Add automatic spell checking to TriBITS #571
Comments
bartlettroscoe
added
type: enhancement
component: testing
|
@KyleFromKitware, based on your comment #560 (comment), I have assigned this Issue to you. Please put it in progress and add a comment if you start working on this and let's have a short conversation over MS Teams about this. |
|
FYI: I was able to install codespell on my SNL RHEL 8 machine with: But it is showing more spelling errors. I will fix those. |
|
This brings up a good point: we should make sure that CI is running the same version of codespell that you use locally. Perhaps we should use the PIP version in CI. |
That is a good idea. I will commit and push my changes to this branch to make codespell pass on this TriBITS repo branch in the meantime. Then we can merge and deploy this PR :-) |
…BITSPub#571) While running codespell 2.2.1, it flagged 'nDoes' in: #print("\nDoes ...") which is not a mispelling (and a later version of codespell does not flag this) but I wanted to remove it since it was not really being used. And while I was at it, I remove all of the commented out print() statements (since those were just there for debugging).
These really should be deleted since we don't maintain bare clones anymore, and if we did, we could not care about these hooks. I did this as part of TriBITSPub#571.
I also added a bash script to automate installing the git hook scripts with printed messages.
…BITSPub#571) While running codespell 2.2.1, it flagged 'n{nothing}Does' in: #print("\n{nothing}Does ...") which is not a misspelling (and a later version of codespell does not flag this) but I wanted to remove it since it was not really being used. And while I was at it, I remove all of the commented out print() statements (since those were just there for debugging).
These really should be deleted since we don't maintain bare clones anymore, and if we did, we could not care about these hooks. I did this as part of TriBITSPub#571.
I also added a bash script to automate installing the git hook scripts with printed messages.
…BITSPub#571) While running codespell 2.2.1, it flagged 'n{nothing}Does' in: #print("\n{nothing}Does ...") which is not a misspelling (and a later version of codespell does not flag this) but I wanted to remove it since it was not really being used. And while I was at it, I remove all of the commented out print() statements (since those were just there for debugging).
These really should be deleted since we don't maintain bare clones anymore, and if we did, we could not care about these hooks. I did this as part of TriBITSPub#571.
I also added a bash script to automate installing the git hook scripts with printed messages.
|
With the merge of PR #573, this is complete. |
As shown here it looks like GHA is currently installing codespell version 2.1.0-1 while pip installed version 2.2.1 on my machine (and the newer version caught more possible misspelled words that I had to account for in a9fe9b7) |
|
Let's just call this done. Having different versions of codespell has not been a big problem. |
Description
I has been brought up several times that there are spelling errors in TriBITS documentation and source code. Examples include #404 and comments in may PRs (e.g. #560 (comment)). I try to remember to turn on flyspell in Emacs when editing files but I don't always remember to do it.
Possible Solutions
The best long-term solution would be to add support to TriBITS to run spell checking (with managed ignores) as part of local and GitHub Actions tests. That way, problems would be caught early and PR could not be merged until the spelling errors (the more known words are added) are addressed.
Tasks
The text was updated successfully, but these errors were encountered: