Files
diarization-ui/app.py

474 lines
18 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import json
import os
import sqlite3
from datetime import datetime
from typing import Optional
import requests
from fastapi import FastAPI, File, Form, HTTPException, UploadFile
from fastapi.responses import HTMLResponse, PlainTextResponse
API_BASE = os.getenv("API_BASE", "http://gx10.aquantico.lan:8093").rstrip("/")
OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL", "http://gx10.aquantico.lan:11434").rstrip("/")
OLLAMA_MODEL = os.getenv("OLLAMA_MODEL", "qwen3.5:9b")
DB_PATH = os.getenv("DB_PATH", "/data/ui.db")
app = FastAPI(title="Diarization UI")
def db():
conn = sqlite3.connect(DB_PATH)
conn.row_factory = sqlite3.Row
return conn
def now_iso() -> str:
return datetime.utcnow().isoformat()
def init_db():
os.makedirs(os.path.dirname(DB_PATH), exist_ok=True)
with db() as c:
c.execute(
"""
CREATE TABLE IF NOT EXISTS projects (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name TEXT UNIQUE NOT NULL,
created_at TEXT NOT NULL
)
"""
)
c.execute(
"""
CREATE TABLE IF NOT EXISTS prompts (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name TEXT UNIQUE NOT NULL,
prompt TEXT NOT NULL,
created_at TEXT NOT NULL,
updated_at TEXT NOT NULL
)
"""
)
c.execute(
"""
CREATE TABLE IF NOT EXISTS documents (
id INTEGER PRIMARY KEY AUTOINCREMENT,
project_id INTEGER NOT NULL,
kind TEXT NOT NULL, -- transcript|analysis
title TEXT NOT NULL,
content_md TEXT NOT NULL,
source_document_id INTEGER,
prompt_id INTEGER,
raw_json TEXT,
created_at TEXT NOT NULL,
FOREIGN KEY(project_id) REFERENCES projects(id),
FOREIGN KEY(source_document_id) REFERENCES documents(id),
FOREIGN KEY(prompt_id) REFERENCES prompts(id)
)
"""
)
# defaults
c.execute("INSERT OR IGNORE INTO projects(name, created_at) VALUES (?,?)", ("Default", now_iso()))
c.execute(
"INSERT OR IGNORE INTO prompts(name, prompt, created_at, updated_at) VALUES (?,?,?,?)",
(
"Zusammenfassung",
"Erstelle eine prägnante Zusammenfassung des Gesprächs in Stichpunkten.",
now_iso(),
now_iso(),
),
)
c.execute(
"INSERT OR IGNORE INTO prompts(name, prompt, created_at, updated_at) VALUES (?,?,?,?)",
(
"Aufgaben",
"Extrahiere alle Aufgaben. Gib pro Aufgabe: Verantwortlich, Aufgabe, Deadline (falls vorhanden), Priorität.",
now_iso(),
now_iso(),
),
)
def layout(title: str, body: str) -> str:
return f"""
<!doctype html>
<html><head><meta charset='utf-8'><meta name='viewport' content='width=device-width, initial-scale=1'>
<title>{title}</title>
<style>
body{{font-family:Arial;margin:0;display:flex;min-height:100vh}}
nav{{width:240px;background:#111;color:#fff;padding:16px}}
nav a{{display:block;color:#fff;text-decoration:none;padding:8px 10px;border-radius:6px;margin:4px 0}}
nav a:hover{{background:#2a2a2a}}
main{{flex:1;padding:20px;max-width:1200px}}
.card{{border:1px solid #ddd;border-radius:8px;padding:12px;margin:10px 0}}
input,select,textarea,button{{padding:8px;font-size:14px}}
textarea{{width:100%;min-height:140px}}
pre{{white-space:pre-wrap;background:#111;color:#0f0;padding:10px;border-radius:8px}}
.row{{display:flex;gap:8px;flex-wrap:wrap;align-items:center}}
small{{color:#666}}
</style></head>
<body>
<nav>
<h3>Menü</h3>
<a href='/'>Upload</a>
<a href='/library'>Datenbank</a>
<a href='/prompts'>Prompt-Konfig</a>
<a href='/run'>Prompt ausführen</a>
<a href='/healthz'>Health</a>
</nav>
<main>{body}</main>
</body></html>
"""
def get_projects():
with db() as c:
return c.execute("SELECT id,name FROM projects ORDER BY name").fetchall()
def get_project_name(project_id: int) -> str:
with db() as c:
r = c.execute("SELECT name FROM projects WHERE id=?", (project_id,)).fetchone()
return r[0] if r else ""
def get_prompts():
with db() as c:
return c.execute("SELECT id,name,prompt FROM prompts ORDER BY name").fetchall()
@app.on_event("startup")
def startup():
init_db()
@app.get("/healthz")
def healthz():
