html report "pretty_file" prevents browser search

[valumapps]

Describe the bug
When outputting html coverage, a user will try to use browser builtin search to find a file, but the search will fail to find the desired file.

To Reproduce

Answer the questions below:

1. What version of Python are you using? python3.13
2. What version of coverage.py shows the problem? The output of coverage debug sys is helpful.
3. What commands should we run to reproduce the problem?

• run coverage html
• then open htmlcov/index.html
• in browser, type ctrl-f (or / depending on your browser)
• type a file path that is present in the page, e.g. foo/bar/baz.py
• the file is not found, when looking in page, it seems there are spaces in name
• type instead foo<space character>/<space character>bar<space character>/<space character>baz.py
• still not found

Expected behavior
The file should be found in browser search

Additional context
The problem is that coverage html outputs "fancy" unicode characters that nobody can type:

coveragepy/coverage/html.py

Line 860 in 8fba929

return re.sub(r"[/\\]", "\N{THIN SPACE}\\g<0>\N{THIN SPACE}", filename)

it makes browser search unusable and it's tedious to find a file when there are hundreds of source files

best thing would be to plain remove the spaces but if you insist on keeping spaces, there are alternatives:

• use regular U+0020 spaces
• don't put spaces but use css to add spacing around slashes

[nedbat]

Good point! I've changed the report so that there is still a little space around the slashes, but searching works correctly. The best of both worlds.

Fixed in 7996dfd.

[valumapps]

thank you! that was real quick