handle un-parsable colspan values

fixes #126
matthewwithanm · Jun 23, 2024 · 2ec3338 · 2ec3338
1 parent c1672ae
commit 2ec3338
Show file tree

Hide file tree

Showing 2 changed files with 16 additions and 4 deletions.
diff --git a/markdownify/__init__.py b/markdownify/__init__.py
@@ -383,13 +383,13 @@ def convert_figcaption(self, el, text, convert_as_inline):
 
     def convert_td(self, el, text, convert_as_inline):
         colspan = 1
-        if 'colspan' in el.attrs:
+        if 'colspan' in el.attrs and el['colspan'].isdigit():
             colspan = int(el['colspan'])
         return ' ' + text.strip().replace("\n", " ") + ' |' * colspan
 
     def convert_th(self, el, text, convert_as_inline):
         colspan = 1
-        if 'colspan' in el.attrs:
+        if 'colspan' in el.attrs and el['colspan'].isdigit():
             colspan = int(el['colspan'])
         return ' ' + text.strip().replace("\n", " ") + ' |' * colspan
 
@@ -406,7 +406,7 @@ def convert_tr(self, el, text, convert_as_inline):
             # first row and is headline: print headline underline
             full_colspan = 0
             for cell in cells:
-                if "colspan" in cell.attrs:
+                if 'colspan' in cell.attrs and cell['colspan'].isdigit():
                     full_colspan += int(cell["colspan"])
                 else:
                     full_colspan += 1

diff --git a/tests/test_tables.py b/tests/test_tables.py
@@ -215,7 +215,7 @@
         <th>Age</th>
     </tr>
     <tr>
-        <td>Jill</td>
+        <td colspan="1">Jill</td>
         <td>Smith</td>
         <td>50</td>
     </tr>
@@ -226,6 +226,17 @@
     </tr>
 </table>"""
 
+table_with_undefined_colspan = """<table>
+    <tr>
+        <th colspan="undefined">Name</th>
+        <th>Age</th>
+    </tr>
+    <tr>
+        <td colspan="-1">Jill</td>
+        <td>Smith</td>
+    </tr>
+</table>"""
+
 
 def test_table():
     assert md(table) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
@@ -240,3 +251,4 @@ def test_table():
     assert md(table_body) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
     assert md(table_with_caption) == 'TEXT\n\nCaption\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n\n'
     assert md(table_with_colspan) == '\n\n| Name | | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
+    assert md(table_with_undefined_colspan) == '\n\n| Name | Age |\n| --- | --- |\n| Jill | Smith |\n\n'