Tabular data processing methods, apparatus, equipment and storage media

Abstract

This application relates to the field of data processing technology, and discloses a method, apparatus, device, and storage medium for processing tabular data, including: acquiring a target tabular data file and a target object type; determining the target file type of the target tabular data file, and calling the corresponding file reading rules to read its header information and original data rows; extracting the mapping annotation information of the attribute fields of the target object type, and constructing mapping rules between the attribute fields and file columns based on the mapping annotation information and the header information; assigning the original data rows to objects instantiated according to the target object type based on the mapping rules to obtain a set of target data objects; and analyzing and processing the set of target data objects according to business statistical rules to generate a target data analysis report. This can be applied to intelligent reporting scenarios in the fields of fintech and healthcare, reducing the coupling between data parsing logic and file physical structure, and improving system scalability when facing multi-source heterogeneous data.

Description

Technical Field

[0001] This application relates to the field of data processing technology and can be applied to fields such as financial technology and healthcare. In particular, it relates to a method, apparatus, device and storage medium for tabular data processing. Background Technology

[0002] In business scenarios such as fintech and healthcare, there is often a need to frequently import and parse massive amounts of heterogeneous tabular data from multiple sources. Examples include transaction logs and claims invoices generated by financial institutions, and medical settlement lists generated by medical institutions, in various formats such as Excel and CSV files. Currently, the industry's common approach to tabular data processing primarily involves developers writing customized data parsing and field mapping code for specific file formats.

[0003] However, the inventors realized that existing data parsing logic is usually deeply bound to the specific physical structure of the file. When there are minor changes to the header layout of the source data file, or when other heterogeneous file formats are added to the business scenario, the original hard-coded program often fails directly, requiring developers to modify, compile, and deploy the underlying code again. In summary, existing tabular data processing solutions suffer from the technical problem of extremely poor scalability when dealing with multi-source heterogeneous data due to the high dependence of the parsing logic on the physical structure of the file. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and storage medium for processing tabular data, in order to solve the technical problem that existing tabular data processing solutions have poor scalability when dealing with multi-source heterogeneous data because the parsing logic is highly dependent on the physical structure of the file.

[0005] To achieve the above objectives, a first aspect of this application provides a method for processing tabular data, the method comprising: Obtain the target table data file and the type of the target object; Determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows of the target table data file; Extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information; Based on the mapping rules, the original data rows are assigned to objects instantiated according to the target object type to obtain a set of target data objects; The target data object set is analyzed and processed according to preset business statistical rules to generate a target data analysis report.

[0006] To achieve the above objectives, a second aspect of this application provides a tabular data processing apparatus, the apparatus comprising: The acquisition module is used to acquire the target table data file and the type of the target object; The reading module is used to determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows of the target table data file; The rule building module is used to extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules corresponding to the attribute fields and the file columns based on the mapping annotation information and the table header information; The mapping assignment module is used to assign the original data row to an object instantiated according to the target object type based on the mapping rule, so as to obtain a target data object set; The analysis and processing module is used to analyze and process the target data object set according to preset business statistical rules and generate a target data analysis report.

[0007] To achieve the above objectives, a third aspect of the present application provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method described in the first aspect.

[0008] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of the first aspect described above.

[0009] The above-mentioned table data processing method, apparatus, device, and storage medium can obtain the target table data file and the target object type; determine the target file type of the target table data file; call the file reading rules matching the target file type to read the header information and original data rows in the target table data file; extract the mapping annotation information of the attribute fields of the target object type; and construct the mapping rules corresponding to the attribute fields and file columns based on the mapping annotation information and the header information; assign the original data rows to the objects instantiated according to the target object type based on the mapping rules to obtain the target data object set; and analyze and process the target data object set according to the preset business statistical rules to generate a target data analysis report. In this application, for the needs of massive multi-source heterogeneous data flow in business scenarios such as financial technology and healthcare, the unified access of heterogeneous data sources is first achieved by dynamically calling the reading rules matching the file type. Then, by extracting the mapping annotations on the target object attributes and combining them with the actual read header information, the mapping rules are dynamically constructed at runtime, thereby completing the flexible conversion and assignment of the original data rows to the target object instance and subsequent analysis and processing. This annotation-driven and dynamic strategy invocation approach severs the mandatory binding relationship between the data parsing code and the header order and physical column structure of the source file from the underlying logic. When faced with minor adjustments to the data source format or the addition of heterogeneous files, only the annotation configuration or the addition of a reading strategy needs to be adjusted to take effect. Objectively, it achieves highly reusable and low-intrusion automated data parsing, which can solve the technical problem of poor scalability when facing multi-source heterogeneous data due to the high dependence of the parsing logic on the physical structure of the file. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of an application environment for a table data processing method according to an embodiment of the present invention; Figure 2 This is a flowchart of the tabular data processing method provided in the embodiments of this application; Figure 3 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 4 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 5 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 6 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 7 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 8 This is another flowchart of the tabular data processing method provided in the embodiments of this application; Figure 9 This is a schematic diagram of the structure of the tabular data processing device provided in the embodiments of this application; Figure 10 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 11 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0013] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0014] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0015] In business scenarios such as fintech and healthcare, enterprises typically need to frequently import and parse massive amounts of heterogeneous tabular data from multiple sources. Examples include transaction logs and claims invoices generated by financial institutions, and medical settlement lists generated by medical institutions, all in different formats such as Excel and CSV files. Currently, the industry's common approach to tabular data processing primarily involves developers writing customized data parsing and field mapping code for specific file formats.

