;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2012, 2013, 2014, 2015, 2016, 2017 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2020, 2021 Ludovic Courtès <ludo@gnu.org>
;;; Copyright © 2015 Mark H Weaver <mhw@netris.org>
;;; Copyright © 2012, 2015 Free Software Foundation, Inc.
;;; Copyright © 2017 Tobias Geerinckx-Rice <me@tobias.gr>
(define-module (guix http-client)
#:use-module (web uri)
+ #:use-module (web http)
#:use-module ((web client) #:hide (open-socket-for-uri))
+ #:use-module (web request)
#:use-module (web response)
+ #:use-module (srfi srfi-1)
#:use-module (srfi srfi-11)
#:use-module (srfi srfi-19)
#:use-module (srfi srfi-26)
#:use-module (guix ui)
#:use-module (guix utils)
#:use-module (guix base64)
- #:autoload (guix hash) (sha256)
+ #:autoload (gcrypt hash) (sha256)
+ #:autoload (gnutls) (error/invalid-session)
#:use-module ((guix build utils)
#:select (mkdir-p dump-port))
#:use-module ((guix build download)
http-get-error-reason
http-fetch
+ http-multiple-get
%http-cache-ttl
http-fetch/cached))
(reason http-get-error-reason)) ; string
-(define-syntax when-guile<=2.0.5-or-otherwise-broken
- (lambda (s)
- (syntax-case s ()
- ((_ body ...)
- ;; Always emit BODY, regardless of VERSION, because sometimes this code
- ;; might be compiled with a recent Guile and run with 2.0.5---e.g.,
- ;; when using "guix pull".
- #'(begin body ...)))))
-
-(when-guile<=2.0.5-or-otherwise-broken
- ;; Backport of Guile commits 312e79f8 ("Add HTTP Chunked Encoding support to
- ;; web modules."), 00d3ecf2 ("http: Do not buffer HTTP chunks."), and 53b8d5f
- ;; ("web: Gracefully handle premature EOF when reading chunk header.")
-
- (use-modules (ice-9 rdelim))
-
- (define %web-http
- (resolve-module '(web http)))
-
- ;; Chunked Responses
- (define (read-chunk-header port)
- "Read a chunk header from PORT and return the size in bytes of the
- upcoming chunk."
- (match (read-line port)
- ((? eof-object?)
- ;; Connection closed prematurely: there's nothing left to read.
- 0)
- (str
- (let ((extension-start (string-index str
- (lambda (c)
- (or (char=? c #\;)
- (char=? c #\return))))))
- (string->number (if extension-start ; unnecessary?
- (substring str 0 extension-start)
- str)
- 16)))))
-
- (define* (make-chunked-input-port port #:key (keep-alive? #f))
- "Returns a new port which translates HTTP chunked transfer encoded
-data from PORT into a non-encoded format. Returns eof when it has
-read the final chunk from PORT. This does not necessarily mean
-that there is no more data on PORT. When the returned port is
-closed it will also close PORT, unless the KEEP-ALIVE? is true."
- (define (close)
- (unless keep-alive?
- (close-port port)))
-
- (define chunk-size 0) ;size of the current chunk
- (define remaining 0) ;number of bytes left from the current chunk
- (define finished? #f) ;did we get all the chunks?
-
- (define (read! bv idx to-read)
- (define (loop to-read num-read)
- (cond ((or finished? (zero? to-read))
- num-read)
- ((zero? remaining) ;get a new chunk
- (let ((size (read-chunk-header port)))
- (set! chunk-size size)
- (set! remaining size)
- (if (zero? size)
- (begin
- (set! finished? #t)
- num-read)
- (loop to-read num-read))))
- (else ;read from the current chunk
- (let* ((ask-for (min to-read remaining))
- (read (get-bytevector-n! port bv (+ idx num-read)
- ask-for)))
- (if (eof-object? read)
- (begin ;premature termination
- (set! finished? #t)
- num-read)
- (let ((left (- remaining read)))
- (set! remaining left)
- (when (zero? left)
- ;; We're done with this chunk; read CR and LF.
- (get-u8 port) (get-u8 port))
- (loop (- to-read read)
- (+ num-read read))))))))
- (loop to-read 0))
-
- (make-custom-binary-input-port "chunked input port" read! #f #f close))
-
- ;; Chunked encoding support in Guile <= 2.0.11 would load whole chunks in
- ;; memory---see <http://bugs.gnu.org/19939>.
- (when (module-variable %web-http 'read-chunk-body)
- (module-set! %web-http 'make-chunked-input-port make-chunked-input-port))
-
- (define (make-delimited-input-port port len keep-alive?)
- "Return an input port that reads from PORT, and makes sure that
-exactly LEN bytes are available from PORT. Closing the returned port
-closes PORT, unless KEEP-ALIVE? is true."
- (define bytes-read 0)
-
- (define (fail)
- ((@@ (web response) bad-response)
- "EOF while reading response body: ~a bytes of ~a"
- bytes-read len))
-
- (define (read! bv start count)
- ;; Read at most LEN bytes in total. HTTP/1.1 doesn't say what to do
- ;; when a server provides more than the Content-Length, but it seems
- ;; wise to just stop reading at LEN.
