Geoconcept reader fails with some geometries

[gaige]

This may only affect computers with certain compilers, however it definitely did affect mine.

Reading lines with iYP columns would fail because the test for iGr-iYP==1 would fail due to testing iGr-iY instead (which will fail in these circumstances.

Chances are that the parenthesis are all that is necessary here, but the flipping of the conditionals makes it more clear what is going on.

ogr/ogrsf_frmts/geoconcept/geoconcept.c:
1306c1306
<                   if( (iYP!=-1 && iGr-iYP!=1) || (iGr-iY!=1) )
---
> 				  if(!(((iYP==-1) && (iGr-iY==1)) ||(iGr-iYP==1)))

[Even Rouault]

CC'ing Didier Richard, the author of the Geoconcept driver

[dgrichard]

I will give a glance asap

[dgrichard]

Replying to gaige:

ogr/ogrsf_frmts/geoconcept/geoconcept.c:
1306c1306
<                   if( (iYP!=-1 && iGr-iYP!=1) || (iGr-iY!=1) )
---
> 				  if(!(((iYP==-1) && (iGr-iY==1)) ||(iGr-iYP==1)))

I would propose this :

ogr/ogrsf_frmts/geoconcept/geoconcept.c:
1306c1306
<                   if( (iYP!=-1 && iGr-iYP!=1) || (iGr-iY!=1) )
---
> 				  if( !( ((iGr!=-1) && ( (iGr==iY+1) || (iGr==iYP+1) )) || (iGr==-1) ) )

which better reflects that if the Graphics field is set (iGr!=-1) then its rank must follow either the abscissa rank or the last abscissa rank in the fields declaration list.

Does it work better with your compiler ?

[gaige]

Replying to dgrichard:

Replying to gaige:

which better reflects that if the Graphics field is set (iGr!=-1) then its rank must follow either the abscissa rank or the last abscissa rank in the fields declaration list.

Does it work better with your compiler ?

Yep, that seems to function just fine and makes more sense.

Do you want to put it in, or shall I?

[dgrichard]

Replying to gaige:

Yep, that seems to function just fine and makes more sense.

Do you want to put it in, or shall I?

I will put it in

Thanks for the report.

[dgrichard]

fixed in r18980