galaxytools
galaxytools copied to clipboard
bgruening
can we somehow capture and show trimgalore's errors?
original shared history https://usegalaxy.org/u/bradlanghorst/h/custom-adapter-trimming
there was an error fetching fastqs and the inputs contained only HTML so trim galore failed, however it provided this useless message:
Trim Galore! on data 2 and data 1: trimmed reads pair 2
An error occurred with this dataset:
1.8
Cutadapt terminated with exit signal: '256'.
Terminating Trim Galore run, please check error message(s) to get an idea what went wrong...
Not sure how I can improve this. I don't remove any output. Maybe it is written to the repooort? https://github.com/bgruening/galaxytools/blob/master/tools/trim_galore/trim_galore.xml#L253
not sure if warnings will cause any other trouble... seems to be ok in my local instance.
now with invalid input i see:
Input file './input_singles' seems to be completely empty. Consider respecifying!
Path to Cutadapt set as: 'cutadapt' (default)
1.5
Cutadapt seems to be working fine (tested command 'cutadapt --version')
AUTO-DETECTING ADAPTER TYPE
===========================
Attempting to auto-detect adapter type from the first 1 million sequences of the first file (>> ./input_singles <<)
Found perfect matches for the following adapter sequences:
Adapter type Count Sequence Sequences analysed Percentage
@bgruening report will be filled only if trim_galore succeeds, as <command> is written now.
@nsoranzo this is what I'm saying maybe we can improve this. But the supress_warn options can also be a way to solve this.
@bwlang not sure why this is failing now :(
@bwlang my best guess is that without supress_warn trim_galore returns an error code != 0. So we might want to use <stdio> here.
stdio block rescues a couple of tests... the rest have changed output. any preference on whether to change the expected output to match what's generated with warnings on vs. a modification to the tests to be more flexible (e.g. re_match?)
@bwlang @bgruening Is this still in progress? Would be very helpful to have the reasons. Tx!
Maybe I can find some time tonight to wrap this up... sorry about the delay.
Original version reported: 0.4.1 Version on http://usegalaxy.org (that still does not present useful error message): 0.4.2
Latest version in Tool Shed 6/1/2017 that appears to be the one with @bwlang 's change: 0.4.3.1
@davebx @martenson Could we please update Main?
Update - 0.4.3.1 was installed at http://usegalaxy.org.
When given a truncated fastq input, the error is still non-specific.
Cutadapt terminated with exit signal: '256'. Terminating Trim Galore run, please check error message(s) to get an idea what went wrong...
@bwlang Is this still expected? I haven't looked at the test cases yet, but this one is pretty common, so if not included yet, perhaps could include it next update?