All pull requests

Pull requests

• fix(pptx): preserve run-level hyperlinks and character formatting

  #4443·avanish-garg opened Sep 29, 2026·

  ·
  1

  2

• fix(latex): decode \input files the same way as the main file

  #4441·dfedoryshchev opened Sep 29, 2026Contributor·

  ·
  2

• fix(html): keep non-li children of lists in document order

  #4440·morten-lagabote opened Sep 29, 2026·

  ·
  1

  2

• fix(xb): don't dereference memberQname on typed dimensions

  #4439·inchang-ing opened Sep 29, 2026·

  ·
  1

  2

• fix(reading-order): read heads split by a spanning block top-down

  #4438·DanielNg0729 opened Sep 29, 2026Member·

  ·
  1

  4

• fix(odf): keep ODT comments out of the body text

  #4436·devYRPauli opened Sep 29, 2026Contributor·

  ·
  1

  4

• fix(odf): keep embedded chart title in meta and caption (#4368)

  #4432·vishalsahare opened Sep 28, 2026·

  ·
  2

• fix(asciidoc): consolidate the AsciiDoc backend fixes into one PR

  #4431·simpleqt opened Sep 28, 2026Contributor·

  ·
  3

• fix: detect the format of path inputs with an upper-case extension

  #4430·JoeyTan21 opened Sep 28, 2026·

  ·
  1

  2

• fix(latex): typeset non-breaking spaces, dashes and quotes

  #4416·MohammadHijjawi97 opened Sep 28, 2026Contributor·

  ·
  3

• feat(ocr): add RapidOCR TensorRT backend

  #4408·Shizoqua opened Sep 28, 2026Contributor·

  ·
  1

  3

• fix(vlm): discard base prompt

  #4407·Shizoqua opened Sep 28, 2026Contributor·

  ·
  1

  3

• fix(latex): parse multirow and multicolumn arguments in real documents

  #4404·simpleqt opened Sep 28, 2026Contributor·

  ·
  4

• fix(md): keep paragraphs with a single nested inline wrapper in one group

  #4399·simpleqt opened Sep 27, 2026Contributor·

  ·
  5

• fix(reading-order): don't merge text across a column gutter

  #4398·yashikam19 opened Sep 27, 2026Contributor·

  ·
  4

• fix(pdf): preserve requested render scale

  #4395·Shizoqua opened Sep 27, 2026Contributor·

  ·
  1

  5

• fix(asciidoc): read listing blocks as code and skip comments

  #4394·r0h1tb opened Sep 27, 2026Contributor·

  ·
  6

• fix(docx): read table rows and cells wrapped in customXml or sdt

  #4393·ebrahimeolakey opened Sep 27, 2026·

  ·
  1

  4

• fix(vlm): convert palette P+tRNS images via RGBA to preserve transparency

  #4392·romainsc opened Sep 27, 2026·

  ·
  1

  3

• fix(asciidoc): keep a titled paragraph and the text after it

  #4391·r0h1tb opened Sep 26, 2026Contributor·

  ·
  3

• fix(html): keep dt/dd groups wrapped in div elements in description lists

  #4390·LindseyZ1205 opened Sep 26, 2026·

  ·
  1

  4

• fix(mhtml): don't decode the HTML part twice when the page has a meta charset

  #4388·r0h1tb opened Sep 26, 2026Contributor·

  ·
  3

• fix(pdf): keep a line-final hyphen that belongs to the word

  #4387·stanleys12 opened Sep 26, 2026·

  ·
  1

  3

• fix(jats): treat xref/@rid as optional and as IDREFS

  #4381·chrikrah opened Sep 26, 2026Contributor·

  ·
  3

• fix(pptx): degrade an area3DChart to a picture instead of failing the deck

  #4380·chrikrah opened Sep 26, 2026Contributor·

  ·
  3