Created attachment 135267 [details] sample Steps to reproduce: 1. Open attached document Observed behaviour: Ole object is empty Versió: 6.0.0.0.alpha0+ ID de la construcció: 0342c5e8086c8200ecadbe9d52dd4ef6a093effb CPU threads: 4; OS: Linux 4.10; UI render: per defecte; VCL: gtk3; Configuració local: ca-ES (ca_ES.UTF-8); Calc: group [Bug found by office-interoperability-tools]
regression introduced in range https://cgit.freedesktop.org/libreoffice/core/log/?qt=range&q=b93d0cadb79f6652dec4d29ef20813f1d57cc708..208e66185b634ebc131121158f93f4f3ae4bd18e related to the EMF refactoring. Bisected with bibisect-linux64-6.0 Adding Cc: to Armin Le Grand
Created attachment 135268 [details] sample2 Another document affected by the same problem
attachment 83969 [details] from bug 68021 is also affected
Created attachment 135305 [details] sample3 Another document affected
Very strange - the WMF has a read error. There was no change on the principle reader itself. Need to compare with an older version...
CVhecked and found. Very bad trap - two (of several) ReadWindowMetafile usages pass a SvStream that is seeked to a StartPosition other than zero. So that seek position has to be used when constructing the VectorDataArray. OTOH there is one usage in sc that avoids that by copying the data to a MemStream to have it start-aligned, so this 'hidden' parameter passing was not clear to all callers of that function. Solved, put on gerrit.
comitted, done
https://cgit.freedesktop.org/libreoffice/core/commit/?id=b5f2402e023fb438341895ad0f81020571c5ec5a
Verified in Version: 6.0.0.0.alpha0+ Build ID: 78960ad06faca055a6d97afbc764c902d5d07f6f CPU threads: 1; OS: Windows 6.1; UI render: default; TinderBox: Win-x86@39, Branch:master, Time: 2017-08-30_06:31:19 Locale: es-ES (es_ES); Calc: group