87 lines
3.1 KiB
Plaintext
87 lines
3.1 KiB
Plaintext
# Copy this file to .env and fill in your own values.
|
|
# Nothing here is a real credential.
|
|
|
|
USE_OLLAMA=true
|
|
OLLAMA_BASE_URL=http://localhost:11434
|
|
OLLAMA_MODEL_NAME=qwen2.5:1.5b
|
|
|
|
USE_EMBEDDINGS=true
|
|
EMBEDDINGS_MODEL=sentence-transformers/all-MiniLM-L6-v2
|
|
|
|
USE_PGVECTOR=true
|
|
DB_HOST=localhost
|
|
DB_PORT=5432
|
|
DB_NAME=pgvector
|
|
DB_USER=postgres
|
|
DB_PASSWORD=changeme
|
|
|
|
USE_S3=true
|
|
S3_ACCESS_KEY=your-do-spaces-key
|
|
S3_SECRET_KEY=your-do-spaces-secret
|
|
S3_ENDPOINT=https://your-space.region.digitaloceanspaces.com
|
|
S3_BUCKET=your-bucket
|
|
S3_REGION=region
|
|
|
|
# Optional - leave USE_GOOGLE_CSE=false if you don't have a Google CSE key
|
|
USE_GOOGLE_CSE=false
|
|
GOOGLE_API_KEY=
|
|
GOOGLE_CSE_ID=
|
|
|
|
USE_DDG_IMAGES=true
|
|
USE_OPEN_FACTS=true
|
|
USE_WIKIMEDIA=true
|
|
USE_PLAYWRIGHT_FALLBACK=true
|
|
|
|
MIN_IMAGE_BYTES=3000
|
|
|
|
# Product SKU: try a live web search for a real marketplace product ID
|
|
# (Amazon ASIN, Flipkart PID, etc.) before falling back to an internal SKU.
|
|
# Set to false to always generate internal SKUs only (faster, offline-safe).
|
|
ENABLE_SKU_WEB_LOOKUP=true
|
|
|
|
# Deterministic product validation (see app/services/product_validator.py).
|
|
# Final gate applied to every catalog row before it's kept; rows scoring
|
|
# below VALIDATION_REJECT_THRESHOLD are dropped (with reasons recorded),
|
|
# rows below VALIDATION_REVIEW_THRESHOLD are kept but flagged for human review.
|
|
ENABLE_PRODUCT_VALIDATION=true
|
|
VALIDATION_REJECT_THRESHOLD=0.35
|
|
VALIDATION_REVIEW_THRESHOLD=0.70
|
|
|
|
# Per-pack-size product images (see docs/PER_VARIANT_IMAGES.md). When true,
|
|
# each size variant of a product (e.g. 100g/200g/500g) gets its own
|
|
# size-targeted image search/upload instead of sharing one generic image
|
|
# set. Falls back to the old shared behaviour automatically whenever no
|
|
# size-specific match is found, so this is safe to leave on. Set to
|
|
# "false" to fully restore the old shared-image behaviour.
|
|
ENABLE_PER_VARIANT_IMAGES=true
|
|
# Max image candidates fetched per size variant (kept smaller than the
|
|
# ~24 used for the general product-level search, since a per-size search
|
|
# only needs a handful of genuinely matching candidates).
|
|
PER_VARIANT_IMAGE_MAX_RESULTS=10
|
|
|
|
# Barcode Retrieval & Product Enrichment (see
|
|
# app/services/enrichment/barcode/ and docs/BARCODE_ENRICHMENT.md). Set to
|
|
# false to disable barcode lookup entirely (rows are stored with
|
|
# barcode=NULL, barcode_lookup_status="disabled").
|
|
ENABLE_BARCODE_LOOKUP=true
|
|
BARCODE_LOOKUP_TIMEOUT_SECONDS=10
|
|
# 30 days, in seconds
|
|
BARCODE_LOOKUP_CACHE_TTL_SECONDS=2592000
|
|
BARCODE_LOOKUP_MAX_CONCURRENCY=5
|
|
BARCODE_COUNTRY_TAG=india
|
|
|
|
# GS1 India tier - leave both blank to skip (no free public API exists at
|
|
# time of writing; the cascade falls through to Open Food Facts). Fill in
|
|
# if/when real GS1 India ("Verified by GS1") API access is provisioned.
|
|
GS1_INDIA_API_BASE_URL=
|
|
GS1_INDIA_API_KEY=
|
|
|
|
# UPCItemDB - works out of the box on the free trial tier (no key needed,
|
|
# rate limited ~100 req/day). Set to switch to the paid production endpoint.
|
|
UPC_DATABASE_API_KEY=
|
|
|
|
# Lowest-trust fallback tier: DuckDuckGo search + manufacturer page text
|
|
# scan for a labelled barcode (uses the same `ddgs` dependency as SKU
|
|
# lookup). Set to false to skip this tier.
|
|
ENABLE_MANUFACTURER_SITE_LOOKUP=true
|