summaryrefslogtreecommitdiff
path: root/chart2/source/model/main/DataSeriesProperties.cxx
diff options
context:
space:
mode:
authorJonathan Clark <jonathan@libreoffice.org>2025-07-09 14:49:31 -0600
committerJonathan Clark <jonathan@libreoffice.org>2025-07-10 01:10:08 +0200
commit7ef1c437f30b0869a5b9fa33809bac2c6665ace3 (patch)
tree71da7c8c89761383c8c420d7de497651ad1c6ac5 /chart2/source/model/main/DataSeriesProperties.cxx
parent744b0b4634171841df24ec7a8094f95e12058048 (diff)
tdf#64991 Remove unnecessary special case from IdctHint hackHEADmaster
In order to emulate sprmCIdctHint support, the DOC importer uses a hack which splits the input text at script type changes and overrides fonts to make Writer's output match Word's. One special case in this algorithm was to override treatment of ASCII-range weak characters as Latin script. This special case had the unfortunate side effect of causing serious performance problems with non-Latin DOC files, as what would otherwise be treated as normal text was split into an excessive number of spans at any ASCII-range weak character (including space characters). This special case was also unnecessary, as our default implementation already handles these characters the same way Microsoft Word does. This change deletes that special case. This should resolve the performance issue, and shouldn't otherwise affect document layout. Since this hack was added, Writer has gained support for its own version of script type hinting. Currently, that implementation isn't perfectly compatible with DOC/DOCX, but in the future we should be able to delete this entire workaround and support sprmCIdctHint natively. Change-Id: I25ffe38efc963b8d569e2911c64a70852e22c679 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/187602 Tested-by: Jenkins Reviewed-by: Jonathan Clark <jonathan@libreoffice.org>
Diffstat (limited to 'chart2/source/model/main/DataSeriesProperties.cxx')
0 files changed, 0 insertions, 0 deletions