[0016] However, the inventors realized that existing data parsing logic is usually deeply bound to the specific physical structure of the file. When there are minor changes to the header layout of the source data file, or when other heterogeneous file formats are added to the business scenario, the original hard-coded program often fails directly, requiring developers to modify, compile, and deploy the underlying code again. In summary, existing tabular data processing solutions suffer from the technical problem of extremely poor scalability when dealing with multi-source heterogeneous data due to the high dependence of the parsing logic on the physical structure of the file.

[0017] Based on this, embodiments of this application provide a method, apparatus, device, and storage medium for processing tabular data, aiming to solve the technical problem that existing tabular data processing schemes have poor scalability when dealing with multi-source heterogeneous data because the parsing logic is highly dependent on the physical structure of the file.

[0018] The tabular data processing method, apparatus, device, and storage medium provided in this application are specifically described through the following embodiments. First, the tabular data processing method in this application is described.

[0019] The tabular data processing method provided in this application embodiment can be applied to, for example, Figure 1In this application environment, the client communicates with the server via a network. The server can obtain the target table data file and the target object type from the client; determine the target file type of the target table data file, and call the file reading rules matching the target file type to read the header information and original data rows in the target table data file; extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules corresponding to the attribute fields and file columns based on the mapping annotation information and the header information; based on the mapping rules, assign the original data rows to objects instantiated according to the target object type to obtain a set of target data objects; analyze and process the set of target data objects according to preset business statistical rules, generate a target data analysis report, and feed the target data analysis report back to the client. This application addresses the import, parsing, and transfer needs of massive, multi-source, heterogeneous data in business scenarios such as fintech and healthcare. First, it achieves unified access to heterogeneous data sources by dynamically invoking reading rules that match file types. Then, it dynamically constructs mapping rules at runtime by extracting mapping annotations from target object attributes and combining them with the actual read table header information. This completes the flexible conversion and assignment of raw data rows to target object instances and subsequent complex analysis and processing. This annotation-driven and dynamic strategy invocation mechanism severs the deep binding relationship between data parsing code and the physical structure of the source file (such as table header order and column names) at the underlying logic level. Objectively, it achieves highly reusable and low-intrusion automated data parsing, solving the technical problem of poor scalability in existing tabular data processing solutions when facing multi-source heterogeneous data due to the high dependence of parsing logic on the physical structure of the file. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers. The following detailed description of specific embodiments further illustrates this application.

[0020] Figure 2 This is an optional flowchart of the tabular data processing method provided in the embodiments of this application. Figure 2 The method may include, but is not limited to, steps S201 to S205.

[0021] Step S201: Obtain the target table data file and the target object type.

[0022] Step S202: Determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows in the target table data file.

[0023] Step S203: Extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information.

[0024] Step S204: Based on the mapping rules, assign the original data rows to objects instantiated according to the target object type to obtain the target data object set.

[0025] Step S205: Analyze and process the target data object set according to the preset business statistics rules to generate a target data analysis report.

[0026] In step S201 of some embodiments, the target tabular data file is an external tabular data source storing business records to be processed. For example, in business scenarios involving financial claims data or medical settlement data, this file can be an exported entity tabular file named "Claims Complaint Details.csv" or "Patient Settlement List.xlsx". The target object type refers to a class reference of a Data Transfer Object (DTO) that is predefined in the underlying development environment and used to carry a single row of standardized business data in memory. For example, the target object type can be specified as a predefined "UserFaceIdDataDTO" class. By obtaining the above tabular data file and object type, a standard template for the physical data input source and internal data structure transformation is clarified for the subsequent data parsing process.

[0027] In step S202 of some embodiments, the received table files often have diverse formats (such as Excel, CSV, etc.) during actual business data flow. In this step, the specific format attributes of the currently incoming file are first identified to determine its target file type. Then, based on the target file type, the underlying preset file reading rules specifically responsible for parsing this type of format are invoked. By running these file reading rules, the data in the table file is structurally decomposed, extracting the header information (e.g., a set of text labels containing "user identifier," "real name," "verification status," etc.) located at the beginning of the table, and extracting all specific business record data located below the header, which are then used as the raw data rows. Through this step, the diverse external table files are uniformly abstracted and transformed into a set of headers and raw data rows that can be traversed and manipulated by the program.

[0028] Please see Figure 3 In some embodiments, step S202 may include, but is not limited to, steps S301 to S303.

[0029] Step S301: Extract the file feature identifier of the target table data file, and determine the target file type based on the file feature identifier.

[0030] Step S302: Match and select the file reading rule corresponding to the target file type from the pre-configured set of file parsing rules.

[0031] Step S303: According to the file reading rules, perform streaming or batch reading on the target table data file to obtain the table header information and the original data rows.

[0032] In step S301 of some embodiments, since the naming and format of data sources often lack unified standards in actual financial or cross-departmental business processes, it is necessary to accurately identify the physical category of the incoming file. File feature identifiers refer to characteristic parameters that objectively reflect the underlying encoding format of the file. In the specific execution process, the file extension (e.g., ".xlsx" or ".csv") at the end of the target tabular data file can be extracted as the file feature identifier by parsing the file path string. To further improve the accuracy of identification and prevent parsing failure due to malicious tampering or hiding of the file extension, the file feature identifier can also be extracted by reading a specific byte sequence (i.e., the file magic number) from the header of the file data stream. After extracting the above identifier, it is compared with the system's built-in format dictionary to accurately determine the target file type to which the file belongs (e.g., whether it is a standard CSV or Excel file type).

