook computer, a tablet computer, and other mobile devices, which are not limited here. Specifically, the processor 52 is configured to control itself and the memory 51 to implement the steps of any of the above-mentioned embodiments of the automatic modification method of the patent-related text. The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 can be an integrated circuit chip having a processing capability of signals. The processor 52 can also be a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. In addition, the processor 52 can be jointly implemented by integrated circuit chips. Please refer to Fig. 31, which is a structural schematic diagram of an embodiment of the computer readable storage medium provided in the application. The computer readable storage medium 60 of the embodiment of the application stores program instructions 61, which, when executed, implement the method provided by any embodiment of the automatic numbering method for patent-related texts and any non-conflicting combination. The program instructions 61 can form a program file and be stored in the computer readable storage medium 60 in the form of a software product, so that a computer device (which can be a personal computer, a server, or a network device, etc.) executes all or part of the steps of the method of each embodiment of the application. The aforementioned computer readable storage medium 60 includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes, or a terminal device such as a computer, a server, a mobile phone, or a tablet. Please refer to Fig. 32, which is a structural schematic diagram of a client and a server in an embodiment of the automatic modification method for patent-related texts. The application provides an automatic modification method for patent-related texts, which includes that the client 80 receives a revision trigger end instruction for the patent-related texts. The client 80 sends the revision trigger end instruction to the server 70, so that the server 70 determines the pre-revision content in the patent-related texts, and defines the structural name in the scope of claims or the specification, and then matches the remaining part or the selected part of the patent-related texts with the pre-revision content. All target content or part of the target content obtained by matching is revised based on the post-revision content or the structural name, and the post-revision content is the content obtained based on the pre-revision content. The client 80 receives the revision result and displays it in the interface. Specifically, the automatic modification method for patent-related texts can also be executed by the server 70. In this embodiment, the client 80 is mainly used for interaction with the user. The patent-related texts can be located on the server 70 and displayed by the client 80. That is, the client 80 can not store the patent-related texts and only serve as a display function, and the processing of the patent-related texts is performed by the server 70. For example, the patent-related texts can be an online cloud operation client 80 interacting with the server 70. After receiving the revision trigger end instruction from the client 80, the server 70 executes the steps of the automatic modification method for patent-related texts in the embodiment of the application, completes the revision of the patent-related texts, and sends the result to the client 80 for display. All embodiments of the application are applicable to patent-related texts, articles, reports, papers, novels, or scientific and technological literature, etc. In an embodiment, the automatic labeling or modifying method of patent-related texts is executed by program instructions, the program instructions are installed in the form of plug-ins into existing word processing software, and the program instructions act on patent-related texts opened by the word processing software to realize the automatic labeling method of patent-related texts. In an embodiment, the automatic labeling or modifying method of patent-related texts is executed by program instructions, the program instructions are installed in the form of plug-ins into existing word processing software, and the program instructions act on patent-related texts opened by the word processing software to realize the automatic labeling method of patent-related texts. In an embodiment, the automatic labeling or modifying method of patent-related texts is executed by program instructions, the program instructions are installed in the form of plug-ins into existing word processing software, and the program instructions act on patent-related texts opened by the word processing software to realize the automatic labeling method of patent-related texts. In an embodiment, the automatic labeling or modifying method of patent-related texts is executed by program instructions, the program instructions are installed in the form of plug-ins into existing word processing software, and the program instructions act on patent-related texts opened by the word processing software to realize the automatic labeling method of patent-related texts. In an embodiment, the automatic labeling or modifying method of patent-related texts is executed by program instructions, the program instructions are installed in the form of plug-ins into existing word processing software, and the program instructions act on patent-related texts opened by the word processing software to realize the automatic labeling method of patent-related texts. At present, part of the work of patent workers is to write patent-related texts, form patent-related texts, and then submit as patent applications through official channels. Patent applications generally include invention applications, utility model applications, or design applications. Among them, the patent-related texts involved in invention applications and utility model applications generally include abstracts, claims, descriptions, and description drawings. Since product-related patent-related texts generally include description drawings, the drawings in the description drawings generally need to be labeled. Usually, labeling the structure in the drawing is a time-consuming and careful process, which requires manually drawing lines, constantly adjusting the position and shape of the lines, manually inserting a text box at one end of the line after the line is appropriate, manually inputting the label in the text box, and finally adjusting the size of the drawing and unifying the size of the labels between different drawings. This process is very time-consuming, especially in the case of complex product structure and more labels, which requires several hours of labeling and adjustment, and the process is tedious and prone to errors. On the other hand, corresponding to the corresponding structure name in the specification, the label is also a time-consuming and careful process, especially in the case of complex product structure, the structure name is various, and the label also needs to follow the relationship between the structure names to distinguish. The existing label is generally found by manual searching of the label one by one, which is tedious and prone to errors. The label method, device...

