-- -- HTTP client for use with io-streams -- -- Copyright © 2012-2018 Operational Dynamics Consulting, Pty Ltd -- -- The code in this file, and the program it is a part of, is -- made available to you by its authors as open source software: -- you can redistribute it and/or modify it under the terms of -- the BSD licence. -- {-# LANGUAGE BangPatterns #-} {-# LANGUAGE CPP #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE MagicHash #-} {-# LANGUAGE OverloadedStrings #-} {-# OPTIONS_GHC -fno-warn-orphans #-} {-# OPTIONS_HADDOCK hide, not-home #-} module Network.Http.Inconvenience ( URL, modifyContextSSL, establishConnection, get, post, postForm, encodedFormBody, put, baselineContextSSL, concatHandler', jsonHandler, TooManyRedirects(..), HttpClientError(..), -- for testing splitURI, parseURL ) where import Blaze.ByteString.Builder (Builder) import qualified Blaze.ByteString.Builder as Builder (fromByteString, fromWord8, toByteString) import qualified Blaze.ByteString.Builder.Char8 as Builder (fromString) import Control.Exception (Exception, bracket, throw) import Data.Aeson (FromJSON, Result (..), fromJSON, json') import Data.Bits (Bits (..)) import Data.ByteString.Char8 (ByteString) import qualified Data.ByteString.Char8 as S import Data.ByteString.Internal (c2w, w2c) import Data.Char (intToDigit) import Data.HashSet (HashSet) import qualified Data.HashSet as HashSet import Data.IORef (IORef, newIORef, readIORef, writeIORef) import Data.List (intersperse) import qualified Data.Text as T import qualified Data.Text.Encoding as T import Data.Typeable (Typeable) import Data.Word (Word16) import GHC.Exts import GHC.Word (Word8 (..)) import Network.URI (URI (..), URIAuth (..), isAbsoluteURI, parseRelativeReference, parseURI, escapeURIString, isUnescapedInURI, uriToString) import OpenSSL (withOpenSSL) import OpenSSL.Session (SSLContext) import qualified OpenSSL.Session as SSL import System.IO.Streams (InputStream, OutputStream) import qualified System.IO.Streams as Streams import qualified System.IO.Streams.Attoparsec as Streams import System.IO.Unsafe (unsafePerformIO) #if !MIN_VERSION_base(4,8,0) import Data.Monoid (Monoid (..), mappend) #endif import Network.Http.Connection import Network.Http.RequestBuilder import Network.Http.Types -- (see also http://downloads.haskell.org/~ghc/8.4.2/docs/html/users_guide/phases.html#standard-cpp-macros -- for a list of predefined CPP macros provided by GHC and/or Cabal; see also the cabal user's guide) #if defined(linux_HOST_OS) || defined(freebsd_HOST_OS) import System.Directory (doesDirectoryExist) #endif type URL = ByteString ------------------------------------------------------------------------------ -- -- | URL-escapes a string (see -- <http://tools.ietf.org/html/rfc2396.html#section-2.4>) -- urlEncode :: ByteString -> URL urlEncode = Builder.toByteString . urlEncodeBuilder {-# INLINE urlEncode #-} -- -- | URL-escapes a string (see -- <http://tools.ietf.org/html/rfc2396.html#section-2.4>) into a 'Builder'. -- urlEncodeBuilder :: ByteString -> Builder urlEncodeBuilder = go mempty where go !b !s = maybe b' esc (S.uncons y) where (x,y) = S.span (flip HashSet.member urlEncodeTable) s b' = b `mappend` Builder.fromByteString x esc (c,r) = let b'' = if c == ' ' then b' `mappend` Builder.fromWord8 (c2w '+') else b' `mappend` hexd c in go b'' r hexd :: Char -> Builder hexd c0 = Builder.fromWord8 (c2w '%') `mappend` Builder.fromWord8 hi `mappend` Builder.fromWord8 low where !c = c2w c0 toDigit = c2w . intToDigit !low = toDigit $ fromEnum $ c .&. 0xf !hi = toDigit $ (c .&. 0xf0) `shiftr` 4 shiftr (W8# a#) (I# b#) = I# (word2Int# (uncheckedShiftRL# a# b#)) urlEncodeTable :: HashSet Char urlEncodeTable = HashSet.fromList $! filter f $! map w2c [0..255] where f c | c >= 'A' && c <= 'Z' = True | c >= 'a' && c <= 'z' = True | c >= '0' && c <= '9' = True f c = c `elem` ("$-_.!