Data interface identifier analysis method

Abstract

The invention relates to a data interface identification analysis method. The method comprises the following steps: firstly, obtaining identification data to be analyzed and interface scene parameters containing data interaction frequency, data volume level and data privacy sensitivity; a multi-dimensional analysis decision model is constructed based on the parameters, a target analysis node is screened, and an adaptive transmission protocol and a data compression strategy are determined; performing format verification and redundant field elimination on the data, converting according to a target node format, and encrypting by adopting a corresponding encryption rule in combination with privacy sensitivity; and finally, sending an analysis request containing encrypted data to the target node, receiving a result, verifying the integrity, adapting a result format and a transmission mode according to scene parameters, and outputting the result to a request end. The objective of the invention is to solve the technical problems of poor node adaptability of identification analysis, insufficient data security and mismatching of analysis efficiency and scene requirements in different interface scenes.

Description

Technical Field

[0001] This invention belongs to the field of industrial network internet technology and relates to a data interface identifier parsing method. Background Technology

[0002] In the industrial internet technology system, data interface identifier resolution is a core supporting technology for achieving data interconnection and interoperability across devices, systems, and platforms. Its function is similar to the Domain Name System (DNS) in the internet field. By assigning unique identifiers to elements such as equipment, materials, and business data in industrial production and completing resolution, it opens up data flow links and enables precise control over the source, flow process, and purpose of data. Currently, the mainstream identifier resolution systems in China include the Handle system, the MA system, and the VAA system. These systems have been gradually implemented in various industry scenarios. For example, in the automotive manufacturing industry, identifier resolution technology enables full lifecycle data traceability of parts from procurement and production to after-sales service; in the chemical industry, identifier resolution enables real-time interaction and remote monitoring of production equipment operating status data. As industrial production transforms towards intelligence and flexibility, the application scenarios of data interfaces are becoming increasingly diversified. The requirements for data interaction frequency (such as millisecond-level real-time control data and minute-level statistical data), data volume level (such as KB-level equipment status data and GB-level production image data), and data privacy sensitivity (such as publicly available equipment model data and confidential process parameter data) vary significantly in different scenarios. This places higher demands on the scenario adaptability of identifier resolution technology.

[0003] However, in actual production and use, existing data interface identifier parsing technologies still suffer from three major problems, making it difficult to meet the application needs of diverse scenarios. First, node adaptability is poor. Existing technologies mostly use fixed parsing node selection logic, unable to dynamically adjust nodes according to interface scenario parameters. This leads to high node response latency affecting the real-time performance of production control in high-frequency interaction scenarios, or insufficient node bandwidth causing parsing request blocking during large-volume data transmission. Second, data security is insufficient. Existing solutions mostly use a single encryption strategy, without designing differentiated protection mechanisms for data with different privacy sensitivities. This can lead to the risk of leakage of highly sensitive data due to insufficient encryption strength, or waste of system resources due to excessive encryption of ordinary data. Finally, parsing efficiency is mismatched with scenario requirements. Existing technologies lack targeted preprocessing and transmission optimization for data. For example, the failure to remove redundant fields increases the data transmission volume, and the failure to select appropriate transmission protocols and compression strategies according to the scenario results in redundant parsing processes in low-frequency scenarios and low efficiency in high-frequency scenarios, seriously affecting the efficiency of production data flow.

[0004] After reviewing relevant materials, the main solutions to the above problems currently focus on multi-system identifier sharing and resolution. The most representative example is patent CN116233058A, which describes an identifier resolution method, device, and medium supporting multi-system identifier sharing. The core principle of this patent is: by acquiring identifier data, determining registration data and corresponding registration nodes, and saving the registration node addresses to a database; during the resolution phase, determining the resolution node address based on the resolution request and searching in the database, and once found, sending a request to the corresponding node to complete the resolution; simultaneously, using a middleware system to achieve compatibility between multi-system identifier registration and resolution, and supporting data format conversion, identifier data encryption, and authorized access control. From a production application perspective, the advantage of this patent lies in its compatibility with multiple identifier resolution systems such as Handle, MA, and VAA, breaking down data barriers between different systems and reducing the cost for enterprises to connect to multiple systems. However, its disadvantages are also quite obvious: First, node selection still relies on preset address matching and does not dynamically filter nodes based on interface scenario parameters (such as data interaction frequency and volume level), failing to solve the problem of poor node adaptability. Second, the encryption strategy is not related to data privacy sensitivity, using only a uniform encryption method, which is difficult to meet the protection requirements of different security scenarios. Third, it does not design differentiated transmission protocols and compression strategies for data interaction frequency and volume level, failing to achieve a precise match between resolution efficiency and scenario requirements. Therefore, although the existing method has made some breakthroughs in multi-system compatibility, it still cannot simultaneously solve the comprehensive problems of poor node adaptability, insufficient data security, and mismatch between resolution efficiency and scenario requirements. There is an urgent need for an identifier resolution method that can combine interface scenario parameters to achieve dynamic node filtering, differentiated security protection, and precise efficiency optimization to meet the diverse needs of actual production. Summary of the Invention

[0005] This invention provides a data interface identifier parsing method to solve the technical problems of poor node adaptability, insufficient data security, and mismatch between parsing efficiency and scenario requirements in different interface scenarios.

[0006] To solve the above problems, the technical solution adopted by the invention is as follows: A data interface identifier parsing method includes: acquiring identifier data to be parsed and interface scenario parameters, wherein the interface scenario parameters include data interaction frequency, data volume level, and data privacy sensitivity; constructing a multi-dimensional parsing decision model based on the interface scenario parameters, filtering target parsing nodes through the multi-dimensional parsing decision model, and determining the appropriate transmission protocol and data compression strategy; performing format verification and redundant field removal on the identifier data to be parsed, converting it according to the data format requirements of the target parsing node, and encrypting it according to the encryption rules corresponding to the data privacy sensitivity; sending a parsing request carrying the encrypted identifier data to the target parsing node, receiving the parsing result and verifying its integrity, and outputting the parsing result to the requesting end after adapting its format and transmission method according to the interface scenario parameters.

[0007] The principle and advantages of this solution are as follows: The core principle of this solution is to construct a full-process data interface identifier resolution mechanism that integrates scene awareness, dynamic decision-making, and precise adaptation. By deeply integrating interface scene parameters with each stage of the resolution process, it achieves intelligent and personalized resolution services. First, by acquiring the identifier data to be resolved and interface scene parameters including data interaction frequency, data volume level, and data privacy sensitivity, it provides a foundation for subsequent decision-making. Second, based on the scene parameters, a multi-dimensional resolution decision model is constructed, comprehensively considering factors such as node performance, transmission efficiency, and security compliance to dynamically select the optimal resolution node and match it with suitable transmission protocols and data compression strategies, ensuring a precise match between resolution resources and scene requirements. Next, the identifier data undergoes format verification, redundancy removal, and format conversion, followed by differentiated encryption processing based on data privacy sensitivity, optimizing data transmission volume while ensuring data security. Finally, the resolution request is sent and the result is received. After integrity verification, the resolution result is adapted to the format and transmission method according to the scene parameters and finally output to the requesting end, achieving full-process scene-based adaptation from data input to result output.

[0008] Compared to existing technologies, which only match parsing nodes based on preset addresses and cannot adapt to diverse scenario requirements, this solution uses a multi-dimensional decision model to prioritize low-latency nodes for high-frequency interaction scenarios and high-bandwidth nodes for large-volume data scenarios. For example, in industrial real-time control scenarios, it can reduce parsing response latency. Secondly, data security protection is upgraded from "single encryption" to "sensitivity-level encryption." Existing technologies often use a uniform encryption strategy, resulting in insufficient security for highly sensitive data or redundant encryption for ordinary data. This solution uses dynamic key encryption for extremely sensitive data, which can resist replay attacks and key leakage. To mitigate risks, this solution employs lightweight encryption for ordinary data, reducing resource consumption. Thirdly, the parsing process has been upgraded from a "fixed process" to "full-link scenario adaptation." Existing technologies lack scenario-based optimization for data preprocessing and result output. This solution reduces data transmission volume by eliminating redundant fields and selects appropriate transmission protocols and compression strategies based on the chosen scenario. For example, in high-frequency reporting scenarios of IoT devices, data transmission efficiency can be improved while ensuring that the format and transmission method of the parsed results meet the requirements of the requesting end. This achieves a perfect balance between parsing efficiency, security, and scenario requirements, solving the technical pain point that existing technologies cannot simultaneously adapt to multiple scenarios.

