Supporting Reference 6: File Reference
This index intentionally lists only files that belong to the GitHub repository. Local-only backup/sanitizer artifacts are excluded from this reference.
R6.1 Core runtime files (GitHub)
main_set.py- orchestrator UI, stage sequencing, and session flow.batch_image_quality_score.py- quality feature computation and score writes.caption_review_local.py- local caption/alt/keyword generation logic.review_editor.py- manual decision UI and row edits.db_uploader.py- FTP upload + MySQL upsert + mirror sync.init_db.py- local DB initialization from SQL contracts.metadata_evidence_pipeline.py- shared metadata evidence/proof pipeline entry used by V1.0 quality checks.run_metadata_quality_production.ps1- PowerShell wrapper for the production metadata-quality proof stage.simple_inference.py- local inference helper used by runtime workflows.amir2000_config.py- environment/config contract.
R6.2 Contract and data files (GitHub)
data/init/review_queue.sql- queue schema bootstrap.data/init/photos_info_revamp.sql- mirror schema bootstrap.data/init/metadata_quality.sql- metadata proof, repair audit, and ML feedback bootstrap.data/used_filenames.json- filename reservation ledger.data/folder_map.json- folder taxonomy mapping.data/location_list.json- location taxonomy list.data/autofix_dict.json- token normalization dictionary.data/spellcheck_exceptions.json- spellcheck allowlist for valid domain terms.data/new_taxonomy_log.json- append log of taxonomy additions from UI flows.data/ui_state.json- persisted UI/editor runtime state.data/prefill_qc_last.json- latest prefill duplicate/suspicious QC report.
R6.3 V1.0 identifier and metadata-quality scripts (GitHub)
scripts/apply_identifier_router_result_to_db.py- applies identifier router output back to SQLite rows.scripts/download_identifier_models.py- prepares optional local identifier model assets.scripts/evidence_subject_pipeline.py- evidence-first subject/metadata helper pipeline.scripts/identifier_biology_bioclip.py- local biology identifier route.scripts/identifier_biology_inaturalist.py- optional biology reference route for taxon candidate support.scripts/identifier_biology_runner.py- production biology runner used by the identifier router.scripts/identifier_consensus.py- combines identifier candidates into a safer consensus result.scripts/identifier_db_setup.py- initializes identifier support tables/assets.scripts/identifier_general_vision.py- general visual identifier route for non-specialized subjects.scripts/identifier_router.py- routes a row or set to the appropriate identifier strategy.scripts/identifier_vehicle_aircraft.py- specialized route for vehicle and aircraft subjects, including visible airline/model/registration evidence.scripts/identifier_visual_evidence.py- extracts visual evidence used by subject/metadata proof.scripts/metadata_auto_repair_loop.py- bounded metadata repair loop for caption, alt text, and keywords.scripts/metadata_quality_production.py- production metadata proof stage that writesmetadata_quality.scripts/series_versioning.py- detects series/variant context for related images.scripts/subject_identifier_engine.py- shared subject identifier engine used by router flows.scripts/subject_identifier_production.py- production subject suggestion/identifier script with aircraft group consolidation guardrails.
R6.4 Helper scripts (GitHub)
helpers/setup_venv313_full.ps1- environment bootstrap and dependency installation.helpers/preflight_multiset.ps1- build-readiness checks and report generation.helpers/build_multiset.ps1- executable packaging flow.helpers/runtime_hook_samevenv_classifier.py- frozen-runtime hook that exposes same-venv classifier dependencies/DLL paths to the EXE.
Note: local-only backup/sanitizer scripts that are not in the public GitHub branch are intentionally not listed.
R6.5 Repository docs (GitHub)
README.md- setup and runtime overview.docs/init/review_queue.sql- docs copy of queue schema.docs/init/photos_info_revamp.sql- docs copy of mirror schema.docs/init/metadata_quality.sql- docs copy of metadata proof schema.