Apache OpenOffice (AOO) Bugzilla – Issue 96083
goodies: dubious && and ||
Last modified: 2017-05-20 11:43:17 UTC
X && Y || Z binds operator precedence-wise as... (X && Y) || Z There are three suspicious cases of this in goodies i.e. source/filter.vcl/eps/eps.cxx while ( ( --nSecurityCount ) && ( *pDest == ' ' ) || ( *pDest == 0x9 ) ) pDest++; i.e. means... while ( (( --nSecurityCount ) && ( *pDest == ' ' )) || ( *pDest == 0x9 ) ) pDest++; source/filter.vcl/ieps/ieps.cxx while ( ( --nSecurityCount ) && ( **pBuf == ' ' ) || ( **pBuf == 0x9 ) ) (*pBuf)++; i.e. means while ( (( --nSecurityCount ) && ( **pBuf == ' ' )) || ( **pBuf == 0x9 ) ) (*pBuf)++; and if( bEnlarge && rCropLeftTop.Width() < 0 || rCropLeftTop.Height() < 0 || rCropRightBottom.Width() < 0 || rCropRightBottom.Height() < 0 ) i.e. means if( (bEnlarge && rCropLeftTop.Width() < 0) || rCropLeftTop.Height() < 0 || rCropRightBottom.Width() < 0 || rCropRightBottom.Height() < 0 ) Those meanings look very suspicious to me, I suspect that the intention is as the attached patch changes them to be.
Created attachment 57865 [details] I bet this is what we really mean
KA=>SJ: could you verify this patch, please?
This patch has been applied to cws[sjfixes10]
This patch didn't end up in integrated-DEV300m40 sjfixes10 is it in some other workspace ? (eis doesn't list it anywhere)
sj@cmc: I am helpless, the patch does not make it into the master. I committed the patch again in cws[impress169] rev 269872.
The patch has been verified by af in cws[impress169]