fix: ensure that bytes is actually bytes (fixes: pdfminer#1059)

dhdaines · Nov 27, 2024 · 7e1e281 · 7e1e281
1 parent 1a8bd2f
commit 7e1e281
Show file tree

Hide file tree

Showing 3 changed files with 8 additions and 2 deletions.
diff --git a/pdfminer/pdfdevice.py b/pdfminer/pdfdevice.py
@@ -167,7 +167,7 @@ def render_string_horizontal(
             if isinstance(obj, (int, float)):
                 x -= obj * dxscale
                 needcharspace = True
-            else:
+            elif isinstance(obj, bytes):
                 for cid in font.decode(obj):
                     if needcharspace:
                         x += charspace
@@ -207,7 +207,7 @@ def render_string_vertical(
             if isinstance(obj, (int, float)):
                 y -= obj * dxscale
                 needcharspace = True
-            else:
+            elif isinstance(obj, bytes):
                 for cid in font.decode(obj):
                     if needcharspace:
                         y += charspace

diff --git a/samples/contrib/issue-1059-cmap-decode.pdf b/samples/contrib/issue-1059-cmap-decode.pdf
diff --git a/tests/test_tools_pdf2txt.py b/tests/test_tools_pdf2txt.py
@@ -68,6 +68,12 @@ def test_contrib_issue_350(self):
         """
         run("contrib/issue-00352-asw-oct96-p41.pdf")
 
+    def test_contrib_issue_1059_textseq(self):
+        """Ensure that CMaps are robust to non-strings in text
+        sequences
+        (https://github.com/pdfminer/pdfminer.six/issues/1059)."""
+        run("contrib/issue-1059-cmap-decode.pdf")
+
     def test_scancode_patchelf(self):
         """Regression test for https://github.com/euske/pdfminer/issues/96"""
         run("scancode/patchelf.pdf")