8000 gh-133722: Add Difflib theme to `_colorize` and 'color' option to `difflib.unified_diff` by dougthor42 · Pull Request #133725 · python/cpython · GitHub
[go: up one dir, main page]

Skip to content

gh-133722: Add Difflib theme to _colorize and 'color' option to difflib.unified_diff 10000 #133725

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 21 commits into
base: main
Choose a base branch
from

Conversation

dougthor42
Copy link
Contributor
@dougthor42 dougthor42 commented May 9, 2025

Add a color arg (defaulting to False) to difflib.unified_diff. When True, ANSI color codes are injected to the diff lines so that the printed result looks like git diff --color:

>>> import difflib
>>> args = [['one', 'three'], ['two', 'three'], 'Original', 'Current']
>>> for line in difflib.unified_diff(*args, lineterm='', color=True):
...     print(line)
...     
--- Original
+++ Current
@@ -1,2 +1,2 @@
-one
+two
 three

image

Fixes #133722.

It's a bummer, I just missed the feature freeze window for 3.14 (it was yesterday! 😭), but c'est la vie!


📚 Documentation preview 📚: https://cpython-previews--133725.org.readthedocs.build/

@dougthor42 dougthor42 marked this pull request as ready for review May 9, 2025 03:40
@tomasr8
Copy link
Member
tomasr8 commented May 9, 2025

Thanks for the PR!

With the basic theming support in #133347 merged, perhaps we could define a separate difflib theme and use that here?

@hugovk
Copy link
Member
hugovk commented May 9, 2025

Thanks for the PR! I had a play with something similar but didn't get very far before the 3.14 freeze.

With the basic theming support in #133347 merged, perhaps we could define a separate difflib theme and use that here?

Yes, let's add theming support.

See this docstring for instructions.

For a couple of standalone examples of adding theme support to a module, see 230d658 and 5c5c3e1 from that PR.

@dougthor42
Copy link
Contributor Author

Ah, TIL! Thanks, I'll take a look at those docs and update accordingly. It'll be a few days a least before I can get to it, so for now I'll revert this back to draft.

@dougthor42 dougthor42 marked this pull request as draft May 9, 2025 17:14
@hugovk
Copy link
Member
hugovk commented May 9, 2025

Absolutely no rush for this, we have 12 months until the 3.15 feature freeze. Don't hesitate if you have questions :)

Copy link
Contributor Author
@dougthor42 dougthor42 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Wow that was really easy! The examples helped a lot too, thanks! PTAL at your leisure.

@dougthor42 dougthor42 marked this pull request as ready for review May 11, 2025 04:38
@dougthor42 dougthor42 changed the title gh-133722: Add 'color' option to 'difflib.unified_diff' gh-133722: Add Difflib theme to _colorize and 'color' option to difflib.unified_diff May 11, 2025
Lib/_colorize.py Outdated
Comment on lines 215 to 217
equal: str = ANSIColors.RESET # context lines
insert: str = ANSIColors.GREEN
delete: str = ANSIColors.RED
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I opted to use the difflib.context_diff internal terminology here. Should I use the more git-like terms context, added, and removed instead?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hmm, maybe the Git-like terms if they're going to be more familiar with people and if the difflib ones are all internal.

Where does Git mention them?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overall they are defined in the GNU unified diff detailed description.

The git diff man page mentions them in a variety of places (emphasis mine):

-U
--unified=

Generate diffs with lines of context instead of the usual three. Implies --patch.

Generating patch text with -p

  1. Hunk headers mention the name of the function to which the hunk applies.

plain

Any line that is added in one location and was removed in another location will be colored with color.diff.newMoved.

Though sometimes it uses "deleted" instead of "removed":

--numstat

Similar to --stat, but shows number of added and deleted lines


I've gone ahead and switched to the GNU unified diff terms prior to the requested sorting. Also, to keep consistent with the other dataclass attributes (which are not currently sorted), the reset value was left at the end.

Copy link
Member
@ZeroIntensity ZeroIntensity left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hm, I would expect to see this in the difflib CLI rather than a library function. But, as it turns out, difflib doesn't have a working CLI? That's different than what's documented.

I personally think it's worth adding one if we're going through the effort to add colorization.

Copy link
Member
@hugovk hugovk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's also add a What's New entry, perhaps something similar to https://docs.python.org/3.14/whatsnew/3.14.html#whatsnew314-color-argparse

Lib/_colorize.py Outdated
Comment on lines 215 to 217
equal: str = ANSIColors.RESET # context lines
insert: str = ANSIColors.GREEN
delete: str = ANSIColors.RED
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hmm, maybe the Git-like terms if they're going to be more familiar with people and if the difflib ones are all internal.

