fetch_lesswrong.py

"""Fetch LessWrong posts via GraphQL. Two-step approach:
  1. Resolve tag IDs for the names we care about.
  2. Fetch posts by tagId using filterSettings.

Writes: data/lesswrong_posts.jsonl
"""
import json
import os
import re
import sys
import time
from pathlib import Path
import requests

PROJECT_DIR = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, PROJECT_DIR)
from config import LW_TAGS, START_DATE

API = "https://www.lesswrong.com/graphql"
OUT = Path(PROJECT_DIR) / "data" / "lesswrong_posts.jsonl"
MIN_WORDS = 200


def strip_html(html):
    text = re.sub(r"<[^>]+>", " ", html or "")
    text = re.sub(r"&nbsp;", " ", text)
    text = re.sub(r"&amp;", "&", text)
    text = re.sub(r"&lt;", "<", text)
    text = re.sub(r"&gt;", ">", text)
    text = re.sub(r"&quot;", '"', text)
    text = re.sub(r"&#39;", "'", text)
    text = re.sub(r"\s+", " ", text)
    return text.strip()


def gql(query, variables=None):
    r = requests.post(API, json={"query": query, "variables": variables or {}}, timeout=60)
    r.raise_for_status()
    j = r.json()
    if "errors" in j:
        raise RuntimeError(f"GraphQL errors: {j['errors']}")
    return j["data"]


def resolve_tag_ids(names):
    """Walk the alphabetical tag list and pick out the ones matching names."""
    print("Resolving tag IDs...", flush=True)
    name_set = {n.lower().strip() for n in names}
    found = {}
    offset = 0
    while offset < 10000 and len(found) < len(name_set):
        data = gql("""
        query GetTags($limit: Int, $offset: Int) {
          tags(input: {terms: {view: "allTagsAlphabetical", limit: $limit, offset: $offset}}) {
            results { _id name slug }
          }
        }
        """, {"limit": 500, "offset": offset})
        results = data.get("tags", {}).get("results", [])
        if not results: break
        for t in results:
            if t["name"].lower().strip() in name_set:
                found[t["name"]] = t["_id"]
                print(f"  found {t['name']} -> {t['_id']}", flush=True)
        offset += len(results)
        time.sleep(0.3)
    missing = name_set - {n.lower().strip() for n in found.keys()}
    if missing:
        print(f"  MISSING: {missing}", flush=True)
    return found


def fetch_posts_for_tag(tag_id, tag_name, max_posts=5000, batch=200):
    """Fetch posts requiring a given tag, paginated by offset."""
    all_posts = []
    offset = 0
    while offset < max_posts:
        data = gql("""
        query GetPosts($tagId: String!, $limit: Int, $offset: Int) {
          posts(input: {terms: {
            view: "new",
            filterSettings: {tags: [{tagId: $tagId, filterMode: "Required"}]},
            limit: $limit, offset: $offset, after: "%s"
          }}) {
            results {
              _id title slug postedAt baseScore author wordCount
              contents { html }
              tags { name }
              user { displayName }
            }
          }
        }
        """ % START_DATE, {"tagId": tag_id, "limit": batch, "offset": offset})
        results = data.get("posts", {}).get("results", [])
        if not results: break
        all_posts.extend(results)
        print(f"  '{tag_name}': {len(all_posts)} fetched (offset {offset})", flush=True)
        if len(results) < batch: break
        offset += batch
        time.sleep(0.4)
    return all_posts


def main():
    OUT.parent.mkdir(parents=True, exist_ok=True)
    tag_ids = resolve_tag_ids(LW_TAGS)
    if not tag_ids:
        print("No tags resolved; abort.")
        return

    seen = set()
    kept = 0
    with open(OUT, "w") as f:
        for tag_name, tag_id in tag_ids.items():
            print(f"\n=== {tag_name} ===", flush=True)
            try:
                posts = fetch_posts_for_tag(tag_id, tag_name)
            except Exception as e:
                print(f"  FAIL: {e}", flush=True)
                continue
            for p in posts:
                pid = p.get("_id")
                if not pid or pid in seen: continue
                body = strip_html((p.get("contents") or {}).get("html") or "")
                wc = p.get("wordCount") or len(body.split())
                if wc < MIN_WORDS or not body: continue
                f.write(json.dumps({
                    "id": pid,
                    "title": p.get("title"),
                    "slug": p.get("slug"),
                    "posted_at": p.get("postedAt"),
                    "body": body,
                    "word_count": wc,
                    "tags": [t["name"] for t in (p.get("tags") or [])],
                    "base_score": p.get("baseScore"),
                    "author": (p.get("user") or {}).get("displayName") or p.get("author"),
                }) + "\n")
                seen.add(pid)
                kept += 1
    print(f"\n=== done: {kept} unique posts ≥{MIN_WORDS} words → {OUT} ===", flush=True)


if __name__ == "__main__":
    main()