| Name | flac-raster JSON |
| Version |
0.1.2
JSON |
| download |
| home_page | None |
| Summary | Experimental CLI tool for converting TIFF raster data to FLAC format |
| upload_time | 2025-07-20 16:12:18 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.9 |
| license | MIT |
| keywords |
raster
flac
geospatial
compression
tiff
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# FLAC-Raster: Experimental Raster to FLAC Converter
[](https://github.com/Youssef-Harby/flac-raster/actions/workflows/ci.yml)
[](https://badge.fury.io/py/flac-raster)
[](https://pypi.org/project/flac-raster/)
[](https://opensource.org/licenses/MIT)
An experimental CLI tool that converts TIFF raster data files into FLAC audio format while preserving all geospatial metadata, CRS, and bounds information. This proof-of-concept explores using FLAC's lossless compression for geospatial data storage and introduces **revolutionary HTTP range streaming** for efficient geospatial data access - **"Netflix for Geospatial Data"**.
## 🚀 **NEW: Netflix-Style Streaming for Geospatial Data**
FLAC-Raster now supports **true streaming** exactly like Netflix and Spotify - each tile is a complete, self-contained FLAC file that can be decoded independently!
### 🎵 **Two Streaming Formats:**
1. **Raw Frames Format** (15MB) - High compression, full file download only
2. **🆕 Streaming Format** (185MB) - Netflix-style independent tiles, perfect for HTTP range streaming
### **✨ Streaming Features:**
- **🎬 Netflix-style tiles**: Each tile is a complete, independent FLAC file
- **🌐 HTTP range streaming**: Stream individual tiles via precise byte range requests
- **⚡ Instant access**: Decode any tile without downloading the full file
- **💰 99%+ bandwidth savings**: Download only what you need (0.8MB vs 185MB)
- **🗺️ Geographic precision**: Query specific areas with pixel-perfect accuracy
- **📱 Web-native**: Works with any HTTP server, CDN, or browser
- **🔗 URL support**: Query remote FLAC files directly via HTTPS URLs
- **🎯 Smart indexing**: Spatial metadata for instant tile discovery
## Features
- **Bidirectional conversion**: TIFF → FLAC and FLAC → TIFF
- **Complete metadata preservation**: CRS, bounds, transform, data type, nodata values
- **🆕 Embedded metadata**: All geospatial metadata stored directly in FLAC files (no sidecar files!)
- **🆕 Spatial tiling**: Convert rasters to tiled FLAC with bbox metadata per tile
- **🆕 HTTP range streaming**: Query and stream data by bounding box with 90%+ bandwidth savings
- **🆕 Exceptional compression**: 7-15× file size reduction while maintaining lossless quality
- **Intelligent audio parameters**: Automatically selects sample rate and bit depth based on raster properties
- **Multi-band support**: Seamlessly handles multi-band rasters (RGB, multispectral) as multi-channel audio
- **Lossless compression**: Perfect reconstruction verified - no data loss
- **FLAC chunking**: Uses FLAC's frame-based compression (4096 samples/frame)
- **Comprehensive logging**: Verbose mode with detailed progress tracking
- **Colorful CLI**: Built with Typer and Rich for an intuitive experience
## Installation
### Prerequisites
First, install pixi:
```bash
# Install pixi (cross-platform package manager)
curl -fsSL https://pixi.sh/install.sh | bash
# or via conda: conda install -c conda-forge pixi
```
### Clone and Setup
```bash
git clone https://github.com/Youssef-Harby/flac-raster.git
cd flac-raster
pixi install # Install all dependencies
```
### Install the CLI tool
```bash
# For regular use:
pixi run pip install .
# For development (editable install):
pixi run pip install -e .
```
### Alternative: Direct pip installation
```bash
pip install rasterio numpy typer rich tqdm pyflac mutagen
# Or install from PyPI (when published):
pip install flac-raster
```
## Usage
### Basic Commands
After installation, you can use the CLI directly:
1. **Convert TIFF to FLAC**:
```bash
flac-raster convert input.tif -o output.flac
```
2. **Convert FLAC back to TIFF**:
```bash
flac-raster convert input.flac -o output.tif
```
3. **Get file information**:
```bash
flac-raster info file.tif
flac-raster info file.flac
```
4. **Compare two TIFF files**:
```bash
flac-raster compare original.tif reconstructed.tif
```
### 🆕 Spatial Tiling & Netflix-Style Streaming
5. **Create spatial FLAC with tiling**:
```bash
# Raw frames format (high compression, 15MB)
flac-raster convert input.tif --spatial -o spatial.flac
# 🆕 Streaming format (Netflix-style tiles, 185MB)
flac-raster create-streaming input.tif --tile-size=1024 --output=streaming.flac
# Custom tile size (256x256)
flac-raster convert input.tif --spatial --tile-size 256 -o spatial.flac
```
6. **Query spatial FLAC by bounding box**:
```bash
# Query local file (raw frames)
flac-raster query spatial.flac --bbox "xmin,ymin,xmax,ymax"
# 🆕 Query streaming FLAC (local or remote)
python test_streaming.py local_streaming.flac --bbox "34.1,28.6,34.3,28.8"
