fix(datasource): avoid over-conservative transformation of num_rows statistics in file scan config

[tschwarzinger]

Which issue does this PR close?

Minor fix I stumbled across. There is no issue (I can create one if this warrants discussion).

Rationale for this change

If the file scan is guaranteed to have no results (num_rows == Precision::Exact(0)), filtering them does not change the result. It will be zero anyway.

What changes are included in this PR?

Only change Exact -> Inexact if there are non-zero rows

Are these changes tested?

Yes.

Are there any user-facing changes?

It can happen that the optimizer can now better optimize the query plans (which we are looking forward to).

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.67%. Comparing base (fb8fe7a) to head (76897f2).
⚠️ Report is 4 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #23670      +/-   ##
==========================================
- Coverage   80.67%   80.67%   -0.01%     
==========================================
  Files        1086     1086              
  Lines      366691   366829     +138     
  Branches   366691   366829     +138     
==========================================
+ Hits       295832   295934     +102     
- Misses      53235    53264      +29     
- Partials    17624    17631       +7     

☔ View full report in Codecov by Harness.
📢 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.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.