Pseuodocode review

[adler-sudo]

Great work Mr. Persse!

Generating a dictionary at the beginning that stores expected combinations of indexes may be useful in determining where to send records based on results.

What will your method be for getting records to new fastq files? I have seen a couple options: open all output files, generate dictionary, or append as you go. Not sure what the best practice is going to be here but may be good to think about CPU and memory usage.

In the last piece regarding the handling of indexes with N, are you looking thinking about potentially doing your best to estimate the missing value? I think this may be above and beyond. My understanding is we just send this to the low-qual/missing index file. Though maybe you'll get bonus points for posing a potential alternative. Just be sure not to write the estimated index to your 'matched index' file.

Functions all look great!

Keep up the good work!