[0009] Furthermore, the construction and node selection process of the multi-dimensional parsing decision model includes: determining parsing decision evaluation indicators, which include the node's historical parsing success rate, the node's current concurrency, the average data transmission time, and the node's security compliance level; assigning dynamic weights to each evaluation indicator based on interface scenario parameters, wherein the weight of data interaction frequency is positively correlated with the average data transmission time, the weight of data volume level is negatively correlated with the node's current concurrency, and the weight of data privacy sensitivity is positively correlated with the node's security compliance level; calculating the decision score of each candidate parsing node using a weighted comprehensive scoring algorithm, and determining the candidate node with the highest score as the target parsing node, wherein the formula for the weighted comprehensive scoring algorithm is: In the formula: S is the decision score of the j-th candidate parsing node, and its value ranges from [0,1]. Let be the dynamic weight of the m-th evaluation indicator, and satisfy . m=1 corresponds to the historical parsing success rate of the node, m=2 corresponds to the current concurrency of the node, m=3 corresponds to the average data transmission time, and m=4 corresponds to the security compliance level of the node. This represents the original value of the m-th evaluation metric for the j-th candidate node; The minimum value of the m-th evaluation index among all candidate nodes; The m-th evaluation metric is the maximum value among all candidate nodes; This normalization process, which normalizes the evaluation result of the m-th evaluation metric for the j-th candidate node, eliminates the imbalance in numerical weights caused by differences in the attributes of different evaluation metrics by unifying the units and value ranges. This provides a foundation for the accuracy of subsequent weighted comprehensive scoring. Metrics such as the node's historical parsing success rate (range 0%-100%), the node's current concurrency (range 0-10000 times / second), average data transmission time, and the node's security compliance level (range 1-5) have completely different original values ​​in terms of magnitude and unit. Directly substituting these values ​​into the scoring calculation can lead to an overemphasis on the impact of large-value metrics and a weakening of the role of small-value metrics. For example, a node's original concurrency value of 8000, far higher than other metrics, might dominate the scoring, while masking the influence of key metrics that truly reflect security and reliability, such as compliance level and parsing success rate. After normalization, the values ​​of all indicators are mapped to the range of [0,1]. The value of each indicator only represents its relative merits in that indicator dimension. For example, a normalized value of 0.8 means that the node is better than 80% of the candidate nodes in that indicator. When combined with dynamic weight calculation of the comprehensive score, it can be ensured that each evaluation indicator contributes its influence according to the preset weight. This avoids decision-making bias caused by differences in scale and accurately reflects the weight requirements of interface scenario parameters for each indicator. Ultimately, it achieves the scientific nature and scenario adaptability of the target parsing node selection.

[0010] Furthermore, the dynamic weight allocation process includes: pre-setting a correlation coefficient matrix between interface scenario parameters and evaluation indicators; data interaction frequency is divided into high frequency (≥100 times / minute), medium frequency (10-100 times / minute), and low frequency (<10 times / minute), with corresponding correlation coefficients of 0.35, 0.20, and 0.05 with average data transmission time, respectively; data volume is divided into large (≥100MB), medium (10-100MB), and small (<10MB), with corresponding correlation coefficients of 0.30, 0.15, and 0.05 with current node concurrency, respectively; data privacy sensitivity is divided into extremely high, high, medium, and low, with corresponding correlation coefficients of 0.40, 0.30, 0.20, and 0.10 with node security compliance level, respectively; and normalizing the correlation coefficients using a weight calibration formula to obtain the dynamic weights of each evaluation indicator. The weight calibration formula is as follows: In the formula: Let be the correlation coefficient between the m-th evaluation metric and the interface scenario parameters; Let be the inherent importance coefficient of the m-th evaluation indicator, with a value range of [0.8, 1.2]. The default inherent importance coefficients are: node security compliance level (1.2), node historical parsing success rate (1.1), average data transmission time (1.0), and current node concurrency (0.9). These can be adjusted according to the actual application scenario. The preset correlation coefficient matrix allows the weight allocation to directly respond to the differentiated needs of the interface scenario: for example, in high-frequency data interaction scenarios, the correlation coefficient of average data transmission time (0.35) is significantly higher than in low- and medium-frequency scenarios, ensuring the priority consideration of low-latency characteristics in high-frequency scenarios; when transmitting large volumes of data, the correlation coefficient of current node concurrency (0.30) is increased, avoiding parsing blockage caused by excessive node load; for highly privacy-sensitive data, the correlation coefficient of node security compliance level (0.40) is increased, strengthening the security of data transmission. Based on this, the weight calibration formula introduces inherent importance coefficients... This further optimizes the rationality of weight allocation: for example, the node security compliance level. The default value is 1.2, higher than other indicators. This reflects the fundamental role of security and compliance in industrial data interaction while preventing the security weight from being marginalized due to low scenario correlation coefficients. The current concurrency of the node... A value of 0.9 ensures load consideration while preventing it from excessively affecting the overall score of the parsing node. Finally, through normalization, the sum of the weights of each indicator is made up to 1, ensuring the effectiveness of the weighted overall score. This achieves an organic combination of scenario-driven requirements and the essential importance of the indicators, making the selection of parsing nodes more closely aligned with the complex scenario requirements in actual applications.

[0011] Furthermore, the encryption process based on the encryption rules corresponding to the data privacy sensitivity includes: if the data privacy sensitivity is extremely high, an SM4 block encryption + dynamic key derivation combination encryption scheme is adopted, wherein the dynamic key is generated through a key generation formula, and the key generation formula is: In the formula: K is a dynamic encryption key with a length of 128 bits; T represents the current timestamp, accurate to milliseconds, in the format "YYYYMMDDHHMMSSmmm"; The ID is a unique identifier for the target parsing node, using the UUIDv4 format, and is 128 bits long; R is a random number, with a value range of [100000, 999999], and a new random number is generated each time the encryption is performed; ⊕ represents the XOR operation; The pre-shared root key is configured offline by the system administrator and is 256 bits long. HMAC-SM3 is a key derivation function based on the SM3 hash algorithm, used to convert concatenated raw data into a 128-bit key. For high data privacy sensitivity, the SM2 asymmetric encryption algorithm is used, encrypting the data using the public key of the target parser node. For medium data privacy sensitivity, the AES-256 symmetric encryption algorithm is used. For low data privacy sensitivity, the DES symmetric encryption algorithm is used. This encryption strategy adapts the encryption scheme according to the level of data privacy sensitivity: for extremely sensitive data, a combination of "SM4 block cipher + dynamic key derivation" is used, with the key generation formula... By integrating timestamps (T), unique node identifiers (ID), random numbers (R), and a pre-shared root key (Kroot), this approach ensures that the encryption key is unique and dynamically changes each time, effectively resisting security risks such as replay attacks and key leaks. Furthermore, the SM4 algorithm, as a national cryptographic standard, possesses high-strength encryption performance. For highly sensitive data, the SM2 asymmetric encryption algorithm uses a public-private key separation mechanism to guarantee identity authentication and data confidentiality during data transmission. Medium- and low-sensitivity data employ AES-256 and DES symmetric encryption algorithms respectively, reducing computational resource consumption and time latency during encryption and decryption while meeting basic security requirements. This hierarchical encryption strategy covers data security needs in different scenarios and optimizes overall encryption efficiency through differentiated design, making it more flexible and practical than traditional single encryption schemes.

[0012] Furthermore, the process of performing format validation and redundant field removal on the data to be parsed includes: obtaining the standard data format specifications supported by the target parsing node, which includes field names, data types, field lengths, and required / optional attributes; performing syntax validation on the original data to be parsed based on the specifications, marking abnormal fields that do not conform to the specifications, and if the abnormal field is a required field, providing a completion prompt to the requesting end; if the abnormal field is an optional field, automatically removing the field; filtering redundant fields using a field importance assessment model, which marks fields with a call frequency <5% as redundant based on their historical parsing frequency and the correlation between fields and parsing results, marking fields with a correlation <10% as redundant, determining the range of redundant fields, automatically removing redundant fields and retaining core data fields; and obtaining the standard data format specifications of the target parsing node, performing syntax validation on the data and processing abnormal fields, providing a completion prompt when required fields are abnormal to ensure the integrity of the data required for parsing; and automatically removing optional fields when they are abnormal to avoid invalid data interfering with the parsing process and reduce the risk of parsing failure from the source. Building upon this foundation, a field importance assessment model is used to filter redundant fields based on clear criteria such as "historical call frequency <5%" and "correlation with parsing results <10%". This accurately removes useless or low-value fields, such as redundant identifiers used only for local records and additional information unrelated to the parsing target, while retaining core data fields. This approach not only solves problems such as inconsistent original data formats, excessive invalid fields leading to unrecognizable parsing nodes, and excessively long parsing times, but also reduces bandwidth consumption and computational resource consumption of parsing nodes during data transmission. For example, for production data containing dozens of fields, redundancy removal can reduce the data volume, thereby shortening data transmission time and parsing node processing time. It also avoids format conflicts that may arise from redundant fields, making the parsing process smoother and more efficient.