return {"ok": True, "api_base": API_BASE, "ollama_base_url": OLLAMA_BASE_URL, "ollama_model": OLLAMA_MODEL, "db_path": DB_PATH}
@app.get("/", response_class=HTMLResponse)
def upload_page(msg: str = ""):
projects = get_projects()
opts = "".join([f"<option value='{p['id']}'>{p['name']}</option>" for p in projects])
body = f"""
<h2>Audio Upload</h2>
<p>Audio wird transkribiert + mit Sprechern angereichert und als Dokument gespeichert.</p>
{f"<p><b>{msg}</b></p>" if msg else ""}
<form action='/upload' method='post' enctype='multipart/form-data' class='card'>
<div class='row'>
<label>Projekt:</label>
<select name='project_id'>{opts}</select>
<input name='title' placeholder='Titel (optional)'>
</div>
<div class='row' style='margin-top:8px'>
<input type='file' name='file' accept='audio/*' required>
<button type='submit'>Verarbeiten & speichern</button>
</div>
</form>
"""
return layout("Upload", body)
@app.post("/projects", response_class=HTMLResponse)
def add_project(name: str = Form(...)):
with db() as c:
c.execute("INSERT INTO projects(name, created_at) VALUES (?,?)", (name.strip(), now_iso()))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.post("/projects/update", response_class=HTMLResponse)
def rename_project(id: int = Form(...), name: str = Form(...)):
with db() as c:
c.execute("UPDATE projects SET name=? WHERE id=?", (name.strip(), id))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.post("/projects/{project_id}/delete", response_class=HTMLResponse)
def delete_project(project_id: int):
with db() as c:
default = c.execute("SELECT id FROM projects WHERE name='Default'").fetchone()
if not default:
c.execute("INSERT INTO projects(name, created_at) VALUES (?,?)", ("Default", now_iso()))
default_id = c.execute("SELECT id FROM projects WHERE name='Default'").fetchone()[0]
else:
default_id = default[0]
if project_id == default_id:
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
c.execute("UPDATE documents SET project_id=? WHERE project_id=?", (default_id, project_id))
c.execute("DELETE FROM projects WHERE id=?", (project_id,))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.post("/upload", response_class=HTMLResponse)
async def upload(project_id: int = Form(...), title: str = Form(""), file: UploadFile = File(...)):
data = await file.read()
if not data:
raise HTTPException(400, "Leere Datei")
files = {"file": (file.filename or "audio.bin", data, file.content_type or "application/octet-stream")}
r = requests.post(f"{API_BASE}/transcribe-diarize", files=files, timeout=1800)
if r.status_code >= 400:
raise HTTPException(r.status_code, r.text)
payload = r.json()
content_md = payload.get("formatted_text", "")
doc_title = (title or "").strip() or (file.filename or "Transkript")
with db() as c:
cur = c.execute(
"INSERT INTO documents(project_id, kind, title, content_md, raw_json, created_at) VALUES (?,?,?,?,?,?)",
(project_id, "transcript", doc_title, content_md, json.dumps(payload, ensure_ascii=False), now_iso()),
)
doc_id = cur.lastrowid
return HTMLResponse(f"<meta http-equiv='refresh' content='0; url=/document/{doc_id}'>")
@app.get("/library", response_class=HTMLResponse)
def library(project_id: Optional[int] = None):
with db() as c:
projects = c.execute("SELECT id,name FROM projects ORDER BY name").fetchall()
if project_id:
docs = c.execute(
"""
SELECT d.id,d.kind,d.title,d.created_at,p.name AS project
FROM documents d JOIN projects p ON p.id=d.project_id
WHERE d.project_id=? ORDER BY d.id DESC
""",
(project_id,),
).fetchall()
else:
docs = c.execute(
"""
SELECT d.id,d.kind,d.title,d.created_at,p.name AS project
FROM documents d JOIN projects p ON p.id=d.project_id
ORDER BY d.id DESC LIMIT 200
"""
).fetchall()
p_opts = "<option value=''>Alle</option>" + "".join(
[f"<option value='{p['id']}' {'selected' if project_id==p['id'] else ''}>{p['name']}</option>" for p in projects]
)
proj_opts = "".join([f"<option value='{p['id']}'>{p['name']}</option>" for p in projects])
items = "".join(
[
