idea to add error catching to avoid core dump

[yinshiyi]

I played around with the test case in Make file
when i use the -k barcode.file, then everything is fine.

flexiplex -b GGGG -k "tests/barcodes.txt" tests/input.fastq
# worked as expected

I tested without -k

flexiplex -b "????" tests/input.fastq
FLEXIPLEX 1.02.5
Setting barcode to search for: ????
For usage information type: flexiplex -h
Searching for barcodes...
Segmentation fault (core dumped)

I tested with -k ? or -k ??

flexiplex -b GGGG -k ? tests/input.fastq
FLEXIPLEX 1.02.5
Setting barcode to search for: GGGG
Setting known barcodes from ?
Number of known barcodes: 1
For usage information type: flexiplex -h
File flexiplex_reads_barcodes.txt already exists, overwriting.
Searching for barcodes...
Segmentation fault (core dumped)

following cases I can avoid core dump

if not using -k only when restrict -f, I can avoid core dump

# when -f <=2
flexiplex -b "????" -f 2 tests/input.fastq
# run without error

if using k ??? (at least 3), it is not finding anything but at least no core dump

flexiplex  -b ???? -k ???  tests/input.fastq
flexiplex  -b ???? -k ???? tests/input.fastq
flexiplex  -b GGGG -k ??? tests/input.fastq

summary

In short, I think add error catching to avoid core dump. Not sure if this is a good idea, but I would like to share my testing to get some feedbacks. Thanks!

[olliecheng]

Hi Yin, you may have stumbled upon a known bug (#70). A fix will be rolled out in the coming days. In the meantime, I believe that a modern compiler will avoid the segfault. Thanks!