# 🆕 Stream from remote URL (Netflix-style!)
python test_streaming.py "https://example.com/streaming.flac" --tile-id=120
# Example with real coordinates
flac-raster query spatial.flac --bbox "-105.3,40.3,-105.1,40.5"
```
7. **🆕 Extract tiles from Netflix-style streaming FLAC**:
```bash
# Extract center tile from remote streaming FLAC
flac-raster extract-streaming "https://example.com/streaming.flac" --center --output=center.tif
# Extract last tile
flac-raster extract-streaming "local_streaming.flac" --last --output=last_tile.tif
# Extract by tile ID
flac-raster extract-streaming "streaming.flac" --tile-id=60 --output=tile_60.tif
# Extract by bounding box
flac-raster extract-streaming "https://cdn.example.com/data.flac" --bbox="602380,3090240,609780,3097640" --output=bbox_tile.tif
```
8. **View spatial index information**:
```bash
flac-raster spatial-info spatial.flac # Raw frames format only
# For streaming format, use extract-streaming with analysis
```
### 🌐 Live Demo: Real Remote Streaming
Try our live streaming FLAC files hosted on Storj DCS with real Sentinel-2 B04 band data:
#### **📍 Single Tile Extraction (99%+ Bandwidth Savings)**
```bash
# 🎯 Extract center tile (coordinates: 554,880, 3,145,140)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--center --output=center_1km.tif
# → Downloads: 1.5 MB | Result: 1024×1024 center tile
# 📦 Extract last tile (southeast corner)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--last --output=southeast_corner.tif
# → Downloads: 0.8 MB | Result: 740×740 edge tile
# 🎬 Extract specific tile by ID (northwest corner)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--tile-id=0 --output=northwest_corner.tif
# → Downloads: 1.5 MB | Result: 1024×1024 first tile
```
#### **🗺️ Geographic Bounding Box Extraction**
```bash
# 📍 Extract specific geographic area (1km² in center)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--bbox="554380,3144640,555380,3145640" \
--output=center_1km_bbox.tif
# → Downloads: 1.5 MB | Result: Exact geographic area
# 📍 Extract southeast corner area (last tile region)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--bbox="602380,3090240,609780,3097640" \
--output=southeast_bbox.tif
# → Downloads: 0.8 MB | Result: 740×740 edge region
# 📍 Extract northwest area (first tile region)
flac-raster extract-streaming \
"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac" \
--bbox="499980,3189800,510220,3200040" \
--output=northwest_bbox.tif
# → Downloads: 1.5 MB | Result: 1024×1024 corner region
```
#### **🌍 Full Dataset Access**
```bash
# 📥 Download full dataset (use raw frames format for efficiency)
flac-raster convert \
"https://link.storjshare.io/raw/juxc544kagtqgkvhezix6wzia5yq/truemaps-public/flac-raster/B04_spatial.flac" \
--output=full_sentinel_B04.tif
# → Downloads: 15 MB | Result: Complete 10,980×10,980 Sentinel-2 dataset
# ⚠️ Note: For full datasets, use the raw frames format (15MB) instead of
# streaming format (177MB) for better compression efficiency
```
#### **📊 Performance Comparison**
| **Use Case** | **Command** | **Download Size** | **Output** | **Savings** |
|--------------|-------------|-------------------|------------|-------------|
| **Single tile** | `--center` | 1.5 MB | 1024×1024 | **99.2%** |
| **Corner tile** | `--last` | 0.8 MB | 740×740 | **99.5%** |
| **Bbox query** | `--bbox="..."` | 0.8-1.5 MB | Exact area | **99%+** |
| **Full dataset** | Raw frames format | 15 MB | 10,980×10,980 | **91.5%** |
| **Full streaming** | All 121 tiles | 177 MB | 10,980×10,980 | **0%** ❌ |
**Netflix-Style Benefits:**
- ⚡ **Instant metadata**: 21KB spatial index loaded once
- 🎯 **Precision targeting**: Download only needed geographic areas
- 🗺️ **Perfect quality**: Pixel-perfect GeoTIFF output with full metadata
- 💰 **Massive savings**: 99%+ bandwidth reduction for area-specific queries
**Alternative**: Use `python main.py` if you haven't installed the package:
```bash
python main.py convert input.tif # Direct script usage
```
### Options
#### Convert command:
- `--output, -o`: Specify output file path (auto-generates if not provided)
- `--compression, -c`: FLAC compression level 0-8 (default: 5)
- `--force, -f`: Overwrite existing output files
- `--verbose, -v`: Enable verbose logging for detailed progress
- `--spatial, -s`: 🆕 Enable spatial tiling (raw frames format)
- `--tile-size`: 🆕 Size of spatial tiles in pixels (default: 512x512)
#### 🆕 Extract-tile command (for raw frames format):
- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax'
- `--output, -o`: Output TIFF file path (required)
#### 🚀 Extract-streaming command (for Netflix-style streaming format):
- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax'
- `--tile-id`: Extract specific tile by ID number
- `--center`: Extract center tile automatically
- `--last`: Extract last tile
- `--output, -o`: Output TIFF file path (required)
#### Query command:
- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax' (required)
- `--output, -o`: Output file for extracted data
- `--format, -f`: Output format: 'ranges' (default) or 'data'
#### 🆕 Streaming test commands:
- `--tile-id`: Extract specific tile by ID number
- `--bbox`: Extract tile by geographic bounding box
- `--last`: Extract last tile (default)
- `--savings`: Show bandwidth savings analysis
#### Compare command:
- `--show-bands/--no-bands`: Show per-band statistics (default: True)
- `--export, -e`: Export comparison results to JSON file
- `--help`: Show help message
### Example Workflow
```bash
# Create sample data
python examples/create_test_data.py
# Convert DEM to FLAC
flac-raster convert test_data/sample_dem.tif -v
# Check the FLAC file info
flac-raster info test_data/sample_dem.flac
# Convert back to TIFF
flac-raster convert test_data/sample_dem.flac -o test_data/dem_reconstructed.tif
# Compare original and reconstructed
flac-raster compare test_data/sample_dem.tif test_data/dem_reconstructed.tif
# Export comparison to JSON
flac-raster compare test_data/sample_dem.tif test_data/dem_reconstructed.tif --export comparison.json
# Test with multi-band data
flac-raster convert test_data/sample_rgb.tif
flac-raster convert test_data/sample_rgb.flac -o test_data/rgb_reconstructed.tif
flac-raster compare test_data/sample_rgb.tif test_data/rgb_reconstructed.tif
# Open in QGIS to verify
# The reconstructed files should be viewable in QGIS with all metadata intact
```
## How It Works
### TIFF to FLAC Conversion
1. **Read raster data** and extract all metadata (CRS, bounds, transform, etc.)
