@memorax/memorax-code 0.1.0

package/lib/memorax-code-claude-adapter/skills/memorax-code/scripts/github_resource_facets.py

@@ -0,0 +1,504 @@
+#!/usr/bin/env python3
+"""Extract GitHub PR/issue resource facets with only Python stdlib + gh CLI."""
+
+from __future__ import annotations
+
+import argparse
+import json
+import re
+import shutil
+import subprocess
+import sys
+import time
+from concurrent.futures import ThreadPoolExecutor
+from pathlib import Path
+from typing import Any, Iterable
+
+
+PR_LIST_FIELDS = "number,title,state,url,updatedAt,baseRefName,headRefName,headRepository,headRepositoryOwner,isDraft"
+PR_VIEW_FIELDS = "number,title,body,state,url,updatedAt,createdAt,closedAt,mergedAt,author,comments,reviews,latestReviews,reviewDecision,files,commits,closingIssuesReferences,mergeCommit,baseRefName,headRefName,headRepository,headRepositoryOwner,isDraft,additions,deletions,changedFiles"
+ISSUE_LIST_FIELDS = "number,title,state,url,updatedAt,labels"
+ISSUE_VIEW_FIELDS = "number,title,body,state,url,updatedAt,labels,comments"
+DEFAULT_GH_RETRIES = 3
+DEFAULT_GH_RETRY_DELAY_MS = 1000
+TRANSIENT_GH_ERROR = re.compile(
+    r"EOF|timeout|timed out|connection reset|connection refused|temporarily unavailable|"
+    r"TLS handshake timeout|502|503|504|rate limit|secondary rate limit|abuse detection",
+    re.IGNORECASE,
+)
+
+
+def is_transient_gh_error(stderr: str, stdout: str = "") -> bool:
+    return TRANSIENT_GH_ERROR.search(f"{stderr}\n{stdout}") is not None
+
+
+def run_gh(args: list[str], retries: int = DEFAULT_GH_RETRIES, retry_delay_ms: int = DEFAULT_GH_RETRY_DELAY_MS) -> Any:
+    if shutil.which("gh") is None:
+        raise SystemExit("GitHub CLI 'gh' is required. Install it and run: gh auth login")
+    attempts = max(1, retries + 1)
+    last_result: subprocess.CompletedProcess[str] | None = None
+    for attempt in range(1, attempts + 1):
+        result = subprocess.run(["gh", *args], text=True, capture_output=True)
+        last_result = result
+        if result.returncode == 0:
+            text = result.stdout.strip()
+            return json.loads(text) if text else None
+        if attempt < attempts and is_transient_gh_error(result.stderr, result.stdout):
+            delay = max(0, retry_delay_ms) / 1000 * (2 ** (attempt - 1))
+            print(
+                f"gh {' '.join(args)} failed with transient error; retrying "
+                f"{attempt}/{retries} after {delay:.1f}s: {result.stderr.strip()}",
+                file=sys.stderr,
+            )
+            time.sleep(delay)
+            continue
+        break
+    assert last_result is not None
+    raise SystemExit(f"gh {' '.join(args)} failed:\n{last_result.stderr.strip()}")
+
+
+def auth_status_failure(result: subprocess.CompletedProcess[str], login_hint: str) -> str:
+    details: list[str] = []
+    stderr = result.stderr.strip()
+    stdout = result.stdout.strip()
+    if stderr:
+        details.append(f"stderr: {stderr}")
+    if stdout:
+        details.append(f"stdout: {stdout}")
+    detail = "\n".join(details) if details else "gh auth status returned no diagnostic output"
+    return f"gh auth check failed with exit code {result.returncode}. Run: {login_hint}\n{detail}"
+
+
+def run_git(repo_path: Path, args: list[str]) -> subprocess.CompletedProcess[str]:
+    if shutil.which("git") is None:
+        raise SystemExit("git is required for PR snapshot filtering")
+    return subprocess.run(["git", "-C", str(repo_path), *args], text=True, capture_output=True)
+
+
+def resolve_snapshot_sha(repo_path: Path, snapshot_ref: str) -> str:
