http://bugs.winehq.org/show_bug.cgi?id=10716
Alex Villacís Lasso a_villacis@palosanto.com changed:
What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|wine-bugs@winehq.org |a_villacis@palosanto.com Status|NEW |ASSIGNED
--- Comment #9 from Alex Villacís Lasso a_villacis@palosanto.com 2007-12-11 15:04:16 --- Created an attachment (id=9603) --> (http://bugs.winehq.org/attachment.cgi?id=9603) FIX for regression in 1.0 emulation on WM_GETTEXTLENGTH
Commit f945f16de201dbd834da10701025665bfa8f05ee introduced a regression in richedit 1.0 emulation, in which WM_GETTEXTLENGTH reports a length that is too long. Due to the way the text is represented in memory, the previous behavior (not counting CR->CRLF conversions) was actually correct for a richedit 1.0 control, because the offset markers at the start of each paragraph already include previous CRLF lengths. The attached patch fixes it by using the GTL_USECRLF only for richedit 2.0. Also included is a test for the behavior in riched32 so this does not happen again.
Patch already sent to wine-patches.