turboquant/scripts/weekly_update.py

#!/usr/bin/env python3
"""
TurboQuant Weekly Progress Update Generator

Generates a structured weekly update from:
- Git log (commits since last week)
- Open/closed issues and PRs
- Benchmark results
- Blockers (open issues labeled 'blocker')

Usage:
    python3 scripts/weekly_update.py                    # This week
    python3 scripts/weekly_update.py --since 2026-04-08 # Custom range
    python3 scripts/weekly_update.py --post             # Post as Gitea comment on tracking issue
"""

import argparse
import json
import os
import subprocess
import sys
from datetime import datetime, timedelta
from pathlib import Path

try:
    import requests
    HAS_REQUESTS = True
except ImportError:
    HAS_REQUESTS = False


REPO_ROOT = Path(__file__).resolve().parent.parent
GITEA_URL = "https://forge.alexanderwhitestone.com"
REPO_PATH = "Timmy_Foundation/turboquant"
TRACKING_ISSUE = 76  # This issue


def git_log(since: str, until: str = None) -> list[dict]:
    """Get commits since a date."""
    until = until or datetime.now().strftime("%Y-%m-%d")
    cmd = [
        "git", "-C", str(REPO_ROOT), "log",
        f"--since={since}", f"--until={until}",
        "--format=%H|%an|%ae|%aI|%s",
        "--all"
    ]
    result = subprocess.run(cmd, capture_output=True, text=True)
    commits = []
    for line in result.stdout.strip().split("\n"):
        if not line:
            continue
        parts = line.split("|", 4)
        if len(parts) == 5:
            commits.append({
                "hash": parts[0][:8],
                "author": parts[1],
                "email": parts[2],
                "date": parts[3][:10],
                "subject": parts[4],
            })
    return commits


def git_diff_stats(since: str) -> dict:
    """Get file change stats."""
    cmd = [
        "git", "-C", str(REPO_ROOT), "diff",
        f"--stat", f"{since}..HEAD"
    ]
    result = subprocess.run(cmd, capture_output=True, text=True)
    lines = result.stdout.strip().split("\n")
    summary = lines[-1] if lines else "No changes"
    return {"summary": summary, "files_changed": len([l for l in lines if "|" in l])}


def find_benchmarks() -> list[dict]:
    """Scan benchmark results directory for recent results."""
    bench_dir = REPO_ROOT / "benchmarks"
    results = []
    if not bench_dir.exists():
        return results

    for f in bench_dir.glob("*.json"):
        try:
            data = json.loads(f.read_text())
            results.append({"file": f.name, "data": data})
        except (json.JSONDecodeError, Exception):
            pass

    # Also check for markdown reports
    for f in bench_dir.glob("*.md"):
        if f.name != "README.md":
            stat = f.stat()
            results.append({
                "file": f.name,
                "type": "report",
                "modified": datetime.fromtimestamp(stat.st_mtime).strftime("%Y-%m-%d"),
                "size": stat.st_size,
            })

    return results


def get_gitea_state(token: str = None) -> dict:
    """Fetch issue/PR state from Gitea API."""
    if not HAS_REQUESTS or not token:
        return {"available": False}

    H = {"Authorization": f"token {token}"}
    base = f"{GITEA_URL}/api/v1/repos/{REPO_PATH}"

    try:
        # Open issues
        r = requests.get(f"{base}/issues?state=open&limit=100", headers=H)
        open_issues = r.json() if r.status_code == 200 else []

        # Closed issues (recent)
        r = requests.get(f"{base}/issues?state=closed&limit=50&sort=updated&order=desc", headers=H)
        closed_issues = r.json() if r.status_code == 200 else []

        # PRs
        r = requests.get(f"{base}/pulls?state=open&limit=50", headers=H)
        open_prs = r.json() if r.status_code == 200 else []

        return {
            "available": True,
            "open_issues": open_issues,
            "closed_issues": closed_issues,
            "open_prs": open_prs,
        }
    except Exception as e:
        return {"available": False, "error": str(e)}


def categorize_commits(commits: list[dict]) -> dict:
    """Categorize commits by conventional prefix."""
    categories = {
        "feat": [], "fix": [], "bench": [], "docs": [],
        "test": [], "refactor": [], "chore": [], "other": []
    }
    for c in commits:
        subject = c["subject"].lower()
        if subject.startswith("feat") or subject.startswith("feature"):
            categories["feat"].append(c)
        elif subject.startswith("fix"):
            categories["fix"].append(c)
        elif subject.startswith("bench") or subject.startswith("perf"):
            categories["bench"].append(c)
        elif subject.startswith("doc"):
            categories["docs"].append(c)
        elif subject.startswith("test"):
            categories["test"].append(c)
        elif subject.startswith("refactor"):
            categories["refactor"].append(c)
        elif subject.startswith("chore") or subject.startswith("ci"):
            categories["chore"].append(c)
        else:
            categories["other"].append(c)
    return {k: v for k, v in categories.items() if v}


def generate_update(since: str, gitea_state: dict = None) -> str:
    """Generate the weekly update markdown."""
    now = datetime.now()
    until = now.strftime("%Y-%m-%d")
    week_label = f"Week of {since} to {until}"

    commits = git_log(since, until)
    diff_stats = git_diff_stats(since)
    categories = categorize_commits(commits)
    benchmarks = find_benchmarks()

    lines = [
        f"## {week_label}",
        "",
        f"**Generated:** {now.strftime('%Y-%m-%d %H:%M UTC')}",
        f"**Commits:** {len(commits)} | **Files changed:** {diff_stats['files_changed']}",
        "",
    ]