[0013] Furthermore, the method also includes: generating a unique request identifier (RequestID) when sending a parsing request to the target parsing node, and initiating a timeout monitoring mechanism. The timeout threshold is determined by a timeout calculation formula, which is: In the formula: This is the timeout threshold, in milliseconds. The average data transmission time of the target parsing node over the past hour was obtained through historical log statistics. The standard deviation of data transmission time for the target node over the past hour is used to cover fluctuations in transmission time; if in If no resolution result is received, the resolution request is resent to the candidate resolution node with the second-highest score, and the timeout event of the target resolution node is recorded. After accumulating 3 timeout events, the node is marked as "temporarily unavailable" and removed from the candidate node list. The node status is re-probeed every 30 minutes. The timeout threshold is determined by the formula. The calculation, combining the average transmission time and standard deviation of the target parsing node over the past hour, not only closely reflects the actual transmission performance of the node but also covers unexpected situations such as network fluctuations and changes in node load through the standard deviation. This avoids accidental timeout retries triggered by occasional delays and also prevents requests from being blocked for extended periods due to excessively high thresholds. When no result is received after a timeout, a request is automatically resent to the next highest-scoring candidate node to ensure uninterrupted parsing. After three timeouts, the node is marked as "temporarily unavailable" and removed. Simultaneously, the node status is periodically monitored, avoiding resource waste caused by frequently sending requests to faulty nodes and ensuring timely reinstatement of the node into the candidate list once it recovers. This mechanism solves the problems of inflexible timeout control and node fault tolerance in existing technologies. For example, in high-frequency data parsing scenarios in the industrial internet, it can reduce the parsing request failure rate by more than 40% while reducing invalid waiting time, ensuring the continuity and efficiency of the parsing service. It is particularly suitable for scenarios with high requirements for parsing timeliness and reliability, such as production control and equipment collaboration.

[0014] Furthermore, the process of receiving the parsing result and verifying its integrity includes: the parsing result carries a checksum generated based on the SM3 algorithm; after receiving the parsing result, the original data portion is extracted and the SM3 hash value is recalculated; the recalculated hash value is compared with the checksum in the parsing result; if they match, the parsing result is determined to be complete; if they do not match, a retransmission request is sent to the target parsing node, carrying the identifier of the parsing result segment received in the first transmission, and only missing or erroneous segments are retransmitted; if the verification still does not match after two consecutive retransmissions, the process switches to the second-highest score candidate node to re-initiate the parsing request, and records the verification failure event of the target parsing node. The parsing result carries a checksum based on the SM3 algorithm, which has high security and uniqueness. By recalculating the hash value and comparing it, it can accurately identify whether there are problems such as data loss or tampering during transmission, fundamentally preventing invalid or erroneous data from entering subsequent application processes. When inconsistencies arise during verification, instead of retransmitting all data directly, a retransmission request carrying a fragment identifier is sent to retrieve only the missing or erroneous fragments, significantly reducing bandwidth consumption and transmission time during retransmission. If two consecutive retransmissions still fail to obtain complete data, the request is switched to the next highest-scoring candidate node to re-initiate the request, avoiding prolonged blocking of the parsing process due to a single node failure. Simultaneously, verification failure events are recorded to provide a basis for node status evaluation. This mechanism solves the problems of "low efficiency of full retransmission" and "no alternative solution for node failure" in traditional integrity verification, while ensuring the reliability of verification results through the strong security of the SM3 algorithm. For example, in large-volume production data parsing scenarios, fragment retransmission can reduce the amount of retransmitted data by more than 60%, significantly improving fault tolerance efficiency while ensuring the accuracy of parsing results, providing reliable assurance for subsequent data applications.

[0015] Furthermore, the output after adapting the parsing results according to the interface scenario parameters includes: if the data interaction frequency in the interface scenario parameters is high, converting the parsing results to Protocol Buffers binary format to reduce data transmission volume; if the data interaction frequency is medium, converting the parsing results to JSON format and compressing them using the Gzip compression algorithm with a compression level of 6; if the data interaction frequency is low, converting the parsing results to XML format and adding a digital signature to ensure the immutability of the data during transmission to the requesting end; simultaneously, adjusting the output caching strategy according to the data volume level, where the data volume is... Large data volumes are output in segments, with each segment not exceeding 10MB. For medium or small data volumes, data is output all at once. For high-frequency interaction scenarios, Protocol Buffers binary format is used to minimize data size and accelerate transmission speed, adapting to scenarios with high timeliness requirements, such as real-time control. For medium-frequency scenarios, JSON format with 6-level Gzip compression is selected to balance data readability and transmission efficiency, meeting the needs of regular business data interaction. For low-frequency scenarios, XML format with digital signature is used to ensure data format standardization and enhance tamper-proofness, adapting to non-real-time scenarios with high data security requirements. Meanwhile, by adjusting the output caching strategy according to the data volume level, large-volume data is output in segments to avoid transmission interruptions caused by excessive pressure in a single transmission. Medium and small-volume data are output at once to reduce process redundancy. For example, after 100MB of production image data is output in segments, the transmission success rate can be improved by more than 35%, while outputting device status data of less than 10MB at once can shorten the interaction time by 20%. This multi-dimensional adaptation solution not only solves the problem that the traditional fixed output method cannot take into account the efficiency and security of different scenarios, but also flexibly matches the format requirements of various requesting parties, making the subsequent application of parsing results smoother.

[0016] Furthermore, the method also includes: constructing a health assessment system for parsing nodes, sending health probe requests to each candidate parsing node every 15 minutes, and probing the node's CPU utilization, memory usage, network bandwidth utilization, and remaining storage space; and assessing the node's health status using a health calculation formula, wherein the health calculation formula is: In the formula: H is the health score of the j-th candidate node, with a value range of [0,1]. A score ≥ 0.8 is considered healthy, 0.5 ≤ score < 0.8 is considered sub-healthy, and a score < 0.5 is considered unhealthy. The CPU utilization (%) of the j-th node; The memory usage rate (%) of the j-th node; The network bandwidth utilization rate (%) of the j-th node; The percentage of remaining storage space for the j-th node is denoted as %; the candidate parsing node list is updated based on the health score, unhealthy nodes are removed, and the weight coefficient of sub-healthy nodes in the multi-dimensional parsing decision model is reduced.

[0017] Furthermore, the method also includes: establishing a full-link log system for the parsing process, recording content including request identifier, characteristics of the data to be parsed, interface scenario parameters, target parsing node information, encryption method, request sending time, parsing result receiving time, integrity verification result, and output adaptation method; performing statistical analysis on the log data weekly, and calculating the optimal parsing parameter combination for each scenario using a parsing efficiency optimization formula, wherein the parsing efficiency optimization formula is: In the formula: The optimal combination of parsing parameters includes target node selection, transmission protocol, encryption method, and compression strategy; θ is the set of combinations of analytical parameters; To analyze the total process time (in milliseconds); Parsing error rate (%); Resource consumption during the parsing process, including CPU usage time and network bandwidth usage, is normalized to a value within the range of [0,1]. The optimal parsing parameter combination is stored in a parameter recommendation library, and subsequent parsing requests with the same interface scenario parameters will prioritize the use of recommended parameters. Every 15 minutes, the CPU utilization, memory usage, network bandwidth utilization, and remaining storage space of candidate nodes are probed to comprehensively and timely grasp the node's operating load and resource status, avoiding parsing request failures or excessive latency due to node overload or resource exhaustion. A health score is quantified into a health score within the range of [0,1] using a health calculation formula, and divided into three levels: healthy, sub-healthy, and unhealthy, achieving accurate assessment and intuitive classification of node status. The candidate node list is updated based on the assessment results. Removing unhealthy nodes directly avoids sending requests to faulty or overloaded nodes, reducing the risk of parsing failure. The weight coefficient of sub-healthy nodes in the decision model is reduced, which neither wastes node resources nor fails to prioritize healthy nodes, ensuring the quality of parsing services. This mechanism solves the problem of lack of dynamic health monitoring and intelligent scheduling of nodes in existing technologies. For example, in the multi-node parsing scenario of the industrial Internet, it can reduce the failure rate of parsing requests by more than 40%. At the same time, by reasonably allocating request traffic, it can improve the throughput and response speed of the overall parsing system, providing users with more stable and efficient parsing services. Attached Figure Description

[0018] Figure 1 This is a flowchart of the steps in this solution. Detailed Implementation

[0019] Example 1 As attached Figure 1 As shown in the attached document Figure 1 As shown, a data interface identifier parsing method includes: acquiring identifier data to be parsed and interface scenario parameters, wherein the interface scenario parameters include data interaction frequency, data volume level, and data privacy sensitivity; constructing a multi-dimensional parsing decision model based on the interface scenario parameters, filtering target parsing nodes through the multi-dimensional parsing decision model, and determining the appropriate transmission protocol and data compression strategy; performing format verification and redundant field removal on the identifier data to be parsed, completing the conversion according to the data format requirements of the target parsing node, and encrypting it in combination with the encryption rules corresponding to the data privacy sensitivity; sending a parsing request carrying the encrypted identifier data to the target parsing node, receiving the parsing result and verifying its integrity, and outputting the parsing result to the requesting end after adapting its format and transmission method according to the interface scenario parameters.

