Fixed HTML context detection with <html> tags.

Reported in aantron/lambdasoup#4.
aantron · Feb 6, 2016 · c8d4332 · c8d4332
1 parent df2dafb
commit c8d4332
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 0 deletions.
diff --git a/src/html_parser.ml b/src/html_parser.ml
@@ -15,6 +15,7 @@ let detect_context tokens throw k =
       | _, `Char c when not @@ is_whitespace c -> k (`Fragment "body")
       | _, `Char _ -> scan ()
       | _, `EOF -> k (`Fragment "body")
+      | _, `Start {name = "html"} -> k `Document
       | _, `Start {name = "head" | "body" | "frameset"} ->
         k (`Fragment "html")
       | _, `Start {name =

diff --git a/test/test_html_parser.ml b/test/test_html_parser.ml
@@ -519,6 +519,14 @@ let tests = [
     expect ~context:None "<p>foo</p>"
       [ 1,  1, S (start_element "p");
         1,  4, S (`Text ["foo"]);
+        1,  7, S  `End_element];
+
+    expect ~context:None "<html></html>"
+      [ 1,  1, S (start_element "html");
+        1,  7, S (start_element "head");
+        1,  7, S  `End_element;
+        1,  7, S (start_element "body");
+        1,  7, S  `End_element;
         1,  7, S  `End_element]);
 
   ("html.parser.foreign-context" >:: fun _ ->