2. **Spatial tiling** (if enabled): Divide raster into configurable tile sizes
3. **Calculate audio parameters**:
- Sample rate: Based on image resolution (44.1kHz to 192kHz)
- Bit depth: Matches the raster's bit depth (16 or 24-bit, minimum 16-bit due to FLAC decoder limitations)
4. **Normalize data** to audio range (-1 to 1)
5. **Reshape data**: Bands become audio channels, pixels become samples
- Single-band → Mono audio
- Multi-band (RGB, multispectral) → Multi-channel audio
6. **Encode to FLAC** with configurable compression
7. **Embed metadata** directly in FLAC using VORBIS_COMMENT blocks
8. **Generate spatial index** with bbox and byte range information for each tile
### FLAC to TIFF Conversion
1. **Decode FLAC** file and extract audio samples
2. **Load metadata** from embedded FLAC metadata (with JSON sidecar fallback)
3. **Reconstruct spatial index** for tiled data
4. **Reshape audio** back to raster dimensions
- Mono → Single-band raster
- Multi-channel → Multi-band raster
5. **Denormalize** to original data range
6. **Write GeoTIFF** with all original metadata preserved
## Metadata Preservation
The tool preserves all geospatial metadata **directly embedded in FLAC files**:
- Width and height dimensions
- Number of bands
- Data type (uint8, int16, float32, etc.)
- Coordinate Reference System (CRS)
- Geospatial transform (affine transformation matrix)
- Bounding box coordinates
- Original data min/max values
- NoData values
- **Spatial index**: Compressed tile bbox and byte range information
- Original driver information
### Embedded Metadata Format
Metadata is stored in FLAC VORBIS_COMMENT blocks:
```
GEOSPATIAL_CRS=EPSG:4326
GEOSPATIAL_WIDTH=1201
GEOSPATIAL_HEIGHT=1201
GEOSPATIAL_SPATIAL_INDEX=<base64(gzip(spatial_index_json))>
...
```
## Lazy Loading & HTTP Range Streaming for Web GIS
### Concept: "Zarr for Geospatial Data using Audio Compression"
The lazy loading feature transforms FLAC-Raster into a **web-native geospatial format** that enables efficient HTTP range request streaming:
```
FLAC URL: https://cdn.example.com/elevation.flac
↓
🏃♂️ Lazy Load: Download first 1MB for metadata only
↓
Query Spatial Index: Find intersecting tiles for bbox
↓
HTTP Range Request: bytes=48152-73513,87850-113211
↓
⬇️ Smart Download: Only 76KB instead of 189KB (60% savings!)
↓
Decode FLAC: Get pixels for visible area only
```
### Lazy Loading Workflow
1. **Metadata First**: Download only 1MB to read embedded spatial index
2. **On-Demand Streaming**: Query specific geographic areas
3. **Precise Downloads**: HTTP Range requests for intersecting tiles only
4. **Progressive Loading**: Cache tiles for repeated access
### Use Cases
1. **Interactive Web Maps**
- Progressive loading as users pan/zoom
- Only download visible area data
- Works with any HTTP server/CDN
2. **Cloud-Native GIS**
- Stream large rasters without specialized servers
- Compatible with S3, CloudFront, etc.
- No need for complex tiling servers
3. **Bandwidth-Constrained Applications**
- Mobile mapping apps
- Satellite/field data collection
- IoT sensor networks
### Web Server Integration
```javascript
// JavaScript lazy loading client example
async function loadRasterData(bbox) {
const flacUrl = '/data/elevation.flac';
// 1. Lazy load: get metadata only (first 1MB)
const metadataResponse = await fetch(flacUrl, {
headers: { 'Range': 'bytes=0-1048575' }
});
const spatialIndex = extractEmbeddedMetadata(metadataResponse);
// 2. Find byte ranges for bbox
const ranges = calculateRanges(bbox, spatialIndex);
// 3. Stream only needed tiles via HTTP ranges
const rangeHeader = ranges.map(r => `${r.start}-${r.end}`).join(',');
const dataResponse = await fetch(flacUrl, {
headers: { 'Range': `bytes=${rangeHeader}` }
});
// 4. Decode FLAC data for bbox
return decodeFLACTiles(dataResponse.body, bbox);
}
```
## Technical Details
### 🎵 **Netflix-Style Streaming Architecture**
FLAC-Raster implements two distinct formats for different use cases:
#### **Raw Frames Format** (Legacy)
- **FLAC frames**: Raw frame chunks within single FLAC file
- **Compression**: Exceptional compression (15MB for 185MB streaming equivalent)
- **Use case**: Full file downloads, highest compression ratio
- **Limitation**: Cannot stream individual tiles
#### **🆕 Streaming Format** (Netflix-Style)
- **Self-contained tiles**: Each tile is a complete, independent FLAC file
- **HTTP range ready**: Perfect byte boundaries for range requests
- **Instant decode**: Any tile can be decoded without full file context
- **Format structure**:
```
[4 bytes index size][JSON spatial index][Complete FLAC Tile 1][Complete FLAC Tile 2]...[Complete FLAC Tile N]
```
### **Core Technologies**
- **🆕 Complete FLAC segments**: Each tile includes full FLAC headers and metadata
- **🆕 HTTP byte ranges**: Precise byte offsets enable partial downloads
- **🆕 Embedded metadata**: All geospatial info stored in FLAC VORBIS_COMMENT blocks
- **🆕 Spatial indexing**: JSON metadata with bbox coordinates and byte ranges
- **Multi-band support**: Each raster band becomes an audio channel (up to 8 channels supported by FLAC)
- **Lossless conversion**: Data is normalized but the process is completely reversible
- **Exceptional compression**: Leverages FLAC's compression algorithms (7-15× size reduction)
- **Self-contained files**: No external dependencies or sidecar files required
- **Data type mapping**:
- uint8 → 16-bit FLAC (due to decoder limitations)
- int16/uint16 → 16-bit FLAC
- int32/uint32/float32 → 24-bit FLAC