[0020] The core principle of this solution is to construct a full-process data interface identifier resolution mechanism that integrates scene awareness, dynamic decision-making, and precise adaptation. By deeply integrating interface scene parameters with each stage of the resolution process, it achieves intelligent and personalized resolution services. First, by acquiring the identifier data to be resolved and interface scene parameters including data interaction frequency, data volume level, and data privacy sensitivity, it provides a foundation for subsequent decision-making. Second, based on the scene parameters, a multi-dimensional resolution decision model is constructed, comprehensively considering factors such as node performance, transmission efficiency, and security compliance to dynamically select the optimal resolution node and match it with suitable transmission protocols and data compression strategies, ensuring a precise match between resolution resources and scene requirements. Next, the identifier data undergoes format verification, redundancy removal, and format conversion, followed by differentiated encryption processing based on data privacy sensitivity, optimizing data transmission volume while ensuring data security. Finally, the resolution request is sent and the result is received. After integrity verification, the resolution result is adapted to the format and transmission method according to the scene parameters and finally output to the requesting end, achieving full-process scene-based adaptation from data input to result output.

[0021] Compared to existing technologies, which only match parsing nodes based on preset addresses and cannot adapt to diverse scenario requirements, this solution uses a multi-dimensional decision model to prioritize low-latency nodes for high-frequency interaction scenarios and high-bandwidth nodes for large-volume data scenarios. For example, in industrial real-time control scenarios, it can reduce parsing response latency. Secondly, data security protection is upgraded from "single encryption" to "sensitivity-level encryption." Existing technologies often use a uniform encryption strategy, resulting in insufficient security for highly sensitive data or redundant encryption for ordinary data. This solution uses dynamic key encryption for extremely sensitive data, which can resist replay attacks and key leakage. To mitigate risks, this solution employs lightweight encryption for ordinary data, reducing resource consumption. Thirdly, the parsing process has been upgraded from a "fixed process" to "full-link scenario adaptation." Existing technologies lack scenario-based optimization for data preprocessing and result output. This solution reduces data transmission volume by eliminating redundant fields and selects appropriate transmission protocols and compression strategies based on the chosen scenario. For example, in high-frequency reporting scenarios of IoT devices, data transmission efficiency can be improved while ensuring that the format and transmission method of the parsed results meet the requirements of the requesting end. This achieves a perfect balance between parsing efficiency, security, and scenario requirements, solving the technical pain point that existing technologies cannot simultaneously adapt to multiple scenarios.

[0022] The construction and node selection process of the multi-dimensional parsing decision model includes: determining parsing decision evaluation indicators, which include the node's historical parsing success rate, current node concurrency, average data transmission time, and node security compliance level; assigning dynamic weights to each evaluation indicator based on interface scenario parameters, wherein the weight of data interaction frequency is positively correlated with the average data transmission time, the weight of data volume level is negatively correlated with the node's current concurrency, and the weight of data privacy sensitivity is positively correlated with the node's security compliance level; calculating the decision score of each candidate parsing node using a weighted comprehensive scoring algorithm, and determining the candidate node with the highest score as the target parsing node. The formula for the weighted comprehensive scoring algorithm is: In the formula: S is the decision score of the j-th candidate parsing node, and its value ranges from [0,1]. Let be the dynamic weight of the m-th evaluation indicator, and satisfy . m=1 corresponds to the historical parsing success rate of the node, m=2 corresponds to the current concurrency of the node, m=3 corresponds to the average data transmission time, and m=4 corresponds to the security compliance level of the node. This represents the original value of the m-th evaluation metric for the j-th candidate node; The minimum value of the m-th evaluation index among all candidate nodes; The m-th evaluation metric is the maximum value among all candidate nodes; This normalization process, which normalizes the evaluation result of the m-th evaluation metric for the j-th candidate node, eliminates the imbalance in numerical weights caused by differences in the attributes of different evaluation metrics by unifying the units and value ranges. This provides a foundation for the accuracy of subsequent weighted comprehensive scoring. Metrics such as the node's historical parsing success rate (range 0%-100%), the node's current concurrency (range 0-10000 times / second), average data transmission time, and the node's security compliance level (range 1-5) have completely different original values ​​in terms of magnitude and unit. Directly substituting these values ​​into the scoring calculation can lead to an overemphasis on the impact of large-value metrics and a weakening of the role of small-value metrics. For example, a node's original concurrency value of 8000, far higher than other metrics, might dominate the scoring, while masking the influence of key metrics that truly reflect security and reliability, such as compliance level and parsing success rate. After normalization, the values ​​of all indicators are mapped to the range of [0,1]. The value of each indicator only represents its relative merits in that indicator dimension. For example, a normalized value of 0.8 means that the node is better than 80% of the candidate nodes in that indicator. When combined with dynamic weight calculation of the comprehensive score, it can be ensured that each evaluation indicator contributes its influence according to the preset weight. This avoids decision-making bias caused by differences in scale and accurately reflects the weight requirements of interface scenario parameters for each indicator. Ultimately, it achieves the scientific nature and scenario adaptability of the target parsing node selection.

[0023] The dynamic weight allocation process includes: pre-setting a correlation coefficient matrix between interface scenario parameters and evaluation indicators; data interaction frequency is divided into high frequency (≥100 times / minute), medium frequency (10-100 times / minute), and low frequency (<10 times / minute), with corresponding correlation coefficients of 0.35, 0.20, and 0.05 with average data transmission time, respectively; data volume is divided into large (≥100MB), medium (10-100MB), and small (<10MB), with corresponding correlation coefficients of 0.30, 0.15, and 0.05 with current node concurrency, respectively; data privacy sensitivity is divided into extremely high, high, medium, and low, with corresponding correlation coefficients of 0.40, 0.30, 0.20, and 0.10 with node security compliance level, respectively; and normalizing the correlation coefficients using a weight calibration formula to obtain the dynamic weights of each evaluation indicator. The weight calibration formula is as follows: In the formula: Let be the correlation coefficient between the m-th evaluation metric and the interface scenario parameters; Let be the inherent importance coefficient of the m-th evaluation indicator, with a value range of [0.8, 1.2]. The default inherent importance coefficients are: node security compliance level (1.2), node historical parsing success rate (1.1), average data transmission time (1.0), and current node concurrency (0.9). These can be adjusted according to the actual application scenario. The preset correlation coefficient matrix allows the weight allocation to directly respond to the differentiated needs of the interface scenario: for example, in high-frequency data interaction scenarios, the correlation coefficient of average data transmission time (0.35) is significantly higher than in low- and medium-frequency scenarios, ensuring the priority consideration of low-latency characteristics in high-frequency scenarios; when transmitting large volumes of data, the correlation coefficient of current node concurrency (0.30) is increased, avoiding parsing blockage caused by excessive node load; for highly privacy-sensitive data, the correlation coefficient of node security compliance level (0.40) is increased, strengthening the security of data transmission. Based on this, the weight calibration formula introduces inherent importance coefficients... This further optimizes the rationality of weight allocation: for example, the node security compliance level. The default value is 1.2, higher than other indicators. This reflects the fundamental role of security and compliance in industrial data interaction while preventing the security weight from being marginalized due to low scenario correlation coefficients. The current concurrency of the node... A value of 0.9 ensures load consideration while preventing it from excessively affecting the overall score of the parsing node. Finally, through normalization, the sum of the weights of each indicator is made up to 1, ensuring the effectiveness of the weighted overall score. This achieves an organic combination of scenario-driven requirements and the essential importance of the indicators, making the selection of parsing nodes more closely aligned with the complex scenario requirements in actual applications.

[0024] The encryption process, which combines encryption rules corresponding to data privacy sensitivity, includes: if the data privacy sensitivity is extremely high, using an SM4 block encryption + dynamic key derivation combination encryption scheme, wherein the dynamic key is generated using a key generation formula, which is: , In the formula: K is a dynamic encryption key with a length of 128 bits; T represents the current timestamp, accurate to milliseconds, in the format "YYYYMMDDHHMMSSmmm"; The ID is a unique identifier for the target parsing node, using the UUIDv4 format, and is 128 bits long; R is a random number, with a value range of [100000, 999999], and a new random number is generated each time the encryption is performed; ⊕ represents the XOR operation; The pre-shared root key is configured offline by the system administrator and is 256 bits long. HMAC-SM3 is a key derivation function based on the SM3 hash algorithm, used to convert concatenated raw data into a 128-bit key. For high data privacy sensitivity, the SM2 asymmetric encryption algorithm is used, encrypting the data using the public key of the target parser node. For medium data privacy sensitivity, the AES-256 symmetric encryption algorithm is used. For low data privacy sensitivity, the DES symmetric encryption algorithm is used. This encryption strategy adapts the encryption scheme according to the level of data privacy sensitivity: for extremely sensitive data, a combination of "SM4 block cipher + dynamic key derivation" is used, with the key generation formula... By integrating timestamps (T), unique node identifiers (ID), random numbers (R), and a pre-shared root key (Kroot), this approach ensures that the encryption key is unique and dynamically changes each time, effectively resisting security risks such as replay attacks and key leaks. Furthermore, the SM4 algorithm, as a national cryptographic standard, possesses high-strength encryption performance. For highly sensitive data, the SM2 asymmetric encryption algorithm uses a public-private key separation mechanism to guarantee identity authentication and data confidentiality during data transmission. Medium- and low-sensitivity data employ AES-256 and DES symmetric encryption algorithms respectively, reducing computational resource consumption and time latency during encryption and decryption while meeting basic security requirements. This hierarchical encryption strategy covers data security needs in different scenarios and optimizes overall encryption efficiency through differentiated design, making it more flexible and practical than traditional single encryption schemes.