Claims

1. A revision assisting method of patent-related texts, wherein, Comprising: receiving a first instruction for a patent-related text, the first instruction comprising a revision end instruction; determining pre-revision content in the patent-related text in response to the first instruction; matching the pre-revision content to a remaining portion or a selected portion of the patent-related text; revising all or part of the target content matched based on post-revision content, the post-revision content being content obtained based on the pre-revision content.

2. The method of claim 1, wherein, The matching the pre-revision content to the remaining portion or the selected portion of the patent-related text before or simultaneously with the matching comprises: de-noising the pre-revision content, the remaining portion, the selected portion for matching, the de-noising comprising at least one of removing a punctuation mark, removing a reference numeral, removing a definite article, removing an indicative pronoun.

3. The method of claim 2, wherein, The revising all or part of the target content matched based on the post-revision content comprises: replacing a first content of the post-revision content that is de-noised with a second content of the target content, without modifying a remaining content of the target content other than the second content.

4. The method of claim 1, wherein, The determining the pre-revision content in the patent-related text comprises: determining a third content with a revision format in the patent-related text; determining a fourth content of a minimum matching unit based on the third content, the fourth content having a length greater than or equal to a length of the third content, the fourth content being the pre-revision content.

5. The method of claim 4, wherein, The determining the fourth content of the minimum matching unit based on the third content comprises: finding a first delimiter appearing in a preceding text and a following text from the third content as a starting point, taking the first delimiter appearing in the preceding text and the following text as an end point, and taking content between the end point of the preceding text and the end point of the following text as the fourth content.

6. The method of claim 4, wherein, The determining the fourth content of the minimum matching unit comprises: taking the third content as a center, and taking a preset length of a preceding text and a following text as a boundary, and determining the fourth content of the minimum matching unit as the preset length of the preceding text and the following text and the third content.

7. The method of claim 1, wherein, The determining the pre-revision content in the patent-related text comprises: taking content in a selected range in the patent-related text as the pre-revision content, or taking a third content determined after comparing a current version of the patent-related text and a historical version of the patent-related text as the pre-revision content.

8. The method of claim 1, wherein, The matching the pre-revision content to the remaining portion or the selected portion of the patent-related text comprises: in response to not matching the target content based on the pre-revision content, finding a second delimiter in the pre-revision content; in response to the second delimiter not existing, replacing a conjunction in the pre-revision content with a subject of a preceding sentence, or inserting the subject of the preceding sentence at a beginning of the pre-revision content; in response to the second delimiter existing, splitting the pre-revision content into a preceding part and a plurality of following parts based on the second delimiter, and replacing a key word in each of the following parts with a subject of the preceding part, or inserting the subject of the preceding part at a beginning of each of the following parts.

9. The method of claim 8, wherein, The pre-revision content is located in the claim of the patent-related text, and the remaining part or selected part of the patent-related text is located at least partially in the specification of the patent-related text; or The pre-revision content is located in the specification of the patent-related text, and the remaining part or selected part of the patent-related text is located at least partially in the claim of the patent-related text.

10. The method of claim 1, wherein, The receiving the first instruction for the patent-related text comprises: The revision end instruction is triggered in response to at least one of clicking an operation item in the interface, pressing a preset key, or inputting a preset instruction.

11. The method of claim 1, wherein, The receiving the first instruction for the patent-related text comprises: The revision end instruction is triggered in response to a preset change in the cursor position.

12. The method of any one of claims 1-11, wherein, The revision of all target contents or part of the target contents obtained by matching based on the post-revision content comprises: Generating a history record, the history record comprising at least part of the information related to the revision of all target contents or part of the target contents obtained by matching; In response to clicking the history record, restoring the target content to the pre-revision state corresponding to the history record.

13. A revision assisting method of patent-related texts, wherein, Comprise: The client receives a first instruction for a patent-related text, the first instruction comprising a revision end instruction; The client sends the first instruction to the server to enable the server to determine pre-revision content in the patent-related text, and to match the remaining part or selected part of the patent-related text with the pre-revision content, revise all target contents or part of the target contents obtained by matching based on post-revision content, the post-revision content being content obtained based on the pre-revision content; The client receives the result of the revision and displays the patent-related text that has been revised in the interface.

14. A revision assisting method of a patent-related document, wherein, Comprise: The client receives a first instruction for a patent-related text, the first instruction comprising a revision end instruction; The client sends the first instruction and the patent-related text to the server to enable the server to determine pre-revision content in the patent-related text, and to match the remaining part or selected part of the patent-related text with the pre-revision content, determine revision position and revision manner based on post-revision content for all target contents obtained by matching, the post-revision content being content obtained based on the pre-revision content; The client receives the revision position and the revision manner, and revises the patent-related text based on the revision position and the revision manner.

15. A revision assisting method of a text, wherein, Comprise: Receiving a first instruction, the first instruction comprising a revision end instruction of a text; In response to the first instruction, determining pre-revision content in the text; Matching the remaining part or selected part of the text with the pre-revision content; Revising all target contents or part of the target contents obtained by matching based on post-revision content, the post-revision content being content obtained based on the pre-revision content.

