[client] cannot properly parse <head> when using CRLF #1067
Description
Expected Behavior
parse(
`<html>\r\n<head>\r\n<title>Test</title>\r\n</head>\r\n<body>\r\n</body>\r\n</html>`
)should produce:
- element (html)
- element (head)
- text (\r\n)
- element (title)
- text (Test)
- text (\r\n)
- element (body)
- text (\r\n)Actual Behavior
parse(
`<html>\r\n<head>\r\n<title>Test</title>\r\n</head>\r\n<body>\r\n</body>\r\n</html>`
)instead produces:
- element (html)
- element (head)
- element (body)
- text (\r\n)
- element (title)
- text (Test)
- text (\r\n\r\n)Steps to Reproduce
parse(
`<html>\r\n<head>\r\n<title>Test</title>\r\n</head>\r\n<body>\r\n</body>\r\n</html>`
)I believe the issue is that, by replacing \r with __HTML_DOM_PARSER_CARRIAGE_RETURN_PLACEHOLDER_timestamp__, the head is no longer a valid head and therefore domParser.parseFromString flattens everything into a new body.
The issue first presented itself with version 5.0.11