[0025] The process of performing format validation and redundant field removal on the data to be parsed includes: obtaining the standard data format specifications supported by the target parsing node, which includes field names, data types, field lengths, and required / optional attributes; performing syntax validation on the original data to be parsed based on the specifications, marking abnormal fields that do not conform to the specifications, and if the abnormal field is a required field, providing a completion prompt to the requesting end; if the abnormal field is an optional field, automatically removing the field; filtering redundant fields using a field importance assessment model, which marks fields with a call frequency <5% as redundant based on their historical parsing frequency and the correlation between fields and parsing results, marking fields with a correlation <10% as redundant, determining the range of redundant fields, automatically removing redundant fields and retaining core data fields; and obtaining the standard data format specifications of the target parsing node, performing syntax validation on the data and processing abnormal fields, providing a completion prompt when required fields are abnormal to ensure the integrity of the data required for parsing; and automatically removing optional fields when they are abnormal to avoid invalid data interfering with the parsing process and reduce the risk of parsing failure from the source. Building upon this foundation, a field importance assessment model is used to filter redundant fields based on clear criteria such as "historical call frequency <5%" and "correlation with parsing results <10%". This accurately removes useless or low-value fields, such as redundant identifiers used only for local records and additional information unrelated to the parsing target, while retaining core data fields. This approach not only solves problems such as inconsistent original data formats, excessive invalid fields leading to unrecognizable parsing nodes, and excessively long parsing times, but also reduces bandwidth consumption and computational resource consumption of parsing nodes during data transmission. For example, for production data containing dozens of fields, redundancy removal can reduce the data volume, thereby shortening data transmission time and parsing node processing time. It also avoids format conflicts that may arise from redundant fields, making the parsing process smoother and more efficient.

[0026] The method further includes: generating a unique request identifier RequestID when sending a parsing request to the target parsing node, and starting a timeout monitoring mechanism. The timeout threshold is determined by a timeout calculation formula, which is: In the formula: This is the timeout threshold, in milliseconds. The average data transmission time of the target parsing node over the past hour was obtained through historical log statistics. The standard deviation of data transmission time for the target node over the past hour is used to cover fluctuations in transmission time; if in If no resolution result is received, the resolution request is resent to the candidate resolution node with the second-highest score, and the timeout event of the target resolution node is recorded. After accumulating 3 timeout events, the node is marked as "temporarily unavailable" and removed from the candidate node list. The node status is re-probeed every 30 minutes. The timeout threshold is determined by the formula. The calculation, combining the average transmission time and standard deviation of the target parsing node over the past hour, not only closely reflects the actual transmission performance of the node but also covers unexpected situations such as network fluctuations and changes in node load through the standard deviation. This avoids accidental timeout retries triggered by occasional delays and also prevents requests from being blocked for extended periods due to excessively high thresholds. When no result is received after a timeout, a request is automatically resent to the next highest-scoring candidate node to ensure uninterrupted parsing. After three timeouts, the node is marked as "temporarily unavailable" and removed. Simultaneously, the node status is periodically monitored, avoiding resource waste caused by frequently sending requests to faulty nodes and ensuring timely reinstatement of the node into the candidate list once it recovers. This mechanism solves the problems of inflexible timeout control and node fault tolerance in existing technologies. For example, in high-frequency data parsing scenarios in the industrial internet, it can reduce the parsing request failure rate by more than 40% while reducing invalid waiting time, ensuring the continuity and efficiency of the parsing service. It is particularly suitable for scenarios with high requirements for parsing timeliness and reliability, such as production control and equipment collaboration.

[0027] The process of receiving and verifying the integrity of the parsing result includes: the parsing result carries a checksum generated based on the SM3 algorithm; after receiving the parsing result, the original data portion is extracted and the SM3 hash value is recalculated; the recalculated hash value is compared with the checksum in the parsing result; if they match, the parsing result is considered complete; if they do not match, a retransmission request is sent to the target parsing node, carrying the identifier of the parsing result segment received in the first transmission, and only missing or erroneous segments are retransmitted; if the verification still does not match after two consecutive retransmissions, the process switches to the second-highest score candidate node to re-initiate the parsing request, and the verification failure event of the target parsing node is recorded. The parsing result carries a checksum based on the SM3 algorithm, which has high security and uniqueness. By recalculating the hash value and comparing it, it can accurately identify whether there are problems such as data loss or tampering during transmission, fundamentally preventing invalid or erroneous data from entering subsequent application processes. When inconsistencies arise during verification, instead of retransmitting all data directly, a retransmission request carrying a fragment identifier is sent to retrieve only the missing or erroneous fragments, significantly reducing bandwidth consumption and transmission time during retransmission. If two consecutive retransmissions still fail to obtain complete data, the request is switched to the next highest-scoring candidate node to re-initiate the request, avoiding prolonged blocking of the parsing process due to a single node failure. Simultaneously, verification failure events are recorded to provide a basis for node status evaluation. This mechanism solves the problems of "low efficiency of full retransmission" and "no alternative solution for node failure" in traditional integrity verification, while ensuring the reliability of verification results through the strong security of the SM3 algorithm. For example, in large-volume production data parsing scenarios, fragment retransmission can reduce the amount of retransmitted data by more than 60%, significantly improving fault tolerance efficiency while ensuring the accuracy of parsing results, providing reliable assurance for subsequent data applications.

[0028] The process of adapting the parsed results to interface scenario parameters and then outputting the results includes: if the data interaction frequency in the interface scenario parameters is high, converting the parsed results to Protocol Buffers binary format to reduce data transmission volume; if the data interaction frequency is medium, converting the parsed results to JSON format and compressing them using the Gzip compression algorithm with a compression level of 6; if the data interaction frequency is low, converting the parsed results to XML format and adding a digital signature to ensure the immutability of the data during transmission to the requesting end; and simultaneously adjusting the output caching strategy according to the data volume level, for large data volumes... Segmented output is adopted, with each segment not exceeding 10MB in size; data volumes of medium or hourly are output in one go. For high-frequency interaction scenarios, Protocol Buffers binary format is used to minimize data volume and accelerate transmission speed, adapting to scenarios with high timeliness requirements such as real-time control; for medium-frequency scenarios, JSON format with 6-level Gzip compression is selected to balance data readability and transmission efficiency, meeting the needs of regular business data interaction; for low-frequency scenarios, XML format is used with digital signatures added, ensuring data format standardization while enhancing tamper-proofness, adapting to non-real-time scenarios with high data security requirements. Meanwhile, by adjusting the output caching strategy according to the data volume level, large-volume data is output in segments to avoid transmission interruptions caused by excessive pressure in a single transmission. Medium and small-volume data are output at once to reduce process redundancy. For example, after 100MB of production image data is output in segments, the transmission success rate can be improved by more than 35%, while outputting device status data of less than 10MB at once can shorten the interaction time by 20%. This multi-dimensional adaptation solution not only solves the problem that the traditional fixed output method cannot take into account the efficiency and security of different scenarios, but also flexibly matches the format requirements of various requesting parties, making the subsequent application of parsing results smoother.

[0029] The method further includes: constructing a health assessment system for parsing nodes, sending health probe requests to each candidate parsing node every 15 minutes, and probing the node's CPU utilization, memory usage, network bandwidth utilization, and remaining storage space; and assessing the node's health status using a health calculation formula, wherein the health calculation formula is: In the formula: H is the health score of the j-th candidate node, with a value range of [0,1]. A score ≥ 0.8 is considered healthy, 0.5 ≤ score < 0.8 is considered sub-healthy, and a score < 0.5 is considered unhealthy. The CPU utilization (%) of the j-th node; The memory usage rate (%) of the j-th node; The network bandwidth utilization rate (%) of the j-th node; The percentage of remaining storage space for the j-th node is denoted as %; the candidate parsing node list is updated based on the health score, unhealthy nodes are removed, and the weight coefficient of sub-healthy nodes in the multi-dimensional parsing decision model is reduced.