+    result = run_git(repo_path, ["rev-parse", "--verify", f"{snapshot_ref}^{{commit}}"])
+    if result.returncode != 0:
+        raise SystemExit(f"git could not resolve snapshot ref {snapshot_ref!r} in {repo_path}:\n{result.stderr.strip()}")
+    return result.stdout.strip()
+
+
+def is_git_ancestor(repo_path: Path, ancestor_sha: str, descendant_sha: str) -> bool:
+    result = run_git(repo_path, ["merge-base", "--is-ancestor", ancestor_sha, descendant_sha])
+    if result.returncode == 0:
+        return True
+    if result.returncode == 1:
+        return False
+    raise SystemExit(
+        f"git could not compare {ancestor_sha} with snapshot {descendant_sha} in {repo_path}:\n{result.stderr.strip()}"
+    )
+
+
+def login_hint(hostname: str) -> str:
+    return f"gh auth login --hostname {hostname}" if hostname and hostname != "github.com" else "gh auth login"
+
+
+def repo_selector(repo: str, hostname: str) -> str:
+    if hostname and hostname != "github.com" and not repo.startswith(f"{hostname}/"):
+        return f"{hostname}/{repo}"
+    return repo
+
+
+def assert_gh_ready(hostname: str) -> None:
+    if shutil.which("gh") is None:
+        raise SystemExit("GitHub CLI 'gh' is required. Install it and run: gh auth login")
+    args = ["auth", "status"]
+    if hostname:
+        args.extend(["--hostname", hostname])
+    attempts = max(1, DEFAULT_GH_RETRIES + 1)
+    last_result: subprocess.CompletedProcess[str] | None = None
+    for attempt in range(1, attempts + 1):
+        result = subprocess.run(["gh", *args], text=True, capture_output=True)
+        last_result = result
+        if result.returncode == 0:
+            return
+        if attempt < attempts and is_transient_gh_error(result.stderr, result.stdout):
+            delay = max(0, DEFAULT_GH_RETRY_DELAY_MS) / 1000 * (2 ** (attempt - 1))
+            print(
+                f"gh auth status failed with transient error; retrying "
+                f"{attempt}/{DEFAULT_GH_RETRIES} after {delay:.1f}s: {result.stderr.strip()}",
+                file=sys.stderr,
+            )
+            time.sleep(delay)
+            continue
+        break
+    assert last_result is not None
+    raise SystemExit(auth_status_failure(last_result, login_hint(hostname)))
+
+
+def unique_strings(values: Iterable[str]) -> list[str]:
+    seen: set[str] = set()
+    out: list[str] = []
+    for value in values:
+        if value and value not in seen:
+            seen.add(value)
+            out.append(value)
+    return out
+
+
+def unique_numbers(values: Iterable[int]) -> list[int]:
+    seen: set[int] = set()
+    out: list[int] = []
+    for value in values:
+        if isinstance(value, int) and value not in seen:
+            seen.add(value)
+            out.append(value)
+    return out
+
+
+def text_field(value: Any, *keys: str) -> str:
+    if isinstance(value, str):
+        return value.strip()
+    if not isinstance(value, dict):
+        return ""
+    for key in keys:
+        item = value.get(key)
+        if isinstance(item, str) and item.strip():
+            return item.strip()
+    return ""
+
+
+def int_field(value: Any, key: str = "number") -> int | None:
+    if isinstance(value, int):
+        return value
+    if isinstance(value, dict) and isinstance(value.get(key), int):
+        return value[key]
+    return None
+
+
+def bounded_summary(parts: Iterable[str], max_chars: int) -> str:
+    text = "\n\n".join(part.strip() for part in parts if isinstance(part, str) and part.strip())
+    return text if len(text) <= max_chars else text[: max_chars - 3] + "..."
