Apache OpenOffice (AOO) Bugzilla – Issue 125476
Pictures in .xls spreadsheets no longer displayed
Last modified: 2015-10-05 11:28:45 UTC
Created attachment 83862 [details] Sample XLS with inserted picture Steps to reproduce: 1. Create a new blank workbook in Excel 2. Insert -> picture 3. Save as .XLS 4. Open file with a recent build of Calc 4.2 Expected Result: Inserted picture is displayed in Calc like Excel Actual Result: “Picture X …” is displayed where the picture should be. On Ubuntu 14.4 Version: 4.1.0 AOO410m18(Build:9764) - Rev. 1589052 2014-04-22 12:08 - Linux i686 Result: PASS Version: 4.2.0-dev AOO420m1(Build:9800) - Rev. 1616944 2014-08-20_04:11:31 - Rev. 1619030 Result:FAIL I believe this issue was caused by: #i125187# more precision where the BLIP graphic is located"
Created attachment 83863 [details] Comparison of Calc 4.1 vs 4.2
Created attachment 83864 [details] Simpler .xls illustrating the issue
Reproduced with the following steps: - New Calc document - Open the Gallery, select "Computers" theme, drag the image "Computer-Laptop-Black" and drag it inside the Calc document - Save as XLS - Reload the document Result: the picture is replaced by an empty "frame" The picture can be seen if the XLS created with 4.2 is loaded with 4.1.0
It is OK in r1601656. It fails in r1608485.
(In reply to Regina Henschel from comment #4) > It is OK in r1601656. > It fails in r1608485. slacka was right, reverting Revision 1607057 from Issue 125187 (which only affects filter/source/msfilter/msdffimp.cxx) solves this regression. To test, if using git (svn): git log --oneline main/filter/source/msfilter/msdffimp.cxx Get the previous revision, say 34ad115 git checkout 34ad115 main/filter/source/msfilter/msdffimp.cxx
Should be one for me, grepping...
Checked with the given examples, these indeed have embedded pics/blobs and the flag I had identified works inverse -> my conclusion using that bit as flag was wrong (see issue https://issues.apache.org/ooo/show_bug.cgi?id=125187). I tend to take back the fix from issue 125187 completely - It seems to trigger quite some errors but fixes only one known example where it was different, thus it is more probable from today's view that in that file the bit was eventually corrupted.
Okay, reverted but kept comments so that someone else about doing the same fix/changes gets warned about it.
"alg" committed SVN revision 1629890 into trunk: i125476 Reverted i125187 but kept comments for the case someone else is about...
Setting target milestone
Confirmed fixed with AOO412m1(Build:9780) - Rev. 1705625 2015-09-28 12:45:04 (Mo, 28 Sep 2015) under Windows XP Pro x86 SP3