[0030] Furthermore, the method also includes: establishing a full-link log system for the parsing process, recording content including request identifier, characteristics of the data to be parsed, interface scenario parameters, target parsing node information, encryption method, request sending time, parsing result receiving time, integrity verification result, and output adaptation method; performing statistical analysis on the log data weekly, and calculating the optimal parsing parameter combination for each scenario using a parsing efficiency optimization formula, wherein the parsing efficiency optimization formula is: In the formula: The optimal combination of parsing parameters includes target node selection, transmission protocol, encryption method, and compression strategy; θ is the set of combinations of analytical parameters; To analyze the total process time (in milliseconds); Parsing error rate (%); Resource consumption during the parsing process, including CPU usage time and network bandwidth usage, is normalized to a value within the range of [0,1]. The optimal parsing parameter combination is stored in a parameter recommendation library, and subsequent parsing requests with the same interface scenario parameters will prioritize the use of recommended parameters. Every 15 minutes, the CPU utilization, memory usage, network bandwidth utilization, and remaining storage space of candidate nodes are probed to comprehensively and timely grasp the node's operating load and resource status, avoiding parsing request failures or excessive latency due to node overload or resource exhaustion. A health score is quantified into a health score within the range of [0,1] using a health calculation formula, and divided into three levels: healthy, sub-healthy, and unhealthy, achieving accurate assessment and intuitive classification of node status. The candidate node list is updated based on the assessment results. Removing unhealthy nodes directly avoids sending requests to faulty or overloaded nodes, reducing the risk of parsing failure. The weight coefficient of sub-healthy nodes in the decision model is reduced, which neither wastes node resources nor fails to prioritize healthy nodes, ensuring the quality of parsing services. This mechanism solves the problem of lack of dynamic health monitoring and intelligent scheduling of nodes in existing technologies. For example, in the multi-node parsing scenario of the industrial Internet, it can reduce the failure rate of parsing requests by more than 40%. At the same time, by reasonably allocating request traffic, it can improve the throughput and response speed of the overall parsing system, providing users with more stable and efficient parsing services.

[0031] I. Implementation Prerequisites and Environmental Configuration (a) Hardware environment Request parsing requirements: CPU frequency ≥ 2.0GHz, memory ≥ 8GB, network bandwidth ≥ 100Mbps, supports TCP / IP protocol stack, and has hardware acceleration capabilities for data encryption / decryption (such as a security chip that supports Chinese cryptographic algorithms).

[0032] Resolution node server: CPU frequency ≥ 3.0GHz, memory ≥ 32GB, storage capacity ≥ 1TB, network bandwidth ≥ 1Gbps, supports concurrent connections ≥ 10000, equipped with redundant power supply and cooling system to ensure stable operation 24 / 7.

[0033] Network environment: latency ≤50ms, packet loss rate ≤0.1%, supports IPv4 / IPv6 dual protocol stack, and has security protection devices such as firewalls and intrusion detection to ensure data transmission security.

[0034] (ii) Software Environment Operating System: The requesting end supports Windows 10 / 11, Linux CentOS 7 and above; the resolving node supports Linux Ubuntu 20.04LTS, CentOS 8 and above.

[0035] Development and runtime environment: Java JDK 1.8 or above (or Python 3.8 or above), supports SM2, SM3, and SM4 national cryptographic algorithm libraries (such as BouncyCastle 1.72 or above), and has ProtocolBuffers, JSON, and XML format parsing toolkits, and Gzip compression tool (version ≥ 1.6).

[0036] Database: MySQL 8.0 or above or PostgreSQL 12 or above is used to store parsing node information, interface scenario parameter configuration, full-link log data, and optimal parsing parameter recommendation library.

[0037] (III) Pre-configured parameters Pre-shared root key (Kroot): A 256-bit random string generated offline by the system administrator, encrypted with AES-256 and stored in a secure storage area between the parsing node and the requesting end, accessible only to authorized processes.

[0038] Interface scenario parameter classification standard: Data interaction frequency: high frequency (≥100 times / minute), medium frequency (10-100 times / minute), low frequency (<10 times / minute).

[0039] Data size levels: Large (≥100MB), Medium (10-100MB), Small (<10MB).

[0040] Data privacy sensitivity: extremely high (e.g., confidential process parameters, core business data), high (e.g., user privacy data, equipment key information), medium (e.g., production statistics, equipment operating status), low (e.g., publicly available equipment models, product names).

[0041] Inherent importance coefficient of assessment metrics (δm): Node security compliance level Success rate of historical node parsing Average data transmission time Current concurrency of the node It can be adjusted according to industry needs.

[0042] Correlation coefficient matrix (γm): Preset in the system configuration file, as follows: Correlation coefficient matrix (γ) m Configuration table

[0043] II. Core Parameters and Formula Explanation (I) Definition of key parameters Multidimensional analysis of decision model parameters Evaluation metrics (m=1~4): m=1 is the node's historical parsing success rate (%), m=2 is the node's current concurrency (times / second), m=3 is the average data transmission time (ms), and m=4 is the node's security compliance level (level 1-5, with level 5 being the highest).

[0044] Dynamic weights ( The sum of the weights of the four evaluation indicators is 1, reflecting the importance of each indicator in different scenarios.

[0045] Decision Score (S): The overall score of the candidate node, with a value range of [0,1]. The higher the score, the better the node's suitability.

[0046] Normalization results ( Mapping the original values ​​of the indicators to the [0,1] interval eliminates dimensional differences. The formula is: ,in This represents the original value of the m-th indicator at the j-th node. This is the minimum value of the indicator. This is the maximum value of the indicator.

[0047] Encryption-related parameters Dynamic key (K): 128 bits, used for encryption of highly privacy-sensitive data, generated by HMAC-SM3 from timestamp (T), node ID, random number (R) and root key (Kroot).

[0048] Timestamp (T): The format is “YYYYMMDDHHMMSSmmm”, accurate to milliseconds, such as 20240520143025123.

[0049] Node ID: A 128-bit unique identifier in UUIDv4 format. For example, 550e8400-e29b-41d4-a716-446655440000.

[0050] Random number (R): 6-bit integer, ranging from [100000, 999999], regenerated after each encryption.

[0051] Timeout and health parameters Timeout threshold (T): unit ms, formula is ,in The average transmission time of the node over the past hour. This represents the standard deviation of transmission time.

[0052] Health score (H): The value ranges from [0,1], ≥0.8 is healthy, 0.5~0.8 is sub-healthy, and <0.5 is unhealthy.

[0053] Parsing efficiency optimization parameters: For the optimal parameter combination, The total time taken for the entire process, For error rate, Normalized resource consumption (0 for minimum, 1 for maximum).

[0054] (II) Calculation Examples of Core Formulas Formula for dynamic weight calculation (ωm): (m=1~4) Example: If the current scenario is "high-frequency interaction (γ3=0.35), small data volume (γ2=0.05), high privacy sensitivity (γ4=0.30)", and the node historical parsing success rate has no scenario correlation (γ1=0, default configuration), then:

[0055]

[0056]

[0057]

[0058]

[0059] Final weights:

[0060] Formula for calculating decision score (S): (m=1~4) Example: The normalized results of the four indicators of a candidate node are as follows: , =0.7、 =0.8、 =0.9, weight is , , , ,but:

[0061] Dynamic key calculation (K) formula: Example: T=20240520143025123, ID=550e8400-e29b-41d4-a716-446655440000, R=123456 XOR operation: T⊕ID⊕R, converts the three values ​​to binary and then performs a bitwise XOR operation. Input the HMAC-SM3 function: using Kroot as the key, perform a hash operation on the XOR result, and output a 128-bit key K.

[0062] Health score (H) calculation formula: Example: If a node has a CPU utilization of 60%, a memory utilization of 70%, a bandwidth utilization of 50%, and a storage space remaining of 80%, then: Determined to be unhealthy III. Detailed Implementation Steps (I) Step 1: Obtain the data to be parsed and the interface scenario parameters Data Acquisition: The requesting end submits identification data to be parsed, such as unique equipment identifiers and material codes, via API interface or file upload. Data formats supported include JSON, XML, and CSV.

[0063] Scenario parameter configuration: The requesting end specifies the interface scenario parameters, which can be entered manually or automatically recognized by the system: Data interaction frequency: The system automatically determines this by counting the number of requests in the past 5 minutes, or the user can manually select high frequency / medium frequency / low frequency.

[0064] Data volume level: The system calculates the size of the data file to be parsed or the total number of bytes in the fields, and automatically matches it to a large / medium / small level.

[0065] Data privacy sensitivity: Users can select extremely high / high / medium / low based on the data type, or the system can automatically determine the sensitivity level by matching keywords such as "confidential" or "key".

[0066] Data validation: The system checks whether the data to be parsed is empty and whether the scene parameters are complete. If any are missing, it returns a completion prompt to the requesting client.

[0067] (II) Step 2: Constructing a multi-dimensional analytical decision-making model and selecting nodes Candidate Node Acquisition: The system retrieves all candidate resolution nodes with a status of "available" from the node list and collects the raw values ​​of the four evaluation metrics for each node: Node historical resolution success rate: Calculate the number of successful resolutions for this node in the past 30 days / total number of requests × 100%.

[0068] Current node concurrency: Real-time query of the number of parsing requests currently being processed by the node (times / second).

[0069] Average data transmission time: The average data transmission time (ms) of each parsing operation of this node in the past hour is calculated.

[0070] Node security compliance level: assessed based on whether the node has passed Level 3 of the Information Security Protection System and whether it supports national cryptographic algorithms (Level 1-5).

[0071] Dynamic weight calculation: Following the method in the "Core Formula Calculation Example", and combining the current interface scenario parameters and correlation coefficient matrix, calculate the dynamic weights ω1~ω4 of the four evaluation indicators.

[0072] Indicator normalization: The original values ​​of the four evaluation indicators for each candidate node are normalized to obtain... (m=1~4).

[0073] Decision score calculation: The decision score S of each candidate node is calculated by a weighted comprehensive scoring algorithm. The scores are sorted from high to low. The node with the highest score is determined as the target parsing node, and the node with the second highest score is used as a backup node.

