[
https://issues.apache.org/jira/browse/PDFBOX-5384?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17501192#comment-17501192
]
Andreas Lehmkühler commented on PDFBOX-5384:
--------------------------------------------
In theory the toUnicode mapping isn't involved at all when it comes to
rendering. It is used for text extraction only.
The given pdf is broken in a couple of ways
* it uses CID type 2 fonts but doesn't embed them
* it claims to use an identity mapping but doesn't
* the toUnicode CMap is named as identity map but it isn't
* the toUnicode CMap doesn't contain any toUnicode mappings
* the toUnicode CMap contains CID-mappings only
There are many workarounds in place to handle malformed pdfs and in the given
case the toUnicode mapping is involved. Obviously the added one from
PDFBOX-4322 broke the rendering of this one. Maybe there is a way to support
both
> Wrong glyphs used
> -----------------
>
> Key: PDFBOX-5384
> URL: https://issues.apache.org/jira/browse/PDFBOX-5384
> Project: PDFBox
> Issue Type: Bug
> Components: Rendering
> Affects Versions: 2.0.25
> Reporter: Oliver Schmidtmer
> Priority: Major
> Labels: regression
> Attachments: DOR-EC E-N20_118345.pdf,
> image-2022-03-02-23-41-15-844.png
>
>
> The attached PDF uses Tahoma fonts.
> It seems the correct font is used, but it uses the wrong glyphs.
> For example the "6" from the screenshot is definitely from Tahoma Glyph 25 /
> CID 54, where it should be "S" Glyph 54 / CID 83.
> The "=" in screenshot is Glyph 32 CID 61 where "Z" Glyph 61 CID 90 should be
> used.
> !image-2022-03-02-23-41-15-844.png!
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
