I wanted to confirm that the "proper" way to retrieve the filename of an uploaded item was by snagging the "x-file-name" header when doing a HEAD on the fpURL. That is what I discovered, but didn't see this documented anywhere. Just want to make sure this is the right way to do it.
Yes, we set this header so that you can get the filename off the headers of the fpURL. A cleaner way to do it would be to use the /metadata endpoint, i.e. https://www.filepicker.io/api/file/5qYoopVTsixCJJiqSWSE/metadata
Related
I'm using an API that requires me to use a header named "m_id" to the request.
When I use
$mech->add_header('m_id' => 'whatever')
WWW::Mechanize (or rather HTTP::Headers) “helpfully” changes the header name to “M-Id”. Which doesn't work.
Is there any way to prevent this from happening?
I thought I RTFMed before posting, but not well enough...
A second read through the HTTP::Headers perldoc told me to use:
$mech->add_header(':m_id'=>'whatever');
and that does the trick.
I am creating an HTTP client downloader in Python. I am able to correctly download a file such as http://www.google.com/images/srpr/logo11w.png just fine. However, I'm not sure what to actually name the thing.
There is of course the filename at the end of the URL, but is this always reliable?
If I recall correctly, wget uses the following heuristic:
If a Content-Disposition header exists, get the filename from there.
If the filename component of the URL exists (e.g. http://myserver/filename), use that.
If there is no filename component (e.g. http://www.google.com), derive the filename from the Content-Type header (such as index.html for text/html)
In all cases, if this filename is already present in the directory use a numerical suffix, such as index (1).html, or overwrite, depending on configuration.
There are plenty of other flags that control other heuristics, such as creating .html for ASP/DHTML content-types.
In short, it really depends how far you want to go. For most people, doing the first two + basic Content-Type->name mapping should be enough.
AFAIK every resource have a url in REST design. for example /user/28 is url of user with id equal to 28 and /users will return all users.
There are some way to represent output format of the resource:
passing a query parameter like format
specify it using extensions(changing /users url to /users.json to get the users in json format)
specifying the requested format(xml, json, xls, ...) by setting Accept http header.
I search the web and it seems the correct way is setting Accept header.
But if you want to have a http link (specified by href) to download list of users in xls format, you can't!Also if you want to download the xls by the browser, you will encounter many problems(you should use ajax so the xls should download using ajax and etc.)
If it is the best way, what is the solution for download link and if its not, which solution is better?
The Accept header is considered 'more correct', but there are plenty examples of all the options you mention. As far as I can tell, none of them is considered "bad". Personally, I'd say that you should honor and prefer the Accept header, but a format query parameter should override it, if present. The downside of the 'extension' method is that each format results in a different resource, which can get ugly.
Let's suppose I have an Group of users and I want to add/delete users from the group. What I am confused about is what will be the best practice to design the urls. Here are the options
OPTION # 1
POST /groups/{groupId}/users -- The request body will contain the userId
DELETE /groups/{groupId}/users/{userId} -- The userId will be in the path and the request body will be empty
OPTION # 2
DELETE /groups/{groupId}/users -- The request body will contain the userId
POST /groups/{groupId}/users/{userId} -- The userId will be in the path and the request body will be empty
I believe both the answers are correct and I am guessing there is no right or wrong answer here, just personal preference.But I would like to know what is used wide-spread. I have been using OPTION # 1 because I read in some book (the name escapes me) that the data you are POSTing shouldn't be a part of the url while using DELETE there is no such best-practice restraint.
All inputs appreciated !
The first option is the most common, but that means nothing, since misconceptions about REST are widespread. As a matter of fact, #1 isn't REST at all, it's RPC pure and simple.
Adding a member to the collection can be done either through a POST to the collection /groups/{groupId}/users, with the location of the created resource returned in the Location response header, or through a PUT request to the final location /groups/{groupId}/users/{userId}. The POST should return a 201 Created response, and the PUT either that or 200 OK, if the resource already existed and was replaced by the new one.
To delete, the correct way is to use DELETE /groups/{groupId}/users/{userId}. It's not a matter of personal preference. POST is a method you use for operations that aren't standardized by the HTTP protocol. Simple deletion is standardized through the DELETE method. Implementing it through the POST method simply means you'll have to document that functionality, instead of relying on the standard itself. You'd use POST only if you are doing something fancy during the deletion, something that already requires the functionality to be documented.
The option 1 seems to be the most common one. I don't have the feeling that the option 2 is valid at all!
Given:
/images: list of all images
/images/{imageId}: specific image
/feed/{feedId}: potentially huge list of some images (not all of them)
How would you query if a particular feed contains a particular image without downloading the full list? Put another way, how would you check whether a resource state contains a component without downloading the entire state? The first thought that comes to mind is:
Alias /images/{imageId} to /feed/{feedId}/images/{imageId}
Clients would then issue HTTP GET against /feed/{feedId}/images/{id} to check for its existence. The downside I see with this approach is that it forces me to hard-code logic into the client for breaking down an image URI to its proprietary id, something that REST frowns upon. Ideally I should be using the opaque image URI. Another option is:
Issue HTTP GET against /feed/{feedId}?contains={imageURI} to check for existence
but that feels a lot closer to RPC than I'd like. Any ideas?
What's wrong with this?
HEAD /images/id
It's unclear what "feed" means, but assuming it contains resources, it'd be the same:
HEAD /feed/id
It's tricky to say without seeing some examples to provide context.
But you could just have clients call HEAD /feed/images/{imageURI} (assuming that you might need to encode the imageURI). The server would respond with the usual HEAD response, or with a 404 error if the resource doesn't exist. You'd need to code some logic on the server to understand the imageURI.
Then the client either uses the image meta info in the head, or gracefully handles the 404 error and does something else (depending on the application I guess)
There's nothing "un-RESTful" about:
/feed/{feedId}?contains={imageURI}[,{imageURI}]
It returns the subset as specified. The resource, /feed/{feedid}, is a list resource containing a list of images. How is the resource returned with the contains query any different?
The URI is unique, and returns the appropriate state from the application. Can't say anything about the caching semantics of the request, but they're identical to whatever the caching semantics are of the original /feed/{feedid}, it simply a subset.
Finally, there's nothing that says that there even exists a /feed/{feedid}/image/{imageURL}. If you want to work with the sub-resources at that level, then fine, but you're not required to. The list coming back will likely just be a list of direct image URLS, so where's the link describing the /feed/{feedid}/image/{imageURL} relationship? You were going to embed that in the payload, correct?
How about setting up a ImageQuery resource:
# Create a new query from form data where you could constrain results for a given feed.
# May or may not redirect to /image_queries/query_id.
POST /image_queries/
# Optional - view query results containing URIs to query resources.
GET /image_queries/query_id
This video demonstrates the idea using Rails.