working scraper workflow using remote ollama

This commit is contained in:
2026-06-13 16:49:20 -07:00
parent 530fd93a0b
commit bfad71e96a
4 changed files with 25 additions and 4 deletions
+5 -1
View File
@@ -8,7 +8,11 @@ import threading
from loguru import logger
# --- Configuration ---
OLLAMA_URL = os.environ.get("OLLAMA_URL", "http://127.0.0.1:11434/api/generate")
from server.config import resolve_ollama_url
OLLAMA_URL = resolve_ollama_url(
os.environ.get("OLLAMA_URL", "http://127.0.0.1:11434/api/generate")
)
MODEL_NAME = os.environ.get("MODEL_NAME", "gemma4:e2b")
PORT = int(os.environ.get("PORT", "8080"))
POLL_INTERVAL = int(os.environ.get("POLL_INTERVAL", "21600")) # 6 hours
+5 -1
View File
@@ -5,7 +5,11 @@ import httpx
from loguru import logger
from typing import List, Dict
OLLAMA_URL = os.environ.get("OLLAMA_URL", "http://127.0.0.1:11434/api/generate")
from server.config import resolve_ollama_url
OLLAMA_URL = resolve_ollama_url(
os.environ.get("OLLAMA_URL", "http://127.0.0.1:11434/api/generate")
)
MODEL_NAME = os.environ.get("MODEL_NAME", "gemma4:e2b")
DATA_DIR = os.environ.get("DATA_DIR", "data")
INPUT_FILE = os.path.join(DATA_DIR, "output.json")
+2 -2
View File
@@ -74,12 +74,12 @@ def _generate_index(mp3_files, scripts_map):
audio_entries += f"""
<div class="episode">
<audio controls preload="none">
<source src="/{f}" type="audio/mpeg">
<source src="/audio/{f}" type="audio/mpeg">
Your browser does not support the audio element.
</audio>
<div class="episode-meta">
<span class="episode-title">{safe_name}</span>
<a href="/{f}" class="download-btn" download>Download</a>
<a href="/audio/{f}" class="download-btn" download>Download</a>
</div>
</div>"""
+13
View File
@@ -0,0 +1,13 @@
"""Shared pipeline configuration helpers."""
import os
_OLLAMA_DEFAULT = "http://127.0.0.1:11434/api/generate"
_OLLAMA_PATH = "/api/generate"
def resolve_ollama_url(raw: str) -> str:
"""Ensure the Ollama URL always ends with the API path."""
if raw.endswith(_OLLAMA_PATH):
return raw
return raw.rstrip("/") + _OLLAMA_PATH