LRriver commented on code in PR #3045: URL: https://github.com/apache/hugegraph/pull/3045#discussion_r3333522525
########## tools/ai/hugegraph-deepwiki-skill/plugins/hugegraph-deepwiki-skill/skills/hugegraph-deepwiki-skill/scripts/deepwiki_mcp.py: ########## @@ -0,0 +1,424 @@ +#!/usr/bin/env python3 +"""Small DeepWiki MCP client for repository-scoped Q&A.""" + +from __future__ import annotations + +import argparse +import json +import os +import re +import sys +import time +import urllib.error +import urllib.request +from pathlib import Path +from typing import Any, Optional + + +DEFAULT_ENDPOINT = "https://mcp.deepwiki.com/mcp"; +SCRIPT_DIR = Path(__file__).resolve().parent +SKILL_DIR = SCRIPT_DIR.parent +REPOS_PATH = SKILL_DIR / "references" / "repos.json" +STOPWORDS = { + "a", + "an", + "and", + "apache", + "are", + "as", + "for", + "hugegraph", + "how", + "in", + "is", + "it", + "of", + "on", + "or", + "the", + "to", + "used", + "what", + "where", + "which", + "why", +} + + +class McpError(RuntimeError): + pass + + +def load_repos() -> dict[str, dict[str, Any]]: + with REPOS_PATH.open("r", encoding="utf-8") as file: + return json.load(file) + + +def resolve_repo(alias_or_name: str) -> str: + repos = load_repos() + profile = repos.get(alias_or_name) + if profile is None: + known = ", ".join(sorted(repos)) + raise McpError(f"Unknown repository alias '{alias_or_name}'. Known aliases: {known}.") + if not profile.get("enabled", False): + raise McpError( + f"Repository alias '{alias_or_name}' is reserved but not enabled yet " + f"({profile.get('repoName')})." + ) + return str(profile["repoName"]) + + +def cache_root() -> Path: + configured = os.environ.get("DEEPWIKI_MCP_CACHE_DIR") + if configured: + return Path(configured).expanduser() + xdg_cache = os.environ.get("XDG_CACHE_HOME") + if xdg_cache: + return Path(xdg_cache).expanduser() / "deepwiki-mcp" + return Path.home() / ".cache" / "deepwiki-mcp" + + +def repo_cache_dir(repo_name: str) -> Path: + return cache_root() / repo_name.replace("/", "__") + + +def contents_cache_path(repo_name: str) -> Path: + return repo_cache_dir(repo_name) / "wiki-contents.md" + + +def write_text_atomic(path: Path, text: str) -> None: + path.parent.mkdir(parents=True, exist_ok=True) + tmp_path = path.with_suffix(path.suffix + ".tmp") + tmp_path.write_text(text, encoding="utf-8") + tmp_path.replace(path) + + +def parse_json(data: str) -> dict[str, Any]: + try: + parsed = json.loads(data) + except json.JSONDecodeError as exc: + raise McpError(f"DeepWiki MCP returned non-JSON content: {data[:500]}") from exc + if not isinstance(parsed, dict): + raise McpError(f"DeepWiki MCP returned an unexpected JSON payload: {data[:500]}") + return parsed + + +def read_sse_response(response: Any, expected_id: Optional[int]) -> dict[str, Any]: + data_lines: list[str] = [] + seen_payloads: list[str] = [] + max_seconds = float(os.environ.get("DEEPWIKI_MCP_STREAM_TIMEOUT", "120")) + deadline = time.monotonic() + max_seconds + + while True: + if time.monotonic() > deadline: + break + raw_line = response.readline() + if not raw_line: + break + + line = raw_line.decode("utf-8", errors="replace").rstrip("\r\n") + if line.startswith("data:"): + data_lines.append(line[5:].lstrip()) + continue + if line: + continue + + if not data_lines: + continue + + data = "\n".join(data_lines) + data_lines = [] + seen_payloads.append(data) + parsed = parse_json(data) + if expected_id is None or parsed.get("id") == expected_id: + return parsed + + if data_lines: + data = "\n".join(data_lines) + seen_payloads.append(data) + parsed = parse_json(data) + if expected_id is None or parsed.get("id") == expected_id: + return parsed + + preview = "\n".join(seen_payloads[-3:]) + raise McpError( + f"DeepWiki MCP stream ended without response id {expected_id} " + f"within {max_seconds:.0f}s: {preview[:500]}" + ) Review Comment: Fixed in `278a7f3f`: updated the DeepWiki MCP client as suggested and re-ran local validation, including `mvn apache-rat:check -DskipTests`, Claude plugin validation, Codex temp install, `structure`, and `context` commands. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
