next-error hangls

[theotherbuzzy]

I'm recently having no end of troubles with the next-error function hanging (I have it bound to a key). I'm running SlickEdit 2017 (v22.0.2.1 64-bit) on Ubuntu and running make against one of my makefiles, which uses gcc (7.3.0) to compile C/C++ code.  Finding the first error works but finding the second always hangs SE (eventually Ubuntu tells me the app is not responding and I have to kill it). I searched the forums for a similar problem but only found a 10 year old post and the fix (to replace errors.e didn't seem like a good choice at this point :-).

This function has worked for months on build outputs of thousands of lines so I'm not really sure what happened. I haven't made any recent changes to the error parsing regular expressions. I know at one point recently my laptop hung and I had to power it off manually, so there may be some state or cache files that got messed up.  Is there any way to reset that (or rebuild the compiled macro files) or does anyone have any other suggestions?

[Graeme]

The "cache file" is called vslick.sta.  Close slick, rename it to something else, re-run slick and it will regenerate it.  This sometimes fixes problems like yours.  It contains "interpretable p-code" amongst other things, including slick c variable/data states.

[theotherbuzzy]

Thank you! That seemed to fix it, although the sta file was dated in October (when I installed SE on this machine). I also played around with the regular expressions I was using for parsing (removing all the ones I didn't need for this code). In the distant past I remember deleting the STA and other files to fix various problems but I was a little hesitant to do so now because I'd read about some folks having problems after doing it.

In case other folks have a problem in the future...if this didn't fix it or if it comes back I was going to export my config, save my work space files, and wipe out the .slickedit directory and reload the config.  I know some recent event caused the problem...just not sure what. Perhaps it as an ancillary file that was also updated when the STA file got rebuilt