from flask import Flask, request, redirect import os, json, time, requests from pathlib import Path from redis import Redis from rq import Queue MEILI_URL = os.getenv("MEILI_URL", "http://meili:7700") MEILI_KEY = os.getenv("MEILI_KEY", "") # from .env REDIS_URL = os.getenv("REDIS_URL", "redis://redis:6379/0") app = Flask(__name__) FEED_LOG = Path(os.getenv("TRANSCRIPT_ROOT", "/transcripts")) / "_feed.log" q = Queue(connection=Redis.from_url(REDIS_URL)) PAGE = """ PodX - unified search

PodX

Batch

Unified search (podcasts + PDFs + EPUB + Kiwix + Web)

Recent jobs

Activity

Loading…





"""

def read_feed_tail(max_lines: int = 200):
    if not FEED_LOG.exists():
        return []
    try:
        with open(FEED_LOG, "rb") as f:
            try:
                f.seek(-65536, 2)  # read last ~64KB
            except OSError:
                f.seek(0)
            data = f.read().decode("utf-8", errors="ignore")
    except Exception:
        return []
    lines = [x.strip() for x in data.splitlines() if x.strip()]
    events = []
    for ln in lines[-max_lines:]:
        try:
            events.append(json.loads(ln))
        except Exception:
            pass
    return events

@app.get("/api/status")
def api_status():
    events = read_feed_tail(200)
    last = events[-1] if events else {}
    summary = {
        "last_status": last.get("status"),
        "last_title": last.get("title") or last.get("path") or last.get("url"),
        "last_time": int(time.time()),
        "count": len(events),
    }
    return {"ok": True, "summary": summary, "events": events}

def meili_search(qstr, limit=30):
    if not qstr.strip():
        return []
    try:
        r = requests.post(
            f"{MEILI_URL}/indexes/library/search",
            headers={"Authorization": f"Bearer {MEILI_KEY}", "Content-Type": "application/json"},
            data=json.dumps({"q": qstr, "limit": limit}),
            timeout=5,
        )
        if r.status_code != 200:
            return []
        return r.json().get("hits", [])
    except Exception:
        return []

@app.get("/health")
def health():
    return "ok"

@app.get("/")
def index():
    return PAGE

@app.post("/enqueue")
def enqueue():
    url = request.form["url"].strip()
    q.enqueue("worker.handle_url", url)
    return redirect("/")

@app.post("/enqueue_batch")
def enqueue_batch():
    urls = [u.strip() for u in request.form["urls"].splitlines() if u.strip()]
    for u in urls: q.enqueue("worker.handle_url", u)
    return redirect("/")

@app.get("/recent")
def recent():
    try:
        with open("/transcripts/_feed.log", "r", encoding="utf-8") as f:
            tail = f.readlines()[-40:]
    except FileNotFoundError:
        tail=[]
    html = []
    for line in reversed(tail):
        try: item = json.loads(line)
        except: continue
        html.append(f"
{item.get('title','')}
{item.get('uploader','')} — {item.get('date','')} — {item.get('status','')}
{item.get('path','')}
")
    return "\n".join(html)

@app.get("/search")
def search():
    qstr = request.args.get("q","")
    hits = meili_search(qstr)
    out=[]
    for h in hits:
        t = h.get("title","")
        src = h.get("source","")
        typ = h.get("type","")
        ctx = h.get("_formatted",{}).get("text", h.get("text","")[:300])
        segs = h.get("segments",[])
        ts = int(segs[0]["start"]) if segs else 0
        if typ == 'podcast':
            open_link = f"/open?file={requests.utils.quote(src)}&t={ts}"
        else:
            open_link = f"/open?file={requests.utils.quote(src)}"
        transcript_link = f" | Transcript" if typ == 'podcast' else ""
        badge = f"{typ}"
        out.append(
          f"
{badge}{t}
{src}"
          f"
{ctx}
"
          f"Open"
          f"{transcript_link}"
          f"
"
        )
    return "\n".join(out) or "No results yet."

@app.get("/open")
def open_local():
    file = request.args.get("file","")
    t = int(request.args.get("t","0"))
    return f"
{file}\nStart at: {t} sec
"

@app.get("/subtitle")
def subtitle():
    file = request.args.get("file","")
    base = os.path.splitext(os.path.basename(file))[0]
    p = f"/transcripts/{base}.vtt"
    if os.path.exists(p):
        with open(p,"r",encoding="utf-8") as f:
            return f"
{f.read()}
"
    return "No VTT found."