Prevent space between merged tags of the same type (#36)

matthiask · Feb 7, 2024 · b755b52 · b755b52
1 parent 17712cb
commit b755b52
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 5 deletions.
diff --git a/html_sanitizer/sanitizer.py b/html_sanitizer/sanitizer.py
@@ -362,12 +362,12 @@ def sanitize(self, html):
                     # tags of a mergeable type.
                     if nx.text:
                         if len(element):
-                            list(element)[-1].tail = "{} {}".format(
+                            list(element)[-1].tail = "{}{}".format(
                                 list(element)[-1].tail or "",
                                 nx.text,
                             )
                         else:
-                            element.text = "{} {}".format(element.text or "", nx.text)
+                            element.text = "{}{}".format(element.text or "", nx.text)
 
                     for child in nx:
                         element.append(child)

diff --git a/html_sanitizer/tests.py b/html_sanitizer/tests.py
@@ -97,12 +97,16 @@ def test_02_a_tag(self):
 
     def test_03_merge(self):
         entries = (
-            ("<h2>foo</h2><h2>bar</h2>", "<h2>foo bar</h2>"),
+            ("<h2>foo</h2><h2>bar</h2>", "<h2>foobar</h2>"),
             ("<h2>foo  </h2>   <h2>   bar</h2>", "<h2>foo bar</h2> "),
         )
 
         self.run_tests(entries)
 
+    def test_no_space_between_same_tags(self):
+        entries = [("<strong>Hel</strong><strong>lo</strong>", "<strong>Hello</strong>")]
+        self.run_tests(entries)
+
     def test_04_p_in_li(self):
         entries = (
             ("<li><p>foo</p></li>", "<li> foo </li>"),
@@ -275,7 +279,7 @@ def test_14_classes(self):
                 (
                     '<p class="centered">Test <span class="bla">span</span>'
                     '<span class="blub">span</span></p>',
-                    '<p class="centered">Test <span class="bla">span span</span></p>',
+                    '<p class="centered">Test <span class="bla">spanspan</span></p>',
                 ),
                 ('<h1 class="centered">Test</h1>', '<h1 class="centered">Test</h1>'),
                 ('<h2 class="centered">Test</h2>', "<h2>Test</h2>"),
@@ -311,7 +315,7 @@ def test_15_classes(self):
                 (
                     '<p class="centered">Test <span class="bla">span</span>'
                     '<span class="bla">span</span></p>',
-                    '<p class="centered">Test <span class="bla">span span</span></p>',
+                    '<p class="centered">Test <span class="bla">spanspan</span></p>',
                 ),
             ],
             sanitizer=sanitizer,