When submitting a GET request to the /search interface, the initial results look correct. However, the page contains a javascript snippet that redirects to a URI of the form /results/<query>/page_0
. This works fine for simple queries, but not if they contain chars that need to be escaped.
https://codesearch.debian.net/search?q=glob+path%3Alib%2Fglob%2Fglob.h
https://codesearch.debian.net/results/glob%20path:lib/glob/glob.h/page_0
where it is clear the slashes have not been escaped correctly. The initial results page contains the javascript snippet:
<script type="text/javascript">
<!--
if (location.pathname.substr(0, '/search'.length) === '/search') {
window.location.replace('/results/glob path:lib\/glob\/glob.h/page_0');
}
-->
</script>