In case a continuation of a run is made starting from the last time step, swak4foam fails to read the sampledGlobalVariables file stored in the latest time step, givin g the error
word 'k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-k��F<9C>?-...'
is too long (max. 1024 characters)
This error is only generated in binary mode, in ascii mode it works fine.
Please have a look at the sample case I have included to generated the error.
I can reproduce this
Seems to be more of a OF-problem than a swak-problem (reading and writing is done with a plain >>).
Will have a look how to switch off binary for that specific file
The attached patch should fix this problem (writing and reading of the global variables is forced back to ASCII)
It is possible that the patch will fail for the README-file. Ignore
This patch fixes the problem. Thanks!
One question: do you also add the bug fixesto the hg repo http://hg.code.sf.net/p/openfoam-extend/swak4Foam openfoam-extend-swak4Foam
because this is the swak version I am tracking. Or is there any other repo where I can track the latest bug fixes? That would be easier than updating all my swak installations (I am running on multiple machine and cluster) with the patch.
Anyway, thanks for this bug fix!
Yes. The patch was generated from my local repository which I push to the public-repository when it is in a state that is fit for public use. The problem is that I did the fix in a feature branch (yeah. I know. One shouldn't do that) that is not yet ready for push
Ok, that explains probably as well why the patch includes a modification to a file which I don't have (Libraries/swakFvOptions/calcResidualFvOption.C). I am on the port_2.0.x branch. Well, the bug was fixed anyway, so it probably is not important.