Error reading link without filename

English:Fess
1

(from github.com/riceming)
When a web page (http://www.abc.com/test.php) consists a link without filename like 2016, normal browser will able to append the current url (http://www.abc.com/test.php?year=2016), but fess crawler wont (http://www.abc.com/?year=2016)

Hope this will be fixed in next version.

2

(from github.com/marevol)
Fess crawls links for tags, could you check if tag is correct?

3

(from github.com/riceming)
Let me rewrite my question, forgot to style the code.

For a webpage (http://www.abc.com/test.php), if there is a link in page:
<a href="?year=2016">2016</a>
Normal browser will prepend the filename test.php and open as http://www.abc.com/test.php?year=2016

But Fess crawler will not prepend the filename and open as http://www.abc.com/?year=2016
which is not the correct url

4

(from github.com/marevol)
It’s not reproduced…
Putting <a href="?abc=123">?abc=123</a>, http://...?abc=123 is crawled.
Does test.php contain <link rel=“canonical”… ?

5

(from github.com/riceming)
No, the page does not have canonical value

6

(from github.com/riceming)
tested with

<html>
<head>
<link rel="canonical" href="http://www.abc.com/test.php">
</head>
<body>
<a href="?year=2016">2016</a>
<a href="?year=2015">2015</a>
<a href="?year=2014">2014</a>
<a href="?year=2013">2013</a>
</body>
</html>

still crawling http://www.abc.com/?year=2016, not http://www.abc.com/test.php?year=2016

7

(from github.com/marevol)
If “http://www.abc.com/test.php?year=2016” page contains <link rel=“canonical”…, the page is not indexed. To disable canonical attribute, set the following setting to empty in fess_config.properties.

crawler.document.html.cannonical.xpath=