Skip to content

Changelog

Changelog#

0.9.3 - 2026-06-04#

  • Added SVG cover image support for EPUB thumbnails (rasterized via resvg)
  • Added IsccThumbExtractionError for recoverable thumbnail extraction failures
  • Changed code_iscc() to handle thumbnail extraction failures gracefully (logs warning, continues without thumbnail instead of raising); genuinely corrupt or invalid source files still raise IsccExtractionError
  • Changed code_iscc() to generate thumbnails early, before heavy content processing
  • Removed EPUB cover fallback to first manifest image (only explicit cover references are used)
  • Fixed EPUB3 cover-image detection for manifests with multiple space-separated property tokens
  • Fixed EPUB cover extraction when archive entries store UTF-8 filename bytes without the ZIP UTF-8 flag (CP437→UTF-8 recovery)
  • Fixed EPUB cover extraction for OPF hrefs containing ../. path segments
  • Fixed PNG cover thumbnail extraction failing on Photoshop-exported covers with large zTXt metadata chunks (raised PngImagePlugin.MAX_TEXT_CHUNK to 4 MB)
  • Wrapped iscc-tika parse failures (TypeError from native bridge, e.g. TIKA-237 on EPUBs with deeply nested XHTML) as IsccExtractionError in text_extract and text_meta_extract
  • Improved API documentation for code_iscc(), code_iscc_mt(), code_content(), and code_text() options
  • Refactored code_iscc_mt() for improved parallelism: text extraction runs before submitting content/semantic futures, thumbnail generation overlaps with sum/meta computation, and result merging follows the same order as code_iscc()
  • Removed redundant onnxruntime from sci and sct optional dependency groups (already a transitive dependency of iscc-sci and iscc-sct)
  • Fixed CI: skip semantic code tests on macOS Python 3.12 (onnxruntime 1.26.0 has no macOS wheels for that version)
  • Updated iscc-schema floor to >=0.8.0 (version-pinned @context/$schema URLs now resolve to 0.8.0)

0.9.2 - 2026-04-21#

  • Replaced pdf-oxide with pypdfium2 for PDF text extraction
  • Replaced pymupdf with pypdfium2 for PDF thumbnail rendering
  • Replaced pymupdf with pypdf for PDF metadata embedding

0.9.1 - 2026-04-14#

  • Pinned pdf-oxide<0.3.20 and iscc-tika<0.5.0 for stability
  • Updated uv_build requirement to >=0.11.2,<0.12.0
  • Removed deprecated PEP 639 license classifier

0.9.0 - 2026-04-12#

  • Added SVG as supported media type with rasterization, metadata, and thumbnail support
  • Added RDF/Dublin Core metadata extraction and embedding for SVG files
  • Added URL support to CLI create and extract commands
  • Added optional file_name parameter to extract_metadata and mediatype_and_mode
  • Added descriptive User-Agent header for URL downloads
  • Added file extension from content sniffing for extensionless downloads
  • Fixed URL scheme validation in download_file for security (S310)
  • Fixed type checking errors in extract_metadata
  • Fixed audio/ogg detection for OGG/OPUS audio files
  • Suppressed noisy pdf-oxide TrueType cmap warnings
  • Replaced defusedxml with lxml safe parser for SVG XML parsing
  • Upgraded ffmpeg/ffprobe from 6.1 to 8.1
  • Upgraded fpcalc (chromaprint) from 1.5.1 to 1.6.0
  • Automated documentation deployment on release

0.8.9 - 2026-03-29#

  • Fixed EPUB cover extraction for URL-encoded paths in OPF manifest
  • Added EPUB3 cover-image property detection for cover extraction

0.8.8 - 2026-03-29#

  • Replaced extractous with iscc-tika for text/metadata extraction
  • Added pre-commit quality gates with prek (formatting, linting, type checking)
  • Resolved all zuban type checking errors
  • Removed unused explicit click dependency (already a transitive dependency of typer)

0.8.7 - 2026-03-18#

  • Migrated from Pydantic v1 compatibility layer to native Pydantic v2 API
  • Replaced pdftext with pdf-oxide for PDF text extraction (~150x faster)
  • Renamed CLI command from idk to iscc-sdk
  • Rounded audio and video duration to integer seconds
  • Fixed CLI tests for Typer 0.24.x (error messages moved to stderr, exit code changes)
  • Dropped Python 3.10 support, added Python 3.14
  • Updated dependencies