[0033] In step S302 of some embodiments, the file parsing rule set is a pre-encapsulated set of strategies in the underlying system that contains various independent file parsing logics. This set registers multiple reading components for different formats (e.g., a "CsvFileReader" rule specifically for handling comma-separated values, and an "ExcelFileReader" rule specifically for handling spreadsheet workbooks). After confirming the target file type, it is used as the key for routing matching, and a search and match is performed in the file parsing rule set to dynamically extract the file reading rules that match the current file type. This dynamic routing-based selection method, when faced with new data format access requirements, only requires registering the corresponding parsing rules in the rule set, without intrusive modifications to the core business code, thus ensuring the open / closed principle of the system architecture.

[0034] In step S303 of some embodiments, when faced with massive amounts of business data (such as a transaction log containing millions of records), the traditional full-load method is prone to causing system memory overflow. Therefore, after mounting the matching file reading rules, an underlying data input stream is established, and the target table data file is scanned using either streaming reading (i.e., iterating row by row and releasing memory space after processing) or batch reading (i.e., setting a fixed buffer size, such as loading 2000 rows of data each time). During the reading process, the first row of text in the located table is parsed first into header information containing column names and their position indices. Then, the read pointer continues to advance, extracting the plain text business data below the header (with format tags removed) row by row to form the original data row set, thereby preparing a standardized data source for subsequent object-oriented mapping.

[0035] By implementing steps S301 to S303 above, this embodiment of the application constructs a multi-format adaptive data access process at the front end of the core mapping processing flow. First, it achieves accurate file type identification by extracting the underlying feature identifiers of the files. Then, it uses a pre-configured set of rules for dynamic policy routing, decoupling the strong binding relationship between file format differences and core business code. Simultaneously, combined with streaming and batch reading mechanisms, it can smoothly and stably parse various heterogeneous table files while maintaining extremely low memory overhead. This mechanism effectively solves the technical problems of code redundancy and expansion difficulties in traditional solutions when facing multi-source heterogeneous data, providing a unified and standardized header and data row input for subsequent dynamic annotation mapping.

[0036] In step S203 of some embodiments, the mapping annotation information refers to the metadata description tags that developers explicitly annotate on various attribute fields of the target object type (such as the aforementioned UserFaceIdDataDTO class) during the code development phase. For example, the "realName" attribute field of this class is configured with mapping annotation information containing "User's Real Name". After extracting these mapping annotation information, they are compared with the text features of the table header information parsed in step S202. If the name of a column in the table header information (for example, the column header of the second column in the physical table is "User's Real Name") matches the above mapping annotation information, then the "realName" attribute field is established to correspond to the second column of the current table file. By summarizing all successfully matched column index relationships, a mapping rule for the current table file can be constructed. Dynamic matching through annotation content avoids hard-coding fixed column index numbers in the underlying code, enabling the data parsing logic to automatically align physical columns based on the table header name.

[0037] Please see Figure 4 In some embodiments, step S203 may include, but is not limited to, steps S401 to S403.

[0038] Step S401: Query whether there is a historical mapping rule corresponding to the target object type in the system cache. If not, use reflection mechanism to obtain all attribute fields of the target object type and scan to extract the target attribute fields configured with mapping annotation information.

[0039] Step S402: Obtain the column description information declared in the mapping annotation information, match the column description information with the table header information, and determine the target column index of the target attribute field in the target table data file.

[0040] Step S403: The association between the target attribute field and the target column index is used as a mapping rule, and the mapping rule is stored in the system cache.

[0041] In step S401 of some embodiments, considering that in scenarios involving massive tabular data processing, frequently performing low-level class resolution operations for each row of data or each new file would increase computational resource consumption and reduce processing speed, before constructing mapping rules, the identifier of the current target object type (such as the fully qualified name of the class) is used as the query key to search the system cache (e.g., using a thread-safe memory cache structure such as ConcurrentHashMap) to determine if a historical mapping rule for that type already exists. If the cache is not found, it indicates that the system is processing the target object type for the first time. At this point, the reflection mechanism at runtime is invoked to dynamically detect and extract all attribute fields defined within the target object type. Subsequently, these extracted attribute fields are scanned one by one, filtering out those target attribute fields that have mapping annotation information configured in their code definition (e.g., with custom annotation tags), which are then used as the core variables to be mapped subsequently.

[0042] In step S402 of some embodiments, the mapping annotation information extracted by the aforementioned scan typically encapsulates a business meaning identifier pre-defined by the developer for the field, i.e., column description information (e.g., "user's real name" or "contact number" in text string form). After extracting the column description information carried by each target attribute field, it is compared and traversed with the set of table header information read from the actual physical table file in the previous steps. When a column description matches the table header name text of a column in the table header information set, the absolute physical location of the business meaning in the currently input table can be established, thereby accurately determining the target column index corresponding to each target attribute field in the current target table data file. For example, it is calculated that the "contact number" attribute field corresponds to the 5th column of the current entity file.