- (let ((count (min count (- len bytes-read))))
- (let loop ((ret (get-bytevector-n! port bv start count)))
- (cond ((eof-object? ret)
- (if (= bytes-read len)
- 0 ; EOF
- (fail)))
- ((and (zero? ret) (> count 0))
- ;; Do not return zero since zero means EOF, so try again.
- (loop (get-bytevector-n! port bv start count)))
- (else
- (set! bytes-read (+ bytes-read ret))
- ret)))))
-
- (define close
- (and (not keep-alive?)
- (lambda ()
- (close-port port))))
-
- (make-custom-binary-input-port "delimited input port" read! #f #f close))
-
- (define (read-header-line port)
- "Read an HTTP header line and return it without its final CRLF or LF.
-Raise a 'bad-header' exception if the line does not end in CRLF or LF,
-or if EOF is reached."
- (match (%read-line port)
- (((? string? line) . #\newline)
- ;; '%read-line' does not consider #\return a delimiter; so if it's
- ;; there, remove it. We are more tolerant than the RFC in that we
- ;; tolerate LF-only endings.
- (if (string-suffix? "\r" line)
- (string-drop-right line 1)
- line))
- ((line . _) ;EOF or missing delimiter
- ((@@ (web http) bad-header) 'read-header-line line))))
-
- (unless (guile-version>? "2.0.11")
- ;; Guile <= 2.0.9 had a bug whereby 'response-body-port' would read more
- ;; than what 'content-length' says. See Guile commit 802a25b.
- ;; Guile <= 2.0.11 had a bug whereby the 'close' method of the response
- ;; body port would fail with wrong-arg-num. See Guile commit 5a10e41.
- (module-set! (resolve-module '(web response))
- 'make-delimited-input-port make-delimited-input-port)
-
- ;; Guile <= 2.0.11 was affected by <http://bugs.gnu.org/22273>. See Guile
- ;; commit 4c7732c.
- (when (module-variable %web-http 'read-line*)
- (module-set! %web-http 'read-line* read-header-line))))
-
-
(define* (http-fetch uri #:key port (text? #f) (buffered? #t)
- keep-alive? (verify-certificate? #t)
- (headers '((user-agent . "GNU Guile"))))
+ (open-connection guix:open-connection-for-uri)
+ (keep-alive? #f)
+ (verify-certificate? #t)
+ (headers '((user-agent . "GNU Guile")))
+ (log-port (current-error-port))
+ timeout)
"Return an input port containing the data at URI, and the expected number of
bytes available or #f. If TEXT? is true, the data at URI is considered to be
textual. Follow any HTTP redirection. When BUFFERED? is #f, return an
-unbuffered port, suitable for use in `filtered-port'. When KEEP-ALIVE? is
-true, send a 'Connection: keep-alive' HTTP header, in which case PORT may be
-reused for future HTTP requests. HEADERS is an alist of extra HTTP headers.
+unbuffered port, suitable for use in `filtered-port'. HEADERS is an alist of
+extra HTTP headers.
+
+When KEEP-ALIVE? is true, the connection is marked as 'keep-alive' and PORT is
+not closed upon completion.
When VERIFY-CERTIFICATE? is true, verify HTTPS server certificates.
+TIMEOUT specifies the timeout in seconds for connection establishment; when
+TIMEOUT is #f, connection establishment never times out.
+
+Write information about redirects to LOG-PORT.
+
Raise an '&http-get-error' condition if downloading fails."
(let loop ((uri (if (string? uri)
(string->uri uri)
uri)))
- (let ((port (or port (guix:open-connection-for-uri uri
- #:verify-certificate?
- verify-certificate?)))
+ (let ((port (or port (open-connection uri
+ #:verify-certificate?
+ verify-certificate?
+ #:timeout timeout)))
(headers (match (uri-userinfo uri)
((? string? str)
(cons (cons 'Authorization
headers))
(_ headers))))
(unless (or buffered? (not (file-port? port)))
- (setvbuf port _IONBF))
+ (setvbuf port 'none))
(let*-values (((resp data)
(http-get uri #:streaming? #t #:port port
- #:keep-alive? #t
+ #:keep-alive? keep-alive?
#:headers headers))
((code)
(response-code resp)))
308) ; permanent redirection
(let ((uri (resolve-uri-reference (response-location resp) uri)))
(close-port port)
- (format #t (G_ "following redirection to `~a'...~%")
+ (format log-port (G_ "following redirection to `~a'...~%")
(uri->string uri))
(loop uri)))
(else
(uri->string uri) code
(response-reason-phrase resp))))))))))))
+(define-syntax-rule (false-if-networking-error exp)
+ "Return #f if EXP triggers a network related exception as can occur when
+reusing stale cached connections."
+ ;; FIXME: Duplicated from 'with-cached-connection'.
