Capturing image files from a URL?

JF1980

Hi, I am trying to capture images from a website, and save them locally. An example would be 'http://www.mywebsite.com/weather/today/001.jpg' so I would want my script to connect to this site, and save the image to a local folder. the images are always xxx.jpg so for example 103.jpg but some numbers will not exist. If the file does not exist the browser will load the website homepage. What I am trying to do is set a loop which trys the image number, if it is there it will save the image, if the image is not there, it just goes to the next number.

I have two problems, firstly, how do I take the image and save it without having to load with GD, create a new image and export? Can I somehow just download the file? Second, how do I detect if the image is actually there, or whether the server trys to redirect to the homepage (therefore I would need to make the script just jump to the next number)?

Sorry if these are simple things to solve but I don't have much experiance with this sort of operation!

Many Thanks.

planetsim

There not exactly easy things to solve. my question is why are you copying images from sites though? do you have any permission?

BuzzLY

Retrieve the web page in question, then parse through, using preg_match, to find all of the images. Then, once you have the URL path for each image, you can run [man]fopen/man to retrieve the files, and write them to the directory of your choice.

Note that this will only work if allow_url_fopen is enabled on your server.

BuzzLY

I don't think there is a permission problem here, planetsim. If the page is browsable, then he could manually get all of the images. All he is doing it making a script that pulls them all. The permissions part comes if he turns around and uses these images as part of his own website, or in some other commercial way. Then he's got issues.

JF1980

Yes quite, I didn't ask about the leagal implications.

Thanks for the help BuzzLY. Can I just use the fopen() to get the image directly? The structure seems to be that they have a folder full of images, and everyday a new image is picked from the folder and shown on the webpage. Now I want to grab the whole archive of images, but the only issue is that while the file numbers are sequential e.g. 001.jpg, 002.jpg .. 101.jpg .. 145.jpg and so on, some are missing, so, if I request a non-existant image, it redirects to the sites homepage. Maybe using the fopen() command, it will simply skip and try the next number?

BuzzLY

Read the manual on [man]fopen/man. In there it tells you what fopen() returns when it can't find the file in question.