[0074] Matching transmission protocols and compression strategies: Determine the adaptation scheme based on interface scenario parameters. High-frequency interaction: Protocol Buffers binary format, uncompressed, prioritizing transmission speed.

[0075] Mid-frequency interaction: JSON format + Gzip compression, 6 compression levels, balancing speed and compression ratio.

[0076] Low-frequency interaction: XML format + digital signature, the signature is generated based on the SM3 algorithm and appended to the end of the data.

[0077] (III) Step 3: Data Preprocessing and Encryption Format validation: The system obtains the standard data format specifications supported by the target parsing node, including field names, data types, lengths, and required / optional attributes, and stores them in the database.

[0078] Perform syntax validation on the original data to be parsed: check whether the field names match, whether the data types are correct, such as whether numeric fields are integers / floating-point numbers, and whether the field length exceeds the limit.

[0079] Handling of abnormal fields: If a required field is abnormal, a prompt for completion will be sent to the requesting client, such as "Field 'Device Number' is a required field, please complete it"; if an optional field is abnormal, the field will be automatically removed.

[0080] Redundant field removal: The field importance assessment model is started, and the historical call frequency of each field is counted. The number of times the field is called by the parsing process in the past 30 days / the total number of parsing times × 100% and the correlation with the parsing results are calculated by Pearson correlation coefficient, with a value range of [0,1].

[0081] Mark redundant fields: Fields with a call frequency of <5% or a correlation of <10% are considered redundant and are automatically removed, while core data fields are retained.

[0082] Format conversion: Convert the preprocessed core data fields into a standard data format supported by the target parsing node.

[0083] Tiered encryption processing: Select the appropriate encryption scheme based on the data privacy sensitivity. Extremely high sensitivity: SM4 block encryption + dynamic key derivation. A 128-bit dynamic key K is generated according to the dynamic key calculation example. The data is encrypted with SM4 using K. The block length is 128 bits and the encryption mode is CBC.

[0084] High sensitivity: SM2 asymmetric encryption, obtains the public key of the target parsing node, stores it in the system in advance, and uses the public key to encrypt the data.

[0085] Medium sensitivity: AES-256 symmetric encryption, with a 128-bit fixed key preset by the system.

[0086] Low sensitivity: DES symmetric encryption, with a 64-bit fixed key preset by the system.

[0087] (iv) Step 4: Parsing Request Sending and Result Processing Request sending and timeout monitoring: Generate a unique request identifier RequestID (UUIDv4 format), and encapsulate it together with the encrypted data and request timestamp into a parsed request packet.

[0088] Calculate the timeout threshold T using the formula: Received, among which The average transmission time of the target resolution node over the past hour. The standard deviation of transmission time is calculated using historical log statistics.

[0089] Start the timeout monitoring mechanism: Send a parsing request to the target parsing node and start timing at the same time. If the parsing result is not received within time T, trigger timeout processing.

[0090] Timeout handling: Resend the request to the second highest-scoring backup node and record the timeout event of the target node; if there are 3 timeouts, mark the node as "temporarily unavailable" and remove it from the candidate node list. Send a health probe request every 30 minutes to check whether the node status has recovered.

[0091] Parsing result reception and integrity verification: After receiving the request, the target parsing node decrypts the data according to the encryption method, generates the parsing result after parsing, and calculates the integrity checksum (32 bytes) of the parsing result based on the SM3 algorithm. The parsing result and the checksum are then returned to the requesting end.

[0092] After receiving the data, the requesting end extracts the original parsed result, recalculates the SM3 hash value, and compares it with the returned checksum. Matching results: The analysis result is determined to be complete, proceed to the next step of processing.

[0093] Inconsistent comparison: Send a retransmission request to the target node, carrying the identifier of the result segment received the first time, such as the data segment number, and only retransmit the missing / erroneous segment; if the two consecutive retransmissions are still inconsistent, switch to the backup node to re-initiate the request, and record the verification failure event of the target node.

[0094] Result adaptation output: Format adaptation: Convert the parsing results according to the transmission protocol and format determined in step 2, such as converting high-frequency scenarios to Protocol Buffers format.

[0095] Output caching strategy adaptation: For large data volumes: output in segments, each segment ≤ 10MB, sent sequentially, each segment carrying a segment number, such as "1 / 5" or "2 / 5", which is then assembled by the requesting end.

[0096] For medium / small data volumes: output all data at once, directly returning the complete parsing results to the requesting client.

[0097] Feedback: After the requesting end receives the parsing result, the system returns a success message. If the parsing fails, it returns the reason for the failure, such as "node timeout" or "data verification failed".

[0098] (v) Step 5: Node health monitoring and parameter optimization Node health assessment: Every 15 minutes, the system sends a health probe request to all candidate resolution nodes to obtain the CPU utilization, memory usage, network bandwidth utilization, and remaining storage space percentage of each node.

[0099] Calculate the health score H for each node using the health calculation formula, and update the candidate node list: Healthy (H≥0.8): Retained in the list with unchanged weighting coefficient.

[0100] Sub-health (0.5≤H<0.8): Retained in the list, but its weight coefficient is reduced in the multidimensional decision model (multiplied by a decay coefficient of 0.8).

[0101] Unhealthy (H<0.5): Remove from the candidate node list and re-probe every 30 minutes.

[0102] End-to-end logging: The system establishes a logging system to record the entire process information for each parsing step. Basic information: RequestID, requesting IP, request time, response time.

[0103] Data information: characteristics of the data to be parsed (data volume, privacy sensitivity), interface scenario parameters, and number of redundant fields to be removed.

[0104] Node information: Target parsing node ID, health score, and backup node ID (if used).

[0105] Security information: encryption method, integrity verification result (success / failure).

[0106] Output information: parsing result format, transmission method, and whether to output in segments.

[0107] Parsing parameter optimization: The system automatically compiles end-to-end log data weekly. For different interface scenario parameter combinations, such as "high frequency + large volume + high sensitivity" or "medium frequency + small to medium volume + medium sensitivity", the optimal parsing parameter combination is calculated using a parsing efficiency optimization formula. : .

[0108] The optimal parameter combination, target node selection preference, transmission protocol, encryption method, and compression strategy are stored in the parameter recommendation library. Subsequent parsing requests for the same scenario will prioritize the use of the recommended parameters to improve parsing efficiency.

[0109] This is merely an embodiment of the present invention. Commonly known structures and characteristics of the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the filing date or priority date, are aware of all prior art in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the description of specific embodiments in the specification can be used to interpret the content of the claims.

Claims

1. A data interface identification resolution method, characterized by, The method comprises the following steps: Obtain the identification data to be parsed and interface scenario parameters, which include data interaction frequency, data volume level, and data privacy sensitivity; Construct a multi-dimensional parsing decision model based on the interface scenario parameters, filter the target parsing node through the multi-dimensional parsing decision model, and determine the adaptive transmission protocol and data compression strategy; Perform format checking and redundant field elimination on the identification data to be parsed, convert according to the data format requirements of the target parsing node, encrypt according to the encryption rules corresponding to the data privacy sensitivity, send the parsing request carrying the encrypted identification data to the target parsing node, receive the parsing result and verify the integrity, and output the result to the request end after format and transmission mode adaptation according to the interface scenario parameters.

2. The method of claim 1, wherein, The construction and node screening process of the multi-dimensional resolution decision model comprises: determining resolution decision evaluation indexes, the evaluation indexes comprising a node historical resolution success rate, a node current concurrency, data transmission average time consumption and a node safety compliance level; assigning dynamic weights to each evaluation index based on interface scene parameters, wherein the weights of data interaction frequency and data transmission average time consumption are positively correlated, the weights of data volume level and node current concurrency are negatively correlated, and the weights of data privacy sensitivity and node safety compliance level are positively correlated; calculating the decision score of each candidate resolution node through a weighted comprehensive score algorithm, and determining the candidate node with the highest score as the target resolution node, and the formula of the weighted comprehensive score algorithm is: In the formula, S is the decision score of the jth candidate parsing node, with a value range of [0, 1]; is the dynamic weight of the mth evaluation index, and satisfies , m=1 corresponds to the node historical analysis success rate, m=2 corresponds to the node current concurrency, m=3 corresponds to the data transmission average time consumption, and m=4 corresponds to the node safety compliance level. original value of the mth evaluation index of the jth candidate node; minm is the minimum value of the mth evaluation index among all candidate nodes; max value of the mth evaluation index in all candidate nodes; is the normalized result of the mth evaluation index of the jth candidate node, used to eliminate the dimensional differences of different indexes.

