Fix HTML for root tag

rushter · Sep 21, 2021 · 243b205 · 243b205
1 parent 482fa59
commit 243b205
Show file tree

Hide file tree

Showing 3 changed files with 13 additions and 1 deletion.
diff --git a/selectolax/parser.pxd b/selectolax/parser.pxd
@@ -440,6 +440,7 @@ cdef extern from "myhtml/tree.h" nogil:
     myhtml_tree_node_t * myhtml_tree_node_clone(myhtml_tree_node_t* node)
     myhtml_tree_node_t * myhtml_tree_node_insert_root(myhtml_tree_t* tree, myhtml_token_node_t* token,
                                                       myhtml_namespace ns)
+    void myhtml_tree_node_add_child(myhtml_tree_node_t* root, myhtml_tree_node_t* node)
 
 cdef extern from "myhtml/serialization.h" nogil:
     mystatus_t myhtml_serialization(myhtml_tree_node_t* scope_node, mycore_string_raw_t* str)

diff --git a/selectolax/parser.pyx b/selectolax/parser.pyx
@@ -385,7 +385,7 @@ cdef class HTMLParser:
             raise RuntimeError("Can't init MyHTML Tree object.")
 
         node = myhtml_node_clone_deep(html_tree, self.html_tree.node_html)
-        myhtml_tree_node_insert_root(html_tree, NULL, MyHTML_NAMESPACE_HTML)
+        myhtml_tree_node_add_child(html_tree.document, node)
         html_tree.node_html = node
 
         cls = HTMLParser.from_tree(

diff --git a/tests/test_parser.py b/tests/test_parser.py
@@ -108,3 +108,14 @@ def test_tags(parser):
     """)
     assert len(html_parser.tags('div')) == 5
 
+
+@pytest.mark.parametrize(*_PARSERS_PARAMETRIZER)
+def test_preserves_doctype(parser):
+    html_parser = parser("""
+    <!DOCTYPE html>
+    <html>
+        <head><title>Test</title></head>
+        <body><p>Hello World</p></body>
+    </html>
+    """)
+    assert '<!DOCTYPE html>' in html_parser.html