Replaced newline-dependent CMS parser with less breakable regex parser - #52
Replaced newline-dependent CMS parser with less breakable regex parser#52maxnbk wants to merge 59 commits into
Conversation
Also prints to std error correctly, and exits with 1.
Obeys config.HALT_ON_ERROR. Updates tests.
Check python versions in setup.py No longer require argparse for versions > 2.6
Fix import in test_ale.py (now that __init__.py is gone)
|
Most excellent. While the CMX parser doesn't currently have unittests (which is why there hasn't been a mainline release yet), do you have a CMX example this would fail on? Would love to replicate the failure in testing. Really appreciate this change, CMX was never tested heavily. I'll merge it in later tonight. |
|
Hold off on it, I just received a couple EDLs it'll fail on and I'll update On Oct 15, 2015, at 4:55 PM, Sean Wallitsch notifications@github.com Most excellent. While the CMX parser doesn't currently have unittests Really appreciate this change, CMX was never tested heavily. I'll merge it in later tonight. — |
|
Added addl commit to cover more situations and care less what order things are in |
|
|
|
|
Fixes shidarin#54, where we had blank lines after the Column and Data headers. Added a test case for this.
Adds 3.5 and pypy3 to testing
…n generic number w config entry
…alities arising from FROM/TO CLIP fields, missing fields, whitespace problems, ASC_SOP field newline wrapping. Introduced more regex with reduced complexity of each.
5 participants
No description provided.