16. A computer device, wherein, The computer device comprises: A processor; a memory connected to the processor for storing program instructions executable by the processor to implement the revision assistance method of any of claims 1-15.

17. A computer readable storage medium, wherein, The computer readable storage medium stores program instructions executable to implement the revision assistance method of any of claims 1-15.

18. A method of assisting in the drafting of a patent-related document, wherein, comprising: reading a data structure defining a number of elements of a protected subject matter and relationships thereof, and defining a set of the elements and a set of the relationships constituting each claim; forming a corresponding claim according to a first preset rule from the set of the elements and the set of the relationships corresponding to each of the claims.

19. The method of claim 18, wherein, the reading the data structure comprises: determining a first claim under a determined protected subject matter according to a second preset rule, and reading the set of the elements and the set of the relationships associated with the first claim.

20. The method of claim 19, wherein, the forming a corresponding claim according to a first preset rule from the set of the elements and the set of the relationships corresponding to each of the claims comprises: determining an order of appearance of the elements in the set of the elements associated with the first claim in the first claim using the set of the relationships associated with the first claim; forming an organization structure of the elements and an organization structure of the relationships of the first claim according to the order of appearance of the elements in the first claim using a template corresponding to the first claim.

21. The method of claim 20, wherein, the set of the elements comprises a set of product elements, and the set of the relationships comprises a set of element relationships, and the determining an order of appearance of the elements in the set of the elements associated with the first claim in the first claim using the set of the relationships associated with the first claim comprises: forming all of the elements in the first claim into a subject-predicate order link using the set of the relationships associated with the first claim, and performing the ordering of the elements in a reverse order of the elements in the subject-predicate order link.

22. The method of claim 20, wherein, the reading the data structure comprises: determining a second claim under the determined protected subject matter, and reading the set of the elements and the set of the relationships associated with the second claim; the forming a corresponding claim according to a first preset rule from the set of the elements and the set of the relationships corresponding to each of the claims comprises: The element ordering method according to the first claim, determining the order of appearance of the elements in the second claim, applying a template corresponding to the second claim to form the organization structure of the elements and the organization structure of the relations of the second claim, traversing the determined protected subject matter to form all the claims under the determined protected subject matter.

23. The method of claim 22, wherein, The reading data structure comprises: traversing the data structure, sequentially determining the remaining protected subject matter in the data structure according to a second preset rule, and forming all the claims under each protected subject matter; Wherein, the second preset rule comprises ordering the protected subject matters from small to large or from large to small according to the containing relationship between the protected subject matters.

24. The method of claim 21, wherein, Automatically check whether a closed loop is realized in each of the claims, the condition of the closed loop comprising whether all the elements are associated with at least another element within one of the claims or between different claims, whether there is only one subject-predicate ordering link, otherwise giving a prompt that the closed loop cannot be realized.

25. The method of claim 20, wherein, The set of elements comprises a set of method steps, the set of relations comprises a set of time precedence relations between the steps, and determining the order of appearance of the elements in the set of elements associated with the first claim in the first claim comprises: Determining the time precedence relations of all the steps in the first claim using the set of relations associated with the first claim.

26. The method of claim 19, wherein, Determining the first claim under the determined protected subject matter comprises: Searching for the claim with the highest serial number under the determined protected subject matter as the first claim; or For each of the elements, searching for the set of elements and the set of relations of all the claims under the determined protected subject matter to obtain all the elements following the vocabulary in the form of "comprising", "containing", "having", "forming", "consisting of", "consisting essentially of", or "provided with", defining all the obtained elements as the set of first appearance elements, defining the remaining elements as the set of later appearance elements, and checking the set of elements and the set of relations of all the claims to obtain the set of elements and the set of relations of the claims in which only the set of first appearance elements appear as the first claim, wherein "**" refers to a character or symbol, formula.

27. The method of claim 26, wherein, Determining the first claim under the determined protected subject matter comprises: continuing the same method to process the set of elements, the set of relations of the remaining set of claims, determining the set of elements, the set of relations of the claims which only appear the first time in the remaining set of claims as a second claim, and determining the second claim to refer to the first claim; or in the remaining set of claims, if there is no set of elements, the set of relations of the claims which appear the first time, taking the set of elements, the set of relations of the claims with the same elements as the referred claim as a second claim or a claim which is sequentially arranged behind, referring to the previous referred claim.

28. The method of claim 25, wherein, the reading data structure comprises: determining a second claim under the determined protected subject, and reading the set of steps, the set of relations associated with the second claim; the forming the set of elements, the set of relations corresponding to each claim according to the first predetermined rule comprises: determining the order of the steps appearing in the second claim according to the step order method of the first claim, applying the template corresponding to the second claim to form the organization structure of the steps of the second claim, traversing the determined protected subject to form all claims under all determined protected subjects.