0.8.6 - 2026-03-03#

  • Replaced iscc-sum with iscc-lib for Data/Instance/Sum code generation
  • Simplified code_data, code_instance, and code_sum using iscc-lib's optimized generators
  • Removed iscc-sum dependency
  • Migrated from iscc-core to iscc-lib for core ISCC algorithms
  • Removed iscc-core from dependencies (blake3, loguru, bitarray, xxhash now explicit)
  • Added optional outpath parameter to embed_metadata for custom output file paths (closes #71)
  • Added support for passing dict as metadata to embed_metadata
  • Fixed code_iscc_mt to accept name, description, and meta parameters like code_iscc
  • Migrated from Poetry to uv for dependency management and build backend
  • Replaced IPFS binary tool with pure Python CIDv1 computation (no external dependency)
  • Fixed installer script referencing removed IPFS tool functions
  • Fixed memory usage in IPFS CID computation for large files (stream-hash per chunk)
  • Updated CI GitHub Actions to latest versions
  • Updated dependencies

0.8.5 - 2025-05-12#

  • Fixed TypeError with Semantic-Code image in granular processing mode.

0.8.4 - 2025-05-12#

  • Fixed issue with passing incompatible options to iscc-sct in experimental mode
  • Improved test coverage

0.8.3 - 2025-05-12#

  • Added code_sum function for efficient combined Data-Code and Instance-Code generation
  • Updated code_iscc & code_meta to accept custom metadata and optionally skip file metadata extraction
  • Updated code_iscc to compute Data & Instance Code in one go and process data sequentially
  • Added separate code_iscc_mt for multihreaded processing
  • Updated dependencies

0.8.2 - 2025-05-02#

  • Added experimental semantic codes for text and image with optional iscc-sct and iscc-sci packages
  • Added byte_offsets option to support UTF-8 byte positions in granular simprints
  • Improved image transparency handling with more robust conversion logic
  • Improved text sanitization
  • Updated lock file dependencies

0.8.1 - 2025-04-09#

  • Added text_keep option to store extracted text on IsccMeta.text property
  • Changed PDF text extraction to use pypdfium2 with reading order reconstruction
  • Re-added pillow-avif-plugin (Pillow dropped bundling libavif)

0.8.0 - 2025-04-02#

  • Added support for optional ISCC-UNITS and configurable bit size (256-bit).
  • Added extract command to CLI for text extraction with error handling.
  • Added AVIF thumbnail format support and metadata stripping for thumbnails.
  • Added robust EPUB cover image extraction with multiple fallback strategies.
  • Added fixed layout EPUB detection with namespace support.
  • Added container processing support for embedded elements in epub files.
  • Added min_image_size option to filter images during EPUB processing.
  • Added text sanitization and HTML cleaning for metadata extraction.
  • Added image_strip_metadata function for thumbnail generation.
  • Added AVIF thumbnail format support.
  • Modified code_iscc to conditionally generate meta-code based on create_meta option.
  • Updated text features generation to use 256-bit MinHash algorithm.
  • Updated video feature extraction with improved hash bits and metadata structure.
  • Simplified media processing functions by removing redundant optional parameters.
  • Updated dependencies.

0.7.0 - 2025-03-23#

  • Added Python 3.13 support
  • Added support for scene based granular video code processing
  • Added ffprobe video metadata extraction (duration, fps, width, height, language)
  • Added support str and Path objects for all file inpunts
  • Added generator (name and version of software) to ISCC metadata
  • Added optional fallback to ISCC-SUM for unsupported media types
  • Handle unsupported SVG files gracefully
  • Replaced python-magic with puremagic wrapper for cross-platform compatibility
  • Replaced exiv2 CLI tool with native python bindings
  • Replaced Tika Java dependency with native extractous package
  • Fix bug with thumbnail generation for image modes

0.6.2 - 2024-06-13#

  • Update and relax dependencies
  • Update ISO project status

0.6.1 - 2024-02-05#

  • Improved robustness of granular text features
  • Added extract_meta & create_thumb parameters
  • Fixed redundant metadata extraction
  • Updated dependencies

0.6.0 - 2024-01-22#

  • Add avif and heic image format support
  • Update to Exiv2 0.27.7
  • Update to Tika 2.9.1
  • Update to FFMPEG 6.1
  • Add Python 3.12 Support
  • Add Pydantic v2 Support
  • Dropped Python 3.8 Support

