Network.Http.Client

A connection to a web server.

In order to make a request you first establish the TCP connection to the server over which to send it.

Ordinarily you would supply the host part of the URL here and it will be used as the value of the HTTP 1.1 Host: field. However, you can specify any server name or IP addresss and set the Host: value later with setHostname when building the request.

Usage is as follows:

    c <- openConnection "localhost" 80
    ...
    closeConnection c

More likely, you'll use withConnection to wrap the call in order to ensure finalization.

HTTP pipelining is supported; you can reuse the connection to a web server, but it's up to you to ensure you match the number of requests sent to the number of responses read, and to process those responses in order. This is all assuming that the server supports pipelining; be warned that not all do. Web browsers go to extraordinary lengths to probe this; you probably only want to do pipelining under controlled conditions. Otherwise just open a new connection for subsequent requests.

Open a connection to a Unix domain socket.

main :: IO ()
main = do
    c <- openConnectionUnix "/var/run/docker.sock"
    ...
    closeConnection c

HTTP Methods, as per RFC 2616

The RequestBuilder monad allows you to abuse do-notation to conveniently setup a Request object.

Run a RequestBuilder, yielding a Request object you can use on the given connection.

    let q = buildRequest1 $ do
                http POST "/api/v1/messages"
                setContentType "application/json"
                setHostname "clue.example.com" 80
                setAccept "text/html"
                setHeader "X-WhoDoneIt" "The Butler"

Obviously it's up to you to later actually send JSON data.

Run a RequestBuilder from within a monadic action.

Older versions of this library had buildRequest in IO; there's no longer a need for that, but this code path will continue to work for existing users.

    q <- buildRequest $ do
             http GET "/"

Begin constructing a Request, starting with the request line.

Set the [virtual] hostname for the request. In ordinary conditions you won't need to call this, as the Host: header is a required header in HTTP 1.1 and is set directly from the name of the server you connected to when calling openConnection.

Indicate the content type you are willing to receive in a reply from the server. For more complex Accept: headers, use setAccept'.

Indicate the content types you are willing to receive in a reply from the server in order of preference. A call of the form:

        setAccept' [("text/html", 1.0),
                    ("application/xml", 0.8),
                    ("*/*", 0)]

will result in an Accept: header value of text/html; q=1.0, application/xml; q=0.8, */*; q=0.0 as you would expect.

Set username and password credentials per the HTTP basic authentication method.

        setAuthorizationBasic "Aladdin" "open sesame"

will result in an Authorization: header value of Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ==.

Basic authentication does not use a message digest function to encipher the password; the above string is only base-64 encoded and is thus plain-text visible to any observer on the wire and all caches and servers at the other end, making basic authentication completely insecure. A number of web services, however, use SSL to encrypt the connection that then use HTTP basic authentication to validate requests. Keep in mind in these cases the secret is still sent to the servers on the other side and passes in clear through all layers after the SSL termination. Do not use basic authentication to protect secure or user-originated privacy-sensitve information.

Set the MIME type corresponding to the body of the request you are sending. Defaults to "text/plain", so usually you need to set this if PUTting.

Specify the length of the request body, in bytes.

RFC 2616 requires that we either send a Content-Length header or use Transfer-Encoding: chunked. If you know the exact size ahead of time, then call this function; the body content will still be streamed out by io-streams in more-or-less constant space.

This function is special: in a PUT or POST request, http-streams will assume chunked transfer-encoding unless you specify a content length here, in which case you need to ensure your body function writes precisely that many bytes.

Generate a random string to be used as an inter-part boundary in RFC 7578 multipart form data. You pass this value to setContentMultipart and subsequently to multipartFormBody.

If sending multipart form data (RFC 7578), you need to set the MIME type to "multipart/form-data" and specify the boundary separator that will be used.

This function is special: you must subsequently use multipartFormBody to sequence the individual body parts. When sending the request it will separate the individual parts by the boundary value set by this function.

Specify that this request should set the expectation that the server needs to approve the request before you send it.

This function is special: in a PUT or POST request, http-streams will wait for the server to reply with an HTTP/1.1 100 Continue status before sending the entity body. This is handled internally; you will get the real response (be it successful 2xx, client error, 4xx, or server error 5xx) in receiveResponse. In theory, it should be 417 if the expectation failed.

Only bother with this if you know the service you're talking to requires clients to send an Expect: 100-continue header and will handle it properly. Most servers don't do any precondition checking, automatically send an intermediate 100 response, and then just read the body regardless, making this a bit of a no-op in most cases.

Override the default setting about how the entity body will be sent.

This function is special: this explicitly sets the Transfer-Encoding: header to chunked and will instruct the library to actually tranfer the body as a stream ("chunked transfer encoding"). See setContentLength for forcing the opposite. You really won't need this in normal operation, but some people are control freaks.

