{"id":29644235,"url":"https://github.com/ccoskrnl/cof-python","last_synced_at":"2025-08-09T03:56:33.283Z","repository":{"id":302993219,"uuid":"1013732104","full_name":"ccoskrnl/cof-python","owner":"ccoskrnl","description":"Code Optimization Framework","archived":false,"fork":false,"pushed_at":"2025-07-18T03:23:14.000Z","size":398,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-07-18T07:20:59.949Z","etag":null,"topics":["compiler","data-flow-analysis","python"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ccoskrnl.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-07-04T11:25:15.000Z","updated_at":"2025-07-18T03:23:18.000Z","dependencies_parsed_at":"2025-07-05T15:45:50.928Z","dependency_job_id":null,"html_url":"https://github.com/ccoskrnl/cof-python","commit_stats":null,"previous_names":["ccoskrnl/cof-python"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/ccoskrnl/cof-python","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ccoskrnl%2Fcof-python","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ccoskrnl%2Fcof-python/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ccoskrnl%2Fcof-python/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ccoskrnl%2Fcof-python/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ccoskrnl","download_url":"https://codeload.github.com/ccoskrnl/cof-python/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ccoskrnl%2Fcof-python/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":266404940,"owners_count":23923490,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-07-21T11:47:31.412Z","response_time":64,"last_error":null,"robots_txt_status":null,"robots_txt_updated_at":null,"robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["compiler","data-flow-analysis","python"],"created_at":"2025-07-22T00:31:03.908Z","updated_at":"2025-08-09T03:56:33.225Z","avatar_url":"https://github.com/ccoskrnl.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Code Optimization Framework\n\n**这并不是一个可用的工具，只是出于学习的目的**\n\n\n\n## 介绍\n\n由Python实现的程序分析框架，通过对传入的MIR进行解析转化成控制流图，对控制流图进行静态分析为后续的优化或者反编译做指导。已经支持数据流分析框架，并包含一些经典的数据流分析问题，如到达定值，活跃变量分析等。同时支持使用符号执行的稀疏条件的常量传播。后续会逐渐添加更强大的分析算法。\n\n在可视化方面，框架内置了一个根据PyQt6开发的控制流图可视化器。在布局方面，可视化器使用 [Reingold-Tilford算法](https://github.com/ccoskrnl/tree_layout_demo) 对控制流图进行初始的布局，但每个基本块都被设置为可拖动的节点。在控制流图边绘制这方面并没有进行很好的实现，对于特殊边的处理效果更是混乱。\n\n\n\n## 特性\n\n### MIR\n\n框架使用自定义的MIR\n\n```\n# test/example.ir\n\n    %entry\n\t%init n\n\n    k := false\n    i := 1\n    j := 2\nL1:\n    cond1 := i \u003c= n\n    %if cond1 %goto \u0026L2\n    %if k %goto \u0026L4\n    i := i + 1\n    %goto \u0026L5\nL2:\n    j := j * 2\n    k := %true\n    cond2 := i \u003c= 5\n    %if cond2 %goto \u0026L3\n    i := i - 1\n    %goto \u0026L1\nL3:\n    i := i + 1\n    %goto \u0026L1\nL4:\n    printf ( i j k )\n    %goto \u0026L5\n\nL5:\n    %exit\n```\n\n| **元素类型**              | **示例代码**               | **功能说明**                                                 |\n| ------------------------- | -------------------------- | ------------------------------------------------------------ |\n| **基本块（Basic Block）** | `%entry`、`L1:`、`L2:`     | 以标签定义代码块，包含顺序语句和终结指令；`%entry` 为入口块。 |\n| **变量赋值**              | `k := false`、`i := 1`     | 用 `:=` 初始化或更新变量，支持布尔值、整数等基本类型。       |\n| **条件跳转**              | `%if cond1 %goto \u0026L2`      | 若条件成立（如 `cond1` 为真），跳转到目标块（`L2`）。        |\n| **无条件跳转**            | `%goto \u0026L1`                | 强制跳转到指定块，用于循环或分支合并。                       |\n| **算术运算**              | `j := j * 2`、`i := i + 1` | 支持加减乘除等运算，直接操作变量。                           |\n| **函数调用**              | `printf(i j k)`            | 嵌入外部函数，参数以空格分隔。                               |\n| **程序边界标记**          | `%exit`                    | 标识程序终止点，辅助编译器优化资源释放。                     |\n\n### 控制流图可视化\n\n上面的MIR在被转换为控制流图后可以被控制流图可视化器绘制。效果如下图所示：\n\n![cfg_demo](./tmp/readme_ref_img01.png)\n\n### 数据流框架\n\n在数据流分析算法的实现中，框架并没有遵循严格意义上的“交半格”或者“并半格“，而是统一使用”并半格“的概念。这样的好处是只要我们选择合适的meet操作以及初始值和安全值就可以统一的对各种常规的数据流问题进行求解。框架的参数如下：\n\n```python\nclass DataFlowAnalysisFramework(Generic[T, B]):\n    \"\"\"Generic data flow analysis framework supporting forward and backward analyses.\n\n    Attributes:\n        cfg: Control flow graph of the program\n        lattice: Abstract semilattice defining the value domain and operations\n        transfer: Transfer function implementation for basic blocks\n        direction: Analysis direction ('forward' or 'backward')\n        merge: Function for merging values at control flow joins\n        on_state_change: Optional callback for state change events\n    \"\"\"\n    def __init__(self,\n                 cfg: 'ControlFlowGraphForDataFlowAnalysis',\n                 lattice: 'Semilattice[T]',\n                 transfer: 'TransferFunction[T, B]',\n                 direction: str = 'forward', # 'forward' or 'backward'\n                 init_value: T = None,\n                 safe_value: T = None,\n                 merge_operation: Optional[Callable[[Iterable[T]], T]] = None,\n                 on_state_change: Optional[Callable[[B, T, T], None]] = None):\n        \"\"\"\n        Initialize the data flow analysis framework.\n\n        Args:\n            cfg: Control flow graph\n            lattice: Semilattice defining the value domain\n            transfer: Transfer function implementation\n            direction: Analysis direction ('forward' or 'backward')\n            init_value: Initial value for the entry/exit block\n            safe_value: Safe value for other blocks\n            merge_operation: Function for merging values (default: lattice.meet)\n            on_state_change: Callback for state change events\n        \"\"\"\n```\n\n### 稀疏条件的常量传播\n\n稀疏条件的常量传播（Sparse Conditional Constant Propagation，SCCP）是普通常量传播的增强版，结合了**常量传播**与**条件分支分析**，能更精确地处理控制流和稀疏数据流。框架首先通过迭代必经边界方法将流图转换为最小的SSA形式，并对基本块进行扁平化（或原子化），并使用流图的边和SSA边来传递信息实现程序的符号执行。\n\n\n\n## 路标\n\n- [x] 数据流分析框架\n- [x] 稀有条件常量传播(符号执行)\n- [ ] 循环依赖分析\n- [ ] 别名分析\n- [ ] 高级数据流分析\n- [ ] 控制树规约\n\n\n\n## License\n\n无","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fccoskrnl%2Fcof-python","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fccoskrnl%2Fcof-python","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fccoskrnl%2Fcof-python/lists"}