Curl on an image returning text/html instead of image/jpeg

Question

I created a script to load an image from an URL using CURL. To test that the file is a JPG, I'm using ($info['content_type'] == 'image/jpeg'). It works on most of the urls, but some return a html type : e.g http://www.super-fond.fr/IMAGES_VOITURES/lexus/2005-Lexus-LF-A-Concept-R-1280x960.jpg Could anyone tell me why this URL is html ?

Thanks

The only person who could tell you for sure is the owner of super-fond.fr, as this is a server configuration problem. I suspect Browsers looks at the extension as well to determine file type, which is why they can see it... — Mike Caron
– Mike Caron, Commented Jun 27, 2011 at 18:26
What is the content of the HTML? It might just say that hot-linking is not allowed on the image. — Ed Manet
– Ed Manet, Commented Jun 27, 2011 at 18:26
I don't know why, but if you receive a text/html, you can use imagecreatefromstring($string) to check if it's an image. — Ravan Scafi
– Ravan Scafi, Commented Jun 27, 2011 at 18:26
Actually, looking at that specific image, it appears to be sending the proper content type, so I'm not sure... — Mike Caron
– Mike Caron, Commented Jun 27, 2011 at 18:26

hakre · Accepted Answer · 2011-06-27 18:31:38Z

4

It is because you not set a user-agent in request headers and this host rejects request with specific user-agent.

See CURLOPT_USERAGENT in curl_setopt() on how to set the user-agent:

curl_setopt($c, CURLOPT_USERAGENT, 'Mozilla');

hakre

200k55 gold badges454 silver badges868 bronze badges

answered Jun 27, 2011 at 18:27

azat

3,5551 gold badge30 silver badges31 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

That should solve it. Checked with CURL on commandline, -A Mozilla serves a 200 OK instead of 403 Forbidden.