[
https://issues.apache.org/jira/browse/TIKA-3127?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17156637#comment-17156637
]
chenshuming commented on TIKA-3127:
-----------------------------------
run test in 2.0.0-SNAPSHOT , this problem exist.
use this code to test:
{code:java}
@Test
public void testParseHref() throws IOException, TikaException, SAXException {
String path = "/test-documents/testHTMLHref.html";
LinkContentHandler link = new LinkContentHandler();
Metadata metadata = new Metadata();
try (InputStream stream = HtmlParserTest.class.getResourceAsStream(path)) {
new HtmlParser().parse(stream, link, metadata, new ParseContext());
}
assertEquals("", link.getLinks().get(0).getUri());
}{code}
The test fail when test file‘s content is :
{code:java}
<html>
<a href>link</a>
</html>
{code}
The test pass when test file‘s content is :
{code:java}
<html>
<a href=>link</a>
</html>
{code}
The problem seems to occurs in method
org.ccil.cowan.tagsoup.HTMLScanner.scan(Reader, ScanHandler) .
> When using html parser any empty attribute sets value to attribute name e.g.
> <a href>link</a> gives href="href"
> ---------------------------------------------------------------------------------------------------------------
>
> Key: TIKA-3127
> URL: https://issues.apache.org/jira/browse/TIKA-3127
> Project: Tika
> Issue Type: Bug
> Components: parser
> Affects Versions: 1.16, 1.24
> Reporter: Milan Vereščák
> Priority: Major
>
> Shouldn't it be rather empty string? It was present in 1.16 version but also
> in 1.24.1
> Thank you for your response.
> [html
> specification|https://html.spec.whatwg.org/multipage/syntax.html#attributes-2]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