f"<div class='card'><b>#{d['id']}</b> [{d['kind']}] {d['title']}<br><small>{d['project']} · {d['created_at']}</small><br>"
f"<a href='/document/{d['id']}'>Ansehen</a> | <a href='/document/{d['id']}/download.md'>Download .md</a>"
f"<form method='post' action='/document/{d['id']}/rename' class='row' style='margin-top:8px'>"
f"<input name='title' value='{d['title']}'><button>Umbenennen</button></form>"
f"<form method='post' action='/document/{d['id']}/move' class='row' style='margin-top:6px'>"
f"<select name='project_id'>{proj_opts}</select><button>Verschieben</button></form>"
f"<form method='post' action='/document/{d['id']}/delete' onsubmit=\"return confirm('Dokument löschen?')\" style='margin-top:6px'>"
f"<button>Löschen</button></form></div>"
for d in docs
]
)
body = f"""
<h2>Datenbank / Dokumente</h2>
<form method='get' class='row card'>
<label>Projekt:</label>
<select name='project_id'>{p_opts}</select>
<button type='submit'>Filtern</button>
</form>
{items or '<p>Keine Einträge.</p>'}
"""
return layout("Library", body)
@app.get("/document/{doc_id}", response_class=HTMLResponse)
def view_document(doc_id: int):
with db() as c:
d = c.execute(
"""
SELECT d.*, p.name AS project, pr.name AS prompt_name
FROM documents d
JOIN projects p ON p.id=d.project_id
LEFT JOIN prompts pr ON pr.id=d.prompt_id
WHERE d.id=?
""",
(doc_id,),
).fetchone()
if not d:
raise HTTPException(404, "not found")
body = f"""
<h2>Dokument #{d['id']} {d['title']}</h2>
<p><small>Projekt: {d['project']} · Typ: {d['kind']} · {d['created_at']}</small></p>
<p><a href='/document/{doc_id}/download.md'>Download .md</a></p>
<pre>{(d['content_md'] or '').replace('<','&lt;')}</pre>
"""
return layout("Dokument", body)
@app.get("/document/{doc_id}/download.md", response_class=PlainTextResponse)
def download_md(doc_id: int):
with db() as c:
d = c.execute("SELECT title,content_md FROM documents WHERE id=?", (doc_id,)).fetchone()
if not d:
raise HTTPException(404, "not found")
return PlainTextResponse(d["content_md"], headers={"Content-Disposition": f"attachment; filename=document_{doc_id}.md"})
@app.post("/document/{doc_id}/rename", response_class=HTMLResponse)
def rename_document(doc_id: int, title: str = Form(...)):
with db() as c:
c.execute("UPDATE documents SET title=? WHERE id=?", (title.strip(), doc_id))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/library'>")
@app.post("/document/{doc_id}/move", response_class=HTMLResponse)
def move_document(doc_id: int, project_id: int = Form(...)):
with db() as c:
c.execute("UPDATE documents SET project_id=? WHERE id=?", (project_id, doc_id))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/library'>")
@app.post("/document/{doc_id}/delete", response_class=HTMLResponse)
def delete_document(doc_id: int):
with db() as c:
c.execute("DELETE FROM documents WHERE id=?", (doc_id,))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/library'>")
@app.get("/prompts", response_class=HTMLResponse)
def prompts_page():
with db() as c:
prompts = c.execute("SELECT * FROM prompts ORDER BY name").fetchall()
projects = c.execute("SELECT id,name FROM projects ORDER BY name").fetchall()
p_list = "".join(
[
f"<div class='card'><b>{p['name']}</b><pre>{(p['prompt'] or '').replace('<','&lt;')}</pre>"
f"<form method='post' action='/prompts/update'><input type='hidden' name='id' value='{p['id']}'><input name='name' value='{p['name']}'><br><textarea name='prompt'>{p['prompt']}</textarea><br><button>Speichern</button></form>"
f"<form method='post' action='/prompts/{p['id']}/delete' onsubmit=\"return confirm('Prompt löschen?')\" style='margin-top:6px'><button>Löschen</button></form>"
f"</div>"
for p in prompts
]
)
project_opts = "".join([f"<option value='{p['name']}'>{p['name']}</option>" for p in projects])
project_list = "".join([
f"<div class='card'><b>{p['name']}</b>"
f"<form method='post' action='/projects/update' class='row' style='margin-top:6px'>"
f"<input type='hidden' name='id' value='{p['id']}'><input name='name' value='{p['name']}'><button>Umbenennen</button></form>"