0.5.9 - 2023-06-21#

  • Fix stdout decoding

0.5.8 - 2023-06-21#

  • Add install cli command
  • Ignore signature files in batch cli command
  • Fix encoding of result files
  • Updated dependencies

0.5.7 - 2023-06-21#

  • Added parallel processing of ISCC-UNITs
  • Handle video thumbnail extraction errors gracefully
  • Add basic command line interface for batch processing
  • Add option to keep MP7 Video Signature
  • Support concurrent audio metadata extraction
  • Updated dependencies
  • Fixed mkdocstrings

0.5.6 - 2023-04-28#

  • Don´t install tika more than once per session
  • Don´t pin poetry build requirement version
  • Publish wheels
  • Use latest poetry with default settings in CI
  • Update dependencies

0.5.5 - 2023-03-26#

  • Switch to official pytaglib distribution
  • Update dependencies

0.5.4 - 2023-03-15#

  • Added global extract_metadata option
  • Removed taglib installation from CI

0.5.3 - 2023-03-12#

  • Update to iscc-core 1.0.3 using binary wheels

0.5.2 - 2023-03-11#

  • Added Python 3.11 support
  • Added WavPack support
  • Added docx metadata embedding support
  • Improved error messages for unsupported mediatypes
  • Improved robustness of audio metadata extraction
  • Normalize mediatype application/xml to text/xml
  • Nomralize mediatype application/vnd.ms-asf video/x-ms-asf
  • Normalize mediatype application/vnd.adobe.flash.movie to application/x-shockwave-flash
  • Add docx and xlsx extensions to mediatypes
  • Updated to iscc-core 1.0.1
  • Updated various dependencies

0.5.1 - 2022-12-09#

  • Fix audio title embedding

0.5.0 - 2022-12-09#

  • Add audio cover art thumbnailing
  • Integrate EPUB metadata embedding

0.4.9 - 2022-12-09#

  • Add EPUB cover image thumbnail extraction
  • Add EPUB metadata embedding
  • Retain existing PDF DocInfo metadata when embedding

0.4.8 - 2022-12-09#

  • Separeta ISCC_CORE and ISCC_SDK options
  • Document ISCC_SDK option
  • Integrate PDF thumbnail extraction

0.4.7 - 2022-12-08#

  • Add metadata embedding support for PDFs
  • Add thumbnail support for PDFs
  • Improve text metadata extraction mapping
  • Improve documentation
  • Update dependencies

0.4.6 - 2022-11-24#

  • Add support for earlier Pillow versions
  • Add option to configure image_max_pixels
  • Don´t embed empty titles in images
  • Handle extracted metadata gracefully
  • Fixed CVE-2007-4559
  • Update to iscc-schema v0.4.0
  • Update to iscc-core v0.2.12
  • Update to tika 2.6.0
  • Update dependencies

0.4.5 - 2022-07-04#

  • Update to iscc-schema v0.3.9
  • Update to iscc-core v0.2.11
  • Fix issue with embedded identifiers
  • Add support for granular text features

0.4.4 - 2022-06-08#

  • Update to iscc-schema v0.3.8
  • Updated dependencies

0.4.3 - 2022-05-08#

  • Moved changelog to seperate file
  • Fixed pillow resampling deprecation warning
  • Fixed exiv2 error on older linux versions
  • Added support for lazy installation of cli-tools
  • Added cleanup of cli-tool archives after installation
  • Removed obsolete ffprobe tool
  • Updated dependencies

0.4.2 - 2022-04-27#

  • Fix metadata extraction failure with long texts
  • Embed Dublin Core metadata in images
  • Support path object inputs for text_name_from_uri

0.4.1 - 2022-03-26#

  • Added video thumbnail support
  • Added support for IPFS wrap-with-directory

0.4.0 - 2022-03-21#

  • Added Text-Code generation
  • Added Video-Code generation
  • Added text document metadata extraction
  • Added video metadata embedding and extraction
  • Added custom exeptions
  • Changed embedding to create a new mediafile
  • Set filename on IsccMeta
  • Set @type on IsccMeta

0.3.0 - 2022-03-12#

  • Added support for Audio-Code with metadata embedding/extraction

0.2.0 - 2022-03-10#

  • Added IPFS support

0.1.0 - 2022-03-09#

  • Initial release with support for ISCC Content-Code Image