[0043] In step S403 of some embodiments, after locating the column numbers of all target attribute fields in the current class, the internal variable name of each target attribute field is bound one-to-one with its found target column index. Summarizing all these relationships constitutes a complete mapping rule for the current target object type and a specific table header. After generating this mapping rule, a cache update mechanism is triggered, storing the mapping rule and the corresponding target object type identifier in the system cache. When subsequent processing continues to traverse the remaining tens of thousands of original data rows, or when a new parsing task encounters the same target object type again, the mapping rule can be directly read from the system cache and assigned a value, thus saving the tedious reflection and matching addressing process.

[0044] By comprehensively implementing steps S401 to S403 above, this embodiment first avoids most of the repetitive low-level reflection calls through the system cache interception and verification mechanism. Only when the cache expires is the annotation configuration extracted using reflection and combined with dynamic alignment of the table header. Finally, the newly generated rules are written back to the cache for global reuse. This not only eliminates the hard-coded dependency on fixed column numbers, giving the system a high degree of adaptability and scalability in the face of changes in table header position, but also effectively overcomes the severe performance degradation caused by simply using a full reflection mechanism. This solution addresses the technical problem that existing dynamic data processing solutions often sacrifice underlying runtime efficiency in pursuit of low code coupling and high scalability. In high-concurrency parsing scenarios with massive amounts of multi-source heterogeneous table data, it achieves a balance between system flexibility and processing performance.

[0045] In step S204 of some embodiments, after establishing the mapping rules between attribute fields and file column positions, the multiple raw data rows extracted earlier are traversed row by row. When traversing to a certain row of raw data, a blank instance object is dynamically created in memory according to the target object type (i.e., a UserFaceIdDataDTO object is instantiated). Subsequently, according to the column index mapping relationship recorded in the mapping rules, the cell text value at the corresponding position is extracted from the raw data of the current row, and the extracted field value is dynamically written to the corresponding attribute field of the instance object (e.g., writing the content of the second column to the "realName" attribute). After all the raw data rows have completed this transformation and assignment operation, these filled instance objects are gathered together to form a target data object set composed of multiple structured business objects. At this point, the external unstructured table text is converted into a normalized object set that can be directly manipulated by the underlying business logic.

[0046] Please see Figure 5 In some embodiments, step S204 may include, but is not limited to, steps S501 to S504.

[0047] Step S501: Traverse the original data rows and instantiate the corresponding number of target objects based on the target object type.

[0048] Step S502: For a single row of original data, extract the corresponding target field value based on the mapping rules, and dynamically set the target field value to the target attribute field of the target object.

[0049] In step S503, during the process of dynamically setting the target field value to the target attribute field of the target object, if a data conversion anomaly is detected in a single row of data, the data conversion anomaly is isolated in a single row, and a structured anomaly data containing the original data content and the reason for the error is recorded.

[0050] Step S504: Skip the single row of original data where the exception occurred and continue to assign values ​​to subsequent rows of original data, and add the target objects that have been assigned values ​​to the target data object collection.

[0051] In step S501 of some embodiments, after successfully parsing and separating the header of the physical file, a set of original data rows carrying business data is obtained. To transform these text rows into an object-oriented structure, the original data row set needs to be traversed line by line. During this traversal, for each row of original data read, the reflection mechanism of the underlying runtime environment (such as class object reflection in Java) is used to dynamically allocate space in system memory and instantiate a corresponding blank target object based on the previously obtained target object type (e.g., the "UserFaceIdDataDTO" class). In this way, an equal number of standard data-carrying entities are prepared on demand for the massive amount of data rows, laying the structural foundation for subsequent accurate data mapping.

[0052] In step S502 of some embodiments, for the currently processed single row of raw data (typically parsed as a string array or list), the previously constructed and cached mapping rules are retrieved. Based on the association between the target attribute field and the target column index recorded in the mapping rules, the cell text at the specific column index position in the row of data is located and extracted as the target field value. Subsequently, reflection or attribute descriptors are used again to call the assignment method of the corresponding target attribute field inside the target object, dynamically writing the extracted target field value into the target object. For example, if the mapping rules indicate that "verification status" corresponds to the 3rd column, the text value of the 3rd column of the current row is extracted and dynamically set to the "status" property of the instantiated object, thereby completing the accurate mapping of the single row of data.

[0053] In step S503 of some embodiments, in actual business scenarios, due to non-standard upstream data entry, tables are often filled with dirty data. For example, Chinese characters are mixed into the "Contact Number" column, which is required to be pure numbers, or the date format is filled incorrectly. When the system attempts to forcibly convert these non-standard text target field values ​​and set them to target attribute fields with strict data type constraints, a data conversion exception will inevitably be thrown. To address this, this embodiment encapsulates an independent exception capture mechanism around the single-row mapping logic. Once such an exception is detected, it will not block the operation of the entire batch processing task, but will isolate the impact of the data conversion exception within the current single row of data. At the same time, it will automatically capture the entire row of original data text that caused the exception, the specific field position where the mapping error occurred, and the specific error reason returned by the underlying layer (such as "type conversion failed"), and encapsulate and record this information as a structured exception data for subsequent tracing and investigation.

[0054] In step S504 of some embodiments, the single row of original data that caused the exception is skipped, and the assignment of subsequent rows of original data continues. The target objects that have been assigned are added to the target data object set. After the structured recording and isolation of the current exception row are completed, the defective semi-finished instance object is actively discarded, the subsequent processing logic of the row is skipped, and the traversal pointer is driven to move to the next row of original data to continue the new round of instantiation and assignment process. For target objects that have not triggered any exceptions and whose attribute assignments are all successful, they are uniformly collected and appended to a pre-initialized memory set. After all rows of the entire target table data file have been traversed, this set constitutes the target data object set containing all valid business data, which is then sent to the downstream business statistics engine.

