tree f83c95d75ab24a88a6848f7336cbbad040e79e44
parent 1443e5ac0acfb54ef2b6c2da3b26a8758bac1e6a
author Hui Yingst <nigi@chromium.org> 1604341196 +0000
committer Chromium commit bot <commit-bot@chromium.org> 1604341196 +0000

Fix issue that some characters from the embedded font don't render.

Due to character composition, different unicodes could represent the
same character, which means they are mapped to the same CID. However the
current implementation for CID to unicode mapping is always 1:1, which
leads to the situation that certain characters cannot be recognized by
one of its valid unicodes during rendering.

- To fix the embedding process, this CL changes |to_unicode| in
  LoadCompositeFont() to std::multimap so that when embedding a font to
  a PDF file, all valid unicodes for the same CID won't overwrite each
  other.

- To fix the rendering process, this CL replaces |m_Map| in
  CPDF_ToUnicodeMap with |m_Multimap| of type std::multimap so that it
  can store multiple entries with the same CID as keys.

- This CL also adds a matching embedder test which embeds a subset of
  of NotoSansSC-Regular font into a PDF and tests the text rendering
  result.

Bug: pdfium:1608
Change-Id: Ifadc2aa4df0de14e9d5a7c38da209f81769c0b3b
Reviewed-on: https://pdfium-review.googlesource.com/c/pdfium/+/75830
Reviewed-by: Daniel Hosseinian <dhoss@chromium.org>
Commit-Queue: Hui Yingst <nigi@chromium.org>