fix(pdf): extend ligature map with Dutch IJ and PUA glyph U+F0A0

[Smeet23]

Summary

Closes #2882

Extends the PDF text sanitizer's ligature normalization map with two entries that were missing:

Code point	Character	Normalized to	Reason
U+0132	IJ	IJ	Latin capital ligature IJ, used in Dutch (e.g. IJssel, IJmuiden)
U+0133	ij	ij	Latin small ligature IJ, used in Dutch
U+F0A0	(PUA)	`` (discarded)	Private-Use Area glyph emitted by some PDF fonts as a spurious character with no textual meaning

The existing map already handles U+FB00–U+FB06 (ff, fi, fl, ffi, ffl, ſt, st). This PR fills the remaining gaps reported in the issue.

Note: U+0152 (Œ) and U+0153 (œ) are intentionally left as-is — they are legitimate characters in French (e.g. œuvre, cœur) and normalizing them to OE/oe would corrupt French text.

Changes

• page_assemble_model.py: add U+0132, U+0133, U+F0A0 to _LIGATURE_MAP; update _LIGATURE_RE to match new code points
• test_page_assemble_model.py: add 4 new test cases covering the new entries

Test plan

• All page assemble model tests pass
• test_ij_capital_ligature — U+0132 IJ → IJ
• test_ij_small_ligature — U+0133 ij → ij
• test_private_use_glyph_stripped — U+F0A0 discarded
• test_private_use_glyph_with_spurious_space_stripped — U+F0A0 + spurious space discarded

[github-actions]

✅ DCO Check Passed

Thanks @Smeet23, all your commits are properly signed off. 🎉

[mergify]

Merge Protections

Your pull request matches the following merge protections and will not be merged until they are valid.

🟢 Enforce conventional commit

Wonderful, this rule succeeded.

Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/

• title ~= ^(fix|feat|docs|style|refactor|perf|test|build|ci|chore|revert)(?:\(.+\))?(!)?:

[dosubot]

Related Documentation

1 document(s) may need updating based on files changed in this PR:

Docling

What are the detailed pipeline options and processing behaviors for PDF, DOCX, PPTX, and XLSX files in the Python SDK?

View Suggested Changes

@@ -107,12 +107,36 @@
 - **Key Options**:
     - `treat_singleton_as_text` (default False): Treat 1x1 cells as TextItem
     - `gap_tolerance` (default 0): Table merging tolerance for empty cells
+    - `sheet_names` (default None): An optional list of sheet names to include in conversion. When set, only sheets whose names appear in this list will be processed. Sheet names are matched case-sensitively. Set to None (default) to include all sheets.
     - Enrichment options (image description, chart extraction)
 - **Processing**:
     - Each sheet is treated as a page
     - Table detection via flood-fill, image extraction (bounding box based on cell anchor)
     - Includes provenance (location info), auto page size calculation
+    - When `sheet_names` is specified, only sheets with names in the list are converted. If a sheet name in the list doesn't exist in the workbook, a warning is logged but processing continues.
 - **Notes**: Table detection algorithm and singleton cell handling are configurable. [Backend options code](https://github.com/docling-project/docling/blob/ae4fdbbb09fd377bb271e9b2efe541873eeb2990/docling/datamodel/backend_options.py#L80-L97).
+
+**Usage Example**:
+
+```python
+from docling.datamodel.backend_options import MsExcelBackendOptions
+from docling.datamodel.base_models import InputFormat
+from docling.document_converter import DocumentConverter, XlsxFormatOption
+
+# Process only specific sheets from an Excel file
+excel_options = MsExcelBackendOptions(
+    sheet_names=["Summary", "Data", "Q4 Results"]
+)
+
+converter = DocumentConverter(
+    format_options={
+        InputFormat.XLSX: XlsxFormatOption(backend_options=excel_options)
+    }
+)
+
+result = converter.convert("workbook.xlsx")
+doc = result.document
+```
 
 ---
 

[Accept] [Decline]

Note: You must be authenticated to accept/decline updates.

^{How did I do? Any feedback?}  

[Smeet23]

Hi @PeterStaar-IBM — this is a reopened version of #3247, now submitted from the correct account.

Two things were fixed compared to the previous PR:

1. Wrong account — the original was opened from Smeet-Agrawal by mistake. This PR is from Smeet23.

2. Stray test file changes removed — the previous submission accidentally included additions to tests/test_backend_msexcel.py (sheet-names filter tests from an unrelated branch in progress). Those caused the CI failures. This PR touches only the files relevant to the ligature fix:

   • docling/models/stages/page_assemble/page_assemble_model.py
   • tests/test_page_assemble_model.py

All page assemble model tests pass cleanly. Sorry for the noise on the previous PR.

[codecov]

Codecov Report

❌ Patch coverage is 70.00000% with 6 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
docling/backend/msexcel_backend.py	66.66%	6 Missing ⚠️

📢 Thoughts on this report? Let us know!