3. The method of claim 2, wherein, The dynamic weight allocation process comprises: preset interface scene parameter and evaluation index correlation coefficient matrix, data interaction frequency is divided into high frequency (≥100 times / min), medium frequency (10-100 times / min), low frequency (<10 times / min), and the corresponding correlation coefficient with the average time of data transmission is 0.35, 0.20 and 0.05 respectively; the data volume level is divided into large (≥100 MB), medium (10-100 MB) and small (<10 MB), and the corresponding correlation coefficient with the current concurrency of the node is 0.30, 0.15 and 0.05 respectively; the data privacy sensitivity is divided into extremely high, high, medium and low, and the corresponding correlation coefficient with the node security compliance level is 0.40, 0.30, 0.20 and 0.10 respectively; the correlation coefficient is normalized by a weight calibration formula to obtain the dynamic weight of each evaluation index, and the weight calibration formula is: In the formula: is the correlation coefficient of the mth evaluation index and the interface scene parameter; The intrinsic importance coefficient of the mth evaluation index, the value range is [0.8, 1.2], wherein the intrinsic importance coefficient of the node security compliance level is 1.2 by default, the historical analysis success rate of the node is 1.1 by default, the average time consumption of data transmission is 1.0 by default, the current concurrency of the node is 0.9 by default, and can be adjusted according to the actual application scene.

4. The method of claim 1, wherein, The encryption rule corresponding to the data privacy sensitivity is encrypted, including: if the data privacy sensitivity is extremely high, an SM4 block encryption + dynamic key derivation combined encryption scheme is used, and a dynamic key is generated through a key generation formula, wherein the key generation formula is: , wherein: K is a dynamic encryption key with a length of 128 bits; T is the current timestamp, accurate to the millisecond level, in the format "YYYYMMDDHHMMSSmmm"; ID is the unique identifier of the target parsing node, in the UUIDv4 format, with a length of 128 bits; R is a random number, with a value range of [100000, 999999], and a new random number is generated each time; ⊕ is the XOR operation; Pre-shared root key, configured by system administrator offline, length is 256 bits; HMAC-SM3 is a key derivation function based on the SM3 hash algorithm, which is used to convert the concatenated original data into a 128-bit key; if the data privacy sensitivity is high, use the SM2 asymmetric encryption algorithm to encrypt the data using the public key of the target parsing node; if the data privacy sensitivity is medium, use the AES-256 symmetric encryption algorithm; if the data privacy sensitivity is low, use the DES symmetric encryption algorithm.

5. The method of claim 1, wherein, The method comprises the following steps: Obtain the identification data to be parsed and interface scenario parameters, which include data interaction frequency, data volume level, and data privacy sensitivity; Construct a multi-dimensional parsing decision model based on the interface scenario parameters, filter the target parsing node through the multi-dimensional parsing decision model, and determine the adaptive transmission protocol and data compression strategy; Perform format checking and redundant field elimination on the identification data to be parsed, convert according to the data format requirements of the target parsing node, encrypt according to the encryption rules corresponding to the data privacy sensitivity, send the parsing request carrying the encrypted identification data to the target parsing node, receive the parsing result and verify the integrity, and output the result to the request end after format and transmission mode adaptation according to the interface scenario parameters. S is the decision score of the jth candidate parsing node, with a value range of [0, 1]; K is a dynamic encryption key with a length of 128 bits; T is the current timestamp, accurate to the millisecond level, in the format "YYYYMMDDHHMMSSmmm"; ID is the unique identifier of the target parsing node, in the UUIDv4 format, with a length of 128 bits; R is a random number, with a value range of [100000, 999999], and a new random number is generated each time; ⊕ is the XOR operation; HMAC-SM3 is a key derivation function based on the SM3 hash algorithm, which is used to convert the concatenated original data into a 128-bit key; if the data privacy sensitivity is high, use the SM2 asymmetric encryption algorithm to encrypt the data using the public key of the target parsing node; if the data privacy sensitivity is medium, use the AES-256 symmetric encryption algorithm; if the data privacy sensitivity is low, use the DES symmetric encryption algorithm. The method comprises the following steps: Obtain the identification data to be parsed and interface scenario parameters, which include data interaction frequency, data volume level, and data privacy sensitivity; Construct a multi-dimensional parsing decision model based on the interface scenario parameters, filter the target parsing node through the multi-dimensional parsing decision model, and determine the adaptive transmission protocol and data compression strategy; Perform format checking and redundant field elimination on the identification data to be parsed, convert according to the data format requirements of the target parsing node, encrypt according to the encryption rules corresponding to the data privacy sensitivity, send the parsing request carrying the encrypted identification data to the target parsing node, receive the parsing result and verify the integrity, and output the result to the request end after format and transmission mode adaptation according to the interface scenario parameters. S is the decision score of the jth candidate parsing node, with a value range of [0, 1]; K is a dynamic encryption key with a length of 128 bits; T is the current timestamp, accurate to the millisecond level, in the format "YYYYMMDDHHMMSSmmm"; ID is the unique identifier of the target parsing node, in the UUIDv4 format, with a length of 128 bits; R is a random number, with a value range of [100000, 999999], and a new random number is generated each time; ⊕ is the XOR operation; HMAC-SM3 is a key derivation function based on the SM3 hash algorithm, which is used to convert the concatenated original data into a 128-bit key; if the data privacy sensitivity is high, use the SM2 asymmetric encryption algorithm to encrypt the data using the public key of the target parsing node; if the data privacy sensitivity is medium, use the AES-256 symmetric encryption algorithm; if the data privacy sensitivity is low, use the DES symmetric encryption algorithm.

6. The method of claim 1, wherein, The method further comprises: when sending the resolution request to the target resolution node, generating a unique request identification RequestID, and starting a timeout monitoring mechanism, wherein a timeout threshold is determined by a timeout calculation formula, and the timeout calculation formula is: In the formula, timeout is a timeout threshold, in milliseconds; To resolve the target node for nearly 1 hour average data transmission time, through the history log statistics; To analyze the standard deviation of data transmission time within 1 hour for the target resolution node, to cover the transmission time fluctuation; if no resolution result is received within the analysis request is re-sent to the candidate resolution node with the second highest score, and the timeout event of the target resolution node is recorded; after 3 timeout events, the node is marked as "temporary unavailable" and removed from the candidate node list, and the node state is re-detected every 30 minutes.

7. The method of claim 1, wherein, The receiving and analyzing the result and verifying the integrity comprises: carrying an integrity check value Checksum generated based on an SM3 algorithm in the result, after receiving the result, extracting the original data part and recalculating an SM3 hash value; comparing the recalculated hash value with the check value in the result, if consistent, determining that the result is complete; if inconsistent, sending a retransmission request to the target analysis node, the retransmission request carrying the first received analysis result segment identifier, only retransmitting the missing or incorrect segment; if the check is still inconsistent after two consecutive retransmissions, switching to the second highest score candidate node to reinitiate the analysis request, and recording the check failure event of the target analysis node.

8. The method of claim 1, wherein, The outputting the result after adapting the result according to the interface scene parameter comprises: if the data interaction frequency in the interface scene parameter is high, converting the result into a ProtocolBuffers binary format to reduce the data transmission volume; if the data interaction frequency is medium, converting the result into a JSON format and compressing the result by using a Gzip compression algorithm, and setting the compression level to 6 levels; if the data interaction frequency is low, converting the result into an XML format and adding a digital signature to ensure the non-tamperability of the data in the process of transmission to the request end; and simultaneously adjusting the output cache strategy according to the data volume level, using segmented output when the data volume is large, and the size of each segment of data is not more than 10 MB; using one-time output when the data volume is medium or small.

9. The method of claim 1, wherein, The method further comprises: constructing an analysis node health degree evaluation system, sending a health probe request to each candidate analysis node every 15 minutes, and the probe content comprises node CPU usage, memory occupation rate, network bandwidth utilization rate and storage remaining space; evaluating the node health state by using a health degree calculation formula, the health degree calculation formula is: In the formula: H is the health degree score of the jth candidate node, the value range is [0, 1], the score is greater than or equal to 0.8, the node is determined to be healthy, the score is greater than or equal to 0.5 and less than 0.8, the node is determined to be sub-healthy, and the score is less than 0.5, the node is determined to be unhealthy. CPU usage (%) for the jth node; Memory usage (%) for the jth node; Network bandwidth utilization (%) for the jth node; Storage remaining space proportion (%) of the jth node; update the candidate analysis node list according to the health score, eliminate unhealthy nodes, and reduce the weight coefficient of sub-healthy nodes in the multi-dimensional analysis decision model.

10. The method of claim 1, wherein, The method further comprises: establishing a whole-link log system of the analysis process, and the recording content comprises a request identifier, data characteristics to be analyzed, interface scene parameters, target analysis node information, an encryption mode, a request sending time, an analysis result receiving time, an integrity check result and an output adaptation mode; statistically analyzing the log data every week, and calculating the optimal analysis parameter combination in each scene by using an analysis efficiency optimization formula, the analysis efficiency optimization formula is: In the formula: θopt is the optimal analytic parameter combination, including target node selection, transmission protocol, encryption method and compression strategy; θ is the analytic parameter combination set; is the analytic whole-process time consumption (milliseconds); is the analytic error rate (%); is the resource consumption of the analytic process, including CPU occupation time and network bandwidth occupation, normalized to [0, 1] interval value; the optimal analytic parameter combination is stored to the parameter recommendation library, and the subsequent analytic request of the same interface scene parameter preferentially calls the recommended parameter.

Images