Skip to content

Configuration

Library-wide settings via environment variables or programmatic configuration.

PyGAEBSettings

Setting Env Var Default Description
default_model PYGAEB_DEFAULT_MODEL anthropic/claude-sonnet-4-6 LLM model for classification and extraction
classifier_concurrency PYGAEB_CLASSIFIER_CONCURRENCY 5 Max parallel LLM calls
xsd_dir PYGAEB_XSD_DIR None Directory containing XSD schemas for validation
log_level PYGAEB_LOG_LEVEL WARNING Logging level applied to all pygaeb.* loggers
large_file_threshold_mb PYGAEB_LARGE_FILE_THRESHOLD_MB 50 Files above this size trigger large-file optimisations
large_file_item_threshold PYGAEB_LARGE_FILE_ITEM_THRESHOLD 10000 Item count above which large-file heuristics apply
max_file_size_mb PYGAEB_MAX_FILE_SIZE_MB 100 Hard limit on input file size (0 = disabled)

PyGAEBSettings

Bases: BaseSettings

Library-wide configuration — supports env vars, .env files, and constructor kwargs.

configure

configure

configure(default_model=None, classifier_concurrency=None, xsd_dir=None, log_level=None, large_file_threshold_mb=None, large_file_item_threshold=None, max_file_size_mb=None)

Override settings for the current session. Only supplied values are changed.

get_settings

get_settings

get_settings()

Get or create the shared settings instance.

reset_settings

reset_settings

reset_settings()

Reset to default settings. Useful in tests.