Set a generic header to be sent in the HTTP request. The other methods in the RequestBuilder API are expressed in terms of this function, but we recommend you use them where offered for their stronger types.

A description of the request that will be sent to the server. Note unlike other HTTP libraries, the request body is not a part of this object; that will be streamed out by you when actually sending the request with sendRequest.

Request has a useful Show instance that will output the request line and headers (as it will be sent over the wire but with the \r characters stripped) which can be handy for debugging.

Note that the actual Host: header is not set until the request is sent, so you will not see it in the Show instance (unless you call setHostname to override the value inherited from the Connection).

A description of the response received from the server. Note unlike other HTTP libraries, the response body is not a part of this object; that will be streamed in by you when calling receiveResponse.

Like Request, Response has a Show instance that will output the status line and response headers as they were received from the server.

Get the virtual hostname that will be used as the Host: header in the HTTP 1.1 request. Per RFC 2616 § 14.23, this will be of the form hostname:port if the port number is other than the default, ie 80 for HTTP.

Having composed a Request object with the headers and metadata for this connection, you can now send the request to the server, along with the entity body, if there is one. For the rather common case of HTTP requests like GET that don't send data, use emptyBody as the output stream:

    sendRequest c q emptyBody

For PUT and POST requests, you can use fileBody or inputStreamBody to send content to the server, or you can work with the io-streams API directly:

    sendRequest c q (\o ->
        Streams.write (Just (Builder.fromString "Hello World\n")) o)

Use this for the common case of the HTTP methods that only send headers and which have no entity body, i.e. GET requests.

Sometimes you just want to send some bytes to the server as a the body of your request. This is easy to use, but if you're doing anything massive use inputStreamBody; if you're sending a file use fileBody; if you have an object that needs to be sent as JSON use jsonBody

Specify a local file to be sent to the server as the body of the request.

You use this partially applied:

    sendRequest c q (fileBody "/etc/passwd")

Note that the type of (fileBody "/path/to/file") is just what you need for the third argument to sendRequest, namely

>>> :t filePath "hello.txt":: OutputStream Builder -> IO ()

Read from a pre-existing InputStream and pipe that through to the connection to the server. This is useful in the general case where something else has handed you stream to read from and you want to use it as the entity body for the request.

You use this partially applied:

    i <- getStreamFromVault                    -- magic, clearly
    sendRequest c q (inputStreamBody i)

This function maps "Builder.fromByteString" over the input, which will be efficient if the ByteString chunks are large.

Specify name/value pairs to be sent to the server in the manner used by web browsers when submitting a form via a POST request. Parameters will be URL encoded per RFC 2396 and combined into a single string which will be sent as the body of your request.

You use this partially applied:

    let nvs = [("name","Kermit"),
               ("type","frog")]
               ("role","stagehand")]

    sendRequest c q (encodedFormBody nvs)

Note that it's going to be up to you to call setContentType with a value of "application/x-www-form-urlencoded" when building the Request object; the postForm convenience (which uses this encodedFormBody function) takes care of this for you, obviously.

Build a list of parts into an upload body.

You use this partially applied:

    boundary <- randomBoundary

    let q = buildRequest1 $ do
          http POST "/api/v1/upload"
          setContentMultipart boundary

    let parts =
            [ simplePart "metadata" Nothing metadata
            , filePart "submission" (Just "audio/wav") filepath
            ]

    sendRequest c q (multipartFormBody boundary parts)

You must have called setContentMultipart when forming the request or the request body you are sending will be invalid and (obviously) you must pass in that same Boundary value when calling this function.

Information about each of the parts of a multipart/form-data form upload. Build these with simplePart, filePart, or inputStreamPart.

Given a simple static set of bytes, send them as a part in a multipart form upload. You need to specify the name of the field for the form, and optionally can supply a MIME content-type.

The most common case in using multipart form data is to upload a file. Specify the name of the field, optionally a MIME content-type, and then the path to the file to be transmitted. The filename (without directory) will be used to name the file to the server.

Build a piece of a multipart submission from an InputStream. You need to specify a field name for this piece of the submission, and can optionally indicate the MIME type and a filename (if what you are sending is going to be interpreted as a file).

If you've got an object of a type with a ToJSON instance and you need to send that object as JSON up to a web service API, this can help.

You use this partially applied:

   sendRequest c q (jsonBody thing)

Handle the response coming back from the server. This function hands control to a handler function you supply, passing you the Response object with the response headers and an InputStream containing the entity body.

For example, if you just wanted to print the first chunk of the content from the server:

    receiveResponse c (\p i -> do
        m <- Streams.read i
        case m of
            Just bytes -> putStr bytes
            Nothing    -> return ())

Obviously, you can do more sophisticated things with the InputStream, which is the whole point of having an io-streams based HTTP client library.

