Refactor ingest_from_query to run_ingest_query, add process_notebook function, and add tests for notebook processing

Author: filipchristiansen

src/gitingest/__init__.py

@@ -2,7 +2,7 @@
 
 from gitingest.clone import clone_repo
 from gitingest.ingest import ingest
-from gitingest.ingest_from_query import ingest_from_query
+from gitingest.ingest_from_query import run_ingest_query
 from gitingest.parse_query import parse_query
 
-__all__ = ["ingest_from_query", "clone_repo", "parse_query", "ingest"]
+__all__ = ["run_ingest_query", "clone_repo", "parse_query", "ingest"]

src/gitingest/ingest.py

@@ -6,7 +6,7 @@
 
 from config import TMP_BASE_PATH
 from gitingest.clone import CloneConfig, clone_repo
-from gitingest.ingest_from_query import ingest_from_query
+from gitingest.ingest_from_query import run_ingest_query
 from gitingest.parse_query import parse_query
 
 
@@ -75,7 +75,7 @@ def ingest(
             else:
                 raise TypeError("clone_repo did not return a coroutine as expected.")
 
-        summary, tree, content = ingest_from_query(query)
+        summary, tree, content = run_ingest_query(query)
 
         if output is not None:
             with open(output, "w", encoding="utf-8") as f:

src/gitingest/ingest_from_query.py

@@ -7,6 +7,7 @@
 import tiktoken
 
 from gitingest.exceptions import AlreadyVisitedError, MaxFileSizeReachedError, MaxFilesReachedError
+from gitingest.notebook_utils import process_notebook
 
 MAX_FILE_SIZE = 10 * 1024 * 1024  # 10 MB
 MAX_DIRECTORY_DEPTH = 20  # Maximum depth of directory traversal
@@ -158,7 +159,10 @@ def _read_file_content(file_path: Path) -> str:
         The content of the file, or an error message if the file could not be read.
     """
     try:
-        with file_path.open(encoding="utf-8", errors="ignore") as f:
+        if file_path.suffix == ".ipynb":
+            return process_notebook(file_path)
+
+        with open(file_path, encoding="utf-8", errors="ignore") as f:
             return f.read()
     except OSError as e:
         return f"Error reading file: {e}"
@@ -819,7 +823,7 @@ def _ingest_directory(path: Path, query: dict[str, Any]) -> tuple[str, str, str]
     return summary, tree, files_content
 
 
-def ingest_from_query(query: dict[str, Any]) -> tuple[str, str, str]:
+def run_ingest_query(query: dict[str, Any]) -> tuple[str, str, str]:
     """
     Main entry point for analyzing a codebase directory or single file.
 

src/gitingest/notebook_utils.py

@@ -0,0 +1,48 @@
+""" Utilities for processing Jupyter notebooks. """
+
+import json
+from pathlib import Path
+
+
+def process_notebook(file: Path) -> str:
+    """
+    Process a Jupyter notebook file and return an executable Python script as a string.
+
+    Parameters
+    ----------
+    file : Path
+        The path to the Jupyter notebook file.
+
+    Returns
+    -------
+    str
+        The executable Python script as a string.
+
+    Raises
+    ------
+    ValueError
+        If an unexpected cell type is encountered.
+    """
+    with file.open(encoding="utf-8") as f:
+        notebook = json.load(f)
+
+    result = []
+
+    for cell in notebook["cells"]:
+        cell_type = cell.get("cell_type")
+
+        # Validate cell type and handle unexpected types
+        if cell_type not in ("markdown", "code", "raw"):
+            raise ValueError(f"Unknown cell type: {cell_type}")
+
+        str_ = "".join(cell.get("source", []))
+        if not str_:
+            continue
+
+        # Convert Markdown and raw cells to multi-line comments
+        if cell_type in ("markdown", "raw"):
+            str_ = f'"""\n{str_}\n"""'
+
+        result.append(str_)
+
+    return "\n\n".join(result)

src/process_query.py

@@ -8,7 +8,7 @@
 
 from config import EXAMPLE_REPOS, MAX_DISPLAY_SIZE
 from gitingest.clone import CloneConfig, clone_repo
-from gitingest.ingest_from_query import ingest_from_query
+from gitingest.ingest_from_query import run_ingest_query
 from gitingest.parse_query import parse_query
 from server_utils import Colors, log_slider_to_size
 
@@ -91,7 +91,7 @@ async def process_query(
             branch=query.get("branch"),
         )
         await clone_repo(clone_config)
-        summary, tree, content = ingest_from_query(query)
+        summary, tree, content = run_ingest_query(query)
         with open(f"{clone_config.local_path}.txt", "w", encoding="utf-8") as f:
             f.write(tree + "\n" + content)
     except Exception as e:

tests/conftest.py

@@ -1,5 +1,6 @@
 """ This module contains fixtures for the tests. """
 
+import json
 from pathlib import Path
 from typing import Any
 
@@ -72,3 +73,18 @@ def temp_directory(tmp_path: Path) -> Path:
     (dir2 / "file_dir2.txt").write_text("Hello from dir2")
 
     return test_dir
+
+
+@pytest.fixture
+def write_notebook(tmp_path: Path):
+    """
+    A helper fixture that returns a function for writing arbitrary notebook content to a temporary .ipynb file.
+    """
+
+    def _write_notebook(name: str, content: dict) -> Path:
+        notebook_path = tmp_path / name
+        with notebook_path.open("w", encoding="utf-8") as f:
+            json.dump(content, f)
+        return notebook_path
+
+    return _write_notebook