Update PDF-A checker to fix mismatched filenames in error messages

[al-niessner]

🗒️ Summary

Changed PDFUtils to not record target as unchanging value.

⚙️ Test Data and/or Report

Beyond reach of regression tests. Was this:

  FAIL: test3_invalid_pdfa_3a.xml
      ERROR  [error.pdf.file.not_pdfa_compliant]   Invalid PDF/A version detected for test3_invalid_pdfa_3a.pdf. Expected: 1a or 1b. Actual: 3a
        1 product validation(s) completed
  FAIL: test1_invalid_pdfa_2b.xml
      ERROR  [error.pdf.file.not_pdfa_compliant]   Invalid PDF/A version detected for test3_invalid_pdfa_3a.pdf. Expected: 1a or 1b. Actual: 2b

Notice that on the second test it still shows the first PDF file as the problem. This matches the reported issue. Updating to not use target in constructor:

  FAIL: test3_invalid_pdfa_3a.xml
      ERROR  [error.pdf.file.not_pdfa_compliant]   Invalid PDF/A version detected for test3_invalid_pdfa_3a.pdf. Expected: 1a or 1b. Actual: 3a
        1 product validation(s) completed
  FAIL: test1_invalid_pdfa_2b.xml
      ERROR  [error.pdf.file.not_pdfa_compliant]   Invalid PDF/A version detected for test1_invalid_pdfa_2b.pdf. Expected: 1a or 1b. Actual: 2b
        2 product validation(s) completed

Notice that the PDF names reflect the test label names better.

♻️ Related Issues

Closes #936

[nutjob4life]

More Vera Greenfield! 👍