[
https://issues.apache.org/jira/browse/PDFBOX-619?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Villu Ruusmann updated PDFBOX-619:
----------------------------------
Attachment: FontBox.patch
When testing with finer logging level (debug instead on info), it turned out
that class org.apache.pdfbox.pdmodel.font.PDType1CFont was still complaining
about not being able to handle some codes. Closer inspection revealed that all
the problematic codes were encoded as supplements.
I have enhanced the patch file to extract mapping information also from
supplements.
> Reapplying parts of PDFBOX-542
> ------------------------------
>
> Key: PDFBOX-619
> URL: https://issues.apache.org/jira/browse/PDFBOX-619
> Project: PDFBox
> Issue Type: Bug
> Components: FontBox
> Affects Versions: 1.0.1
> Reporter: Villu Ruusmann
> Attachments: FontBox.patch
>
>
> The application of PDFBOX-542 was accompanied with several optimizations.
> Later testing has revealed that one of the optimizations - the replacement of
> List-type field org.apache.fontbox.cff.encoding.CFFEncoding#entries with two
> Map-type fields #code2sid and #sid2code - may come short in situations where
> the mapping between codes and SIDs is not one-to-one (AFAIK, it is not
> covered by the CFF specification, but there are real-life examples for that).
> I've compiled a patch file that restores the code in the original state.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
