TST: Add tests for page labels (#1642)

Author: pubpub-zz

pypdf/_page_labels.py

@@ -145,7 +145,7 @@ def index2label(reader: PdfReaderProtocol, index: int) -> str:
         start_index = 0
         while i < len(nums):
             start_index = nums[i]
-            value = nums[i + 1]
+            value = nums[i + 1].get_object()
             if i + 2 == len(nums):
                 break
             if nums[i + 2] > index:
@@ -159,8 +159,7 @@ def index2label(reader: PdfReaderProtocol, index: int) -> str:
             "/A": number2uppercase_letter,
             "/a": number2lowercase_letter,
         }
-        if not isinstance(value, dict):
-            value = reader.get_object(value)
+        # if /Nums array is not following the specification or if /Nums is empty
         if not isinstance(value, dict):
             return str(index + 1)  # Fallback
         start = value.get("/St", 1)
@@ -176,7 +175,7 @@ def index2label(reader: PdfReaderProtocol, index: int) -> str:
             __name__,
         )
     # TODO: Implement /Kids and /Limits for number tree
-    return str(index + 1)  # Fallback
+    return str(index + 1)  # Fallback if /Nums is not in the number_tree
 
 
 def nums_insert(

tests/test_page_labels.py

@@ -1,11 +1,27 @@
+from io import BytesIO
+
 import pytest
 
+from pypdf import PdfReader
 from pypdf._page_labels import (
+    index2label,
     number2lowercase_letter,
     number2lowercase_roman_numeral,
+    number2uppercase_letter,
     number2uppercase_roman_numeral,
+    nums_clear_range,
+    nums_insert,
+    nums_next,
+)
+from pypdf.generic import (
+    DictionaryObject,
+    NameObject,
+    NullObject,
+    NumberObject,
 )
 
+from . import get_pdf_from_url
+
 
 @pytest.mark.parametrize(
     ("number", "expected"),
@@ -44,3 +60,46 @@ def test_number2lowercase_roman_numeral():
 )
 def test_number2lowercase_letter(number, expected):
     assert number2lowercase_letter(number) == expected
+
+
+def test_number2uppercase_letter():
+    with pytest.raises(ValueError):
+        number2uppercase_letter(-1)
+
+
+@pytest.mark.external
+def test_index2label(caplog):
+    url = "https://github.com/py-pdf/pypdf/files/10773829/waarom-meisjes-het-beter-doen-op-HAVO-en-VWO-ROA.pdf"
+    name = "waarom-meisjes-het-beter-doen-op-HAVO-en-VWO-ROA.pdf"
+    r = PdfReader(BytesIO(get_pdf_from_url(url, name=name)))
+    assert index2label(r, 1) == "ii"
+    assert index2label(r, 9) == "6"
+    # very silly data to get test cover
+    r.trailer["/Root"]["/PageLabels"]["/Nums"].append(8)
+    r.trailer["/Root"]["/PageLabels"]["/Nums"].append(NullObject())
+    assert index2label(r, 9) == "10"
+
+    with pytest.raises(ValueError):
+        nums_clear_range(
+            NumberObject(10), 8, r.trailer["/Root"]["/PageLabels"]["/Nums"]
+        )
+    r.trailer["/Root"]["/PageLabels"]["/Nums"].append(8)
+    with pytest.raises(ValueError):
+        nums_next(NumberObject(10), r.trailer["/Root"]["/PageLabels"]["/Nums"])
+    with pytest.raises(ValueError):
+        nums_clear_range(
+            NumberObject(10), 8, r.trailer["/Root"]["/PageLabels"]["/Nums"]
+        )
+    with pytest.raises(ValueError):
+        nums_insert(
+            NumberObject(10),
+            DictionaryObject(),
+            r.trailer["/Root"]["/PageLabels"]["/Nums"],
+        )
+
+    del r.trailer["/Root"]["/PageLabels"]["/Nums"]
+    assert index2label(r, 1) == "2"
+    caplog.clear()
+    r.trailer["/Root"]["/PageLabels"][NameObject("/Kids")] = NullObject()
+    assert index2label(r, 1) == "2"
+    assert caplog.text != ""

tests/test_protocols.py

@@ -0,0 +1,14 @@
+from pypdf._protocols import PdfObjectProtocol
+
+
+class iPdfObjectProtocol(PdfObjectProtocol):
+    pass
+
+
+def test_pdfobjectprotocol():
+    o = iPdfObjectProtocol()
+    assert o.clone(None, False, None) is None
+    assert o._reference_clone(None, None) is None
+    assert o.get_object() is None
+    assert o.hash_value() is None
+    assert o.write_to_stream(None, None) is None