f"<form method='post' action='/projects/{p['id']}/delete' onsubmit=\"return confirm('Projekt löschen? Dokumente werden auf Default verschoben.')\" style='margin-top:6px'><button>Löschen</button></form>"
f"</div>" for p in projects
])
body = f"""
<h2>Prompt-Konfiguration</h2>
<div class='card'>
<form method='post' action='/prompts/add'>
<h4>Neuer Prompt</h4>
<input name='name' placeholder='Name' required>
<br><textarea name='prompt' placeholder='Prompttext' required></textarea>
<br><button type='submit'>Anlegen</button>
</form>
</div>
<div class='card'>
<form method='post' action='/projects'>
<h4>Neues Projekt</h4>
<input name='name' list='projectNames' placeholder='Projektname' required>
<datalist id='projectNames'>{project_opts}</datalist>
<button type='submit'>Anlegen</button>
</form>
</div>
<h3>Projekte</h3>
{project_list}
<h3>Prompts</h3>
{p_list}
"""
return layout("Prompts", body)
@app.post("/prompts/add", response_class=HTMLResponse)
def prompt_add(name: str = Form(...), prompt: str = Form(...)):
with db() as c:
c.execute(
"INSERT INTO prompts(name,prompt,created_at,updated_at) VALUES (?,?,?,?)",
(name.strip(), prompt.strip(), now_iso(), now_iso()),
)
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.post("/prompts/update", response_class=HTMLResponse)
def prompt_update(id: int = Form(...), name: str = Form(...), prompt: str = Form(...)):
with db() as c:
c.execute("UPDATE prompts SET name=?, prompt=?, updated_at=? WHERE id=?", (name.strip(), prompt.strip(), now_iso(), id))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.post("/prompts/{prompt_id}/delete", response_class=HTMLResponse)
def prompt_delete(prompt_id: int):
with db() as c:
c.execute("DELETE FROM prompts WHERE id=?", (prompt_id,))
return HTMLResponse("<meta http-equiv='refresh' content='0; url=/prompts'>")
@app.get("/run", response_class=HTMLResponse)
def run_page():
with db() as c:
docs = c.execute("SELECT id,title,kind,created_at FROM documents ORDER BY id DESC LIMIT 200").fetchall()
prompts = c.execute("SELECT id,name FROM prompts ORDER BY name").fetchall()
d_opts = "".join([f"<option value='{d['id']}'>#{d['id']} [{d['kind']}] {d['title']}</option>" for d in docs])
p_opts = "".join([f"<option value='{p['id']}'>{p['name']}</option>" for p in prompts])
body = f"""
<h2>Prompt ausführen</h2>
<form method='post' action='/run' class='card'>
<label>Dokument:</label><br>
<select name='document_id' style='width:100%'>{d_opts}</select><br><br>
<label>Prompt:</label><br>
<select name='prompt_id' style='width:100%'>{p_opts}</select><br><br>
<button type='submit'>Ausführen (Qwen)</button>
</form>
"""
return layout("Run", body)
@app.post("/run", response_class=HTMLResponse)
def run_prompt(document_id: int = Form(...), prompt_id: int = Form(...)):
with db() as c:
doc = c.execute("SELECT * FROM documents WHERE id=?", (document_id,)).fetchone()
prm = c.execute("SELECT * FROM prompts WHERE id=?", (prompt_id,)).fetchone()
if not doc or not prm:
raise HTTPException(404, "Dokument oder Prompt nicht gefunden")
llm_prompt = (
"Du bist ein präziser Assistent. Antworte auf Deutsch.\n"
f"AUFTRAG:\n{prm['prompt']}\n\n"
f"TEXT:\n{doc['content_md']}\n"
)
r = requests.post(
f"{OLLAMA_BASE_URL}/api/generate",
json={"model": OLLAMA_MODEL, "prompt": llm_prompt, "stream": False},
timeout=1200,
)
if r.status_code >= 400:
raise HTTPException(r.status_code, r.text)
answer = r.json().get("response", "")
with db() as c:
cur = c.execute(
"""
INSERT INTO documents(project_id, kind, title, content_md, source_document_id, prompt_id, raw_json, created_at)
VALUES (?,?,?,?,?,?,?,?)
""",
(
doc["project_id"],
"analysis",
f"Analyse: {prm['name']} · {doc['title']}",
answer,
doc["id"],
prm["id"],
json.dumps({"ollama_response": r.json()}, ensure_ascii=False),
now_iso(),
),
)
new_id = cur.lastrowid
return HTMLResponse(f"<meta http-equiv='refresh' content='0; url=/document/{new_id}'>")