## Performance Examples
From comprehensive testing against `report.md` analysis:
### Compression Results
- **DEM file** (1201×1201, int16): 2.8 MB → 185 KB FLAC (**15.25× compression**)
- **Multispectral** (200×200×6, uint8): 235 KB → 32 KB FLAC (**7.38× compression**)
- **RGB** (256×256×3, uint8): 193 KB → 27 KB FLAC (**7.26× compression**)
### HTTP Range Streaming Efficiency
- **Small area queries**: Up to **98.8% bandwidth savings** vs full download
- **Geographic precision**: Query exact areas with pixel-perfect accuracy
- **Optimized ranges**: Smart merging of contiguous tiles reduces HTTP requests
All conversions are perfectly lossless (verified with numpy array comparison)
## Limitations
- Maximum 8 bands (FLAC channel limitation)
- Minimum 16-bit encoding (pyflac decoder limitation)
- Large rasters may take time to process
- FLAC format limitations apply (specific bit depths: 16, 24-bit)
- Requires mutagen library for embedded metadata support
- Experimental: Not recommended for production use without thorough testing
## Project Structure
```
flac-raster/
├── src/flac_raster/ # Main package
│ ├── __init__.py # Package initialization
│ ├── cli.py # Command-line interface
│ ├── converter.py # Core conversion logic
│ ├── spatial_encoder.py # 🆕 Spatial tiling & HTTP range streaming
│ ├── metadata_encoder.py # 🆕 Embedded metadata handling
│ └── compare.py # Comparison utilities
├── examples/ # Example scripts
│ ├── create_test_data.py # Generate test datasets
│ └── spatial_streaming_example.py # 🆕 HTTP range streaming demo
├── test_data/ # Test datasets
│ ├── dem-raw.tif # Large DEM for testing
│ ├── sample_multispectral.tif # 6-band multispectral
│ └── sample_rgb.tif # RGB test data
├── report.md # 🆕 Comprehensive analysis & benchmarks
├── main.py # Main entry point
├── pyproject.toml # Project configuration
├── README.md # This file
└── pixi.toml # Pixi package configuration
```
## CI/CD & Publishing
This project uses GitHub Actions for:
- **Continuous Integration**: Tests on Python 3.9-3.12 across Windows, macOS, and Linux
- **Automated Building**: Package building and validation
- **PyPI Publishing**: Automatic publishing on release creation
- **Quality Assurance**: Integration testing via CLI commands
### Publishing to PyPI
See [PUBLISHING.md](PUBLISHING.md) for detailed instructions on publishing releases.
## Contributing
1. Fork the repository
2. Create a feature branch: `git checkout -b feature-name`
3. Make your changes and test them
4. Commit your changes: `git commit -am 'Add feature'`
5. Push to the branch: `git push origin feature-name`
6. Create a Pull Request
## Future Improvements
- **Adaptive tiling**: Variable tile sizes based on data complexity
- **Temporal support**: Time-series data with temporal indexing
- **Band selection**: Spectral subsetting for multispectral data
- **Compression tuning**: Automatic optimization of FLAC parameters
- **Caching strategy**: Intelligent tile caching for frequently accessed areas
- **JavaScript client**: Browser-based FLAC decoder for web mapping
- **Parallel processing**: Multi-threaded encoding/decoding
- **More formats**: Support for HDF5, NetCDF, Zarr integration
- **Performance optimization**: Memory usage and processing speed improvements
Raw data
{
"_id": null,
"home_page": null,
"name": "flac-raster",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "raster, flac, geospatial, compression, tiff",
"author": null,
"author_email": "Youssef Harby <me@youssefharby.com>",
"download_url": "https://files.pythonhosted.org/packages/9d/5e/bc6877635c8651debef65fba9bd0bbb1a7093e2b0b9c2f6eb75014aa07a2/flac_raster-0.1.2.tar.gz",
"platform": null,
"description": "# FLAC-Raster: Experimental Raster to FLAC Converter\n\n[](https://github.com/Youssef-Harby/flac-raster/actions/workflows/ci.yml)\n[](https://badge.fury.io/py/flac-raster)\n[](https://pypi.org/project/flac-raster/)\n[](https://opensource.org/licenses/MIT)\n\nAn experimental CLI tool that converts TIFF raster data files into FLAC audio format while preserving all geospatial metadata, CRS, and bounds information. This proof-of-concept explores using FLAC's lossless compression for geospatial data storage and introduces **revolutionary HTTP range streaming** for efficient geospatial data access - **\"Netflix for Geospatial Data\"**.\n\n## \ud83d\ude80 **NEW: Netflix-Style Streaming for Geospatial Data**\n\nFLAC-Raster now supports **true streaming** exactly like Netflix and Spotify - each tile is a complete, self-contained FLAC file that can be decoded independently! \n\n### \ud83c\udfb5 **Two Streaming Formats:**\n\n1. **Raw Frames Format** (15MB) - High compression, full file download only\n2. **\ud83c\udd95 Streaming Format** (185MB) - Netflix-style independent tiles, perfect for HTTP range streaming\n\n### **\u2728 Streaming Features:**\n- **\ud83c\udfac Netflix-style tiles**: Each tile is a complete, independent FLAC file\n- **\ud83c\udf10 HTTP range streaming**: Stream individual tiles via precise byte range requests\n- **\u26a1 Instant access**: Decode any tile without downloading the full file\n- **\ud83d\udcb0 99%+ bandwidth savings**: Download only what you need (0.8MB vs 185MB)\n- **\ud83d\uddfa\ufe0f Geographic precision**: Query specific areas with pixel-perfect accuracy\n- **\ud83d\udcf1 Web-native**: Works with any HTTP server, CDN, or browser\n- **\ud83d\udd17 URL support**: Query remote FLAC files directly via HTTPS URLs\n- **\ud83c\udfaf Smart indexing**: Spatial metadata for instant tile discovery\n\n## Features\n\n- **Bidirectional conversion**: TIFF \u2192 FLAC and FLAC \u2192 TIFF\n- **Complete metadata preservation**: CRS, bounds, transform, data type, nodata values\n- **\ud83c\udd95 Embedded metadata**: All geospatial metadata stored directly in FLAC files (no sidecar files!)