GitParse error: Expected binary line to be in 'Binary files a/fileA and b/fileB differ' format.

[rgmz]

Please review the Community Note before submitting

TruffleHog Version

3.67.1

Trace Output

2024-02-05T13:27:57-05:00       info-2  trufflehog      Expected binary line to be in 'Binary files a/fileA and b/fileB differ' format. {"got": "QmluYXJ5IGZpbGVzIC9kZXYvbnVsbCBhbmQgImIvYXNzZXRzL3JldGFpbGVycy9PTi1pa29ueS1QbGF0Zm9ybWEtZWNvbSBcMzQyXDIwMFwyMjQga29waWEucG5nIiBkaWZmZXIK"}
2024-02-05T13:27:57-05:00       info-2  trufflehog      Expected binary line to be in 'Binary files a/fileA and b/fileB differ' format. {"got": "QmluYXJ5IGZpbGVzIC9kZXYvbnVsbCBhbmQgImIvYXNzZXRzL3JldGFpbGVycy9PTi1pa29ueS1Qcm9zdGEtc2thbGEgXDM0MlwyMDBcMjI0IGtvcGlhLnBuZyIgZGlmZmVyCg=="}

Expected Behavior

The diff is parsed.

Actual Behavior

The diff is not parsed correctly, because file B is wrapped in quotes:

Binary files /dev/null and "b/assets/retailers/ON-ikony-Platforma-ecom \342\200\224 kopia.png" differ

Steps to Reproduce

N/A

Environment

N/A

Additional Context

That weird encoding, \342\200\224, is apparently an em dash.
https://apple.stackexchange.com/questions/413832/hyphen-is-being-replaced-by-342-200-224-in-bash

References

N/A

[rosecodym]

What are the circumstances that trigger this? Is it only when one of the file paths includes weird stuff like em dashes?

[rgmz]

I believe so. Whitespace in the file name doesn't seem to result in quoting (#2201).

I think the specific logic is here:

https://github.com/git/git/blob/2a540e432fe5dff3cfa9d3bf7ca56db2ad12ebb9/diff.c#L3648 ->
https://github.com/git/git/blob/2a540e432fe5dff3cfa9d3bf7ca56db2ad12ebb9/diff.c#L3563-L3570 ->
https://github.com/git/git/blob/2a540e432fe5dff3cfa9d3bf7ca56db2ad12ebb9/diff.c#L523