post to .profile

scripts/language_stats.py

@@ -2,9 +2,9 @@
 """
 Gitea Language Stats
 Calculates the percentage of every programming language used
-in all repositories of a given account and prints a markdown
+in all repositories of a given account and updates the README.md
-table to stdout, delimited by the
+in the .profile repository between the markers
-markers  <!--LANG_STATS_START--> … <!--LANG_STATS_END-->.
+<!--LANG_STATS_START--> and <!--LANG_STATS_END-->.
 
 When run as a systemd job, it checks for new commits since the last run
 and only regenerates if there are changes. Use --force to always regenerate.
@@ -12,7 +12,7 @@ and only regenerates if there are changes. Use --force to always regenerate.
 Required environment variables
 --------------------------------
 GITEA_BASE_URL   e.g. https://gitea.example.com
-GITEA_TOKEN      personal access-token with `repo` scope
+GITEA_TOKEN      personal access-token with `repo` scope (needs write access)
 GITEA_USERNAME   the account to analyse
 
 Optional environment variables
@@ -20,8 +20,10 @@ Optional environment variables
 STATE_FILE       path to store last-run state (default: ~/.local/state/language_stats.json)
 """
 import argparse
+import base64
 import json
 import os
+import re
 import sys
 from collections import defaultdict
 from pathlib import Path
@@ -38,6 +40,14 @@ USERNAME = os.getenv("GITEA_USERNAME")
 DEFAULT_STATE_DIR = Path.home() / ".local" / "state"
 STATE_FILE = Path(os.getenv("STATE_FILE", DEFAULT_STATE_DIR / "language_stats.json"))
 
+# Profile repository where README will be updated
+PROFILE_REPO = ".profile"
+README_PATH = "README.md"
+
+# Markers for content replacement
+MARKER_START = "<!--LANG_STATS_START-->"
+MARKER_END = "<!--LANG_STATS_END-->"
+
 # List of files whose byte-counts should be excluded from the totals.
 # Each tuple: (repository_name, file_path, language_name)
 EXCLUDE_FILES: list[tuple[str, str, str]] = [
@@ -80,8 +90,16 @@ def fetch_repos(session: requests.Session) -> list[dict]:
 
 
 def get_repo_timestamps(repos: list[dict]) -> dict[str, str]:
-    """Extract repository names and their updated_at timestamps."""
+    """Extract repository names and their updated_at timestamps.
-    return {repo["name"]: repo.get("updated_at", "") for repo in repos}
+
+    Excludes the profile repo to avoid infinite loops (updating it
+    would trigger another update).
+    """
+    return {
+        repo["name"]: repo.get("updated_at", "")
+        for repo in repos
+        if repo["name"] != PROFILE_REPO
+    }
 
 
 def load_state() -> dict:
@@ -151,11 +169,11 @@ def subtract_excluded_files(session: requests.Session, totals: dict[str, int]) -
     return totals
 
 
-def generate_markdown(totals: dict[str, int]) -> list[str]:
+def generate_markdown(totals: dict[str, int]) -> str:
     """Generate markdown output for language statistics."""
     total_bytes = sum(totals.values())
     if total_bytes == 0:
-        return []
+        return ""
 
     pairs = sorted(totals.items(), key=lambda x: x[1], reverse=True)
 
@@ -165,7 +183,51 @@ def generate_markdown(totals: dict[str, int]) -> list[str]:
         bar = "█" * int(pct / 2)  # up to 50 chars
         lines.append(f"{bar} {lang}{pct:5.1f}%\\")
 
-    return lines
+    return "\n".join(lines)
+
+
+def fetch_readme(session: requests.Session) -> tuple[str, str]:
+    """Fetch the README from the profile repository.
+
+    Returns:
+        Tuple of (content, sha) where sha is needed for updating.
+    """
+    resp = session.get(
+        f"{BASE_URL}/api/v1/repos/{USERNAME}/{PROFILE_REPO}/contents/{README_PATH}",
+        timeout=30,
+    )
+    resp.raise_for_status()
+    data = resp.json()
+    content = base64.b64decode(data["content"]).decode("utf-8")
+    return content, data["sha"]
+
+
+def replace_between_markers(content: str, new_content: str) -> str:
+    """Replace content between LANG_STATS markers."""
+    pattern = re.compile(
+        rf"({re.escape(MARKER_START)})(.*?)({re.escape(MARKER_END)})",
+        re.DOTALL,
+    )
+    replacement = rf"\1\n{new_content}\n\3"
+    new_readme, count = pattern.subn(replacement, content)
+    if count == 0:
+        raise ValueError(f"Markers {MARKER_START} and {MARKER_END} not found in README")
+    return new_readme
+
+
+def update_readme(session: requests.Session, content: str, sha: str) -> None:
+    """Update the README in the profile repository."""
+    encoded_content = base64.b64encode(content.encode("utf-8")).decode("ascii")
+    resp = session.put(
+        f"{BASE_URL}/api/v1/repos/{USERNAME}/{PROFILE_REPO}/contents/{README_PATH}",
+        json={
+            "content": encoded_content,
+            "sha": sha,
+            "message": "Update language statistics",
+        },
+        timeout=30,
+    )
+    resp.raise_for_status()
 
 
 def main() -> int:
@@ -213,9 +275,14 @@ def main() -> int:
     if sum(totals.values()) == 0:
         sys.exit("No language data returned from API.")
 
-    lines = generate_markdown(totals)
+    markdown = generate_markdown(totals)
-    for line in lines:
+
-        print(line)
+    # Fetch current README, update it, and push
+    readme_content, readme_sha = fetch_readme(session)
+    updated_readme = replace_between_markers(readme_content, markdown)
+    update_readme(session, updated_readme, readme_sha)
+
+    print("README.md updated successfully.", file=sys.stderr)
 
     # Save current state for next run
     new_state = {"repo_timestamps": current_timestamps}