+
+
+def extract_symbols(text: str) -> list[str]:
+    symbols: list[str] = []
+    for match in re.finditer(r"\b[$A-Z_a-z][$\w]*\b", text):
+        value = match.group(0)
+        looks_symbolic = (
+            re.search(r"[a-z][A-Z]", value) is not None
+            or re.match(r"^[A-Z0-9_]{2,}$", value) is not None
+            or "_" in value
+            or "$" in value
+        )
+        if looks_symbolic:
+            symbols.append(value)
+    return unique_strings(symbols)
+
+
+def commit_sha(value: Any) -> str:
+    if isinstance(value, dict) and isinstance(value.get("commit"), dict):
+        return text_field(value["commit"], "oid", "sha", "abbreviatedOid")
+    return text_field(value, "oid", "sha", "abbreviatedOid")
+
+
+def commit_headline(value: Any) -> str:
+    if isinstance(value, dict) and isinstance(value.get("commit"), dict):
+        return text_field(value["commit"], "messageHeadline", "message", "title")
+    return text_field(value, "messageHeadline", "message", "title")
+
+
+def author_login(value: Any) -> str:
+    return text_field(value, "login", "name")
+
+
+def review_state(value: Any) -> str:
+    return text_field(value, "state")
+
+
+def file_path(value: Any) -> str:
+    return text_field(value, "path", "filename")
+
+
+def label_name(value: Any) -> str:
+    return text_field(value, "name")
+
+
+def repo_name(value: Any) -> str:
+    if isinstance(value, str):
+        return value.strip()
+    if not isinstance(value, dict):
+        return ""
+    owner = value.get("owner")
+    owner_login = text_field(owner, "login") if isinstance(owner, dict) else ""
+    name = text_field(value, "name", "nameWithOwner")
+    name_with_owner = text_field(value, "nameWithOwner")
+    if name_with_owner and "/" in name_with_owner:
+        return name_with_owner
+    return f"{owner_login}/{name}" if owner_login and name else name
+
+
+def body_text(value: Any) -> str:
+    return text_field(value, "body", "title", "name")
+
+
+def pr_detail_to_facet(repo: str, detail: dict[str, Any], summary_chars: int) -> dict[str, Any]:
+    number = int_field(detail) or 0
+    title = text_field(detail, "title") or f"PR #{number}"
+    body = text_field(detail, "body")
+    comments = [body_text(item) for item in detail.get("comments", []) if body_text(item)]
+    reviews = [body_text(item) for item in detail.get("reviews", []) if body_text(item)]
+    commits = unique_strings(commit_sha(item) for item in detail.get("commits", []) if commit_sha(item))
+    commit_headlines = unique_strings(commit_headline(item) for item in detail.get("commits", []) if commit_headline(item))
+    files = unique_strings(file_path(item) for item in detail.get("files", []) if file_path(item))
+    review_states = unique_strings(review_state(item) for item in detail.get("latestReviews", []) if review_state(item))
+    issues = unique_numbers(
+        number for number in (int_field(item) for item in detail.get("closingIssuesReferences", [])) if number is not None
+    )
+    summary = bounded_summary([body, *comments, *reviews], summary_chars) or title
+    base_ref = text_field(detail, "baseRefName")
+    head_ref = text_field(detail, "headRefName")
+    head_repo = repo_name(detail.get("headRepository")) or repo
+    branch_label = f"{base_ref} <- {head_ref}" if base_ref or head_ref else ""
+    evidence = unique_strings(
+        [
+            f"PR #{number}: {title}",
+            f"PR #{number} changed {', '.join(files)}" if files else "",
+            f"PR #{number} closes issue {', '.join(f'#{issue}' for issue in issues)}" if issues else "",
+        ]
+    )
+    return {
+        "facetId": f"pr.{number}",
+        "sourceType": "pr",
+        "repo": repo,
+        "title": title,
+        "summary": summary,
+        "url": text_field(detail, "url"),
+        "state": text_field(detail, "state"),
+        "createdAt": text_field(detail, "createdAt"),
+        "updatedAt": text_field(detail, "updatedAt"),
+        "closedAt": text_field(detail, "closedAt"),
+        "mergedAt": text_field(detail, "mergedAt"),
+        "author": author_login(detail.get("author")),
+        "isDraft": bool(detail.get("isDraft")),
+        "base_ref": base_ref,
+        "head_ref": head_ref,
+        "head_repo": head_repo,
+        "branch_label": branch_label,
+        "changed_files": detail.get("changedFiles") or len(files),
+        "additions": detail.get("additions") or 0,
+        "deletions": detail.get("deletions") or 0,
+        "commit_headlines": commit_headlines,
+        "review_decision": text_field(detail, "reviewDecision"),
+        "review_states": review_states,
+        "commits": commits,
+        "prs": [number] if number else [],
+        "issues": issues,
+        "files": files,
+        "symbols": extract_symbols("\n".join([title, summary, *files])),
+        "evidence": evidence,
+    }
+
+
+def issue_detail_to_facet(repo: str, detail: dict[str, Any], summary_chars: int) -> dict[str, Any]:
+    number = int_field(detail) or 0
+    title = text_field(detail, "title") or f"Issue #{number}"
+    body = text_field(detail, "body")
+    comments = [body_text(item) for item in detail.get("comments", []) if body_text(item)]