\n- **\ud83c\udd95 Spatial tiling**: Convert rasters to tiled FLAC with bbox metadata per tile\n- **\ud83c\udd95 HTTP range streaming**: Query and stream data by bounding box with 90%+ bandwidth savings\n- **\ud83c\udd95 Exceptional compression**: 7-15\u00d7 file size reduction while maintaining lossless quality\n- **Intelligent audio parameters**: Automatically selects sample rate and bit depth based on raster properties\n- **Multi-band support**: Seamlessly handles multi-band rasters (RGB, multispectral) as multi-channel audio\n- **Lossless compression**: Perfect reconstruction verified - no data loss\n- **FLAC chunking**: Uses FLAC's frame-based compression (4096 samples/frame)\n- **Comprehensive logging**: Verbose mode with detailed progress tracking\n- **Colorful CLI**: Built with Typer and Rich for an intuitive experience\n\n## Installation\n\n### Prerequisites\nFirst, install pixi:\n```bash\n# Install pixi (cross-platform package manager)\ncurl -fsSL https://pixi.sh/install.sh | bash\n# or via conda: conda install -c conda-forge pixi\n```\n\n### Clone and Setup\n```bash\ngit clone https://github.com/Youssef-Harby/flac-raster.git\ncd flac-raster\npixi install # Install all dependencies\n```\n\n### Install the CLI tool\n```bash\n# For regular use:\npixi run pip install .\n\n# For development (editable install):\npixi run pip install -e .\n```\n\n### Alternative: Direct pip installation\n```bash\npip install rasterio numpy typer rich tqdm pyflac mutagen\n\n# Or install from PyPI (when published):\npip install flac-raster\n```\n\n## Usage\n\n### Basic Commands\n\nAfter installation, you can use the CLI directly:\n\n1. **Convert TIFF to FLAC**:\n ```bash\n flac-raster convert input.tif -o output.flac\n ```\n\n2. **Convert FLAC back to TIFF**:\n ```bash\n flac-raster convert input.flac -o output.tif\n ```\n\n3. **Get file information**:\n ```bash\n flac-raster info file.tif\n flac-raster info file.flac\n ```\n\n4. **Compare two TIFF files**:\n ```bash\n flac-raster compare original.tif reconstructed.tif\n ```\n\n### \ud83c\udd95 Spatial Tiling & Netflix-Style Streaming\n\n5. **Create spatial FLAC with tiling**:\n ```bash\n # Raw frames format (high compression, 15MB)\n flac-raster convert input.tif --spatial -o spatial.flac\n \n # \ud83c\udd95 Streaming format (Netflix-style tiles, 185MB)\n flac-raster create-streaming input.tif --tile-size=1024 --output=streaming.flac\n \n # Custom tile size (256x256)\n flac-raster convert input.tif --spatial --tile-size 256 -o spatial.flac\n ```\n\n6. **Query spatial FLAC by bounding box**:\n ```bash\n # Query local file (raw frames)\n flac-raster query spatial.flac --bbox \"xmin,ymin,xmax,ymax\"\n \n # \ud83c\udd95 Query streaming FLAC (local or remote)\n python test_streaming.py local_streaming.flac --bbox \"34.1,28.6,34.3,28.8\"\n \n # \ud83c\udd95 Stream from remote URL (Netflix-style!)\n python test_streaming.py \"https://example.com/streaming.flac\" --tile-id=120\n \n # Example with real coordinates\n flac-raster query spatial.flac --bbox \"-105.3,40.3,-105.1,40.5\"\n ```\n\n7. **\ud83c\udd95 Extract tiles from Netflix-style streaming FLAC**:\n ```bash\n # Extract center tile from remote streaming FLAC\n flac-raster extract-streaming \"https://example.com/streaming.flac\" --center --output=center.tif\n \n # Extract last tile\n flac-raster extract-streaming \"local_streaming.flac\" --last --output=last_tile.tif\n \n # Extract by tile ID\n flac-raster extract-streaming \"streaming.flac\" --tile-id=60 --output=tile_60.tif\n \n # Extract by bounding box\n flac-raster extract-streaming \"https://cdn.example.com/data.flac\" --bbox=\"602380,3090240,609780,3097640\" --output=bbox_tile.tif\n ```\n\n8. **View spatial index information**:\n ```bash\n flac-raster spatial-info spatial.flac # Raw frames format only\n # For streaming format, use extract-streaming with analysis\n ```\n\n### \ud83c\udf10 Live Demo: Real Remote Streaming\n\nTry our live streaming FLAC files hosted on Storj DCS with real Sentinel-2 B04 band data:\n\n#### **\ud83d\udccd Single Tile Extraction (99%+ Bandwidth Savings)**\n\n```bash\n# \ud83c\udfaf Extract center tile (coordinates: 554,880, 3,145,140)\nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --center --output=center_1km.tif\n# \u2192 Downloads: 1.5 MB | Result: 1024\u00d71024 center tile\n\n# \ud83d\udce6 Extract last tile (southeast corner) \nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --last --output=southeast_corner.tif\n# \u2192 Downloads: 0.8 MB | Result: 740\u00d7740 edge tile\n\n# \ud83c\udfac Extract specific tile by ID (northwest corner)\nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --tile-id=0 --output=northwest_corner.tif\n# \u2192 Downloads: 1.5 MB | Result: 1024\u00d71024 first tile\n```\n\n#### **\ud83d\uddfa\ufe0f Geographic Bounding Box Extraction**\n\n```bash\n# \ud83d\udccd Extract specific geographic area (1km\u00b2 in center)\nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --bbox=\"554380,3144640,555380,3145640\" \\\n --output=center_1km_bbox.tif\n# \u2192 Downloads: 1.5 MB | Result: Exact geographic area\n\n# \ud83d\udccd Extract southeast corner area (last tile