+ (catch #t
+ (lambda ()
+ exp)
+ (lambda (key . args)
+ ;; If PORT was cached and the server closed the connection in the
+ ;; meantime, we get EPIPE. In that case, open a fresh connection and
+ ;; retry. We might also get 'bad-response or a similar exception from
+ ;; (web response) later on, once we've sent the request, or a
+ ;; ERROR/INVALID-SESSION from GnuTLS.
+ (if (or (and (eq? key 'system-error)
+ (= EPIPE (system-error-errno `(,key ,@args))))
+ (and (eq? key 'gnutls-error)
+ (eq? (first args) error/invalid-session))
+ (memq key
+ '(bad-response bad-header bad-header-component)))
+ #f
+ (apply throw key args)))))
+
+(define* (http-multiple-get base-uri proc seed requests
+ #:key port (verify-certificate? #t)
+ (open-connection guix:open-connection-for-uri)
+ (keep-alive? #t)
+ (batch-size 1000))
+ "Send all of REQUESTS to the server at BASE-URI. Call PROC for each
+response, passing it the request object, the response, a port from which to
+read the response body, and the previous result, starting with SEED, à la
+'fold'. Return the final result.
+
+When PORT is specified, use it as the initial connection on which HTTP
+requests are sent; otherwise call OPEN-CONNECTION to open a new connection for
+a URI. When KEEP-ALIVE? is false, close the connection port before
+returning."
+ (let connect ((port port)
+ (requests requests)
+ (result seed))
+ (define batch
+ (if (>= batch-size (length requests))
+ requests
+ (take requests batch-size)))
+
+ ;; (format (current-error-port) "connecting (~a requests left)..."
+ ;; (length requests))
+ (let ((p (or port (open-connection base-uri
+ #:verify-certificate?
+ verify-certificate?))))
+ ;; For HTTPS, P is not a file port and does not support 'setvbuf'.
+ (when (file-port? p)
+ (setvbuf p 'block (expt 2 16)))
+
+ ;; Send BATCH in a row.
+ ;; XXX: Do our own caching to work around inefficiencies when
+ ;; communicating over TLS: <http://bugs.gnu.org/22966>.
+ (let-values (((buffer get) (open-bytevector-output-port)))
+ ;; Inherit the HTTP proxying property from P.
+ (set-http-proxy-port?! buffer (http-proxy-port? p))
+
+ (unless (false-if-networking-error
+ (begin
+ (for-each (cut write-request <> buffer) batch)
+ (put-bytevector p (get))
+ (force-output p)
+ #t))
+ ;; If PORT becomes unusable, open a fresh connection and retry.
+ (close-port p) ; close the broken port
+ (connect #f requests result)))
+
+ ;; Now start processing responses.
+ (let loop ((sent batch)
+ (processed 0)
+ (result result))
+ (match sent
+ (()
+ (match (drop requests processed)
+ (()
+ (unless keep-alive?
+ (close-port p))
+ (reverse result))
+ (remainder
+ (connect p remainder result))))
+ ((head tail ...)
+ (match (false-if-networking-error (read-response p))
+ ((? response? resp)
+ (let* ((body (response-body-port resp))
+ (result (proc head resp body result)))
+ ;; The server can choose to stop responding at any time,
+ ;; in which case we have to try again. Check whether
+ ;; that is the case. Note that even upon "Connection:
+ ;; close", we can read from BODY.
+ (match (assq 'connection (response-headers resp))
+ (('connection 'close)
+ (close-port p)
+ (connect #f ;try again
+ (drop requests (+ 1 processed))
+ result))
+ (_
+ (loop tail (+ 1 processed) result)))))
+ (#f
+ (close-port p)
+ (connect #f ; try again
+ (drop requests processed)
+ result)))))))))
+
\f
;;;
;;; Caching.
(base64-encode digest 0 (bytevector-length digest)
#f #f base64url-alphabet))))
-(define* (http-fetch/cached uri #:key (ttl (%http-cache-ttl)) text?)
+(define* (http-fetch/cached uri #:key (ttl (%http-cache-ttl)) text?
+ (write-cache dump-port)
+ (cache-miss (const #t))
+ (log-port (current-error-port))
+ (timeout 10))
"Like 'http-fetch', return an input port, but cache its contents in
-~/.cache/guix. The cache remains valid for TTL seconds."
+~/.cache/guix. The cache remains valid for TTL seconds.
+
+Call WRITE-CACHE with the HTTP input port and the cache output port to write
+the data to cache. Call CACHE-MISS with URI just before fetching data from
+URI.
+
+TIMEOUT specifies the timeout in seconds for connection establishment.
+
+Write information about redirects to LOG-PORT."
(let ((file (cache-file-for-uri uri)))
(define (update-cache cache-port)
(define cache-time
cache-port)
(raise c))))
(let ((port (http-fetch uri #:text? text?
- #:headers headers)))
+ #:log-port log-port
+ #:headers headers #:timeout timeout)))
+ (cache-miss uri)
(mkdir-p (dirname file))
(when cache-port
(close-port cache-port))
(with-atomic-file-output file
- (cut dump-port port <>))
+ (cut write-cache port <>))
(close-port port)
(open-input-file file))))
HCoop / jackhill / guix / guix.git / blobdiff