~*'(),"::String) ------------------------------------------------------------------------------ {- The default SSLContext used by the convenience APIs in the http-streams library. This is a kludge, unsafe bad yada yada. The technique, however, was described on a Haskell Wiki page, so that makes it an officially supported kludge. The justification for doing this is a) the functions accessing this IORef are themselves all in the IO monad, and b) these contortions are necessary to allow the library to be used for https:// URLs *without* requiring the developer to do 'withOpenSSL'. -} global :: IORef SSLContext global = unsafePerformIO $ do ctx <- baselineContextSSL newIORef ctx {-# NOINLINE global #-} -- -- | 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'. -- modifyContextSSL :: (SSLContext -> IO SSLContext) -> IO () modifyContextSSL f = do ctx <- readIORef global ctx' <- f ctx writeIORef global ctx' -- -- | 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 -- > ... -- establishConnection :: URL -> IO (Connection) establishConnection r' = do establish u where u = parseURL r' {-# INLINE establishConnection #-} establish :: URI -> IO (Connection) establish u = case scheme of "http:" -> do openConnection host port "https:" -> do ctx <- readIORef global openConnectionSSL ctx host ports "unix:" -> do openConnectionUnix $ uriPath u _ -> error ("Unknown URI scheme " ++ scheme) where scheme = uriScheme u auth = case uriAuthority u of Just x -> x Nothing -> URIAuth "" "localhost" "" host = S.pack (uriRegName auth) port = case uriPort auth of "" -> 80 _ -> read $ tail $ uriPort auth :: Word16 ports = case uriPort auth of "" -> 443 _ -> read $ tail $ uriPort auth :: Word16 -- -- | 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'. -- {- We would like to turn certificate verification on for everyone, but this has proved contingent on leveraging platform specific mechanisms to reach the certificate store. That logic should probably be in hsopenssl, but feel free to change this as appropriate for your OS. -} baselineContextSSL :: IO SSLContext baselineContextSSL = withOpenSSL $ do ctx <- SSL.context SSL.contextSetDefaultCiphers ctx #if defined(darwin_HOST_OS) SSL.contextSetVerificationMode ctx SSL.VerifyNone #elif defined(mingw32_HOST_OS) SSL.contextSetVerificationMode ctx SSL.VerifyNone #elif defined(freebsd_HOST_OS) SSL.contextSetCAFile ctx "/usr/local/etc/ssl/cert.pem" SSL.contextSetVerificationMode ctx $ SSL.VerifyPeer True True Nothing #elif defined(openbsd_HOST_OS) SSL.contextSetCAFile ctx "/etc/ssl/cert.pem" SSL.contextSetVerificationMode ctx $ SSL.VerifyPeer True True Nothing #else fedora <- doesDirectoryExist "/etc/pki/tls" if fedora then do SSL.contextSetCAFile ctx "/etc/pki/tls/certs/ca-bundle.crt" else do SSL.contextSetCADirectory ctx "/etc/ssl/certs" SSL.contextSetVerificationMode ctx $ SSL.VerifyPeer True True Nothing #endif return ctx parseURL :: URL -> URI parseURL r' = case parseURI r of Just u -> u Nothing -> error ("Can't parse URI " ++ r) where r = escapeURIString isUnescapedInURI $ T.unpack $ T.decodeUtf8 r' ------------------------------------------------------------------------------ {- Account for bug where "http://www.example.com" is parsed with no path element, resulting in an illegal HTTP request line. -} path :: URI -> ByteString path u = case url of "" -> "/" _ -> url where url = T.encodeUtf8 $! T.pack $! concat [uriPath u, uriQuery u, uriFragment u] ------------------------------------------------------------------------------ -- -- | 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. -- get :: URL -- ^ Resource to GET from. -> (Response -> InputStream ByteString -> IO β) -- ^ Handler function to receive the response from the server. -> IO β get r' handler = getN 0 r' handler getN n r' handler = do bracket (establish u) (teardown) (process) where teardown = closeConnection u = parseURL r' q = buildRequest1 $ do http GET (path u) setAccept "*/*" process c = do sendRequest c q emptyBody receiveResponse c (wrapRedirect u n handler) {- This is fairly simple-minded. Improvements could include reusing the Connection if the redirect is to the same host, and closing the original Connection if it is not. These are both things that can be done manually if using the full API, so not worried about it for now. -} wrapRedirect :: URI -> Int -> (Response -> InputStream ByteString -> IO β) -> Response -> InputStream ByteString -> IO β wrapRedirect u n handler p i = do if (s == 301 || s == 302 || s == 303 || s == 307) then case lm of Just l -> getN n' (splitURI u l) handler Nothing -> handler p i else handler p i where s = getStatusCode p lm = getHeader p "Location" !n' = if n < 5 then n + 1 else throw $! TooManyRedirects n splitURI :: URI -> URL -> URL splitURI old new' = let new = S.unpack new' in if isAbsoluteURI new then new' else let rel = parseRelativeReference new in case rel of Nothing -> new' Just x -> S.pack $ uriToString id old { uriPath = uriPath x, uriQuery = uriQuery x, uriFragment = uriFragment x } "" data TooManyRedirects = TooManyRedirects