A low-coupling high-parallel modular software development method

By employing a modular software development approach characterized by low coupling and high parallelism, through system decomposition, module partitioning, batch parallel development, and global optimization, the problems of high coupling and low parallel efficiency in modular software development are solved, enabling rapid development and easy maintenance of software systems.

CN122363684APending Publication Date: 2026-07-10薛梁

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
薛梁
Filing Date
2026-05-22
Publication Date
2026-07-10

Smart Images

  • Figure FT_1
    Figure FT_1
Patent Text Reader

Abstract

The application discloses a low-coupling high-parallel modular software development method, and relates to the technical field of software development.The method first disassembles a software system into minimum function unit modules and defines interfaces and dependencies;then, topological sorting is used to divide batches for parallel development;after each batch is completed, interface standardization, cycle dependency elimination and global optimization for decoupling are performed;the cycle is iterated until module development is completed, and a complete system is obtained through system integration, end-to-end testing, gray release, old system compatibility and global monitoring deployment.The application reduces module coupling degree, improves parallel development efficiency, shortens the development cycle, guarantees system scalability and maintainability, and is suitable for various software rapid development scenarios.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of software development technology, and in particular to a modular software development method. Background Technology

[0002] As software systems grow in scale, traditional waterfall development struggles to keep up with rapid iteration demands. Existing modular development suffers from issues such as unreasonable module division, lack of global optimization in batch development, and inconsistent interfaces, leading to frequent compatibility failures, extended development cycles, and decreased maintainability during the integration phase. Some batch development models focus only on single-batch development, neglecting global collaborative optimization between batches. This leads to a continuous increase in module coupling as development progresses, affecting system stability. Furthermore, the lack of canary releases and compatibility adaptation mechanisms during system integration results in high switching risks. Therefore, a modular software development methodology that can address these shortcomings is urgently needed. Summary of the Invention Summary of the Invention

