[jira] [Commented] (PDFBOX-4115) Problem creating PDF with German text using embedded Type1 (PFB) font

["Tilman Hausherr (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/PDFBOX-4115?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16368506#comment-16368506 ] 

Tilman Hausherr commented on PDFBOX-4115:
-----------------------------------------

Here's a workaround until I've found the cause:
{code:java}
PDType1Font font = new PDType1Font(document, new FileInputStream("n019003l.pfb"), WinAnsiEncoding.INSTANCE);
....
COSDictionary dict = new COSDictionary();
dict.setItem(COSName.BASE_ENCODING, COSName.WIN_ANSI_ENCODING);
font.getCOSObject().setItem(COSName.ENCODING, dict);
{code}
How I found this: I searched for test files with a type1 non subsetted embedded font and looked at the font with PDFDebugger. PDFBOX-1713 has such a file. Note the slight difference in font construction (not relevant to the problem)

> Problem creating PDF with German text using embedded Type1 (PFB) font
> ---------------------------------------------------------------------
>
>                 Key: PDFBOX-4115
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-4115
>             Project: PDFBox
>          Issue Type: Bug
>          Components: FontBox
>    Affects Versions: 2.0.8
>            Reporter: Tamir Hassan
>            Priority: Major
>              Labels: type1, type1font
>         Attachments: n019003l.pfb
>
>
> Hi all,
> When creating a PDF and adding text using a PostScript Type1 font (e.g. the attached n019003l.pfb but also others), an error occurs when the text contains German characters. After a reply from Tilman Hausherr on the users mailing list, I have been advised to submit this as an issue.
> I am using the latest release (2.0.8).
> The error occurs with e.g. the character "ä" (adieresis) and other similar umlaut characters; it does not occur with "ß" (germandbls).
> Using an embedded TTF seems to work fine but when I load the PFB like this:
> InputStream pfb = new FileInputStream(fontFile);
> font = new PDType1Font(document, pfb);
> I get an encoding error whenever I try to print an "ä" to the page:
> java.lang.IllegalArgumentException: U+00E4 ('adieresis') is not available in this font NimbusSanL-Regu (generic: NimbusSanL-Regu) encoding: built-in (Type 1)
> If I specify a different encoding (WinANSI) when loading the font:
> InputStream pfb = new FileInputStream(fontFile);
> font = new PDType1Font(document, pfb, new WinAnsiEncoding());
> then the exception is not thrown, but I just have an empty space in place of the "ä".
> I have tried to look into the code, in particular I have played around with the class PDType1FontEmbedder.
> When the FontBox object Type1Font is created by the parser in the following line of code:
> type1 = Type1Font.createWithPFB(pfbBytes);
> I have tried to look into the charstring dictionary:
> type1.getCharStringsDict()
> and, by iterating through the set keys, can see that "adieresis" is in there.
> However, when using the default encoding from the font (i.e. by passing "null" to the PDType1FontEmbedder), the resulting encoding that is obtained by the following line of code:
> fontEncoding = Type1Encoding.fromFontBox(type1.getEncoding());
> does not contain "adieresis" (or other "compound" characters), but just "dieresis"
> Thanks,
> Tamir



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: dev-help@pdfbox.apache.org