region)\nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --bbox=\"602380,3090240,609780,3097640\" \\\n --output=southeast_bbox.tif \n# \u2192 Downloads: 0.8 MB | Result: 740\u00d7740 edge region\n\n# \ud83d\udccd Extract northwest area (first tile region)\nflac-raster extract-streaming \\\n \"https://link.storjshare.io/raw/ju6tov7vffpleabbilqgxfpxz5cq/truemaps-public/flac-raster/B04_streaming.flac\" \\\n --bbox=\"499980,3189800,510220,3200040\" \\\n --output=northwest_bbox.tif\n# \u2192 Downloads: 1.5 MB | Result: 1024\u00d71024 corner region\n```\n\n#### **\ud83c\udf0d Full Dataset Access**\n\n```bash\n# \ud83d\udce5 Download full dataset (use raw frames format for efficiency)\nflac-raster convert \\\n \"https://link.storjshare.io/raw/juxc544kagtqgkvhezix6wzia5yq/truemaps-public/flac-raster/B04_spatial.flac\" \\\n --output=full_sentinel_B04.tif\n# \u2192 Downloads: 15 MB | Result: Complete 10,980\u00d710,980 Sentinel-2 dataset\n\n# \u26a0\ufe0f Note: For full datasets, use the raw frames format (15MB) instead of \n# streaming format (177MB) for better compression efficiency\n```\n\n#### **\ud83d\udcca Performance Comparison**\n\n| **Use Case** | **Command** | **Download Size** | **Output** | **Savings** |\n|--------------|-------------|-------------------|------------|-------------|\n| **Single tile** | `--center` | 1.5 MB | 1024\u00d71024 | **99.2%** |\n| **Corner tile** | `--last` | 0.8 MB | 740\u00d7740 | **99.5%** |\n| **Bbox query** | `--bbox=\"...\"` | 0.8-1.5 MB | Exact area | **99%+** |\n| **Full dataset** | Raw frames format | 15 MB | 10,980\u00d710,980 | **91.5%** |\n| **Full streaming** | All 121 tiles | 177 MB | 10,980\u00d710,980 | **0%** \u274c |\n\n**Netflix-Style Benefits:**\n- \u26a1 **Instant metadata**: 21KB spatial index loaded once\n- \ud83c\udfaf **Precision targeting**: Download only needed geographic areas \n- \ud83d\uddfa\ufe0f **Perfect quality**: Pixel-perfect GeoTIFF output with full metadata\n- \ud83d\udcb0 **Massive savings**: 99%+ bandwidth reduction for area-specific queries\n\n**Alternative**: Use `python main.py` if you haven't installed the package:\n```bash\npython main.py convert input.tif # Direct script usage\n```\n\n### Options\n\n#### Convert command:\n- `--output, -o`: Specify output file path (auto-generates if not provided)\n- `--compression, -c`: FLAC compression level 0-8 (default: 5)\n- `--force, -f`: Overwrite existing output files\n- `--verbose, -v`: Enable verbose logging for detailed progress\n- `--spatial, -s`: \ud83c\udd95 Enable spatial tiling (raw frames format)\n- `--tile-size`: \ud83c\udd95 Size of spatial tiles in pixels (default: 512x512)\n\n#### \ud83c\udd95 Extract-tile command (for raw frames format):\n- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax'\n- `--output, -o`: Output TIFF file path (required)\n\n#### \ud83d\ude80 Extract-streaming command (for Netflix-style streaming format):\n- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax'\n- `--tile-id`: Extract specific tile by ID number\n- `--center`: Extract center tile automatically\n- `--last`: Extract last tile\n- `--output, -o`: Output TIFF file path (required)\n\n#### Query command:\n- `--bbox, -b`: Bounding box as 'xmin,ymin,xmax,ymax' (required)\n- `--output, -o`: Output file for extracted data\n- `--format, -f`: Output format: 'ranges' (default) or 'data'\n\n#### \ud83c\udd95 Streaming test commands:\n- `--tile-id`: Extract specific tile by ID number\n- `--bbox`: Extract tile by geographic bounding box\n- `--last`: Extract last tile (default)\n- `--savings`: Show bandwidth savings analysis\n\n#### Compare command:\n- `--show-bands/--no-bands`: Show per-band statistics (default: True)\n- `--export, -e`: Export comparison results to JSON file\n- `--help`: Show help message\n\n### Example Workflow\n\n```bash\n# Create sample data\npython examples/create_test_data.py\n\n# Convert DEM to FLAC\nflac-raster convert test_data/sample_dem.tif -v\n\n# Check the FLAC file info\nflac-raster info test_data/sample_dem.flac\n\n# Convert back to TIFF\nflac-raster convert test_data/sample_dem.flac -o test_data/dem_reconstructed.tif\n\n# Compare original and reconstructed\nflac-raster compare test_data/sample_dem.tif test_data/dem_reconstructed.tif\n\n# Export comparison to JSON\nflac-raster compare test_data/sample_dem.tif test_data/dem_reconstructed.tif --export comparison.json\n\n# Test with multi-band data\nflac-raster convert test_data/sample_rgb.tif\nflac-raster convert test_data/sample_rgb.flac -o test_data/rgb_reconstructed.tif\nflac-raster compare test_data/sample_rgb.tif test_data/rgb_reconstructed.tif\n\n# Open in QGIS to verify\n# The reconstructed files should be viewable in QGIS with all metadata intact\n```\n\n## How It Works\n\n### TIFF to FLAC Conversion\n\n1. **Read raster data** and extract all metadata (CRS, bounds, transform, etc.)