[0003] (a) Technical problems to be solved This invention aims to overcome the shortcomings of existing technologies, such as unreasonable module division, lack of global optimization in batch development, high integration risk, high coupling, and low parallel efficiency, and provides a modular software development method with low coupling and high parallelism. (II) Technical Solution To achieve the above objectives, this invention provides a modular software development method with low coupling and high parallelism, comprising the following steps: System decomposition and module division: The software system to be developed is decomposed into several minimum functional unit modules, and the interface specifications of each module and the dependencies between modules are defined; Batch parallel development: Based on topological sorting analysis of module dependencies, modules are divided into multiple batches according to development priorities, and each batch of modules is developed in parallel. Batch global optimization: After each batch of module development is completed, global optimization is performed on the developed modules, including interface standardization verification, elimination of circular dependencies between modules, and adjustment of module coupling. Iterative loop: Repeatedly execute the batch parallel development and batch global optimization steps until all batch module development is completed; System integration and deployment assurance: Perform top-level system integration, and sequentially complete end-to-end full-process testing, gray-scale release control, old system compatibility adaptation and global monitoring system deployment to obtain a complete software system. Furthermore, the minimum functional unit module is a functional unit with independent input and output and capable of independent compilation and testing. Furthermore, the interface specifications include data format specifications, communication protocol specifications, and error code specifications. Furthermore, the gray-scale release control involves gradually increasing the release volume according to user groups or traffic ratios, monitoring operational indicators in real time, and quickly rolling back in case of anomalies. Furthermore, the old system compatibility and adaptation adopts a dual-write and dual-read mechanism to ensure data consistency between the old and new systems. (III) Beneficial Effects Compared with the prior art, the present invention has the following beneficial effects: Low coupling: Globally optimize interfaces, dependencies, and coupling in batches to eliminate circular dependencies, reduce module coupling, and reduce integration failures; High parallelism: Based on topological sorting, batch parallel development improves development efficiency and shortens the development cycle by more than 40% compared with traditional methods; Risks are manageable: During the integration phase, canary releases and compatibility adaptations with legacy systems are used to reduce the risk of system switching. Easy to maintain and expand: The minimal functional unit module design, combined with global monitoring, ensures the maintainability and scalability of the system. Attached Figure Description Figure 1 is a flowchart of a modular software development method with low coupling and high parallelism provided in an embodiment of the present invention. The corresponding relationships of each step in Figure 1 are as follows: S1: System decomposition and module division, used to break down the software system to be developed into the smallest functional unit modules, and define interface specifications and dependencies; S2: Batch parallel development, used to divide development batches based on topological sorting and implement them in parallel; S3: Batch global optimization, used for interface standardization, eliminating circular dependencies and adjusting module coupling; S4: Iterate and judge; if not all batches are completed, return to S2 to continue development; if completed, proceed to system integration. S5: System integration and deployment assurance, including end-to-end full-process testing, canary release management, compatibility and adaptation of legacy systems, and global monitoring and deployment, ultimately forming a complete software system. Detailed Implementation The present invention will be further described in detail below with reference to specific embodiments. Example 1 A modular software development method with loose coupling and high parallelism, taking e-commerce platform development as an example, includes the following steps: S1. System Decomposition and Module Division: The e-commerce platform is decomposed into the smallest functional unit modules such as user management, product management, order management, payment management, and logistics management. Each module has independent input and output and can be independently compiled and tested. Interface specifications are defined: the data format adopts JSON, the communication protocol adopts HTTP / 2, and the error code is a 6-digit unified code. Dependencies are sorted out, such as order management depending on user management and product management, and payment management depending on order management. S2. Batch Parallel Development: Based on topological sorting analysis of dependencies, user management and product management without dependencies are divided into the first batch, and two groups are arranged for parallel development; order management and payment management are divided into the second batch, and two groups are arranged for parallel development; logistics management and auxiliary modules are divided into the third batch, and one group is arranged for development. S3. Batch Global Optimization: After the first batch of development is completed, verify whether the interface conforms to JSON, HTTP / 2, and unified error code specifications, and correct non-compliant interfaces; sort out dependencies, identify and eliminate circular dependencies; adjust coupling, and change strongly coupled calls between modules to standard interface calls. After the second and third batches of development are completed, repeat the above global optimization operations. S4. Iterative Loop: Repeatedly develop in batches in parallel and optimize globally, dynamically adjusting the subsequent batch divisions and priorities during iterations until all modules are developed. S5. System Integration and Deployment Assurance: Integrate all modules to complete top-level integration; conduct end-to-end testing in a simulated production environment to verify the collaboration and data flow of the entire process of registration, browsing, ordering, payment, and logistics; gradually increase traffic in stages (10%, 30%, 50%, 100%), monitor response time and error rate, and quickly roll back from anomalies; enable dual-write and dual-read functionality during the initial deployment to ensure data consistency, and decommission the old system after one month of stability; deploy a performance, anomaly, and business monitoring system covering all modules, collect data in real time, and automatically issue alerts. The result is a loosely coupled, highly stable, and easily maintainable e-commerce platform system.

Claims

1. A modular software development method with low coupling and high parallelism, characterized in that, Includes the following steps: System decomposition and module division: The software system to be developed is decomposed into several minimum functional unit modules, and the interface specifications of each module and the dependencies between modules are defined; Batch parallel development: Based on topological sorting analysis of module dependencies, modules are divided into multiple batches according to development priorities, and each batch of modules is developed in parallel. Batch global optimization: After each batch of module development is completed, global optimization is performed on the developed modules, including interface standardization verification, elimination of circular dependencies between modules, and adjustment of module coupling. Iterative loop: Repeatedly execute the batch parallel development and batch global optimization steps until all batch module development is completed; System integration and deployment assurance: Perform top-level system integration, and sequentially complete end-to-end full-process testing, gray-scale release control, old system compatibility adaptation and global monitoring system deployment to obtain a complete software system.

2. The method according to claim 1, characterized in that, The minimum functional unit module is a functional unit with independent input and output, and capable of independent compilation and testing.

3. The method according to claim 1, characterized in that, The interface specifications include data format specifications, communication protocol specifications, and error code specifications.

4. The method according to claim 1, characterized in that, The gray-scale release control involves gradually increasing the release volume according to user groups or traffic ratios, monitoring operational indicators in real time, and quickly rolling back in case of anomalies.

5. The method according to claim 1, characterized in that, The old system compatibility and adaptation adopts a dual-write and dual-read mechanism to ensure data consistency between the old and new systems.