clms2cdse

Collection of tools and scripts for preparing, converting, and submitting CLMS products to the Copernicus Data Space Ecosystem (CDSE).

As part of the migration of Copernicus Land Monitoring Service products from the legacy distribution infrastructure to CDSE, every dataset must be re-packaged and validated to meet the CDSE ingestion requirements. This repository provides reusable, well-documented scripts that automate those steps — one tool per concern, easy to run locally or in CI/CD pipelines.

---

Scripts

convert_to_cog.py — Raster to Cloud Optimized GeoTIFF

Converts a single raster or every raster in a directory into a Cloud Optimized GeoTIFF (COG). Built on top of GDAL's gdal_translate with the COG driver, it handles the common practical details that CDSE ingestion requires.

What it does

• Reads any GDAL-supported raster format (GeoTIFF, NetCDF, JP2, …).
• Optionally wraps the source in a VRT to strip unwanted palette / color-table metadata.
• Translates to a tiled, compressed, overview-rich COG.
• Validates the output with gdalinfo (dimensions, layout, overviews, size).
• Supports both single-file and batch (whole directory) modes.

Key features

• Palette-to-Gray conversion via VRT (safe — never touches source pixels)
• Configurable compression (LZW, DEFLATE, ZSTD, NONE)
• Adjustable blocksize, overview levels, predictor, and resampling
• BIGTIFF, nodata, and output data type overrides
• JSON config file support (save / load settings)
• Dry-run mode to preview commands before executing
• Verbose per-file progress logging with timing

Quick start

# Single file
python convert_to_cog.py --input data/sample.tif --output output/sample_cog.tif

# All TIFFs in a directory
python convert_to_cog.py --input-dir ./raw_data --output-dir ./cogs

# Custom compression + no validation
python convert_to_cog.py \
    --input scene.tif --output scene_cog.tif \
    --compress ZSTD --predictor 2 --no-validate

# Dry-run to check what would happen
python convert_to_cog.py --input-dir ./raw --output-dir ./cogs --dry-run

# Save/load settings as JSON
python convert_to_cog.py --save-config my_config.json --compress DEFLATE
python convert_to_cog.py --config my_config.json --input scene.tif --output out.tif

Dependencies

• Python ≥ 3.10
• GDAL ≥ 3.6 (the gdal_translate and gdalbuildvrt commands must be on PATH)

---

txt2sld2qml.py — QGIS colour map .txt → SLD + QML

Converts standard QGIS colour map export files (.txt) into Styled Layer Descriptor (SLD) for GeoServer / MapServer and QGIS Layer Style (.qml) files.

What it does

• Reads QGIS colour map .txt files (INTERPOLATED ramps or DISCRETE value maps).
• Generates SLD 1.0.0 files with <RasterSymbolizer> / <ColorMap> — ready for GeoServer, MapServer, or any OGC-compliant server.
• Generates QML files for direct use in QGIS.
• Supports an optional --scale factor to convert real values to raw pixel values when the GeoTIFF stores data with a scale/offset (e.g. ×10000 for CLMS HR-VPP products).

Key features

• Handles both INTERPOLATION:INTERPOLATED (ramp) and DISCRETE colour maps.
• No external dependencies — uses only Python stdlib (xml.etree.ElementTree).
• Keeps the original .txt values as the source of truth (no guessing).
• Valid SLD 1.0.0 XML output.

Quick start

# Convert all .txt files in a directory
python3 txt2sld2qml.py ./txt/ --outdir .

# Apply a scale factor (e.g. raw pixel values = real values × 10000)
python3 txt2sld2qml.py ./txt/ --outdir . --scale 10000

Input format

Each .txt file is a standard QGIS colour map export:

# QGIS Generated Color Map Export File
INTERPOLATION:INTERPOLATED    (or DISCRETE)
value, R, G, B, Alpha, label

Output structure

output/
├── sld/    ← GeoServer / MapServer styles
└── qml/    ← QGIS native styles

---

Layout

Tool	Purpose
convert_to_cog.py	Raster → Cloud Optimized GeoTIFF
analyze_cog_header.py	Inspect/validate COG structure
process_files_new.py	Batch processing pipeline
txt2sld2qml.py	QGIS colour map .txt → SLD + QML

clms2cdse/
├── convert_to_cog.py
├── analyze_cog_header.py
├── process_files_new.py
├── txt2sld2qml.py
├── README.md
└── LICENSE.txt

---

License

EUPL-1.2