Apache OpenOffice (AOO) Bugzilla – Issue 115697
PDF creation: Reader cannot extract embedded font
Last modified: 2017-05-20 10:15:10 UTC
When using the PDF icon or "Export as PDF" approach with the "PDF/A-1a" option and the "Submit format PDF" option to be sure that all fonts are embedded in the PDF, Adobe Reader gives the following error message when I open the PDF: "Cannot extract the embedded font 'Formata Bold'. Some characters may not display or print correctly." The font show up as dots in the PDF. When I look at "Properties" in Reader, it says the following for the Formata Bold font: ("A" for Adobe) Formata Bold Type: Type 1 Encoding: Built in Actual Font: Unknown Same problems with Formata Regular. These are legitimate Adobe fonts that I've had for so long that they came on 3.5" floppies. I tried a simple two-line document, one line using Formata Regular and one using Formata Bold. I had the problem with both fonts. They show up as dots in the resulting PDF. Further problems: If I cut & past from the PDF file and drop the dots into an OO Writer document, they reappear as the text of the original Formata fonts, although the "regular" font is now bolder, but not quite as bold as the actual Bold version of the font. Also, looking at the windows for the font name and point size, the fonts that were just pasted back from Reader are slightly smaller in size than the original text (11.5 pts vs. 12 pts), and the names appear as follows: "Formata;Formata" (it should be "Formata Regular") "Formata;Formata Bold" (It should be "Formata Bold") I have many documents that use the Formata fonts, both Bold and Regular, so I don't want to have to edit all of them to use a different font. These fonts have worked fine with earlier versions of OO. They are legitimate fonts, from Adobe.
The same problems exist with version 3.3 RC6. So I uninstalled it and reinstalled version 3.2.1 and the PDF export now works as it should.
Please provide a sample file as *.odt & *.pdf Thank you
Created attachment 75212 [details] OpenOffice 3.2 test file (.odt)
Created attachment 75213 [details] OpenOffice3.2 text (.pdf)
Created attachment 75214 [details] OpenOffice3.3 RC7 text (.odt)
Created attachment 75215 [details] OpenOffice3.3 RC7 pdf (.pdf)
I am attaching four files: TestOOO320m18.odt TestOOO320m18.pdf TestOOO330m17.odt TestOOO330m17.pdf The first two were created with OpenOffice 3.2. The last two were created with OpenOffice 3.3 RC7. I used the pdf icon in each version of OO to create the pdf file from the .odt file for that version. In addition to the problems listed in my original post, there is another problem with the pdf files created by OO, a problem that has existed for many years: The trademark symbol (I describe it as "circle-R") in the Formata fonts is displayed as the characters "fi" in the resulting pdf. I believe that the trademark symbol officially exists in the Formata fonts, but OO has never handled it properly. I use the Insert/Special Character feature of OO to insert the trademark symbols. They display and print correctly in OO, but not in the resulting pdf files. I have had to use the trademark symbol from a more common font such as Helvetica to get a true trademark symbol to display. I have had the same problem with the ohm symbol and others in the Formata fonts. Regarding the process of attaching files here, I hope I chose the correct mime descriptions. The files are .odt and .pdf files from their respective versions of OO. If you need anything further, please let me know. Thank you.
As mentioned, with Formata Bold, the trademark symbol (circle-R) ends up as the letters "fi" in the resulting pdf that is created from OO. Using the Insert/ Special Character function of OO, I scrolled through all of the characters in the Formata Bold font and found "fi" under the subset described as "Alphabetic Presentation forms". The character address of "fi" was listed as "U+FB01" in the window for inserting special characters. The actual trademark symbol has a character address of "U+00AE (174)", according to that window. Somehow, OO mishandles things when converting to pdf.
Just to clarify: I reported these problems with OO 3.3 RC5, but the problems still exist with RC7.
Same problems exist with OpenOffice 3.3.0 RC8.