29. The method of claim 28, wherein, automatically checking whether a closed loop is realized in each claim, the condition of the closed loop comprising whether all steps in one claim, between different claims are the same as at least part of another step, otherwise giving a prompt that the closed loop cannot be realized.

30. The method of claim 18, wherein, the reading data structure comprises: displaying a first area and a second area, the first area being used for editing the data structure, and the second area being used for displaying the content of the formed claims or technical disclosure.

31. The method of claim 30, wherein, the data structure in the first area is displayed in the form of a tree diagram or a flowchart, and the reading data structure comprises: receiving a first instruction of text editing, node adding, or changing node position of a selected node in the tree diagram or the flowchart; in response to the first instruction, performing text editing, node adding, or changing node position of the selected node in the tree diagram or the flowchart.

32. The method of claim 31, wherein, after the text editing, node adding, or changing node position of the selected node in the tree diagram or the flowchart, the reading data structure comprises: receiving a second instruction of defining relations of a selected node in the tree diagram; In response to the second instruction, lines connecting the selected nodes are drawn in the first area, one end of each line is connected to a corresponding selected node, and the other end of each line is connected to a text input box, and a cursor is located in the text input box, if the number of the selected nodes is greater than or equal to two, a plurality of lines connecting all the selected nodes are drawn in the first area, one end of each line is connected to a corresponding selected node, and the other end of each line is connected to a text input box, and a cursor is located in the text input box.

33. The method of claim 32, wherein, the number of the text input boxes is one or two, if the number is one, a relationship input area and a technical problem input area are divided in the text input box, if the number is two, one text input box is used to input the relationship, and the other text input box is used to input the technical problem.

34. The method of claim 18, wherein, after the corresponding claim is formed, the method comprises: generating a labeling strategy matching the relationship of the elements by using the relationship of the elements; labeling part or all of the elements of the patent-related text according to the labeling strategy.

35. The method of claim 34, wherein, the generating a labeling strategy matching the relationship of the elements by using the relationship of the elements comprises: if the relationship between different elements is an upper-lower layer containing relationship, the generated label of the corresponding upper layer element is smaller than the label of the corresponding lower layer element, if the relationship between different elements is a same layer parallel relationship, the generated labels of the corresponding different elements belong to the same order of magnitude.

36. A method of assisting in the drafting of a patent-related document, wherein, comprises: the client sends a data structure to the server, the data structure defines a plurality of elements and their relationships of a protected subject, and defines a set of elements and a set of relationships constituting each claim; the client receives a claim from the server, the claim is formed by the server according to a first preset rule according to the set of elements and the set of relationships corresponding to each claim.

37. A method of assisting in the drafting of a patent-related document, wherein, comprises: the client receives a first instruction for editing a data structure; the client sends the first instruction to the server to make the server edit the data structure, and the edited data structure defines a plurality of elements and their relationships of a protected subject, and defines a set of elements and a set of relationships constituting each claim; the client receives a claim or a processing completion notification from the server, the claim is formed by the server according to a first preset rule according to the set of elements and the set of relationships corresponding to each claim, and the processing completion notification includes the processing result of forming the claim.

38. A computer device, wherein, the computer device comprises: a processor; a memory connected to the processor, used to store program instructions, the program instructions can be executed by the processor to implement the patent-related text writing assistance method according to any one of claims 18-37.

39. A computer readable storage medium, wherein, The computer readable storage medium is configured to store program instructions executable to implement the patent-related text writing assistance method of any one of claims 18-37.

40. A method of assisting in the drafting of a patent-related document, wherein, The method comprises: reading a file name of the patent-related text, the file name comprising a file identifier associated with writing reference information of the patent-related text; obtaining the writing reference information associated with the file identifier; wherein the obtained writing reference information is displayed simultaneously with the content of the patent-related text.

41. The method of claim 40, wherein the obtaining the writing reference information associated with the file identifier comprises: obtaining customer information associated with the file identifier; obtaining the writing reference information associated with the customer information, the writing reference information associated with different customer information being different.

42. The method of claim 40, wherein the writing reference information comprises writing requirements, and the obtaining the writing reference information associated with the file identifier comprises: automatically modifying the content of the patent-related text according to the writing requirements; or annotating the content of the patent-related text to indicate parts not meeting the writing requirements or prompting in a pop-up window.

43. The method of claim 42, wherein the writing requirements comprise a prohibition of using prohibited words, and the obtaining the writing reference information associated with the file identifier comprises: searching the content of the patent-related text using the prohibited words as keywords; annotating the search results in the patent-related text or replacing the search results in the patent-related text with corresponding replacement words, each prohibited word being preconfigured with at least one corresponding replacement word.

44. The method of claim 43, wherein the replacing the search results with the corresponding replacement words comprises: positioning the current display position of the patent-related text to the search results and displaying at least two selectable replacement words near the search results; receiving a first instruction to select one of the replacement words; and replacing the search results with the selected replacement word in response to the first instruction.

