From 49817ec206f7a1e8dd3990a51404f69d6c86eb28 Mon Sep 17 00:00:00 2001
From: Lacey Henschel <laceynwilliams@gmail.com>
Date: Mon, 3 Jun 2024 22:12:10 -0700
Subject: [PATCH] Create generate_toc_for_subdirectory.md

---
 python/generate_toc_for_subdirectory.md | 114 ++++++++++++++++++++++++
 1 file changed, 114 insertions(+)
 create mode 100644 python/generate_toc_for_subdirectory.md

diff --git a/python/generate_toc_for_subdirectory.md b/python/generate_toc_for_subdirectory.md
new file mode 100644
index 0000000..26d39f2
--- /dev/null
+++ b/python/generate_toc_for_subdirectory.md
@@ -0,0 +1,114 @@
+# Generating a clickable table-of-contents for each directory in my TILs 
+
+## Use case 
+
+I have a GitHub Action that rewrites my main README every time I commit to `main` to regenerate the table of contents, sorted by directory. 
+
+I wanted a README in each directory with its own table of contents. 
+
+## Code 
+
+```python
+# scripts/build_subdir_toc.py
+import os
+
+import click
+
+# Takes a directory as an argument 
+@click.command()
+@click.argument("directory", type=click.Path(exists=True))
+def generate_readme_links(directory):
+    """
+    Generate a README file with clickable links to all files in the specified directory.
+
+    DIRECTORY: The path to the directory containing files.
+    """
+    readme_file = os.path.join(directory, "README.md")
+
+    with open(readme_file, "w") as f:
+        # Use first h1 as the header for the readme 
+        header = os.path.basename(directory)
+        f.write(f"# {header}\n\n")
+
+        for root, dirs, files in os.walk(directory):
+            contents = []
+            for file in files:
+                title = None
+                # Read the file so I can get the file's header, which is the TIL title.
+                with open(os.path.join(root, file), "r") as file_obj:
+                    title = file_obj.readline().replace("#", "").strip()
+
+                # Skip files that don't have a title.
+                if not title:
+                    continue
+
+                # Generate the URL for the file, so the TIL titles are clickable 
+                file_path = os.path.join(root, file)
+                relative_path = os.path.relpath(file_path, directory)
+
+                # Write the next list item with a link to this file 
+                contents.append(f"- [{title}]({relative_path})")
+
+            # Alphabetize by title 
+            sorted_contents = sorted(contents)
+            f.write("\n".join(sorted_contents))
+
+    click.echo(f"README file generated at {readme_file}")
+
+
+if __name__ == "__main__":
+    generate_readme_links()
+```
+
+Example call: 
+
+```bash
+python3 scripts/build_subdir_toc.py django-rest-framework
+```
+
+This generates something like: 
+
+```markdown
+# django-rest-framework
+
+- [Adding a custom pagination class to an action](custom_action_pagination.md)
+- [Passing extra info in `context` to your DRF serializer](pass_to_context.md)
+```
+
+## GitHub Action 
+
+Then, I added calls to this script for each TIL directory I wanted to include to the existing GitHub action. 
+
+```yaml
+name: Build README
+
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      ...
+      - name: Update subdir README's 
+        run: |-
+          python3 scripts/build_subdir_toc.py aws
+          python3 scripts/build_subdir_toc.py celery
+          python3 scripts/build_subdir_toc.py django
+       ...
+```
+
+### Why aren't you just traversing all your directories? 
+
+I tried that, but I was getting some internal directories. As I was coding an EXCLUDED list, I decided I preferred an ALLOW list. 
+
+This does mean that I will need to manually add a new line to the GitHub action whenever I create a new TIL directory, but I am comfortable with that. Even if I don't get around to it, the main TIL README will still update, so it's an inconvenience I can live with. Also, I'm tired. 
+
+--- 
+
+## Notes 
+
+- As soon as I finished, it occured to me that I could have done this in a DRY-er way by re-using more of the code in `update_readme.py`. But I was re-using code from a side project that worked a little differently, so I tried a different pattern. 
+- I tried to move the scripts for the project-level README into `scripts/` but there was a git-related `InvalidGitRepositoryError` error that I didn't feel like debugging, so I just moved them back out.