fix: match GNU error format for unrecognized options

[Ryujiyasu]

Summary

Follow-up to #178 / #179. Adjusts the unrecognized option error message to match the actual GNU diff/cmp output format:

• Remove colon after unrecognized option
• Use single quotes instead of double quotes ('--foobar' instead of "--foobar")
• Use contains instead of starts_with in integration test to handle the command prefix (e.g. cmp: unrecognized option ...)

Before: unrecognized option: "--foobar"
After: unrecognized option '--foobar'

As pointed out by @GunterSchmidt in #178 (comment).

Test plan

• cargo test — all 125 unit tests and 21 integration tests pass
• Verified diffutils diff --foobar outputs unrecognized option '--foobar'
• Verified diffutils cmp --foobar outputs unrecognized option '--foobar'

[GunterSchmidt]

Looks good! Thanks.

[codspeed-hq]

Merging this PR will not alter performance

✅ 26 untouched benchmarks

---

_{Comparing Ryujiyasu:fix/unrecognized-option-format (5398c6d) with main (59e130a)}

[oSoMoN]

The change looks harmless. I haven't been paying much attention to uutils/diffutils lately, and I'm having a hard time parsing the following things:

• The two commits in this PR are co-authored by Claude Opus 4.6, but their contents are really very trivial. Is that a reasonable use of LLMs? Has the use of coding agents been discussed in the context of uutils, do we have some sort of policy/guidelines around their usage in the projects?
• The two commits are authored by @Ryujiyasu, but committed by @sylvestre, in Ryuji Yasukochi's branch. What's up with this? EDIT: is this because Sylvestre rebased and force-pushed?

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 0.00%. Comparing base (59e130a) to head (5398c6d).
⚠️ Report is 6 commits behind head on main.

Additional details and impacted files

@@     Coverage Diff     @@
##   main   #180   +/-   ##
===========================
===========================

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.