Where does Git mention them?

Comment on lines 229 to +232
syntax: Syntax = field(default_factory=Syntax)
traceback: Traceback = field(default_factory=Traceback)
unittest: Unittest = field(default_factory=Unittest)
difflib: Difflib = field(default_factory=Difflib)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

And here:

Suggested change
syntax: Syntax = field(default_factory=Syntax)
traceback: Traceback = field(default_factory=Traceback)
unittest: Unittest = field(default_factory=Unittest)
difflib: Difflib = field(default_factory=Difflib)
difflib: Difflib = field(default_factory=Difflib)
syntax: Syntax = field(default_factory=Syntax)
traceback: Traceback = field(default_factory=Traceback)
unittest: Unittest = field(default_factory=Unittest)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm very slightly concerned about sorting when the dataclasses aren't kw_only.

Should I update the dataclass decorator to include kw_only=True and then sort? Or is that out of scope of this PR?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's ask @ambv about this.

@@ -0,0 +1,2 @@
Added a ``color`` option to :func:`difflib.unified_diff` that injects ANSI color
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
Added a ``color`` option to :func:`difflib.unified_diff` that injects ANSI color
Added a *color* option to :func:`difflib.unified_diff` that injects ANSI color

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Still todo :)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oops!

Copy link
Contributor Author
@dougthor42 dougthor42 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What's New entry

Added, assuming a 3.15 release.


I would expect to see this in the difflib CLI rather than a library function.

My underlying use case and request is specific to the library function. That said, I agree that any CLI should also include coloring.

difflib doesn't have a working CLI? That's different than what's documented.

Hmm... I'm not able to find the difflib CLI docs? The closest thing I found was https://docs.python.org/3/library/difflib.html#difflib-interface but that just looks like an example of how to make a CLI for difflib. Was that what you were thinking of?

If I actually read the comment I would have seen that the docs were indeed linked... And yes, the linked item is just an example of how one might make a CLI for difflib.

I personally think it's worth adding one if we're going through the effort to add colorization.

I'd be happy to modify any existing CLI code to include colors. I can't say I'd have the time to add a CLI though.

Lib/_colorize.py Outdated
Comment on lines 215 to 217
equal: str = ANSIColors.RESET # context lines
insert: str = ANSIColors.GREEN
delete: str = ANSIColors.RED
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overall they are defined in the GNU unified diff detailed description.

The git diff man page mentions them in a variety of places (emphasis mine):

-U
--unified=

Generate diffs with lines of context instead of the usual three. Implies --patch.

Generating patch text with -p

  1. Hunk headers mention the name of the function to which the hunk applies.

plain

Any line that is added in one location and was removed in another location will be colored with color.diff.newMoved.

Though sometimes it uses "deleted" instead of "removed":

--numstat

Similar to --stat, but shows number of added and deleted lines


I've gone ahead and switched to the GNU unified diff terms prior to the requested sorting. Also, to keep consistent with the other dataclass attributes (which are not currently sorted), the reset value was left at the end.

Comment on lines 229 to +232
syntax: Syntax = field(default_factory=Syntax)
traceback: Traceback = field(default_factory=Traceback)
unittest: Unittest = field(default_factory=Unittest)
difflib: Difflib = field(default_factory=Difflib)
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm very slightly concerned about sorting when the dataclasses aren't kw_only.

Should I update the dataclass decorator to include kw_only=True and then sort? Or is that out of scope of this PR?

@@ -0,0 +1,2 @@
Added a ``color`` option to :func:`difflib.unified_diff` that injects ANSI color
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

@dougthor42
Copy link
Contributor Author

Hmm... I'm not able to find the difflib CLI docs? The closest thing I found

@ZeroIntensity Doh! I just re-read your comment and realized you linked to it haha.

Anyway yeah that's just example code. To my knowledge there is no official CLI for difflib.

27A9
Comment on lines 229 to +232
syntax: Syntax = field(default_factory=Syntax)
traceback: Traceback = field(default_factory=Traceback)
unittest: Unittest = field(default_factory=Unittest)
difflib: Difflib = field(default_factory=Difflib)
F438 Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's ask @ambv about this.

@@ -0,0 +1,2 @@
Added a ``color`` option to :func:`difflib.unified_diff` that injects ANSI color
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Still todo :)

@Wulian233
Copy link
Contributor

This PR needs to be updated to resolve the conflicts → #134581

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Add a color: bool arg to difflib.unified_diff
5 participants
0