\n2. **Spatial tiling** (if enabled): Divide raster into configurable tile sizes\n3. **Calculate audio parameters**:\n - Sample rate: Based on image resolution (44.1kHz to 192kHz)\n - Bit depth: Matches the raster's bit depth (16 or 24-bit, minimum 16-bit due to FLAC decoder limitations)\n4. **Normalize data** to audio range (-1 to 1)\n5. **Reshape data**: Bands become audio channels, pixels become samples\n - Single-band \u2192 Mono audio\n - Multi-band (RGB, multispectral) \u2192 Multi-channel audio\n6. **Encode to FLAC** with configurable compression\n7. **Embed metadata** directly in FLAC using VORBIS_COMMENT blocks\n8. **Generate spatial index** with bbox and byte range information for each tile\n\n### FLAC to TIFF Conversion\n\n1. **Decode FLAC** file and extract audio samples\n2. **Load metadata** from embedded FLAC metadata (with JSON sidecar fallback)\n3. **Reconstruct spatial index** for tiled data\n4. **Reshape audio** back to raster dimensions\n - Mono \u2192 Single-band raster\n - Multi-channel \u2192 Multi-band raster\n5. **Denormalize** to original data range\n6. **Write GeoTIFF** with all original metadata preserved\n\n## Metadata Preservation\n\nThe tool preserves all geospatial metadata **directly embedded in FLAC files**:\n- Width and height dimensions\n- Number of bands \n- Data type (uint8, int16, float32, etc.)\n- Coordinate Reference System (CRS)\n- Geospatial transform (affine transformation matrix)\n- Bounding box coordinates\n- Original data min/max values\n- NoData values\n- **Spatial index**: Compressed tile bbox and byte range information\n- Original driver information\n\n### Embedded Metadata Format\n\nMetadata is stored in FLAC VORBIS_COMMENT blocks:\n```\nGEOSPATIAL_CRS=EPSG:4326\nGEOSPATIAL_WIDTH=1201\nGEOSPATIAL_HEIGHT=1201\nGEOSPATIAL_SPATIAL_INDEX=<base64(gzip(spatial_index_json))>\n...\n```\n\n## Lazy Loading & HTTP Range Streaming for Web GIS\n\n### Concept: \"Zarr for Geospatial Data using Audio Compression\"\n\nThe lazy loading feature transforms FLAC-Raster into a **web-native geospatial format** that enables efficient HTTP range request streaming:\n\n```\nFLAC URL: https://cdn.example.com/elevation.flac\n \u2193\n\ud83c\udfc3\u200d\u2642\ufe0f Lazy Load: Download first 1MB for metadata only\n \u2193\nQuery Spatial Index: Find intersecting tiles for bbox\n \u2193\nHTTP Range Request: bytes=48152-73513,87850-113211\n \u2193 \n\u2b07\ufe0f Smart Download: Only 76KB instead of 189KB (60% savings!)\n \u2193\nDecode FLAC: Get pixels for visible area only\n```\n\n### Lazy Loading Workflow\n\n1. **Metadata First**: Download only 1MB to read embedded spatial index\n2. **On-Demand Streaming**: Query specific geographic areas\n3. **Precise Downloads**: HTTP Range requests for intersecting tiles only\n4. **Progressive Loading**: Cache tiles for repeated access\n\n### Use Cases\n\n1. **Interactive Web Maps**\n - Progressive loading as users pan/zoom\n - Only download visible area data \n - Works with any HTTP server/CDN\n\n2. **Cloud-Native GIS**\n - Stream large rasters without specialized servers\n - Compatible with S3, CloudFront, etc.\n - No need for complex tiling servers\n\n3. **Bandwidth-Constrained Applications** \n - Mobile mapping apps\n - Satellite/field data collection\n - IoT sensor networks\n\n### Web Server Integration\n\n```javascript\n// JavaScript lazy loading client example\nasync function loadRasterData(bbox) {\n const flacUrl = '/data/elevation.flac';\n \n // 1. Lazy load: get metadata only (first 1MB)\n const metadataResponse = await fetch(flacUrl, {\n headers: { 'Range': 'bytes=0-1048575' }\n });\n const spatialIndex = extractEmbeddedMetadata(metadataResponse);\n \n // 2. Find byte ranges for bbox\n const ranges = calculateRanges(bbox, spatialIndex);\n \n // 3. Stream only needed tiles via HTTP ranges\n const rangeHeader = ranges.map(r => `${r.start}-${r.end}`).join(',');\n const dataResponse = await fetch(flacUrl, {\n headers: { 'Range': `bytes=${rangeHeader}` }\n });\n \n // 4. Decode FLAC data for bbox\n return decodeFLACTiles(dataResponse.body, bbox);\n}\n```\n\n## Technical Details\n\n### \ud83c\udfb5 **Netflix-Style Streaming Architecture**\n\nFLAC-Raster implements two distinct formats for different use cases:\n\n#### **Raw Frames Format** (Legacy)\n- **FLAC frames**: Raw frame chunks within single FLAC file\n- **Compression**: Exceptional compression (15MB for 185MB streaming equivalent) \n- **Use case**: Full file downloads, highest compression ratio\n- **Limitation**: Cannot stream individual tiles\n\n#### **\ud83c\udd95 Streaming Format** (Netflix-Style)\n- **Self-contained tiles**: Each tile is a complete, independent FLAC file\n- **HTTP range ready**: Perfect byte boundaries for range requests\n- **Instant decode**: Any tile can be decoded without full file context\n- **Format structure**:\n ```\n [4 bytes index size][JSON spatial index][Complete FLAC Tile 1][Complete FLAC Tile 2]...[Complete FLAC Tile N]\n ```\n\n### **Core Technologies**\n- **\ud83c\udd95 Complete FLAC segments**: Each tile includes full FLAC headers and metadata\n- **\ud83c\udd95 HTTP byte ranges**: Precise byte offsets enable partial downloads \n- **\ud83c\udd95 Embedded metadata**: All geospatial info stored in FLAC VORBIS_COMMENT blocks\n- **\ud83c\udd95 Spatial indexing**: JSON metadata with bbox coordinates and byte ranges\n- **Multi-band support**: Each raster band becomes an audio channel (up to 8 channels supported by FLAC)\n- **Lossless conversion**: Data is normalized but the process is completely reversible\n- **Exceptional compression**: Leverages FLAC's compression algorithms (7-15\u00d7 size reduction)\n- **Self-contained files**: No external dependencies or sidecar files required\n- **Data type mapping**:\n - uint8 \u2192 16-bit FLAC (due to decoder limitations)\n - int16/uint16 \u2192 16-bit FLAC\n - int32/uint32/float32 \u2192 24-bit FLAC\n\n## Performance Examples\n\nFrom comprehensive testing against `report.md` analysis:\n\n### Compression Results\n- **DEM file** (1201\u00d71201, int16): 2.8 MB \u2192 185 KB FLAC (**15.25\u00d7 compression**)\n- **Multispectral** (200\u00d7200\u00d76, uint8): 235 KB \u2192 32 KB FLAC (**7.38\u00d7 compression**)\n- **RGB** (256\u00d7256\u00d73, uint8): 193 KB \u2192 27 KB FLAC (**7.26\u00d7 compression**)\n\n### HTTP Range Streaming Efficiency\n- **Small area queries**: Up to **98.8% bandwidth savings** vs full download\n- **Geographic precision**: Query exact areas with pixel-perfect accuracy\n- **Optimized ranges**: Smart merging of contiguous tiles reduces HTTP requests\n\nAll conversions are perfectly lossless (verified with numpy array comparison)\n\n## Limitations\n\n- Maximum 8 bands (FLAC channel limitation)\n- Minimum 16-bit encoding (pyflac decoder limitation)\n- Large rasters may take time to process\n- FLAC format limitations apply (specific bit depths: 16, 24-bit)\n- Requires mutagen library for embedded metadata support\n- Experimental: Not recommended for production use without thorough testing\n\n## Project Structure\n\n```\nflac-raster/\n\u251c\u2500\u2500 src/flac_raster/ # Main package\n\u2502 \u251c\u2500\u2500 __init__.py # Package initialization\n\u2502 \u251c\u2500\u2500 cli.py # Command-line interface\n\u2502 \u251c\u2500\u2500 converter.py # Core conversion logic\n\u2502 \u251c\u2500\u2500 spatial_encoder.py # \ud83c\udd95 Spatial tiling & HTTP range streaming\n\u2502 \u251c\u2500\u2500 metadata_encoder.py # \ud83c\udd95 Embedded metadata handling\n\u2502 \u2514\u2500\u2500 compare.py # Comparison utilities\n\u251c\u2500\u2500 examples/ # Example scripts\n\u2502 \u251c\u2500\u2500 create_test_data.py # Generate test datasets\n\u2502 \u2514\u2500\u2500 spatial_streaming_example.py # \ud83c\udd95 HTTP range streaming demo\n\u251c\u2500\u2500 test_data/ # Test datasets\n\u2502 \u251c\u2500\u2500 dem-raw.tif # Large DEM for testing\n\u2502 \u251c\u2500\u2500 sample_multispectral.tif # 6-band multispectral\n\u2502 \u2514\u2500\u2500 sample_rgb.tif # RGB test data\n\u251c\u2500\u2500 report.md # \ud83c\udd95 Comprehensive analysis & benchmarks\n\u251c\u2500\u2500 main.py # Main entry point\n\u251c\u2500\u2500 pyproject.toml # Project configuration\n\u251c\u2500\u2500 README.md # This file\n\u2514\u2500\u2500 pixi.toml # Pixi package configuration\n```\n\n## CI/CD & Publishing\n\nThis project uses GitHub Actions for:\n- **Continuous Integration**: Tests on Python 3.9-3.12 across Windows, macOS, and Linux\n- **Automated Building**: Package building and validation\n- **PyPI Publishing**: Automatic publishing on release creation\n- **Quality Assurance**: Integration testing via CLI commands\n\n### Publishing to PyPI\nSee [PUBLISHING.md](PUBLISHING.md) for detailed instructions on publishing releases.\n\n## Contributing\n\n1. Fork the repository\n2. Create a feature branch: `git checkout -b feature-name`\n3. Make your changes and test them\n4. Commit your changes: `git commit -am 'Add feature'`\n5. Push to the branch: `git push origin feature-name`\n6. Create a Pull Request\n\n## Future Improvements\n\n- **Adaptive tiling**: Variable tile sizes based on data complexity\n- **Temporal support**: Time-series data with temporal indexing \n- **Band selection**: Spectral subsetting for multispectral data\n- **Compression tuning**: Automatic optimization of FLAC parameters\n- **Caching strategy**: Intelligent tile caching for frequently accessed areas\n- **JavaScript client**: Browser-based FLAC decoder for web mapping\n- **Parallel processing**: Multi-threaded encoding/decoding\n- **More formats**: Support for HDF5, NetCDF, Zarr integration\n- **Performance optimization**: Memory usage and processing speed improvements\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Experimental CLI tool for converting TIFF raster data to FLAC format",
"version": "0.1.2",
"project_urls": {
"Homepage": "https://github.com/Youssef-Harby/flac-raster",
"Issues": "https://github.com/Youssef-Harby/flac-raster/issues",
"Repository": "https://github.com/Youssef-Harby/flac-raster"
},
"split_keywords": [
"raster",
" flac",
" geospatial",
" compression",
" tiff"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "ac61c08b36fc562ab085823a0355924861fa7b7f79d29568be47e1617c4469b6",
"md5": "1a567229327b7849eb33b7c424f6e453",
"sha256": "0889b51812f01c1137add230b73c5323fd77c1452c89206756cb3ae746345ca4"
},
"downloads": -1,
"filename": "flac_raster-0.1.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1a567229327b7849eb33b7c424f6e453",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 35774,
"upload_time": "2025-07-20T16:12:17",
"upload_time_iso_8601": "2025-07-20T16:12:17.052088Z",
"url": "https://files.pythonhosted.org/packages/ac/61/c08b36fc562ab085823a0355924861fa7b7f79d29568be47e1617c4469b6/flac_raster-0.1.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "9d5ebc6877635c8651debef65fba9bd0bbb1a7093e2b0b9c2f6eb75014aa07a2",
"md5": "db4006b459a93a67c25744d5112d3e9a",
"sha256": "7a2efeaddb2e0f0a884a6bf19922f65e55d87e639c788d2cc02ad956cf7d7c75"
},
"downloads": -1,
"filename": "flac_raster-0.1.2.tar.gz",
"has_sig": false,
"md5_digest": "db4006b459a93a67c25744d5112d3e9a",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 45222,
"upload_time": "2025-07-20T16:12:18",
"upload_time_iso_8601": "2025-07-20T16:12:18.500868Z",
"url": "https://files.pythonhosted.org/packages/9d/5e/bc6877635c8651debef65fba9bd0bbb1a7093e2b0b9c2f6eb75014aa07a2/flac_raster-0.1.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-07-20 16:12:18",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "Youssef-Harby",
"github_project": "flac-raster",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "flac-raster"
}
