[
https://issues.apache.org/jira/browse/PDFBOX-5987?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17944444#comment-17944444
]
Tilman Hausherr commented on PDFBOX-5987:
-----------------------------------------
The first strategy didn't work because addAlternatingCurve and addCurve do a
switch between two numbers that may not be known at the time of calling. I
tried a different solution, preproccessing the sequence to replace all "DIV"
operations with their results and that worked immediately. However this is not
a general solution. The best would probably be to get rid of the converter and
do a full implementation of the type 2 charstring renderer. (Which would
produce a lot of double code and possibly be useless if no font in the wild
would use the extra features)
> Can't render some Type1 fonts.
> ------------------------------
>
> Key: PDFBOX-5987
> URL: https://issues.apache.org/jira/browse/PDFBOX-5987
> Project: PDFBox
> Issue Type: Bug
> Components: Rendering
> Affects Versions: 3.0.4 PDFBox
> Reporter: Daniel Persson
> Priority: Minor
> Attachments: PDFBOX-5987.cff, PDFBOX-5987_reduced.pdf,
> font-issue.pdf, screenshot-1.png
>
>
> This PDF seems to have a font rendering issue in PDFBox.
>
> I've used both Poppler and Chrome to view the PDF, and it renders correctly.
> Stepping through the code on the first character, I see that we have a
> rendering list with commands that end in
> -40 41 DIV -41 21 DIV CLOSEPATH ENDCHAR
>
> This is not correct if you look at the byte sequence that ends in
> 12 98 -89 12 12 -117 8 14
>
> Could be interpreted as
> -127 -41 28 -127 -127 0 RRCURVETO ENDCHAR
>
> So perhaps there is an off-by-one error somewhere or something?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@pdfbox.apache.org
For additional commands, e-mail: dev-h...@pdfbox.apache.org
