Deterministic, compliance-first standardization for spatial data.
Attribute mapping · CRS normalization · validation gates · CI audit trails
Geospatial Schema is a free, in-depth reference resource for engineers who have to make heterogeneous spatial data conform — to a schema, to a coordinate reference system, to a regulator's metadata mandate — without silent corruption. It is written for GIS data managers, government technology teams, Python ETL engineers, and open-source maintainers who need runnable pipeline code and measurable thresholds, not slideware.
Every page anchors on a concrete engineering problem and resolves it the way a good design document would: problem framing → declarative spec/contract → executable Python → verification and failure modes. Code targets Python 3.10+ with pinned library versions (geopandas >=0.14, pyproj >=3.6, pyarrow >=14), and each reference standard — INSPIRE, FGDC, OGC, ISO 19115 — is mapped to the exact pipeline stage that satisfies it.
| Area | What it covers |
|---|---|
| Schema Architecture & Standards | INSPIRE / FGDC / ISO 19115 alignment, schema registries, declarative field mapping, and cross-platform schema translation. |
| CRS Normalization & Sync | Projection workflows, datum transformation fallback chains, unit & tolerance thresholds, and multi-CRS harmonization. |
| Attribute Transformation & ETL | Batch schema processing, field renaming & type coercion, nested GeoJSON flattening, and resilient error handling. |
| Compliance Reporting & Audit Trails | Append-only lineage manifests, schema.org Dataset & DCAT-AP JSON-LD publishing, and CI conformance scorecards. |
Across these areas the resource covers 49 in-depth guides — from Batch-Transforming 10k Shapefiles Without Memory Leaks to Publishing a Schema Conformance Scorecard in GitHub Actions — plus head-to-head comparisons such as pyproj vs GDAL ogr2ogr and pyarrow vs pandas.
- Runnable, not hand-wavy. Declarative YAML/JSON manifests, executable Python, and explicit numeric tolerances you can put in a validation gate.
- Compliance is a first-class output. Lineage manifests, ISO 19115
LI_Lineage, and machine-readable dataset metadata make an audit reproducible instead of anecdotal. - Determinism throughout. Pinned PROJ/EPSG snapshots, golden-dataset regression tests, and CI gates so the same input always yields the same output — and the same evidence.
- Densely interlinked. Guides cross-reference each other so you can assemble an end-to-end pipeline rather than reading isolated tips.
A fast static site generated with Eleventy (11ty) and deployed to Cloudflare Pages. Content lives as Markdown under content/; layouts, data, and assets live under src/. Every page ships hand-authored inline SVG diagrams, structured data (JSON-LD: Article, BreadcrumbList, TechArticle, and HowTo/FAQPage where relevant), and passes an accessibility (WCAG 2 A/AA), structured-data, and performance gate suite before publication.
npm install # install dependencies
npm start # local dev server with live reload
npm run build # production build to _site/
npm run deploy # build + deploy to Cloudflare PagesIssues and suggestions are welcome via the issue tracker.
Note on commit authorship: every commit to this repository is authored solely by the
geospatialschemaGitHub account. Commits carry no co-authors and noCo-Authored-Bytrailers — a single, consistent author of record for the entire history.
© Geospatial Schema. Content and code are published for reference and educational use. Please link back to www.geospatialschema.com when citing.