The final value from the handler function is the return value of receiveResponse, if you need it.

Throws UnexpectedCompression if it doesn't know how to handle the compression format used in the response.

This is a specialized variant of receiveResponse that explicitly does not handle the content encoding of the response body stream (it will not decompress anything). Unless you really want the raw gzipped content coming down from the server, use receiveResponse.

Handle the response coming back from the server. This function is the same as receiveResponse, but it does not consume the body for you after the handler is done. This means that it can only be safely used if the handler will fully consume the body, there is no body, or when the connection is not being reused (no pipelining).

Get the HTTP response status code.

Get the HTTP response status message. Keep in mind that this is not normative; whereas getStatusCode values are authoritative.

Lookup a header in the response. HTTP header field names are case-insensitive, so you can specify the name to lookup however you like. If the header is not present Nothing will be returned.

    let n = case getHeader p "Content-Length" of
               Just x' -> read x' :: Int
               Nothing -> 0

which of course is essentially what goes on inside the client library when it receives a response from the server and has to figure out how many bytes to read.

There is a fair bit of complexity in some of the other HTTP response fields, so there are a number of specialized functions for reading those values where we've found them useful.

Print the response headers and response body to stdout. You can use this with receiveResponse or one of the convenience functions when testing. For example, doing:

    c <- openConnection "kernel.operationaldynamics.com" 58080

    let q = buildRequest1 $ do
                http GET "/time"

    sendRequest c q emptyBody

    receiveResponse c debugHandler

would print out:

HTTP/1.1 200 OK
Transfer-Encoding: chunked
Content-Type: text/plain
Vary: Accept-Encoding
Server: Snap/0.9.2.4
Content-Encoding: gzip
Date: Mon, 21 Jan 2013 06:13:37 GMT

Mon 21 Jan 13, 06:13:37.303Z

or thereabouts.

Sometimes you just want the entire response body as a single blob. This function concatonates all the bytes from the response into a ByteString. If using the main http-streams API, you would use it as follows:

   ...
   x' <- receiveResponse c simpleHandler
   ...

The methods in the convenience API all take a function to handle the response; this function is passed directly to the receiveResponse call underlying the request. Thus this utility function can be used for get as well:

   x' <- get "http://www.example.com/document.txt" simpleHandler

Either way, the usual caveats about allocating a single object from streaming I/O apply: do not use this if you are not absolutely certain that the response body will fit in a reasonable amount of memory.

Note that this function makes no discrimination based on the response's HTTP status code. You're almost certainly better off writing your own handler function.

A special case of concatHandler, this function will return the entire response body as a single ByteString, but will throw HttpClientError if the response status code was other than 2xx.

If you're working with a data stream that is in application/json, then chances are you're using aeson to handle the JSON to Haskell decoding. If so, then this helper function might be of use.

    v <- get "http://api.example.com/v1/" jsonHandler

This function feeds the input body to the json' attoparsec Parser in order to get the aeson Value type. This is then marshalled to your type represeting the source data, via the FromJSON typeclass.

The above example was actually insufficient; when working with aeson you need to fix the type so it knows what FromJSON instance to use. Let's say you're getting Person objects, then it would be

    v <- get "http://api.example.com/v1/person/461" jsonHandler :: IO Person

assuming your Person type had a FromJSON instance, of course.

Note

This function parses a single top level JSON object or array, which is all you're supposed to get if it's a valid document. People do all kinds of crazy things though, so beware. Also, this function (like the "concatHander" convenience) loads the entire response into memory; it's not streaming; if you're receiving a document which is (say) a very long array of objects then you may want to implement your own handler function, perhaps using "Streams.parserToInputStream" and the Data.Aeson.Parser combinators directly — with a result type of InputStream Value, perhaps — by which you could then iterate over the Values one at a time in constant space.

Shutdown the connection. You need to call this release the underlying socket file descriptor and related network resources. To do so reliably, use this in conjunction with openConnection in a call to bracket:

--
-- Make connection, cleaning up afterward
--

foo :: IO ByteString
foo = bracket
   (openConnection "localhost" 80)
   (closeConnection)
   (doStuff)

--
-- Actually use Connection to send Request and receive Response
--

doStuff :: Connection -> IO ByteString

or, just use withConnection.

While returning a ByteString is probably the most common use case, you could conceivably do more processing of the response in doStuff and have it and foo return a different type.

Given an IO action producing a Connection, and a computation that needs one, runs the computation, cleaning up the Connection afterwards.

    x <- withConnection (openConnection "s3.example.com" 80) $ (\c -> do
        let q = buildRequest1 $ do
            http GET "/bucket42/object/149"
        sendRequest c q emptyBody
        ...
        return "blah")

which can make the code making an HTTP request a lot more straight-forward.