+    labels = [label_name(item) for item in detail.get("labels", []) if label_name(item)]
+    labels_text = f"Labels: {', '.join(labels)}" if labels else ""
+    summary = bounded_summary([body, *comments, labels_text], summary_chars) or title
+    evidence = unique_strings(
+        [
+            f"issue #{number}: {title}",
+            f"issue #{number} labels: {', '.join(labels)}" if labels else "",
+        ]
+    )
+    return {
+        "facetId": f"issue.{number}",
+        "sourceType": "issue",
+        "repo": repo,
+        "title": title,
+        "summary": summary,
+        "url": text_field(detail, "url"),
+        "state": text_field(detail, "state"),
+        "updatedAt": text_field(detail, "updatedAt"),
+        "labels": labels,
+        "commits": [],
+        "prs": [],
+        "issues": [number] if number else [],
+        "files": [],
+        "symbols": extract_symbols("\n".join([title, summary, *labels])),
+        "evidence": evidence,
+    }
+
+
+def gh_list(kind: str, repo: str, limit: int, state: str, retries: int, retry_delay_ms: int) -> list[dict[str, Any]]:
+    fields = PR_LIST_FIELDS if kind == "pr" else ISSUE_LIST_FIELDS
+    return run_gh([kind, "list", "--repo", repo, "--limit", str(limit), "--state", state, "--json", fields], retries, retry_delay_ms)
+
+
+def gh_view(kind: str, repo: str, number: int, retries: int, retry_delay_ms: int) -> dict[str, Any]:
+    fields = PR_VIEW_FIELDS if kind == "pr" else ISSUE_VIEW_FIELDS
+    return run_gh([kind, "view", str(number), "--repo", repo, "--json", fields], retries, retry_delay_ms)
+
+
+def pr_is_in_snapshot(repo_path: Path, detail: dict[str, Any], snapshot_sha: str) -> bool:
+    if not snapshot_sha:
+        return True
+    merge_sha = commit_sha(detail.get("mergeCommit"))
+    return bool(merge_sha) and is_git_ancestor(repo_path, merge_sha, snapshot_sha)
+
+
+def fetch_pr_facets(
+    query_repo: str,
+    facet_repo: str,
+    repo_path: Path,
+    snapshot_sha: str,
+    limit: int,
+    state: str,
+    concurrency: int,
+    summary_chars: int,
+    retries: int,
+    retry_delay_ms: int,
+) -> list[dict[str, Any]]:
+    if snapshot_sha and state == "open":
+        return []
+
+    facets: list[dict[str, Any]] = []
+    seen_numbers: set[int] = set()
+    candidate_limit = max(1, limit)
+    previous_candidate_count = -1
+
+    while len(facets) < limit:
+        candidates = gh_list("pr", query_repo, candidate_limit, state, retries, retry_delay_ms)
+        if not isinstance(candidates, list):
+            raise SystemExit("gh PR list response was not a JSON list")
+
+        candidate_count = len(candidates)
+        numbers: list[int] = []
+        for item in candidates:
+            number = int_field(item)
+            if number is not None and number not in seen_numbers:
+                seen_numbers.add(number)
+                numbers.append(number)
+
+        def one(number: int) -> dict[str, Any] | None:
+            detail = gh_view("pr", query_repo, number, retries, retry_delay_ms)
+            if snapshot_sha and not pr_is_in_snapshot(repo_path, detail, snapshot_sha):
+                return None
+            return pr_detail_to_facet(facet_repo, detail, summary_chars)
+
+        if numbers:
+            with ThreadPoolExecutor(max_workers=max(1, concurrency)) as pool:
+                for facet in pool.map(one, numbers):
+                    if facet:
+                        facets.append(facet)
+                        if len(facets) >= limit:
+                            break
+
+        if len(facets) >= limit or candidate_count < candidate_limit or candidate_count == previous_candidate_count:
+            break
+        previous_candidate_count = candidate_count
+        candidate_limit *= 2
+
+    return facets[:limit]
+
+
+def fetch_facets(
+    repo: str,
+    repo_path: Path,
+    hostname: str,
+    snapshot_ref: str,
+    pr_limit: int,
+    issue_limit: int,
+    state: str,
+    include: set[str],
+    concurrency: int,
+    summary_chars: int,
+    retries: int,
+    retry_delay_ms: int,
+) -> list[dict[str, Any]]:
+    assert_gh_ready(hostname)
+    selected_repo = repo_selector(repo, hostname)
+    snapshot_sha = resolve_snapshot_sha(repo_path, snapshot_ref) if snapshot_ref and "prs" in include else ""
+    facets: list[dict[str, Any]] = []
+    jobs: list[tuple[str, int]] = []
+    if "prs" in include:
+        facets.extend(fetch_pr_facets(selected_repo, repo, repo_path, snapshot_sha, pr_limit, state, concurrency, summary_chars, retries, retry_delay_ms))
+    if "issues" in include:
+        jobs.extend(("issue", item["number"]) for item in gh_list("issue", selected_repo, issue_limit, state, retries, retry_delay_ms))
+
+    def one(job: tuple[str, int]) -> dict[str, Any] | None:
+        kind, number = job
+        detail = gh_view(kind, selected_repo, number, retries, retry_delay_ms)
+        return issue_detail_to_facet(repo, detail, summary_chars)
+
+    with ThreadPoolExecutor(max_workers=max(1, concurrency)) as pool:
+        facets.extend(facet for facet in pool.map(one, jobs) if facet)
+    return facets
+
+
+def parse_args(argv: list[str]) -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="Extract GitHub PR/issue SourceFacet JSON using gh CLI.")