[0055] By comprehensively implementing steps S501 to S504 above, this embodiment of the application, when faced with massive amounts of heterogeneous table data of varying quality, can automatically identify and isolate dirty data such as data with disordered formats and mismatched types by narrowing the anomaly capture boundary to the single-row level. This avoids the overall task crashing and stopping due to a single data error in traditional batch processing solutions. Simultaneously, by skipping and structurally recording abnormal data, it provides business personnel with a basis for error investigation without interfering with normal data flow. This mechanism effectively solves the technical problems of batch processing interruption risk and difficulty in anomaly tracing caused by the lack of intelligent fault tolerance and isolation in existing data processing solutions.

[0056] Please see Figure 6 In some embodiments, after step S502, steps S601 to S603 may also be included, but are not limited to.

[0057] In step S601, in response to the target object type being configured with a corresponding target data post-processing rule, the target object is intercepted by calling the target data post-processing rule before being included in the target data object set.

[0058] Step S602: Based on the target data post-processing rules, perform business logic verification on the target attribute fields of the target object that has completed the initial assignment, and obtain the logic verification result.

[0059] Step S603: Based on the logical verification results, perform data cleaning or format reconstruction on the target attribute fields, and then include the reconstructed target objects into the target data object set.

[0060] In step S601 of some embodiments, in actual business processing, different scenarios often have specific constraints on data standardization, and simple field mapping assignment may not fully meet the final data standards. Therefore, at the stage where the target object has completed the initial data assignment but has not yet been added to the overall collection, it is detected whether the current target object type (such as the UserFaceIdDataDTO class) has configured specific target data post-processing rules. Target data post-processing rules refer to a set of rules predefined in the underlying logic, used to perform secondary cleaning, format conversion, or business compliance verification on object instances that have completed field mapping but have not yet been formally entered into the database. This set of rules is not limited to simple text replacement or cleaning operations, but can also include advanced business logic units such as regular expression matching, numerical range verification based on business thresholds, and logical consistency comparison across fields. Its core function is to ensure that the target data object collection that finally enters the business analysis engine meets the preset compliance standards in terms of business logic and data quality. If the corresponding rule configuration exists, an interception operation will be triggered, and the target object will not be temporarily stored in the collection, but will be transferred to the post-processing process for targeted verification and processing.

[0061] In step S602 of some embodiments, after the target object is intercepted, the data values ​​of the target attribute fields inside the target object are extracted according to the conditions set in the target data post-processing rules, and business compliance verification is performed. For example, in scenarios involving financial transactions or medical patient data, it can be verified whether the "channel source identifier" or "department of treatment" field in the target object conforms to a preset enumeration range, or whether the "real name" field contains illegal characters. Through the above specific verification operations, it is possible to determine whether the data of the target attribute field conforms to business specifications and generate corresponding logical verification results.

[0062] In step S603 of some embodiments, based on the aforementioned generated logical verification results, if it is confirmed that the data of the current target attribute field is non-standard, it is modified according to the operation logic defined in the post-processing rules. For example, if the "contact number" field is detected to contain spaces or other special symbols, a character replacement operation will be performed to remove non-numeric characters; if fields such as channel name are non-standard, their values ​​will be replaced and reconstructed according to a preset mapping table. After completing the above-mentioned targeted data cleaning or format reconstruction, the target object meets the business requirements, and its interception is then lifted. The reconstructed target object is then formally included in the target data object set for use in subsequent stages.

[0063] By comprehensively implementing steps S601 to S603 above, this embodiment of the application, through configuring interception rules and combining specific business logic verification and data cleaning and reconstruction, can perform secondary correction and standardization processing on the initially mapped data. This mechanism solves the problem that a single mapping operation cannot handle complex non-standard data formats, ensuring that tabular data meets business standards before final aggregation and analysis, thereby improving the overall accuracy and quality of data processing.

[0064] In step S205 of some embodiments, the target data object set in memory already possesses a unified standard structure. Business statistical rules refer to pre-written data filtering, grouping, or aggregation calculation logic designed to meet specific business indicator requirements. For example, business statistical rules can be set to: filter out the object set whose "verification status" field is "passed," and group and count the number of objects according to the "user's channel" field. Using these preset business statistical rules, corresponding data extraction and calculations are performed on the target data object set to obtain statistical results reflecting the business operation status. Finally, the calculated indicator data is output according to a predetermined format to generate a target data analysis report containing business conclusions.

[0065] Please see Figure 7 In some embodiments, the step of analyzing and processing the target data object set according to preset business statistics rules may include, but is not limited to, steps S701 to S704.

[0066] Step S701: Extract the target attribute features of each target object in the target data object set.

[0067] Step S702: Based on preset numerical thresholds or character matching conditions, the target attribute features are compared to filter out the target data set from the target data object set.

[0068] Step S703: Classify and divide the target data set according to the preset data attribute dimensions to obtain multiple data groups.

[0069] Step S704: Perform aggregate calculations on the data volume and proportion of the target objects contained in each data group to obtain the statistical results of the business indicators corresponding to each data group.