I just took the .odt test file that I created in OO 3.3.0 RC8 and brought it into OO 3.2.1, then used the pdf icon to create a pdf file. When I opened the resulting pdf file in Adobe Reader it displayed the main text of the Formata fonts correctly, but still displayed the ligature "fi" in place of the trademark symbol (circle-R).
These problems continue to exist with OpenOffice 3.3.0 RC9. May I respectfully ask, is this being investigated? Thank you.
These problems still exist in the official release of OpenOffice 3.3.0.
@PL: I don't have this formata font so it will be replaced. Therefore I can't reprduce the problem with error message and the dots. Please invetigate, Thanks.
Resetting target accroding to new task handling scheme announced here: http://blogs.sun.com/ratte/entry/some_changes_for_the_openoffice And BTW: what makes you think I would have this font ?
pl: (Sorry if this is a duplicate message - I sent one but it has not shown up) I am the person that reported this problem. I would be happy to send the Formata fonts to you for the purposes of investigating the problem, if it is appropriate to do so. If so, please advise the best way to get them to you. Thank you.
I asked some colleagues and it seems I now have access to a "Formata" OpenType font.
If the Formata "OpenType" fonts that you received do not reproduce the problems, please let me know and I will get these early Formata fonts to you. These are Adobe Postscript fonts that I bought a very long time ago, so long ago that they came on a floppy disc. There may be differences between them and the OpenType versions that you just got. Thank you.
I can confirm this bug with the poster's PDF on the final release of 3.3, also in PDFs created on my system (XP) with, so far, *any* Type 1 font (pfb/pfm), and with any export setting (not just PDF/A-1a). PDFs with postscript-flavour OTFs are fine. Notes: 1. The "cannot extract" error occurs *only* in Adobe Acrobat (in my case version 6.0). The same PDFs open and display just fine in Foxit and PDF-XChange. Both report correctly embedded fonts, and both will still display correctly when the fonts are not installed on the system. 2. PDFs created from within OOo but using a 3rd party PDF printer, in my case PDFCreator, open without problems in Acrobat. 3. Looking at a readable .ps conversion of the .pdf (pdftops.exe from xpdf 3.01), I see that although the font outlines are correctly embedded, the encoding table that follows is completely screwed in the OOo 3.3-generated version, with most glyphs named /.notdef (and hence the dots). Strangely, if I regenerate the pdf using that .ps file (for which PDFCreator obliges), the resulting pdf opens without trouble in Acrobat, with the fonts correctly embedded. In brief, OOo is screwing up the encoding of type 1 fonts but somehow readers other than Acrobat are managing to work around the problem, or never actually use the embedded font's encoding. Definite regression, as it obviously worked before, and I'd say a priority given that the actual PDF, albeit readable in some/most PDF readers, is mis-encoded and therefore corrupt.
Duplicate of issue 114901 (sorry, only just noticed). hdu - although I don't use Type 1 fonts, I know plenty who still do, and who would never in a dozen lifetimes know what to do with Fontforge, assuming they even knew it existed and got a friendly nerd to install it for them... So it's somewhat depressing to see 114901 targeted as merely "OOo Later", especially as it's a regression.
I was unable to reset my password, so I am reregistered under a new identity. I filed the bug. Thank you.
I can confirm this bug with 3.4 too. I'm using it headlessly on FreeBSD 7.4/amd64 to convert documents in PDF in KnowledgeTree. This patch solves: http://cgit.freedesktop.org/libreoffice/libs-gui/commit/?id=ba678b7d44fbde9e9ac1f0e01b49dc0b211530c0 HTH.
This Issue requires more information ('needmoreinfo'), but has not been updated within the last year. Please provide feedback as requested and re-test with the the latest version of OpenOffice - the problem(s) may already be addressed. You can download Apache OpenOffice 3.4.1 from http://www.openoffice.org/download Please report back the outcome of your testing, so this Issue may be closed or progressed as necessary - otherwise the issue may be Resolved as Invalid in the future.
Mark as duplicate per comment from issue author *** This issue has been marked as a duplicate of issue 114901 ***