[sgtBurns7]

Description

added a fallback strategy to _search_in_f_string in occurences.py allowing for backwards compatibility as the initial parsing method will be the same, added unit test case for the bug provided

Fixes #(issue)

Checklist (delete if not relevant):

• [Done] I have added tests that prove my fix is effective or that my feature works
• I have updated CHANGELOG.md
• I have made corresponding changes to user documentation for new features
• I have made corresponding changes to library documentation for API changes

[mcepl]

#906 addresses the same f-string occurrence-discovery problem by replacing regex-selected fragment parsing with tokenisation.

I prefer that direction over the textual fallback here. Searching for identifier text after parsing fails can also find names inside literal content, so it needs particular care before Rename or Inline can safely use those offsets.

Could you contribute your nested-quote reproduction as a regression test for #906 and check that it passes there? That PR currently has Python 3.10 and 3.11 CI failures, so it is not yet a replacement we can merge. Please keep this PR open while that compatibility work is resolved.