[0070] In step S701 of some embodiments, in order to filter out feature indicators with business analysis value from structured data objects, the previously generated set of target data objects is first traversed. Each object in the target object set is an instance formed by mapping and assigning values ​​to the original table data rows, and each instance contains multiple attribute fields. In this step, according to the feature extraction dimensions preset for business analysis, the numerical or character information of the corresponding attribute fields is extracted from each target object instance. For example, if the business statistics target is to analyze "channel verification pass status", then the extracted target attribute features are the "channel source" field and the "verification status" field in each object instance. By extracting and aggregating these key business attributes distributed in different object instances, a basic feature vector is provided for subsequent logical comparison and data filtering.

[0071] In step S702 of some embodiments, after extracting the target attribute features, they are compared one by one with the comparison parameters defined in the preset business statistical rules. These business statistical rules typically include numerical threshold judgments (e.g., counting transaction records with amounts greater than a preset value) or character matching conditions (e.g., matching records with a verification status of "passed"). During execution, the comparison engine performs logical judgments on the attribute features of each object. If the attribute feature meets the preset numerical threshold or character matching condition, the object is identified as an object of interest for analysis and included in the target data set. Through this step, the accurate extraction of specific business samples from massive amounts of basic data is achieved, reducing the data scale of subsequent statistical analysis.

[0072] In step S703 of some embodiments, to enable the statistical analysis results to more accurately reflect business performance, dimensional clustering is performed on the selected target data set. Data attribute dimensions refer to predefined classification criteria based on business needs, such as geographical attributes, departmental affiliation, or time periods. Based on the selected attribute dimension, the values ​​of the corresponding attribute fields for each object in the target data set are scanned. Objects with the same attribute value are grouped into the same logical category, thereby dividing the entire target data set into several independent data groups. For example, if "province" is used as the data attribute dimension, the system will generate data groups corresponding to each province, making the originally mixed business data present a clear hierarchical classification structure.

[0073] In step S704 of some embodiments, after the classification is completed, mathematical aggregation operations are performed on each independent data group. Specifically, firstly, the total number of target object instances contained in each data group is counted, and then the percentage of that group's number in the entire target data set is calculated, serving as an important reference indicator for evaluating the business scale under that dimension. By summarizing these statistics, the final business indicator statistics results corresponding to each data group are formed, and these results intuitively quantify the operational status of each business dimension. By implementing the above steps S701 to S704, the embodiments of this application can effectively solve the problems of poor flexibility and redundant calculation logic in traditional manual report statistics, providing objective and accurate data support for business decisions.

[0074] Please see Figure 8 In some embodiments, the step of generating a target data analysis report may include, but is not limited to, steps S801 to S804.

[0075] Step S801: Obtain the pre-set output structure parameters, which include the worksheet name, worksheet sorting weight, and overview page identifier corresponding to each data group.

[0076] Step S802: Create an overview worksheet based on the overview page identifier, and write the statistical results of the business indicators corresponding to each data group into the overview worksheet.

[0077] Step S803: Create corresponding detail worksheets for each data group, and write the target objects in the target data set into the detail worksheets with corresponding names.

[0078] Step S804: Based on the worksheet sorting weights corresponding to each detailed worksheet, sort and merge the overview worksheet and each detailed worksheet to generate a target data analysis report in the form of multiple worksheets.

[0079] In step S801 of some embodiments, during the data analysis report generation stage, the physical structure of the final output report needs to be determined based on the user's preset template parameters. The output structure parameters are predefined metadata configurations, where "worksheet name" specifies the name of each generated sub-sheet page; "worksheet sorting weight" is a numerical logical parameter used to control the order in which the sub-sheets are arranged in the file during report rendering; and "overview page identifier" is a specific flag used to specify whether a primary worksheet exists for summarizing the results of the global statistical view. By reading these parameters, a report framework conforming to business format specifications can be pre-established.

[0080] In step S802 of some embodiments, after clarifying the output structure, it is determined whether a global statistical view needs to be generated based on the obtained "overview page identifier". If an overview requirement is determined, a summary overview worksheet is created at the beginning or a specified position of the report file. Subsequently, the business indicator statistics results corresponding to each data group calculated in step S704 are sequentially filled into the specified cells of the overview worksheet according to a preset table format. This overview worksheet provides users with an interactive interface that allows them to quickly view the overall business situation by aggregating the data scattered across various dimension groups.

[0081] In step S803 of some embodiments, to achieve a vertical display of data analysis, it is necessary to provide not only a summary overview but also specific business records under each dimension. All previously divided data groups are traversed. For each group, a detailed worksheet named after the corresponding "worksheet name" is created in the report file. Subsequently, the detailed attribute data of all target object instances contained within that group are filled into the corresponding detailed worksheet row by row. This detailed table system based on grouping allows business personnel to directly switch to the corresponding detailed tab after viewing the overall statistical data, enabling them to conduct more in-depth verification of specific business objects.

[0082] In step S804 of some embodiments, after all worksheets have been filled, a preset worksheet sorting weight is used to adjust all worksheets. The worksheets are sorted in ascending or descending order based on the weight value corresponding to each overview worksheet, with smaller weight values ​​typically appearing at the beginning of the file. After sorting, the overview worksheet containing summary indicators and multiple detailed worksheets recording detailed business data are packaged and merged into a single complete spreadsheet file, ultimately exported as a deliverable target data analysis report.

[0083] Through the above steps S801 to S804, this embodiment of the application effectively solves the technical problems of low report generation efficiency and easy typesetting errors under the traditional method by integrating data aggregation calculation with structured form filling, and improves the standardization of data output and the intuitive readability of analysis conclusions.