45. The method of claim 40, wherein the obtaining the writing reference information associated with the file identifier comprises: receiving current content input in a display interface of the patent-related text; searching a document set corresponding to a current user group using at least a part of the current content as a keyword, the document set corresponding to different user groups being different; or searching a current text; displaying matching paragraphs or matching sentences in the searched document set or the current text that match the keyword.

46. The method of claim 45, wherein the number of matching paragraphs or matching sentences is a plurality, and the displaying the matching paragraphs or matching sentences comprises: receiving a second instruction to select one of the matching paragraphs or matching sentences; and replacing the selected matching paragraph or matching sentence with a corresponding replacement word. combining the current content with the selected matching paragraph or matching sentence in response to the second instruction.

47. The method of claim 45, wherein, displaying the matching paragraph or matching sentence comprises: receiving a third instruction to select a portion of the content of the matching paragraph or matching sentence; combining the selected portion of the content of the matching paragraph or matching sentence in response to the third instruction.

48. The method of claim 45, wherein, the searching the set of documents corresponding to the current user group with the at least one portion of the current content as a keyword comprises: the searching the set of documents corresponding to the current user group with the keyword obtained after a de-noising process, the de-noising process comprising removing "the", "this", "of", "above" or "below".

49. The method of claim 45, wherein, the matching paragraph or matching sentence is a plurality of matching paragraphs or matching sentences, and the displaying the matching paragraph or matching sentence comprises: ordering the plurality of matching paragraphs or matching sentences according to their distance from the current cursor position, and placing the closest matching paragraph or matching sentence to the cursor at the top of the order, and displaying the closest matching paragraph or matching sentence below the cursor in a candidate window.

50. The method of claim 49, wherein, the displaying the matching paragraph or matching sentence comprises: distinguishing the closest matching paragraph or matching sentence from the rest of the matching paragraphs or matching sentences, and displaying the current cursor position and the text containing the closest matching paragraph or matching sentence in separate screens or windows.

51. The method of claim 50, wherein, the closest matching paragraph or matching sentence is set as a default selected item, and the displaying the current cursor position and the text containing the closest matching paragraph or matching sentence comprises: receiving a fourth instruction to re-select the closest matching paragraph or matching sentence, and displaying the re-selected closest matching paragraph or matching sentence.

52. The method of claim 49, wherein, the receiving the current content inputted in the display interface of the patent-related text comprises: automatically scrolling the current text backward in response to whether the distance between the current cursor and the bottom of the current text is less than a threshold value, so that the distance between the current cursor and the bottom of the current text is not less than the threshold value.

53. The method of claim 45, wherein, the displaying the matching paragraph or matching sentence in the set of documents and the current text comprises: filling one of the searched document set, the matched paragraph or the matched phrase in the current text matching the keyword into the text input space behind the current cursor, and if there are more than one of the matched paragraph or the matched phrase in the searched document set, the matched paragraph or the matched phrase in the current text matching the keyword, putting the rest of the matched paragraph or the matched phrase into a candidate window, and if a fifth instruction of selecting the next matched paragraph or matched phrase is received, highlighting the next matched paragraph or matched phrase in the candidate window.

54. The method of claim 40, wherein, the displaying the obtained writing reference information comprises: scrolling and highlighting the writing reference information in the display interface of the patent-related text; or the obtaining the writing reference information associated with the file identifier further comprises: receiving a sixth instruction of closing the patent-related text; in response to the sixth instruction, displaying the writing reference information in a pop-up window before closing the patent-related text; the reading the file name of the patent-related text further comprises: obtaining a module position of the current cursor in the patent-related text, the module position comprising at least one of an abstract module position, a claim module position, an invention name position, a technical field module position, a background technology module position, an invention content module position, a drawing description module position, and a specific implementation module position; the obtaining the writing reference information associated with the file identifier comprises: obtaining the writing reference information associated with the file identifier and the module position of the current cursor in the patent-related text.

55. The method of claim 45, wherein, the writing reference information comprises a data structure, the data structure defining a plurality of elements of a protected subject matter and relationships thereof, and defining a set of the elements and a set of the relationships constituting each claim; the obtaining the writing reference information associated with the file identifier further comprises: forming the set of the elements and the set of the relationships corresponding to each of the claims into a corresponding claim according to a first preset rule.

56. The method of claim 55, wherein, the reading the data structure comprises: determining a protected subject matter in the data structure according to a second preset rule, determining a first claim under the determined protected subject matter, and reading the set of the elements and the set of the relationships associated with the first claim.

57. The method of claim 56, wherein, the forming the set of the elements and the set of the relationships corresponding to each of the claims into a corresponding claim according to a first preset rule comprises: determining an order of appearance of the elements in the set of the elements associated with the first claim in the first claim using the set of the relationships associated with the first claim; forming an organization structure of the elements and an organization structure of the relationships of the first claim according to the order of appearance of the elements in the first claim using a template corresponding to the first claim.