+    parser.add_argument("--repo", required=True, help="GitHub repository in owner/name form")
+    parser.add_argument("--hostname", default="", help="GitHub hostname for auth/API commands; defaults to gh's current host")
+    parser.add_argument("--repo-path", default=".", help="Local git repository path used for PR snapshot filtering")
+    parser.add_argument(
+        "--snapshot-ref",
+        default="",
+        help="Optional local git ref; PRs are kept only when their merge commit is an ancestor of this ref",
+    )
+    parser.add_argument("--pr-limit", type=int, default=None, help="Maximum retained PRs after snapshot filtering")
+    parser.add_argument("--issue-limit", type=int, default=None, help="Maximum retained provider issues")
+    parser.add_argument("--state", choices=["all", "open", "closed"], default="all")
+    parser.add_argument("--include", default="prs,issues", help="Comma-separated: prs,issues")
+    parser.add_argument("--concurrency", type=int, default=3)
+    parser.add_argument("--gh-retries", type=int, default=DEFAULT_GH_RETRIES, help="Retries for transient gh CLI failures")
+    parser.add_argument("--gh-retry-delay-ms", type=int, default=DEFAULT_GH_RETRY_DELAY_MS, help="Initial retry delay for transient gh CLI failures")
+    parser.add_argument("--summary-chars", type=int, default=4000)
+    parser.add_argument("--pretty", action="store_true", help="Pretty-print JSON")
+    parser.add_argument("--out", help="Write JSON to this file instead of stdout")
+    args = parser.parse_args(argv)
+    if not re.match(r"^[^/]+/[^/]+$", args.repo):
+        parser.error("--repo must be owner/name")
+    include = {part.strip() for part in args.include.split(",") if part.strip()}
+    if not include or include - {"prs", "issues"}:
+        parser.error("--include must contain prs, issues, or prs,issues")
+    args.include = include
+    args.pr_limit = args.pr_limit if args.pr_limit is not None else 30
+    args.issue_limit = args.issue_limit if args.issue_limit is not None else 30
+    if args.pr_limit < 1 or args.pr_limit > 100:
+        parser.error("--pr-limit must be from 1 to 100")
+    if args.issue_limit < 1 or args.issue_limit > 100:
+        parser.error("--issue-limit must be from 1 to 100")
+    if args.concurrency < 1 or args.concurrency > 10:
+        parser.error("--concurrency must be from 1 to 10")
+    if args.gh_retries < 0 or args.gh_retries > 10:
+        parser.error("--gh-retries must be from 0 to 10")
+    if args.gh_retry_delay_ms < 0:
+        parser.error("--gh-retry-delay-ms must be non-negative")
+    args.repo_path = Path(args.repo_path).expanduser().resolve()
+    return args
+
+
+def main(argv: list[str]) -> int:
+    args = parse_args(argv)
+    facets = fetch_facets(
+        args.repo,
+        args.repo_path,
+        args.hostname,
+        args.snapshot_ref,
+        args.pr_limit,
+        args.issue_limit,
+        args.state,
+        args.include,
+        args.concurrency,
+        args.summary_chars,
+        args.gh_retries,
+        args.gh_retry_delay_ms,
+    )
+    text = json.dumps(facets, ensure_ascii=False, indent=2 if args.pretty else None)
+    if args.out:
+        Path(args.out).write_text(text + "\n", encoding="utf-8")
+    else:
+        print(text)
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main(sys.argv[1:]))