[0084] By comprehensively implementing steps S201 to S205, the mapping annotation information in the target object type is dynamically compared with the actual header information of the read physical table. At runtime, mapping rules between attribute fields and column positions are automatically constructed, and then the original data is dynamically assigned to the instantiated business object based on these rules. This data processing mechanism severs the strong dependency between the underlying parsing code and the physical structure of the source table (such as specific column header ordering or hard-bound column number positions). This allows for adaptive parsing even when the business side adjusts the column order of the imported table on the front end or adds a heterogeneous data source, as long as the header text and mapping annotations correspond. This solution effectively reduces the coupling between the data parsing logic and the file's physical structure, solving the technical problem of poor scalability in existing table data processing solutions when facing multi-source heterogeneous data due to the high dependence of the parsing logic on the physical structure. It significantly reduces the cost of subsequent code maintenance and manual adjustments.

[0085] As can be seen in the above scheme, please refer to Figure 9 This application also provides a tabular data processing apparatus that can implement the above-described tabular data processing method. The apparatus includes: The acquisition module is used to acquire the target table data file and the type of the target object; The reading module is used to determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and raw data rows of the target table data file; The rule building module is used to extract the mapping annotation information of the attribute fields of the target object type, and to build the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information; The mapping assignment module is used to assign the original data rows to objects instantiated according to the target object type based on mapping rules, so as to obtain a collection of target data objects; The analysis and processing module is used to analyze and process the target data object set according to preset business statistical rules and generate a target data analysis report.

[0086] In some embodiments, the reading module is specifically used for: Extract the file feature identifier of the target table data file, and determine the target file type based on the file feature identifier; From the pre-configured set of file parsing rules, match and select the file reading rule corresponding to the target file type; Based on the file reading rules, perform streaming or batch reading on the target table data file to obtain the table header information and the original data rows.

[0087] In some embodiments, the rule building module is specifically used for: Check if there is a historical mapping rule corresponding to the target object type in the system cache. If not, use reflection to obtain all attribute fields of the target object type and scan to extract the target attribute fields configured with mapping annotation information. Retrieve the column description information declared in the mapping annotation information, match the column description information with the table header information, and determine the target column index of the target attribute field in the target table data file; The association between the target attribute field and the target column index is used as the mapping rule, and the mapping rule is stored in the system cache.

[0088] In some embodiments, the mapping assignment module is specifically used for: Iterate through the original data rows and instantiate the corresponding number of target objects based on the target object type; For a single row of raw data, extract the corresponding target field value based on the mapping rules, and dynamically set the target field value to the target attribute field of the target object; During the process of dynamically setting the target field value to the target attribute field of the target object, if a data conversion anomaly is detected in a single row of data, the data conversion anomaly is isolated in a single row, and a structured anomaly data containing the original data content and the reason for the error is recorded. Skip the single row of raw data where the exception occurred and continue to assign values ​​to subsequent rows of raw data, then add the target objects that have been assigned values ​​to the target data object collection.

[0089] In some embodiments, the mapping assignment module is further used for: If a target object type is configured with a corresponding target data post-processing rule, the target object is intercepted by calling the target data post-processing rule before being included in the target data object collection. Based on the target data post-processing rules, business logic validation is performed on the target attribute fields of the target object that has completed the initial assignment, and the logic validation results are obtained. Based on the logical verification results, perform data cleaning or format reconstruction on the target attribute fields, and then include the reconstructed target objects into the target data object collection.

[0090] In some embodiments, the analysis and processing module is specifically used for: Extract the target attribute features of each target object in the target data object set; Based on preset numerical thresholds or character matching conditions, the target attribute features are compared to filter out the target data set from the target data object set. The target data set is classified and divided according to the preset data attribute dimensions to obtain multiple data groups; Aggregate the data volume and proportion of the target objects contained in each data group to obtain the statistical results of the business indicators corresponding to each data group.

[0091] In some embodiments, the analysis and processing module is specifically used for: Obtain the pre-defined output structure parameters, which include the worksheet name, worksheet sort weight, and overview page identifier for each data group. Create an overview worksheet based on the overview page identifier, and write the statistical results of business indicators corresponding to each data group into the overview worksheet. Create corresponding detail worksheets for each data group, and write the target objects in the target data set into the detail worksheets with the corresponding names; Based on the worksheet sorting weights of each detailed worksheet, the overview worksheet and each detailed worksheet are sorted and combined to generate a multi-worksheet target data analysis report.

[0092] This invention provides a tabular data processing device. First, it achieves unified access to heterogeneous data sources by dynamically invoking reading rules that match file types. Then, it dynamically constructs mapping rules at runtime by extracting mapping annotations from target object attributes and combining them with the actual read header information. This completes the flexible conversion and assignment of raw data rows to target object instances and subsequent analysis and processing. This annotation-driven and dynamic strategy invocation severs the mandatory binding relationship between data parsing code and the source file header order and physical column structure at the underlying logic level. When facing minor adjustments to the data source format or the addition of heterogeneous files, only the annotation configuration or the addition of a reading strategy needs to be adjusted to take effect. Objectively, this achieves highly reusable and low-intrusion automated data parsing, solving the technical problem of poor scalability in the face of multi-source heterogeneous data due to the high dependence of parsing logic on the physical structure of the file.

[0093] Specific limitations regarding the tabular data processing device can be found in the limitations of the tabular data processing method described above, and will not be repeated here. Each module in the aforementioned tabular data processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0094] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When executed by the processor, the computer program implements the functions or steps of a tabular data processing method on the server side.

