#2003: piraha_everywhere Zelmani stress test ---------------------+------------------------------------------------------ Reporter: rhaas | Owner: sbrandt Type: defect | Status: assigned Priority: major | Milestone: Component: Cactus | Version: development version Resolution: | Keywords: piraha ---------------------+------------------------------------------------------
Comment (by rhaas):
Replying to [comment:10 sbrandt]:
CST_fatal is new with me, AFAIK. The existing facilities didn't abort
immediately, which led to some strange additional error messages from the system that compares the results of the current parser to the old one. All it does is call CST_PrintErrors() and die.
Oh true, sorry. I only looked at the author information in CSTUtils.pl and not at git-blame for the CST_fatal routine in that file. On top of that I apparently got confused when looking at the patch not realizing in which file things are defined. Ideally it should not abort on the very first syntax error but instead report all (or at least some number) of syntax errors before aborting (same as eg a compiler reports all errors in a source file and not only the first one, same as the runtime parameter file parser does).
This is extra work and not totally trivial (in particular given that one has to recover from a faulty line of input) but I think is needed for a good error reporting.