An efficient program parsing method and system based on file dependencies

Abstract

The invention provides an efficient program parsing method and system based on a file dependency relationship. The method comprises the following steps that: S1: carrying out segmentation processing on a source file, and obtaining a corresponding preprocessed unit set; and S2: executing the following processing on each preprocessed unit in the preprocessed unit set: if the type of the preprocessedunit is confirmed as a file inclusion instruction in a preprocessed instruction, extracting a header file in the preprocessed unit, linking the abstract syntax tree of the header file to the abstractsyntax tree of the source file if a confirmation result shows that the abstract syntax tree obtained through header file pre-parsing is in the presence, and executing S1 and S2 to parse the header file as the source file if the confirmation result shows that the abstract syntax tree obtained through header file pre-parsing is in the absence. By use of the method, whether the header file is parsedin advance or not is judged, the abstract syntax tree obtained by pre-parsing is linked to the abstract syntax tree of the source file, the repeated parsing of the same header file is avoided, and program parsing time is shortened.

Description

technical field [0001] The invention relates to the technical field of program analysis, in particular to an efficient program analysis method and system based on file dependencies. Background technique [0002] Parsing is the first step in program analysis, and the goal is to generate an abstract syntax tree (Abstract Syntax Tree, AST) from the program source code. An abstract syntax tree describes the hierarchy of a program, such as figure 1 As shown, it is an abstract syntax tree of a main function, and its abstract syntax tree specifically includes a root node (TranslationUnit) and a subtree about a function declaration (MethodDeclaration), where the function declaration contains a function return type (PrimitiveType ), the function name (SimpleName) and the function body (Block), the function body includes the return statement (ReturnStatement) and the number literal (NumberLiteral). A resolution unit corresponds to a source file. Since the function in this example is...

AI Technical Summary

Problems solved by technology

Since defect detection first needs to know the grammatical structure of the program, if the abstract syntax tree cannot be parsed quickly, the entire detection speed will be slowed down accordingly

like figure 2 As shown, each source file is parsed into an abstract syntax tree (ie, AST) by a separate process; however, each source file may include header files that are included in different source files Contains and thus is parsed repeatedly throughout the parsing process

[0005] In the prior art, there has been research on the problem of repeated parsing of header files, but the current research only solves this problem in the field of compilation, and there is still a lack of technology to solve this problem in the field of static analysis; secondly, because the current research is optimized for compilation, However, each file is run through an independent process during compilation, so it is difficult to achieve memory sharing for the abstract syntax tree across files, and the existing compilation acceleration technology is difficult to apply in the scenario of full program analysis

Images