CHANGELOG.md

v2.15.1 - 2025-01-21

Fix

• Backward compatible add_text() (#132) (7e45817)

v2.15.0 - 2025-01-21

Feature

• Add CodeItem as pydantic type, update export methods and APIs (#129) (c940aa5)

Fix

• Fix hybrid chunker token constraint (#131) (b741eea)
• Always return a new bbox when changing origin (#128) (841668f)

v2.14.0 - 2025-01-10

Feature

• Dev/add labels for pictures-classes (#127) (078cd61)

v2.13.1 - 2025-01-08

Fix

• Restore proper string serialization of DocItemLabel (#124) (a52bb88)

v2.13.0 - 2025-01-08

Feature

• Add mapping to colors into DocItemLabel (#123) (639f122)

Fix

• Quote referenced URIs in markdown and html (#122) (127dd2f)

v2.12.1 - 2024-12-17

Fix

• Fixes for legacy-doc handling (#115) (b116c46)

v2.12.0 - 2024-12-17

Feature

• Added the new label comment_section in the groups (#114) (5101dd8)

Fix

• Skip labels not included in the allow-list (#113) (d147c25)
• Always write with utf8 encoding (#111) (268c294)

v2.11.0 - 2024-12-16

Feature

• Add group labels for form and key-value areas (#110) (aeaf89d)

v2.10.0 - 2024-12-13

Feature

• Add legacy to DoclingDocument utility (#108) (b31e0a3)
• Add DoclingDocument viewer to CLI (#99) (9628d19)
• Add default tokenizer to HybridChunker (#107) (2591c70)

Fix

• Improve doc item typing (#105) (047a196)
• Set origin when merging chunks (#109) (b546c0a)
• Add REFERENCE to exported labels and remove CAPTION (#106) (a66b0bb)

v2.9.0 - 2024-12-09

Feature

• Utilities converting document formats (#91) (437c498)

Fix

• markdown: Preserve underscores in image URLs during markdown export (#98) (fd7529f)

v2.8.0 - 2024-12-06

Feature

• Add hybrid chunker (#68) (628ab67)

v2.7.1 - 2024-12-06

Fix

• Multimodal output (#96) (2133af6)

v2.7.0 - 2024-12-04

Feature

• Export to OTSL method for docling doc tables (#86) (180e294)

v2.6.1 - 2024-12-02

Fix

• Fix circular import (#87) (63e6c01)

v2.6.0 - 2024-12-02

Feature

• Extend source resolution with streams and workdir (#79) (9a74d13)
• Simple method to load DoclingDocument from .json files (#71) (fc1cfb0)

Fix

• Allow all url types in referenced exports (#82) (3bd83bc)
• Even better style for HTML export (#78) (8422ad4)

v2.5.1 - 2024-11-27

Fix

• Hotfix for TableItem.export_to_html args (#76) (ae2f131)
• Artifacts dir double stem (#75) (f93332b)

v2.5.0 - 2024-11-27

Feature

• Adding HTML export to DoclingDocument, adding export of images in png with links to Markdown & HTML (#69) (ef49fd3)

v2.4.1 - 2024-11-21

Fix

• Temporarily force pydantic < 2.10 (#70) (289b629)

v2.4.0 - 2024-11-18

Feature

• Add get_image for all DocItem (#67) (9d7e831)
• Allow exporting a specific page to md. (#63) (1a201bc)

v2.3.2 - 2024-11-11

Fix

• Fixed selection logic for a slice of the document (#66) (dfdc76b)

v2.3.1 - 2024-11-01

Fix

• Include titles to chunk heading metadata (#62) (bfeb2db)

v2.3.0 - 2024-10-29

Feature

• Added pydantic models to store charts data (pie, bar, stacked bar, line, scatter) (#52) (36b7bea)

v2.2.3 - 2024-10-29

Fix

• Str representation of enum across python versions (#60) (8528918)
• Title for export to markdown and add text_width parameter (#59) (4993c34)

v2.2.2 - 2024-10-26

Fix

• Fix non-string table cell handling in chunker (#58) (b5d07b2)

v2.2.1 - 2024-10-25

Fix

• Escaping underscore characters in md export (#57) (c344d0f)

v2.2.0 - 2024-10-24

Feature

• Add headers argument and a custom user-agents for http requests (#53) (44941b5)

Fix

• Fix resolution in case of URL without path (#55) (2c88e56)

v2.1.0 - 2024-10-22

Feature

• Improve markdown export of DoclingDocument (#50) (328778e)
• Extend chunk meta with schema, version, origin (#49) (d09fe7e)

v2.0.1 - 2024-10-18

Fix

• Fix legacy doc ref (#48) (e12d6a7)
• Add-mimetype-for-asciidocs-markdown (#47) (0aab007)

v2.0.0 - 2024-10-16

Feature

• Expose DoclingDocument as main type, move old typing to legacy (#41) (03df97f)

Breaking

• Expose DoclingDocument as main type, move old typing to legacy (#41) (03df97f)

v1.7.2 - 2024-10-09

Fix

• Loosen pandas version (#40) (aec1a41)

v1.7.1 - 2024-10-07

Fix

• Make doc metadata keys pure strings (#38) (246627f)
• Align chunk ref format with one used in Document (#37) (b5592ad)

v1.7.0 - 2024-10-01

Feature

• (experimental) introduce new document format (#21) (688789e)
• Add doc metadata extractor and ID generator classes (#34) (b76780c)
• Support heading as chunk metadata (#36) (4bde515)

v1.6.3 - 2024-09-26

Fix

• Change order of JSON Schema to search mapper transformations (#32) (a4ddd14)

v1.6.2 - 2024-09-24

Fix

• Remove duplicate captions in markdown (#31) (a334b9f)

v1.6.1 - 2024-09-24

Fix

• Remove unnecessary package dependency (#30) (e706d68)

v1.6.0 - 2024-09-23

Feature

• Add figures in markdown export (#27) (b843ae6)

v1.5.0 - 2024-09-20

Feature

• Add export to doctags for document components (#25) (891530f)
• Add file source resolution utility (#22) (752cbc3)

v1.4.1 - 2024-09-18

Fix

• Add export to xml and html (#17) (9bc256e)

v1.4.0 - 2024-09-18

Feature

• Add table exporters (#20) (2cc2429)

v1.3.0 - 2024-09-11

Feature

• Add hierarchical chunker (#18) (9698d30)

v1.2.0 - 2024-09-10

Feature

• Added the XML export (#16) (acdf816)

v1.1.4 - 2024-09-06

Fix

• Validate_model() could be called with other types rather than dict (#14) (235b2cd)

Documentation

• Update link to Docling (#12) (aaf17fe)

v1.1.3 - 2024-08-28

Fix

• Use same base type for all components (#10) (f450c8c)

v1.1.2 - 2024-07-31

Fix

• Make page number strictly positive (#8) (ec3cff9)

v1.1.1 - 2024-07-23

Fix

• Set type to optional (#7) (faf472c)

Documentation

• Revamp installation instructions (#6) (3f77b2e)

v1.1.0 - 2024-07-18

Feature

• Add document Markdown export (#4) (d0ffc85)

v1.0.0 - 2024-07-17

Feature

• Trigger v1.0.0 release (#3) (daece4c)

Breaking

• trigger v1.0.0 release (#3) (daece4c)

v0.0.1 - 2024-07-17

Fix

• Fix definition issues in record type (#2) (656f563)