58. The method of claim 57, wherein, the set of elements comprises a set of product elements, the set of relations comprises a set of element relations, and determining the order of appearance of the elements in the set of elements associated with the first claim in the first claim using the set of relations associated with the first claim comprises: constructing all of the elements in the first claim into a subject-predicate order link using the set of relations associated with the first claim, and ordering the elements in the reverse order of the elements in the subject-predicate order link.

59. A method of assisting in the composition of a text, wherein, comprising: a server receiving a file name sent by a client, the file name comprising a file identifier, the file identifier being associated with writing reference information of a text; the server obtaining the writing reference information associated with the file identifier; the server sending the writing reference information to the client, so that the client displays the obtained writing reference information while displaying the content of the text.

60. A method of assisting in the composition of a text, wherein, comprising: inputting content in a text using an input method in a text editing application; obtaining the content using a plug-in of the text editing application, and searching for a matching paragraph or a matching sentence in or out of the text using at least part of the characters of the content as a keyword; displaying the matching paragraph or the matching sentence as writing reference information on the same screen as the content.

61. The method of claim 60, wherein, the displaying the matching paragraph or the matching sentence as writing reference information on the same screen as the content comprises: receiving a second instruction of selecting one of the matching paragraphs or the matching sentences; combining the current content with the selected matching paragraph or the matching sentence in response to the second instruction.

62. The method of claim 60, wherein, the searching for a matching paragraph or a matching sentence in or out of the text comprises: arousing a suggestion window to display the matching paragraph or the matching sentence; the inputting content in a text using an input method in a text editing application comprises: refreshing the cache information of the input method so that the input method rebinds the position of the current cursor in the text to be inserted based on the current cursor position; or 63. A composition for use in a method of claim 62, wherein arousing the suggestion window, pressing a key at the moment, and then triggering the candidate window form to close, sending the previously recorded key to the operating system again, triggering a key input event again, and activating the input; or closing the input method after the input method inputs the content for the first time and before the input method obtains the key input message again, so that the input method is not bound to the focus on the suggestion window. comprising: receiving first content input in a text or receiving first content input in an input method interface; searching for a matching paragraph or a matching phrase in the text or a designated document using the first content as a keyword, the matching paragraph or the matching phrase including the first content and the second content, or the matching paragraph or the matching phrase including the second content and the related content of the first content, the matching paragraph or the matching phrase being delimited in the text or the designated document by a delimiting character, the delimiting character being a general character adjacent to the matching paragraph or the matching phrase; displaying the matching paragraph or the matching phrase as writing reference information or recommendation information on the same screen as the text.

64. The method of claim 63, wherein, the delimiting character includes any one of the following general characters appearing in a sentence in the patent-related text: "comprising", "having", "including", "provided with", and any one of the following general characters appearing in a sentence in the patent-related text: a punctuation mark, a symbol.

65. The method of claim 63, wherein, the receiving the first content input in the text or the receiving the first content input in the input method interface includes: receiving the first content input in the text using an input method or receiving the first content input in the input method interface using the input method; the searching for the matching paragraph or the matching phrase in the text or the designated document and displaying the matching paragraph or the matching phrase as the writing reference information or the recommendation information on the same screen as the text includes: searching for the matching paragraph or the matching phrase in the text or the designated document using the input method and displaying the matching phrase or the matching paragraph in the form of a candidate word in the interface of the input method to be selected and filled in the text; or searching for the matching paragraph or the matching phrase in the text or the designated document using a plug-in of a text editing application and displaying the matching phrase or the matching paragraph in the form of a candidate word in the interface of the plug-in to be selected and filled in the text.

66. A computer device, wherein, The computer device includes: a processor; a memory connected to the processor for storing program instructions executable by the processor to implement the writing assistance method of any one of claims 40-65.

67. A computer readable storage medium, wherein, The computer-readable storage medium is used to store program instructions executable to implement the writing assistance method of any one of claims 40-65.

68. A method of automatically labeling patent-related text, wherein, includes: determining a claim scope or a specification scope in the patent-related text; delimiting a structural name in the claim scope or the specification; hierarchically classifying at least two of the delimited structural names; labeling the at least two structural names according to the hierarchical classification.

69. The method of claim 68, wherein, the delimiting the structural name in the claim scope includes: The first content between the first word and the tail symbol, or the second content between adjacent first words, in the scope of the claim is searched for, as the structure name, or further determined from the first content or the second content as the structure name.

70. The method of claim 69, wherein, the first word comprises any one of: comprising, having, provided with, including, and the tail symbol comprises a punctuation mark.

71. The method of claim 69, wherein, the further determining the structure name from the first content or the second content comprises: finding the content after the last "of" and before the tail symbol in the first content or the second content as the structure name.

