add python parsing capabilities

README.md

@@ -71,6 +71,17 @@ npm run start
 ```
 
 9. Create a pull request on one of your repositories and watch the review agent submit a review!
-  - Make sure to create the pull request on a repository that your GitHub app has access to.
-  - Make sure the pull request has at least one changed file that is supported by the review agent. The following file extensions are ignored: ".pdf", ".png", ".jpg", ".jpeg", ".gif", ".mp4", ".mp3", ".md", ".json", ".env", ".toml", and ".svg".
-  - You will have to create new pull requests each time to test the review agent, as it will not work on the same pull request twice.
+
+- Make sure to create the pull request on a repository that your GitHub app has access to.
+- Make sure the pull request has at least one changed file that is supported by the review agent. The following file extensions are ignored: ".pdf", ".png", ".jpg", ".jpeg", ".gif", ".mp4", ".mp3", ".md", ".json", ".env", ".toml", and ".svg".
+- You will have to create new pull requests each time to test the review agent, as it will not work on the same pull request twice.
+
+## Resources
+
+- [Project Diagram](https://excalidraw.com/#json=Tyo1OFlu5AY5b7AiZyPJw,jIJfR3fdbbqsxecfalkkEw)
+- [Abstract Syntax Trees (AST)](https://earthly.dev/blog/python-ast/)
+- [ngrok](https://dashboard.ngrok.com/get-started/setup/windows)
+- [Code Review Automation With LLMs](https://www.sciencedirect.com/science/article/pii/S0950584924001289)
+- [AI-powered Code Review](https://arxiv.org/pdf/2404.18496)
+
+_Used NotebookLM to create highlight podcasts_

src/context/language/python-parser.ts

@@ -1,15 +1,82 @@
 import { AbstractParser, EnclosingContext } from "../../constants";
+import { Python3Parser } from "python-parser";
+
+// Function to process each node and determine if it is the largest enclosing context
+const processNode = (
+  node: any,
+  lineStart: number,
+  lineEnd: number,
+  largestSize: number,
+  largestEnclosingContext: any | null
+) => {
+  const { start, end } = node.loc;
+  if (start.line <= lineStart && lineEnd <= end.line) {
+    const size = end.line - start.line;
+    if (size > largestSize) {
+      largestSize = size;
+      largestEnclosingContext = node;
+    }
+  }
+  return { largestSize, largestEnclosingContext };
+};
+
 export class PythonParser implements AbstractParser {
   findEnclosingContext(
     file: string,
     lineStart: number,
     lineEnd: number
   ): EnclosingContext {
-    // TODO: Implement this method for Python
-    return null;
+    const parser = new Python3Parser(); // Create a new instance of Python3Parser
+    const ast = parser.parse(file); // Parse the Python file to get the AST
+    let largestEnclosingContext: any = null; // Initialize the largest enclosing context
+    let largestSize = 0; // Initialize the largest size
+
+    // Custom listener class to traverse the AST
+    class CustomListener {
+      // What does the enterEveryRule method do?
+      // It makes the listener visit every rule node in the AST
+      // and process each node to determine the largest enclosing context
+      // The ctx parameter is the context object for the rule node
+      enterEveryRule(ctx: any): void {
+        // Process each rule node
+        ({ largestSize, largestEnclosingContext } = processNode(
+          ctx,
+          lineStart,
+          lineEnd,
+          largestSize,
+          largestEnclosingContext
+        ));
+      }
+    }
+
+    const listener = new CustomListener();
+    parser.listen(listener, ast); // Traverse the AST with the custom listener
+
+    return {
+      enclosingContext: largestEnclosingContext, // Return the largest enclosing context
+    } as EnclosingContext;
   }
   dryRun(file: string): { valid: boolean; error: string } {
-    // TODO: Implement this method for Python
-    return { valid: false, error: "Not implemented yet" };
+    try {
+      const parser = new Python3Parser(); // Create a new instance of Python3Parser
+      parser.parse(file); // Parse the Python file
+      return {
+        valid: true,
+        error: "", // No errors
+      };
+    } catch (exc) {
+      return {
+        valid: false,
+        error: exc.message, // Return the error message
+      };
+    }
   }
 }
+/*
+    AST (Abstract Syntax Tree) is a tree representation of the source code
+    It represents the structure of the code in a hierarchical manner
+    The AST is used to analyze and manipulate the code programmatically
+    The Python3Parser class is used to parse Python code and generate the AST
+    --> As the code is parsed, the code is analyzed and broken down into parts
+        --> These parts are represented as nodes in the AST
+  */