Int deriving (Typeable, Show, Eq) instance Exception TooManyRedirects -- -- | Send content to a server via an HTTP POST request. Use this -- function if you have an 'OutputStream' with the body content. -- post :: URL -- ^ Resource to POST to. -> ContentType -- ^ MIME type of the request body being sent. -> (OutputStream Builder -> IO α) -- ^ Handler function to write content to server. -> (Response -> InputStream ByteString -> IO β) -- ^ Handler function to receive the response from the server. -> IO β post r' t body handler = do bracket (establish u) (teardown) (process) where teardown = closeConnection u = parseURL r' q = buildRequest1 $ do http POST (path u) setAccept "*/*" setContentType t process c = do _ <- sendRequest c q body x <- receiveResponse c handler return x -- -- | 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'. -- postForm :: URL -- ^ Resource to POST to. -> [(ByteString, ByteString)] -- ^ List of name=value pairs. Will be sent URL-encoded. -> (Response -> InputStream ByteString -> IO β) -- ^ Handler function to receive the response from the server. -> IO β postForm r' nvs handler = do bracket (establish u) (teardown) (process) where teardown = closeConnection u = parseURL r' q = buildRequest1 $ do http POST (path u) setAccept "*/*" setContentType "application/x-www-form-urlencoded" process c = do _ <- sendRequest c q (encodedFormBody nvs) x <- receiveResponse c handler return x -- -- | 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. -- encodedFormBody :: [(ByteString,ByteString)] -> OutputStream Builder -> IO () encodedFormBody nvs o = do Streams.write (Just b) o where b = mconcat $ intersperse (Builder.fromString "&") $ map combine nvs combine :: (ByteString,ByteString) -> Builder combine (n',v') = mconcat [urlEncodeBuilder n', Builder.fromString "=", urlEncodeBuilder v'] -- -- | 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) -- put :: URL -- ^ Resource to PUT to. -> ContentType -- ^ MIME type of the request body being sent. -> (OutputStream Builder -> IO α) -- ^ Handler function to write content to server. -> (Response -> InputStream ByteString -> IO β) -- ^ Handler function to receive the response from the server. -> IO β put r' t body handler = do bracket (establish u) (teardown) (process) where teardown = closeConnection u = parseURL r' q = buildRequest1 $ do http PUT (path u) setAccept "*/*" setHeader "Content-Type" t process c = do _ <- sendRequest c q body x <- receiveResponse c handler return x -- -- | 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@. -- concatHandler' :: Response -> InputStream ByteString -> IO ByteString concatHandler' p i = if s >= 300 then throw (HttpClientError s m) else concatHandler p i where s = getStatusCode p m = getStatusMessage p data HttpClientError = HttpClientError Int ByteString deriving (Typeable) instance Exception HttpClientError instance Show HttpClientError where show (HttpClientError s msg) = Prelude.show s ++ " " ++ S.unpack msg {- There should probably also be HttpServerError and maybe even HttpRedirectError, but as these names don't seem to show up in the runtime when raised, not sure it's worth the bother. It's not like we'd want anything different in their Show instances. -} -- -- | 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 'Data.Aeson.Parser.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. -- {- This looks simple. It wasn't. The types involved are rediculous to disentangle. The biggest problem is that the Parser type used in [aeson] is *NOT* the Parser type from [attoparsec]. But the parsing function `json` and `json` from Aeson use the attoparsec Parser even though the rest of the top level page is all about Aeson's parser as used in FromJSON! Anyway, `json` and `json'` are [attoparsec] Parser [aeson] Value; we run that using the [io-streams] convenience function `parseFromStream` which gets us a Value which is the intermediate abstract syntax tree for a JSON document. Then (and this was hard to find) to work with that in terms of the FromJSON typeclass, you use the `fromJSON` function which has type (FromJSON α => Value -> Result α). Then finally, pull the result out of it. Why in Bog's name this wasn't just Either I'll never know. -} jsonHandler :: (FromJSON α) => Response -> InputStream ByteString -> IO α jsonHandler _ i = do v <- Streams.parseFromStream json' i -- Value let r = fromJSON v -- Result case r of (Success a) -> return a (Error str) -> error str