- Add Rule 11 for Z.AI Coding Plan API usage (not BigModel)
- Add transliteration standards for non-Latin scripts
- Document GLM model options and Python implementation
- Add emic_name, name_language, and standardized_name to 1,781 custodian files
- Remove 2,239 duplicate files that had name suffixes in filename
- Consolidate data into base GHCID files per PID stability rules
- Part of UNESCO Memory of the World custodian enrichment
Database Panels:
- Add D3.js force-directed graph visualization to Oxigraph and TypeDB panels
- Add 'Explore' tab with class/entity browser, graph/table toggle, and search
- Add data explorer to PostgreSQL panel with table browser, pagination, search, export
- Fix SPARQL variable naming bug in Oxigraph getGraphData() function
- Add node details panel showing selected entity attributes
- Add zoom/pan controls and node coloring by entity type
Map Features:
- Add TimelineSlider component for temporal filtering of institutions
- Support dual-handle range slider with decade histogram
- Add quick presets (Ancient, Medieval, Modern, Contemporary)
- Show institution density visualization by founding decade
Hooks:
- Extend useOxigraph with getGraphData() for graph visualization
- Extend useTypeDB with getGraphData() for graph visualization
- Extend usePostgreSQL with getTableData() and exportTableData()
- Improve useDuckLakeInstitutions with temporal filtering support
Styles:
- Add HeritageDashboard.css with shared panel styling
- Add TimelineSlider.css for timeline component styling
Replace static netherlands_municipalities_simplified.geojson with dynamic
PostGIS API call to /boundaries/countries/NL/admin2/geojson.
Transform API response properties to expected format:
- API: {code, name, name_local, admin1_code, admin1_name}
- Expected: {code, naam, provincieCode, provincieNaam}
This ensures NL boundary data comes from the authoritative PostGIS
database rather than a static file that could become outdated.
- Handle single-letter GLAM type codes (G, L, A, M, O, R, C, etc.)
- Handle legacy GRP.HER.* format
- Support compound types like 'M,F' -> 'MUSEUM,FEATURES'
- Fix type hint syntax for Python 3.10+
- Add search bar to filter table data across all columns
- Filter web archive claims by selected page
- Include source_page in claim queries for filtering
- Fix TypeScript unused parameter warning
New React hook that fetches administrative boundaries from the PostGIS API:
- Supports international boundaries (NL, JP, CZ, DE, BE, CH, AT, etc.)
- Caches admin1, admin2, and GeoJSON data
- Provides point-in-polygon lookup
- Includes utility functions for filtering boundaries by code/name
- Replaces static GeoJSON file loading pattern
Add new endpoints for querying administrative boundaries:
- GET /boundaries/countries - List all countries with boundary data
- GET /boundaries/countries/{code}/admin1 - List provinces/states
- GET /boundaries/countries/{code}/admin2 - List municipalities
- GET /boundaries/lookup?lat=&lon= - Point-in-polygon lookup
- GET /boundaries/admin2/{id}/geojson - Get single boundary GeoJSON
- GET /boundaries/countries/{code}/admin2/geojson - Get all admin2 as GeoJSON
- GET /boundaries/stats - Get boundary statistics
Uses proper joins through country_id/admin1_id foreign keys.
Remove redundant ch_annotator metadata and duplicate ghcid_history entries
that were causing YAML parsing issues. Files now have cleaner, more
consistent structure while preserving all essential data.
- Add fix_yaml_history.py and fix_yaml_history_v2.py for cleaning up
malformed ghcid_history entries with duplicate/redundant data
- Update load_custodians_to_ducklake.py for DuckDB lakehouse loading
- Update migrate_web_archives.py for web archive management
- Update deploy.sh with improvements
- Ignore entire data/ducklake/ directory (generated databases)
Rename 144 custodian files from XXX placeholders to resolved city codes:
- BR (65): ASA, RBR, MAN, FAZ, MAC, VIT, FOR, GUA, BRE, GOI, SLU, BHO, XAN, CGR, CUI, etc.
- CH (24): ZUR, BER, GEN, BAL, LUC, etc.
- MX (23): MEX, GDL, MTY, PUE, etc.
- CL (9): SCL, VAL, etc.
- CZ (5): PRG, BRN, MSV, etc.
- KR (4): SEL, etc.
- GB (4): LON, etc.
- FR (3): PAR, etc.
