| Issue 740: | search api return status code 403 instead of 404 when there is no matching tweets | |
| 1 person starred this issue and may be notified of changes. | Back to list |
Sign in to add a comment
|
As of June 24th 2009, the search API started to return status code:403 when there is no matching tweets. It used to be returning status code 404. ----------- [Wed Jun 24 12:50:31 JST 2009]GET http://search.twitter.com/search.json?q=from%3Atwit4j+doesnothit [Wed Jun 24 12:50:31 JST 2009]X-Twitter-Client-URL: http://yusuke.homeip.net/twitter4j/en/twitter4j-undefined.xml [Wed Jun 24 12:50:31 JST 2009]Accept-Encoding: gzip [Wed Jun 24 12:50:31 JST 2009]User-Agent: twitter4j http://yusuke.homeip.net/twitter4j//undefined [Wed Jun 24 12:50:31 JST 2009]X-Twitter-Client-Version: undefined [Wed Jun 24 12:50:31 JST 2009]Response: [Wed Jun 24 12:50:31 JST 2009]HTTP/1.1 403 Forbidden [Wed Jun 24 12:50:31 JST 2009]Age: 0 [Wed Jun 24 12:50:31 JST 2009]X-Served-From: searchdb014 [Wed Jun 24 12:50:31 JST 2009]Content-Length: 53 [Wed Jun 24 12:50:31 JST 2009]Expires: Wed, 24 Jun 2009 03:55:32 GMT [Wed Jun 24 12:50:31 JST 2009]X-Served-By: searchweb014.twitter.com [Wed Jun 24 12:50:31 JST 2009]Connection: close [Wed Jun 24 12:50:31 JST 2009]Server: hi [Wed Jun 24 12:50:31 JST 2009]X-Cache: MISS [Wed Jun 24 12:50:31 JST 2009]Cache-Control: max-age=60, must- revalidate, max-age=300 [Wed Jun 24 12:50:31 JST 2009]Status: 403 Forbidden [Wed Jun 24 12:50:31 JST 2009]X-Varnish: 120647426 [Wed Jun 24 12:50:31 JST 2009]Date: Wed, 24 Jun 2009 03:50:32 GMT [Wed Jun 24 12:50:31 JST 2009]Vary: Accept-Encoding [Wed Jun 24 12:50:31 JST 2009]Content-Encoding: gzip [Wed Jun 24 12:50:31 JST 2009]Via: 1.1 varnish [Wed Jun 24 12:50:31 JST 2009]X-Cache-Svr: searchweb014.twitter.com [Wed Jun 24 12:50:31 JST 2009]Content-Type: application/json; charset=utf-8 [Wed Jun 24 12:50:31 JST 2009]{"error":"Exceptions::NoResults"} ----------- |
||||||||||||
,
Jun 24, 2009
This seems to depend on the combination of search operators and was therefore not caught by our tests. My test for a search with 0 results worked fine: $ curl -v 'http://search.twitter.com/search.json?q=blagorama' * About to connect() to search.twitter.com port 80 (#0) * Trying 128.121.146.235... connected * Connected to search.twitter.com (128.121.146.235) port 80 (#0) > GET /search.json?q=blagorama HTTP/1.1 > User-Agent: curl/7.19.4 (i386-apple-darwin9.6.2) libcurl/7.19.4 OpenSSL/0.9.8k zlib/1.2.3 > Host: search.twitter.com > Accept: */* > < HTTP/1.1 200 OK < Date: Wed, 24 Jun 2009 14:28:59 GMT < Server: hi < Status: 200 OK < X-Served-From: searchdb016 < X-Timeline-Cache-Hit: Hit < Cache-Control: max-age=60, must-revalidate, max-age=300 < Content-Type: application/json; charset=utf-8 < X-Served-By: searchweb016.twitter.com < Expires: Wed, 24 Jun 2009 14:33:59 GMT < Content-Length: 171 < Vary: Accept-Encoding < X-Varnish: 1350153762 < Age: 0 < Via: 1.1 varnish < X-Cache-Svr: searchweb016.twitter.com < X-Cache: MISS < Connection: close < * Closing connection #0 {"results":[],"since_id":0,"max_id":2309905681,"refresh_url":"? since_id=2309905681&q=blagorama","results_per_page":15,"completed_in":0.021938,"page":1,"query":"blago rama"} Also correctly returning 0 results are searches for: [foo from:mzsanford] [from:mzsanford bar] [from:twit4j foo] It would seem to be something with doesnothit (probably a word never seen). Failure cases: [doesnothit] [doesnothit from:mzsanford] I'll look into this.
Status: Accepted
Owner: m...@twitter.com Labels: Component-Search |
|||||||||||||
,
Jun 24, 2009
Found the root cause and a fix is awaiting review. |
|||||||||||||
,
Jun 24, 2009
Fix deployed and verified with: curl 'http://search.twitter.com/search.json?q=from%3Atwit4j+doesnothit'
Status: Fixed
|
|||||||||||||
 Sign in to add a comment
|
|||||||||||||
