Added logic to preserve carriage returns when updating the manifest file

[ranger-ross]

What does this PR try to resolve?

This is an attempt to preserve /r/n line endings in Cargo.toml during cargo add/remove by checking if it already had at least one /r/n.

See #14863

Additional information

For testing, it appears that snapbox will normalize line ending so the tests will directly compare the file contents.

[rustbot]

r? @epage

rustbot has assigned @epage.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[weihanglo]

Overall this does what we want. However, I am not sure if this should be in toml_edit or in Cargo. Found one relevant comment from an issue: toml-rs/toml#752 (comment).

[weihanglo]

#14863 isn't marked as S-accepted. I would generally recommend discussing there, and leave this for implementation only dicussion.

[ranger-ross]

isn't marked as S-accepted. I would generally recommend discussing there, and leave this for implementation only dicussion.

ah my bad 😅 I was looking through C-bug issues for small bugs to work on to learn/get some experience with the Cargo codebase. I'll be sure to stick to S-accepted or at least discuss further on the issue before starting/opening a PR.

[weihanglo]

This may be auto-converted, if core.autocrlf = true is set in git configuration on Linux. So the test always passes even without this patch. I haven't tested it locally though. Maybe we should have a .gitattributes file and set the file eol=crlf?

Slightly related, in contributing doc we recommend splitting new tests into its own commit, and it still passes CI. Then followed by a fix and a test update. The test change in the second commit tell that that the bug is actually there, and the patch really fixes it.