Initial commit
This commit is contained in:
144
app/app.py
Normal file
144
app/app.py
Normal file
@@ -0,0 +1,144 @@
|
||||
from flask import Flask, request, redirect
|
||||
import os, json, requests
|
||||
from redis import Redis
|
||||
from rq import Queue
|
||||
|
||||
MEILI_URL = os.getenv("MEILI_URL", "http://meili:7700")
|
||||
MEILI_KEY = os.getenv("MEILI_KEY", "")
|
||||
REDIS_URL = os.getenv("REDIS_URL", "redis://redis:6379/0")
|
||||
|
||||
app = Flask(__name__)
|
||||
q = Queue(connection=Redis.from_url(REDIS_URL))
|
||||
|
||||
PAGE = """
|
||||
<!doctype html><html><head><meta charset="utf-8">
|
||||
<title>PodX - unified search</title>
|
||||
<style>
|
||||
body{font-family:system-ui, sans-serif;max-width:880px;margin:2rem auto;padding:0 1rem}
|
||||
form{display:flex;gap:.5rem;margin-bottom:1rem}
|
||||
input[type=url]{flex:1;padding:.7rem}
|
||||
button{padding:.7rem 1rem}
|
||||
.card{border:1px solid #ddd;padding:1rem;border-radius:8px;margin:.5rem 0}
|
||||
small{color:#666}
|
||||
input[type=search]{width:100%;padding:.6rem;margin:.5rem 0 1rem}
|
||||
mark{background: #fff2a8}
|
||||
.badge{display:inline-block;font-size:.75rem;border:1px solid #999;padding:.1rem .4rem;border-radius:999px;margin-right:.4rem}
|
||||
</style></head><body>
|
||||
<h1>PodX</h1>
|
||||
<form action="/enqueue" method="post">
|
||||
<input type="url" name="url" placeholder="Paste podcast/video URL…" required>
|
||||
<button type="submit">Fetch & Transcribe</button>
|
||||
</form>
|
||||
<details><summary>Batch</summary>
|
||||
<form action="/enqueue_batch" method="post">
|
||||
<textarea name="urls" rows="4" style="width:100%" placeholder="One URL per line"></textarea>
|
||||
<button type="submit">Queue All</button>
|
||||
</form>
|
||||
</details>
|
||||
|
||||
<h2>Unified search (podcasts + PDFs + EPUB + Kiwix)</h2>
|
||||
<form id="sform">
|
||||
<input type="search" name="q" placeholder='e.g., "vector database" OR retrieval augmented generation' autofocus />
|
||||
</form>
|
||||
<div id="results"></div>
|
||||
|
||||
<script>
|
||||
const form = document.getElementById('sform');
|
||||
async function doSearch(){
|
||||
const q = new URLSearchParams(new FormData(form)).toString();
|
||||
const r = await fetch('/search?'+q);
|
||||
document.getElementById('results').innerHTML = await r.text();
|
||||
}
|
||||
form.addEventListener('input', doSearch);
|
||||
doSearch();
|
||||
</script>
|
||||
|
||||
<h2>Recent jobs</h2>
|
||||
<div id="feed"></div>
|
||||
<script>
|
||||
(async function poll(){
|
||||
try{
|
||||
const r = await fetch('/recent');
|
||||
document.getElementById('feed').innerHTML = await r.text();
|
||||
}catch(e){}
|
||||
setTimeout(poll, 4000);
|
||||
})();
|
||||
</script>
|
||||
</body></html>
|
||||
"""
|
||||
|
||||
def meili_search(qstr, limit=30):
|
||||
if not qstr.strip(): return []
|
||||
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}))
|
||||
return r.json().get("hits", [])
|
||||
|
||||
@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"<div class='card'><b>{item.get('title','')}</b><br><small>{item.get('uploader','')} — {item.get('date','')} — {item.get('status','')}</small><br><small>{item.get('path','')}</small></div>")
|
||||
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
|
||||
open_link = f"/open?file={{requests.utils.quote(src)}}&t={ts}" if typ=='podcast' else f"/open?file={{requests.utils.quote(src)}}"
|
||||
badge = f"<span class='badge'>{typ}</span>"
|
||||
out.append(
|
||||
f"<div class='card'><b>{badge}{t}</b><br><small>{src}</small>"
|
||||
f"<p>{ctx}</p>"
|
||||
f"<a href='{open_link}'>Open</a>"
|
||||
f"{' | <a href=\"/subtitle?file='+requests.utils.quote(src)+'\">Transcript</a>' if typ=='podcast' else ''}"
|
||||
f"</div>"
|
||||
)
|
||||
return "\n".join(out) or "<small>No results yet.</small>"
|
||||
|
||||
@app.get("/open")
|
||||
def open_local():
|
||||
file = request.args.get("file","")
|
||||
t = int(request.args.get("t","0"))
|
||||
return f"<pre>{file}\nStart at: {t} sec</pre>"
|
||||
|
||||
@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"<pre>{f.read()}</pre>"
|
||||
return "<small>No VTT found.</small>"
|
||||
Reference in New Issue
Block a user