From bfad71e96a04b537dcb26d4af567503567e8e23c Mon Sep 17 00:00:00 2001 From: kfj001 Date: Sat, 13 Jun 2026 16:49:20 -0700 Subject: [PATCH] working scraper workflow using remote ollama --- pipeline.py | 6 +++++- reader.py | 6 +++++- server/app.py | 4 ++-- server/config.py | 13 +++++++++++++ 4 files changed, 25 insertions(+), 4 deletions(-) create mode 100644 server/config.py diff --git a/pipeline.py b/pipeline.py index bbcd712..ffad3f3 100644 --- a/pipeline.py +++ b/pipeline.py @@ -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 diff --git a/reader.py b/reader.py index 1c76ce9..13a29f1 100644 --- a/reader.py +++ b/reader.py @@ -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") diff --git a/server/app.py b/server/app.py index c0c40e8..724f9a0 100644 --- a/server/app.py +++ b/server/app.py @@ -74,12 +74,12 @@ def _generate_index(mp3_files, scripts_map): audio_entries += f"""
{safe_name} - Download + Download
""" diff --git a/server/config.py b/server/config.py new file mode 100644 index 0000000..07b85c5 --- /dev/null +++ b/server/config.py @@ -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