Wraps Control.Exception's bracket.

Issue an HTTP GET request and pass the resultant response to the supplied handler function. This code will silently follow redirects, to a maximum depth of 5 hops.

The handler function is as for receiveResponse, so you can use one of the supplied convenience handlers if you're in a hurry:

    x' <- get "http://www.bbc.co.uk/news/" concatHandler

But as ever the disadvantage of doing this is that you're not doing anything intelligent with the HTTP response status code. If you want an exception raised in the event of a non 2xx response, you can use:

    x' <- get "http://www.bbc.co.uk/news/" concatHandler'

but for anything more refined you'll find it easy to simply write your own handler function.

Throws TooManyRedirects if more than 5 redirects are thrown.

Send content to a server via an HTTP POST request. Use this function if you have an OutputStream with the body content.

Send form data to a server via an HTTP POST request. This is the usual use case; most services expect the body to be MIME type application/x-www-form-urlencoded as this is what conventional web browsers send on form submission. If you want to POST to a URL with an arbitrary Content-Type, use post.

Place content on the server at the given URL via an HTTP PUT request, specifying the content type and a function to write the content to the supplied OutputStream. You might see:

    put "http://s3.example.com/bucket42/object149" "text/plain"
        (fileBody "hello.txt") (\p i -> do
            putStr $ show p
            Streams.connect i stdout)

Open a secure connection to a web server.

import OpenSSL (withOpenSSL)

main :: IO ()
main = do
    ctx <- baselineContextSSL
    c <- openConnectionSSL ctx "api.github.com" 443
    ...
    closeConnection c

If you want to tune the parameters used in making SSL connections, manually specify certificates, etc, then setup your own context:

import OpenSSL.Session (SSLContext)
import qualified OpenSSL.Session as SSL

    ...
    ctx <- SSL.context
    ...

See OpenSSL.Session.

Crypto is as provided by the system openssl library, as wrapped by the HsOpenSSL package and openssl-streams.

/There is no longer a need to call withOpenSSL explicitly; the initialization is invoked once per process for you/

Creates a basic SSL context. This is the SSL context used if you make an "https://" request using one of the convenience functions. It configures OpenSSL to use the default set of ciphers.

On Linux, OpenBSD and FreeBSD systems, this function also configures OpenSSL to verify certificates using the system/distribution supplied certificate authorities' certificates

On other systems, no certificate validation is performed by the generated SSLContext because there is no canonical place to find the set of system certificates. When using this library on such system, you are encouraged to install the system certificates somewhere and create your own SSLContext.

Modify the context being used to configure the SSL tunnel used by the convenience API functions to make https:// connections. The default is that setup by baselineContextSSL.

Given a URL, work out whether it is normal, secure, or unix domain, and then open the connection to the webserver including setting the appropriate default port if one was not specified in the URL. This is what powers the convenience API, but you may find it useful in composing your own similar functions.

For example (on the assumption that your server behaves when given an absolute URI as the request path), this will open a connection to server www.example.com port 443 and request /photo.jpg:

    let url = "https://www.example.com/photo.jpg"

    c <- establishConnection url
    let q = buildRequest1 $ do
                http GET url
    ...

Create a raw Connection object from the given parts. This is primarily of use when teseting, for example:

fakeConnection :: IO Connection
fakeConnection = do
    o  <- Streams.nullOutput
    i  <- Streams.nullInput
    c  <- makeConnection "www.example.com" (return()) o i
    return c

is an idiom we use frequently in testing and benchmarking, usually replacing the InputStream with something like:

    x' <- S.readFile "properly-formatted-response.txt"
    i  <- Streams.fromByteString x'

If you're going to do that, keep in mind that you must have CR-LF pairs after each header line and between the header and body to be compliant with the HTTP protocol; otherwise, parsers will reject your message.

The map of headers in a Request or Response. Note that HTTP header field names are case insensitive, so if you call setHeader on a field that's already defined but with a different capitalization you will replace the existing value.

Get the Headers from a Request or Response.y

Get the headers that will be sent with this request. You likely won't need this but there are some corner cases where people need to make calculations based on all the headers before they go out over the wire.

If you'd like the request headers as an association list, import the header functions:

import Network.Http.Types

then use Network.Http.Types.retreiveHeaders as follows:

>>> let kvs = retreiveHeaders $ getHeadersFull c q>>> :t kvs:: [(ByteString, ByteString)]

If you want to fix the multipart boundary to a known value (for testing purposes) you can use this. The ideal such string, in case you are wondering, is "bEacHV0113YB@ll".

This isn't safe for use in production; you need to use an unpredictable value as the boundary separtor so prefer randomBoundary.