72. The method of claim 69, wherein, the defining the structure name in the scope of the claim comprises: determining at least two identical words in the scope of the claim, and at least one of the identical words is provided with the definite article "the" or "that", and the other is not provided with the definite article "the" or "that", and the identical words are taken as the structure name.

73. The method of claim 72, wherein, the determining at least two identical words in the scope of the claim comprises: determining a two-character word in the scope of the claim, and traversing all identical words as the two-character word in the scope of the claim; in response to finding more than one identical word, adding one character to the left or right of the two-character word as a three-character word, and continuing the step of traversing all identical words as the two-character word in the scope of the claim for the three-character word, and the subsequent steps for the three-character word, and continuing the loop; in response to failing to find more than one identical word, determining another two-character word in the scope of the claim, and continuing the step of traversing all identical words as the two-character word in the scope of the claim for the another two-character word, and the subsequent steps for the another two-character word, and continuing the loop until the scope of the claim is traversed.

74. The method of claim 68, wherein, the hierarchically classifying the at least two defined structure names comprises: classifying two structure names immediately before and after the first word as upper and lower levels; the numbering the at least two structure names according to the hierarchical classification comprises: numbering the two structure names of the upper and lower levels according to the upper and lower levels.

75. The method of claim 68, wherein, the determining the scope of the claim in the patent-related text comprises: searching for "claim" in the patent-related text; in response to finding the "CLAIM", taking the found result as a first boundary, and searching for "TECHNICAL FIELD", "ABSTRACT", "ABSTRACT DRAWINGS", "DRAWINGS", "SPECIFICATION", or a file ender after the "CLAIM"; in response to finding "TECHNICAL FIELD", "ABSTRACT", "ABSTRACT DRAWINGS", "DRAWINGS", or "SPECIFICATION", taking the found result as a second boundary, and taking the area between the first boundary and the second boundary as the claim scope.

76. The method of claim 68, wherein, the marking of the at least two structure names comprises: receiving a first instruction to insert a mark in the patent-related text; in response to the first instruction, inserting the mark in the patent-related text; taking the mark as a third boundary, and determining a structure name immediately preceding the third boundary; searching for the structure name in the patent-related text, and inserting the mark after the found content, or the marking of the at least two structure names comprises: receiving a second instruction to delete or replace a mark; in response to the second instruction, deleting or replacing the mark; taking the deletion or replacement of the mark as a fourth boundary, and determining a structure name immediately preceding the fourth boundary; searching for the structure name in the patent-related text, and deleting or replacing the mark of the found content, or the delimiting of the structure names in the claim scope comprises: receiving a third instruction to revise the patent-related text; in response to the third instruction, determining whether the revision is directed to a structure name, or immediately adjacent to a structure name; in response to the revision being directed to a structure name, or immediately adjacent to a structure name, replacing at least another identical word of the structure name in the patent-related text with the revised structure name.

77. A method of automatically labeling patent-related text, wherein, comprising: delimiting structure names in the claim scope or the specification; marking the at least two delimiting structure names in a random mark manner or in a pre-defined mark manner.

78. A method of automatically modifying patent-related text, wherein, comprising: receiving an artificial revision end instruction directed to the patent-related text; in response to the artificial revision end instruction, or in response to a pre-set condition being met, determining pre-revision content in the patent-related text, and delimiting structure names in the claim scope or the specification; in response to the pre-revision content matching the structure names, matching the remaining or selected part of the patent-related text with the structure names or the pre-revision content; based on the post-revision content, revising all or part of the matched target content.

79. The method of claim 78, wherein, the delimiting of the structure names in the claim scope comprises: The first content between the first word and the tail symbol, or the second content between the adjacent first words, in the scope of the claim, is searched, and the first content or the second content is used as the structure name, or the structure name is further determined from the first content or the second content.

80. A computer device, wherein, The computer device comprises: a processor; a memory connected to the processor, configured to store program instructions executable by the processor to implement the automatic numbering method or the automatic modification method of the patent-related text according to any one of claims 68-79.

81. A method of labeling the figures, wherein, comprising: displaying at least one of a name set and a number set, wherein the names in the name set correspond to the numbers in the number set; receiving a drag start or end instruction of dragging the name itself or a first anchor point thereof, or the number itself or a second anchor point thereof, to a first position in or out of the drawing; in response to the drag start or end instruction, generating a line extending from the first position to one side of the name or the number dragged or copied to a second position in or out of the drawing, or under the name or the number.

82. The method of claim 81, wherein, the displaying the drawing and at least one of the name set and the number set comprises: displaying the drawing, and displaying the name set and the number set in parallel or mixed in the same file.

83. The method of claim 81, wherein, the displaying the drawing and at least one of the name set and the number set comprises: displaying the drawing in a graphic editing software interface, and displaying at least one of the name set and the number set in a text editing software interface, and the displaying in the graphic editing software interface and the text editing software interface in parallel.