- IN (2): DEL, etc.
- PH, JP, EE (1 each)
City codes derived from GeoNames reverse geocoding using institution coordinates.
GHCID format: {COUNTRY}-{REGION}-{CITY}-{TYPE}-{ABBREV}
Enrichment scripts for country-specific city data:
- enrich_austrian_cities.py, enrich_belgian_cities.py, enrich_belgian_v2.py
- enrich_bulgarian_cities.py, enrich_czech_cities.py, enrich_czech_cities_fast.py
- enrich_japanese_cities.py, enrich_swiss_isil_cities.py, enrich_cities_google.py
Location resolution utilities:
- resolve_cities_from_file_coords.py - Resolve cities using coordinates in filenames
- resolve_cities_wikidata.py - Use Wikidata P131 for city resolution
- resolve_country_codes.py - Standardize country codes
- resolve_cz_xx_regions.py - Fix Czech XX region codes
- resolve_locations_by_name.py - Name-based location lookup
- resolve_regions_from_city.py - Derive regions from city data
- update_ghcid_with_geonames.py - Update GHCIDs with GeoNames data
CH-Annotator integration:
- create_custodian_from_ch_annotator.py - Create custodians from annotations
- add_ch_annotator_location_claims.py - Add location claims
- extract_locations_ch_annotator.py - Extract locations from annotations
Migration and fixes:
- migrate_egyptian_from_ch.py - Migrate Egyptian data
- migrate_web_archives.py - Migrate web archive data
- fix_belgian_cities.py - Fix Belgian city data
- Fix polygon rendering with static paint properties instead of data-driven
- Add ensureSourceAndLayers() helper for reliable layer management
- Use setPaintProperty() for historical vs modern styling distinction
- Improve Database page layout with back buttons and cleaner navigation
- Add ResizableNestedTable component for DuckLake data display
- Optimize spacing and layout in Database.css
- Update schema manifest
Remove 229 custodian YAML files containing invalid characters in GHCIDs:
- Ampersand (&) in abbreviations (e.g., BM&HS, UNL&AG, DR&IMSM)
- Parentheses in abbreviations (e.g., WHO(RA, VK(, SL()
- Unicode characters in filenames (Ö, Ä, Å, É, İ, Ż, etc.)
These files are replaced with corrected versions using alphabetic-only
abbreviations per AGENTS.md Rule 8 (Special Characters MUST Be Excluded).
Related scripts updated for location resolution.
- Introduced `llm_extract_archiveslab.py` script for entity and relationship extraction using LLMAnnotator with GLAM-NER v1.7.0.
- Replaced regex-based extraction with generative LLM inference.
- Added functions for loading markdown content, converting annotation sessions to dictionaries, and generating extraction statistics.
- Implemented comprehensive logging of extraction results, including counts of entities, relationships, and specific types like heritage institutions and persons.
- Results and statistics are saved in JSON format for further analysis.
- Implemented `generate_mermaid_with_instances.py` to create ER diagrams that include all classes, relationships, enum values, and instance data.
- Loaded instance data from YAML files and enriched enum definitions with meaningful annotations.
- Configured output paths for generated diagrams in both frontend and schema directories.
- Added support for excluding technical classes and limiting the number of displayed enum and instance values for readability.
- Implemented a Python script to validate KB library YAML files for required fields and data quality.
- Analyzed enrichment coverage from Wikidata and Google Maps, generating statistics.
- Created a comprehensive markdown report summarizing validation results and enrichment quality.
- Included error handling for file loading and validation processes.
- Generated JSON statistics for further analysis.
- Introduced a comprehensive class diagram for the heritage custodian observation reconstruction schema.
- Defined multiple classes including AllocationAgency, ArchiveOrganizationType, AuxiliaryDigitalPlatform, and others, with relevant attributes and relationships.
- Established inheritance and associations among classes to represent complex relationships within the schema.
- Generated on 2025-11-28, version 0.9.0, excluding the Container class.
- Implemented a Python script that fetches and enriches entries from the NDE Register using data from Wikidata.
- Utilized the Wikibase REST API and SPARQL endpoints for data retrieval.
- Added logging for tracking progress and errors during the enrichment process.
- Configured rate limiting based on authentication status for API requests.
- Created a structured output in YAML format, including detailed enrichment data.
- Generated a log file summarizing the enrichment process and results.