[0095] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 11 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements client-side functions or steps of a table-based data processing method.

[0096] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Obtain the target table data file and the type of the target object; Determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows from the target table data file; Extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information; Based on the mapping rules, the original data rows are assigned to objects instantiated according to the target object type to obtain the target data object collection; The target data object set is analyzed and processed according to the preset business statistics rules to generate a target data analysis report.

[0097] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Obtain the target table data file and the type of the target object; Determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows from the target table data file; Extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information; Based on the mapping rules, the original data rows are assigned to objects instantiated according to the target object type to obtain the target data object collection; The target data object set is analyzed and processed according to the preset business statistics rules to generate a target data analysis report.

[0098] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0099] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0100] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0101] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0102] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for processing tabular data, characterized in that, include: Obtain the target table data file and the type of the target object; Determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows of the target table data file; Extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information; Based on the mapping rules, the original data rows are assigned to objects instantiated according to the target object type to obtain a set of target data objects; The target data object set is analyzed and processed according to preset business statistical rules to generate a target data analysis report.

2. The tabular data processing method according to claim 1, characterized in that, The step of determining the target file type of the target table data file and calling the file reading rules matching the target file type to read the header information and original data rows in the target table data file includes: Extract the file feature identifier of the target table data file, and determine the target file type based on the file feature identifier; From the pre-configured set of file parsing rules, match and select the file reading rule corresponding to the target file type; According to the file reading rules, the target table data file is read in a streaming or batch manner to obtain the table header information and the original data rows.

3. The tabular data processing method according to claim 1, characterized in that, The step of extracting the mapping annotation information of the attribute fields of the target object type, and constructing the mapping rules between the attribute fields and the file columns based on the mapping annotation information and the table header information, includes: The system cache is checked to see if there is a historical mapping rule corresponding to the target object type. If not, all attribute fields of the target object type are obtained using reflection, and the target attribute fields configured with the mapping annotation information are scanned and extracted. Obtain the column description information declared in the mapping annotation information, match the column description information with the table header information, and determine the target column index of the target attribute field in the target table data file; The association between the target attribute field and the target column index is used as the mapping rule, and the mapping rule is stored in the system cache.

4. The tabular data processing method according to claim 1, characterized in that, The step of assigning the original data rows to objects instantiated according to the target object type based on the mapping rule to obtain a target data object set includes: Traverse the original data rows and instantiate the corresponding number of target objects based on the target object type; For a single row of original data, the corresponding target field value is extracted based on the mapping rule, and the target field value is dynamically set to the target attribute field of the target object; During the process of dynamically setting the target field value to the target attribute field of the target object, if a data conversion anomaly is detected in a single row of data, the data conversion anomaly is isolated in a single row, and a structured anomaly data containing the original data content and the reason for the error is recorded. Skip the single row of original data where the exception occurred and continue to assign values ​​to subsequent rows of original data, then add the target object that has been assigned a value to the target data object set.

5. The tabular data processing method according to claim 4, characterized in that, After dynamically setting the target field value to the target attribute field of the target object, the method further includes: In response to the fact that the target object type is configured with a corresponding target data post-processing rule, the target object is intercepted by calling the target data post-processing rule before it is included in the target data object set; Based on the target data post-processing rules, business logic verification is performed on the target attribute fields of the target object that has completed the initial assignment, and the logic verification result is obtained. Based on the logical verification results, data cleaning or format reconstruction is performed on the target attribute field, and the reconstructed target object is incorporated into the target data object set.

6. The tabular data processing method according to claim 1, characterized in that, The step of analyzing and processing the target data object set according to preset business statistical rules includes: Extract the target attribute features of each target object in the target data object set; Based on preset numerical thresholds or character matching conditions, the target attribute features are compared to filter out the target data set from the target data object set. The target data set is classified and divided according to preset data attribute dimensions to obtain multiple data groups; The data volume and proportion of the target objects contained in each data group are aggregated and calculated to obtain the business indicator statistics results corresponding to each data group.

7. The tabular data processing method according to claim 6, characterized in that, The generated target data analysis report includes: Obtain pre-defined output structure parameters, which include the worksheet name, worksheet sorting weight, and overview page identifier for each data group; An overview worksheet is created based on the overview page identifier, and the statistical results of the business indicators corresponding to each data group are written into the overview worksheet; Create corresponding detail worksheets for each of the data groups, and write the target objects in the target data set into the detail worksheets with corresponding names; Based on the worksheet sorting weights corresponding to each of the detailed worksheets, the overview worksheet and each of the detailed worksheets are sorted and combined to generate the target data analysis report in a multi-worksheet format.

8. A tabular data processing device, characterized in that, The device includes: The acquisition module is used to acquire the target table data file and the type of the target object; The reading module is used to determine the target file type of the target table data file, and call the file reading rules that match the target file type to read the header information and original data rows of the target table data file; The rule building module is used to extract the mapping annotation information of the attribute fields of the target object type, and construct the mapping rules corresponding to the attribute fields and the file columns based on the mapping annotation information and the table header information; The mapping assignment module is used to assign the original data row to an object instantiated according to the target object type based on the mapping rule, so as to obtain a target data object set; The analysis and processing module is used to analyze and process the target data object set according to preset business statistical rules and generate a target data analysis report.

9. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the tabular data processing method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the tabular data processing method according to any one of claims 1 to 7.

Images