updated normal comments handling

addresses #19 (closed)

requested review from @christophknote

assigned to @FObersteiner

mentioned in issue #19 (closed)

added 1 commit

• 56715efa - updated comments on fileinfo

Compare with previous version

not 100% sure if this is the end of that story... there are still two files in the TODO directory of the test files. I'll have to investigate tomorrow.

@christophknote I've had another look at the two files remaining in TODO directory.

• 'korusaq-mrg10-dc8_merge_20160510_R4.ict' is parsed correctly but warnings are triggered because of variable names that do not agree with the standard (but also #20 (closed) came up)
• 'AROTAL-RAY_DC8_20040715_R1.ict' also triggers warnings and has multiple keywords per line in the normal comments section (which aren't parsed correctly), plus one line of free-form comment goes missing during parsing.

To me, this raises the question what we want to signal with a "warning". Should we parse multiple keywords per line etc. correctly and just raise a warning? If we write that data back to file, the file would look different from the input (but would be correct according to standard). Under which conditions should an exception be raised (stopping import altogether)?

added 1 commit

• 541c485b - two more files in tests

Compare with previous version

added 2 commits

• 229d6c78 - 1 commit from branch mbees:master
• 88d88727 - Merge remote-tracking branch 'upstream/master'

Compare with previous version

approved this merge request

@christophknote sorry, I didn't see this morning that there was a re-base missing. Done now.

merged