    # Completed work by category
    lines.append("### Completed")
    lines.append("")
    if commits:
        for cat, items in categories.items():
            label = {
                "feat": "Features", "fix": "Fixes", "bench": "Benchmarks",
                "docs": "Documentation", "test": "Tests", "refactor": "Refactoring",
                "chore": "Maintenance", "other": "Other"
            }.get(cat, cat)
            lines.append(f"**{label}:**")
            for c in items:
                lines.append(f"- `{c['hash']}` {c['subject']} ({c['author']}, {c['date']})")
            lines.append("")
    else:
        lines.append("- No commits this week")
        lines.append("")

    # Benchmark results
    if benchmarks:
        lines.append("### Benchmark Results")
        lines.append("")
        for b in benchmarks:
            if b.get("type") == "report":
                lines.append(f"- **{b['file']}** (updated {b['modified']}, {b['size']} bytes)")
            else:
                lines.append(f"- **{b['file']}** — see `benchmarks/{b['file']}`")
        lines.append("")

    # Gitea state (if available)
    if gitea_state and gitea_state.get("available"):
        open_issues = gitea_state["open_issues"]
        open_prs = gitea_state["open_prs"]
        closed = gitea_state["closed_issues"]

        lines.append("### In Progress")
        lines.append("")
        blockers = []
        for issue in open_issues:
            labels = [l["name"] for l in issue.get("labels", [])]
            prefix = ""
            if "blocker" in labels:
                blockers.append(issue)
                prefix = "🚧 BLOCKER — "
            assignee = issue.get("assignee", {})
            who = assignee.get("login", "unassigned") if assignee else "unassigned"
            lines.append(f"- {prefix}#{issue['number']}: {issue['title']} ({who})")

        if open_prs:
            lines.append("")
            lines.append("**Open PRs:**")
            for pr in open_prs:
                lines.append(f"- #{pr['number']}: {pr['title']} ({pr['user']['login']})")
        lines.append("")

        # Blockers
        if blockers:
            lines.append("### Blockers")
            lines.append("")
            for b in blockers:
                lines.append(f"- #{b['number']}: {b['title']}")
                if b.get("body"):
                    snippet = b["body"][:200].replace("\n", " ")
                    lines.append(f"  > {snippet}...")
            lines.append("")

        # Recently closed
        recent_closed = [i for i in closed if i.get("closed_at")]
        if recent_closed:
            lines.append("### Closed This Period")
            lines.append("")
            for issue in recent_closed[:10]:
                closed_date = issue.get("closed_at", "")[:10]
                lines.append(f"- #{issue['number']}: {issue['title']} (closed {closed_date})")
            lines.append("")

    # Next week
    lines.append("### Next Week")
    lines.append("")
    lines.append("- _TBD — fill in planned work_")
    lines.append("")

    return "\n".join(lines)


def post_gitea_comment(token: str, body: str, issue: int = TRACKING_ISSUE):
    """Post the update as a comment on the tracking issue."""
    if not HAS_REQUESTS:
        print("ERROR: requests library not available", file=sys.stderr)
        return False

    H = {"Authorization": f"token {token}", "Content-Type": "application/json"}
    url = f"{GITEA_URL}/api/v1/repos/{REPO_PATH}/issues/{issue}/comments"
    r = requests.post(url, headers=H, json={"body": body})

    if r.status_code in (200, 201):
        print(f"Posted comment on issue #{issue}")
        return True
    else:
        print(f"Failed to post: {r.status_code} {r.text}", file=sys.stderr)
        return False


def main():
    parser = argparse.ArgumentParser(description="Generate TurboQuant weekly progress update")
    parser.add_argument("--since", help="Start date (YYYY-MM-DD), default: 7 days ago")
    parser.add_argument("--post", action="store_true", help="Post as Gitea comment on issue #76")
    parser.add_argument("--json", action="store_true", help="Output raw data as JSON")
    args = parser.parse_args()

    since = args.since or (datetime.now() - timedelta(days=7)).strftime("%Y-%m-%d")

    # Try to load Gitea token
    token = None
    token_path = Path.home() / ".config" / "gitea" / "token"
    if token_path.exists():
        token = token_path.read_text().strip()

    gitea_state = get_gitea_state(token) if token else {"available": False}

    if args.json:
        data = {
            "since": since,
            "commits": git_log(since),
            "benchmarks": find_benchmarks(),
            "gitea": {k: v for k, v in gitea_state.items() if k != "available"} if gitea_state.get("available") else None,
        }
        print(json.dumps(data, indent=2, default=str))
        return

    update = generate_update(since, gitea_state)

    if args.post:
        if not token:
            print("ERROR: No Gitea token found at ~/.config/gitea/token", file=sys.stderr)
            sys.exit(1)
        post_gitea_comment(token, update)
    else:
        print(update)


if __name__ == "__main__":
    main()