[JENKINS-68415] Skip archives, jar-files and executables from fingerprinting

[uhafner]

For files with extension "o", "exe", "dll", "so", "a", "lib", "jar", "war", "zip", "7z", "gz", "bz2" it does not make sense to compute a fingerprint.

See JENKINS-68415

[github-actions]

☀️ Quality Monitor

• 🚦 Tests: 1271 tests passed
• 👣 Code Coverage: 92% coverage achieved
• ⚠️ Style: No warnings found
• 🐛 Bugs: 6 warnings found (0 error, 0 high, 0 normal, 6 low)

More details are available in the GitHub Checks Result.

[github-actions]

☀️ Quality Monitor

• 🚦 Tests: 1272 tests passed
• 👣 Code Coverage: 92% coverage achieved
• ⚠️ Style: No warnings found
• 🐛 Bugs: 6 warnings found (0 error, 0 high, 0 normal, 6 low)

More details are available in the GitHub Checks Result.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (0a5a2c8) 93.14% compared to head (217d5c3) 93.14%.

Additional details and impacted files

@@            Coverage Diff            @@
##               main    #1003   +/-   ##
=========================================
  Coverage     93.14%   93.14%           
- Complexity     2409     2413    +4     
=========================================
  Files           353      353           
  Lines          6637     6640    +3     
  Branches        691      692    +1     
=========================================
+ Hits           6182     6185    +3     
  Misses          255      255           
  Partials        200      200           

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

[KalleOlaviNiemitalo]

Alongside each exe or dll file, we typically have a binary pdb (program database) file for debug information such as names of local variables. But I suppose these pdb files typically have no warnings reported for them and then won't be fingerprinted anyway. Also, chemists apparently use the same pdb extension for text-format Protein Data Bank files and might want to fingerprint those.

[KalleOlaviNiemitalo]

A heuristic might work here: read a kilobyte or so from the start of the file and check whether that contains any null bytes. Although this would need to make an exception if the charset is declared as UTF-16.

[uhafner]

Yes, I think people can add more extensions easily if they found such problems now.