84. The method of claim 81, wherein, the number of names in the name set is at least two, and the number of numbers in the number set is at least two, and the dragging or copying the name or the number to the second position in or out of the drawing comprises: dragging or copying the name or the number to the second position in or out of the drawing, wherein the second positions of different names are different, and the names are arranged in a preset arrangement logic and unified in size.

85. The method of claim 81, wherein, the displaying the drawing and at least one of the name set and the number set comprises: determining a claim scope or a specification scope in the patent-related text; defining a structure name in the claim scope or the specification; classifying the at least two defined structure names in a hierarchy; labeling the at least two structure names according to the hierarchy, thereby obtaining the name set and the number set; or After the structure names in the scope of the claims or the specification are defined, at least two of the defined structure names are labeled in a random manner or in a pre-defined manner.

86. The method of claim 85, wherein, the defining of the structure names in the scope of the claims comprises: finding a first content between a first word leading to the structure name in the claims and a tail word following the structure name, or a second content adjacent to the first word, in the scope of the claims, and taking the first content, the second content as the structure name, or further determining the structure name from the first content, the second content.

87. The method of claim 86, wherein, the hierarchical division of the at least two defined structure names comprises: dividing two of the structure names before and after the first word into upper and lower levels; the labeling of the at least two structure names according to the hierarchical division comprises: labeling the two structure names in the upper and lower levels according to the upper and lower levels.

88. The method of claim 87, wherein, the labeling of the at least two structure names further comprises: receiving a first instruction to insert a label in the patent-related text; in response to the first instruction, inserting the label in the patent-related text; determining a structure name immediately preceding a third boundary of the label in the patent-related text; finding the structure name in the patent-related text and inserting the label after the found content, or the labeling of the at least two structure names further comprises: receiving a second instruction to delete or replace a label; in response to the second instruction, deleting or replacing the label; determining a structure name immediately preceding a fourth boundary of the deleted or replaced label in the patent-related text; finding the structure name in the patent-related text and deleting or replacing the label of the found content, or the defining of the structure names in the scope of the claims further comprises: receiving a third instruction to revise the patent-related text; in response to the third instruction, determining whether the revision is directed to a structure name or immediately adjacent to a structure name; in response to the revision being directed to a structure name or immediately adjacent to a structure name, replacing at least another structure name in the patent-related text with the revised structure name.

89. The method of claim 81, wherein, the generating of the line in response to the start or end instruction comprises: determining a component type of the name or the label, the component type comprising a normal type, a face and side type, a cavity and recess type; generating a corresponding preset line style based on the component type.

90. The method of claim 89, wherein, the determining of the component type of the name or the label comprises: displaying at least two type labels near the name or the label; in response to a mouse click, a keyboard shortcut, or a text input, selecting the type label; based on the selected type label, determining the component type of the name or the label.

91. The method of claim 90, wherein, The method further comprises: based on semantic features of the name, determining the component type of the name; based on the component type, displaying the type label corresponding to the component type.

92. The method of claim 81, wherein, The method further comprises: obtaining input data, the input data comprising at least one of the names and / or the labels; adjusting the format of each of the names and the labels in the input data according to a preset rule to form the set of names and the set of labels.

93. The method of claim 92, wherein, The method further comprises: displaying an input box, and inputting content in the input box as the input data; or reading a file, and inputting text content in the file as the input data; or obtaining communication information of an application, and inputting the communication information as the input data.

94. The method of claim 92 or 93, wherein, The method further comprises: in response to the input data comprising only the names, labeling each of the names with a random label or a predefined label to form the set of labels; or in response to the input data comprising only the labels, displaying only the set of labels; or in response to the input data comprising both the names and the labels, and at least one of the names not being associated with any of the labels, displaying a prompt message.

95. A method of labeling the figures, wherein, The method further comprises: displaying the drawing, and displaying at least one of the set of names and the set of labels, the names in the set of names corresponding to the labels in the set of labels; receiving a first instruction of selecting or clicking the name or the label, and receiving a second instruction of clicking a first position in or out of the drawing; in response to the first instruction and the second instruction, generating a line, and moving or copying the name or the label to a second position in or out of the drawing, one end of the line extending to the first position, and the other end of the line extending to a side of the moved or copied name or label; or the line being located below the name or the label.

96. A method of labeling the figures, wherein, The method further comprises: displaying the drawing; receiving a first click instruction of clicking a first position in or out of the drawing; in response to the first click instruction, generating a line, and generating a name or a label at a second position, one end of the line extending to the first position, and the other end of the line extending to a side of the name or the label; or the line being located below the name or the label.

97. The method of claim 96, wherein: the generating a name or a label comprises: displaying a set of names or a set of labels, the set of names comprising the names, and the set of labels comprising the labels; receiving a second click instruction of selecting or clicking the name or the label; in response to the second click instruction, copying or moving the name or the label to the second position.

98. A computer device, wherein, The computer device comprises: a processor; a memory coupled to the processor for storing program instructions that are executable by the processor to implement the token method of any of claims 81-97.

Images