# Devour Example Configuration
# Copy this file to devour.yaml and customize for your needs

version: 1

# Storage paths
storage:
  docs_dir: ./devour_data/docs
  index_dir: ./devour_data/index
  metadata_dir: ./devour_data/metadata
  cache_dir: ./devour_data/cache

# Embedding settings (optional for lexical search; required for future embedding flows)
embeddings:
  provider: openai           # openai, mock
  model: text-embedding-3-small
  dimensions: 1536
  api_key: ${OPENAI_API_KEY} # Use environment variable
  batch_size: 100

# Vector database (optional in current local-first pipeline)
vector_db:
  type: memory               # memory, chromem
  persist: true
  similarity_metric: cosine

# Scraping settings
scraper:
  user_agent: "Devour/1.0 (+https://github.com/yourorg/devour)"
  timeout: 30s
  retry_count: 3
  retry_delay: 1s
  concurrency: 10
  rate_limit: 500ms
  max_depth: 3
  cache_dir: ./devour_data/cache

# Scheduler
scheduler:
  enabled: true
  interval: 72h              # Every 3 days
  check_method: hash         # hash, timestamp
  on_startup: false

# Server settings
server:
  mode: local                # local, remote
  transport: stdio           # stdio, http
  port: 8080
  host: localhost

# Local lexical indexing defaults
indexing:
  enabled: true
  auto_reindex: true
  snippet_length: 220
  max_docs: 10000

# Verification defaults
verification:
  enabled: true
  timeout: 90s

# Example sources
sources:
  # Web documentation
  - name: example-docs
    type: url
    url: https://docs.example.com
    include:
      - ".*\\.md"
      - ".*\\.html"
    exclude:
      - ".*/api/.*"
      - ".*/legacy/.*"
    schedule: 24h

  # OpenAPI specification
  - name: api-spec
    type: openapi
    url: https://api.example.com/openapi.json
    schedule: 168h           # Weekly

  # GitHub repository docs
  - name: github-repo
    type: github
    repo: org/repository
    branch: main
    include:
      - "(?i)(^|/)README\\.md$"
      - "(?i)(^|/)docs?/"

  # Local files
  - name: local-docs
    type: local
    path: ./docs
    include:
      - ".*\\.md"
      - ".*\\.txt"

  # Self-hosted search API (e.g. SearxNG) with no API key
  - name: local-searxng-go
    type: localsearch
    url: http://127.0.0.1:8080/search
    query: golang http client
    result_limit: 8
    domains:
      - pkg.go.dev
      - go.dev

  # New framework examples
  - name: nextjs-docs
    type: url
    url: https://nextjs.org/docs

  - name: express-docs
    type: url
    url: https://expressjs.com/en/guide/routing.html