Fix filename parsing for multiple dots#340
Conversation
9393ce5 to
df5bbee
Compare
|
@pradeeban Worked on the fix, nd thanks for confirming the approach |
|
hey @pradeeban, hve been reading through mkconcore.py to understand the workflow generation nd noticed it's handling a lot in one file (parsing graphml, building adjacency matrix, zmq port setup and generating scripts for docker/posix/windows). It's about 1300 lines now and still growing..... I was thinking it might be easier to maintain if we split it into separate modules. Something like having the graph parser in one file, platform-specific generators in another, and port mapping separate. that way we could test each part independently and add new features without touching everything I've already been working in there for the multi-dot filename fix, so I have a decent sense of how it flows. If this sounds like a useful direction, so you'd want this tackled now or I could include a proper plan for it in my GSoC proposal and work on it during the summer. Just wanted to check with you first if it aligns with the project's goals or if there's a reason to keep it as one file. Thanks 🌸 |
|
Hey @avinxshKD, Pls do not bother with refactoring that one-file. That is too much trouble for now. |
|
Hey @pradeeban sounds good, will keep it as-is. Thanks |
Changed all sourcecode.split(".") to sourcecode.rsplit(".", 1) in mkconcore.py (22 instances)...Now handles filenames like data.processor.py without breaking on the tuple unpack
Splitting from the right gives us exactly two parts every time, everything before the last dot becomes the name, last part is the extension
Fixes #283