web server: do not provide a response body where it is not permitted

[bpt/guile.git] / doc / ref / web.texi

@c -*-texinfo-*-
@c This is part of the GNU Guile Reference Manual.
@c Copyright (C) 2010, 2011, 2012 Free Software Foundation, Inc.
@c See the file guile.texi for copying conditions.

@node Web
@section @acronym{HTTP}, the Web, and All That
@cindex Web
@cindex WWW
@cindex HTTP

It has always been possible to connect computers together and share
information between them, but the rise of the World-Wide Web over the
last couple of decades has made it much easier to do so.  The result is
a richly connected network of computation, in which Guile forms a part.

By ``the web'', we mean the HTTP protocol@footnote{Yes, the P is for
protocol, but this phrase appears repeatedly in RFC 2616.} as handled by
servers, clients, proxies, caches, and the various kinds of messages and
message components that can be sent and received by that protocol,
notably HTML.

On one level, the web is text in motion: the protocols themselves are
textual (though the payload may be binary), and it's possible to create
a socket and speak text to the web.  But such an approach is obviously
primitive.  This section details the higher-level data types and
operations provided by Guile: URIs, HTTP request and response records,
and a conventional web server implementation.

The material in this section is arranged in ascending order, in which
later concepts build on previous ones.  If you prefer to start with the
highest-level perspective, @pxref{Web Examples}, and work your way
back.

@menu
* Types and the Web::           Types prevent bugs and security problems.
* URIs::                        Universal Resource Identifiers.
* HTTP::                        The Hyper-Text Transfer Protocol.
* HTTP Headers::                How Guile represents specific header values.
* Requests::                    HTTP requests.
* Responses::                   HTTP responses.
* Web Client::                  Accessing web resources over HTTP.
* Web Server::                  Serving HTTP to the internet.
* Web Examples::                How to use this thing.
@end menu

@node Types and the Web
@subsection Types and the Web

It is a truth universally acknowledged, that a program with good use of
data types, will be free from many common bugs.  Unfortunately, the
common practice in web programming seems to ignore this maxim.  This
subsection makes the case for expressive data types in web programming.

By ``expressive data types'', we mean that the data types @emph{say}
something about how a program solves a problem.  For example, if we
choose to represent dates using SRFI 19 date records (@pxref{SRFI-19}),
this indicates that there is a part of the program that will always have
valid dates.  Error handling for a number of basic cases, like invalid
dates, occurs on the boundary in which we produce a SRFI 19 date record
from other types, like strings.

With regards to the web, data types are helpful in the two broad phases
of HTTP messages: parsing and generation.

Consider a server, which has to parse a request, and produce a response.
Guile will parse the request into an HTTP request object
(@pxref{Requests}), with each header parsed into an appropriate Scheme
data type.  This transition from an incoming stream of characters to
typed data is a state change in a program---the strings might parse, or
they might not, and something has to happen if they do not.  (Guile
throws an error in this case.)  But after you have the parsed request,
``client'' code (code built on top of the Guile web framework) will not
have to check for syntactic validity.  The types already make this
information manifest.

This state change on the parsing boundary makes programs more robust,
as they themselves are freed from the need to do a number of common
error checks, and they can use normal Scheme procedures to handle a
request instead of ad-hoc string parsers.

The need for types on the response generation side (in a server) is more
subtle, though not less important.  Consider the example of a POST
handler, which prints out the text that a user submits from a form.
Such a handler might include a procedure like this:

@example
;; First, a helper procedure
(define (para . contents)
  (string-append "<p>" (string-concatenate contents) "</p>"))

;; Now the meat of our simple web application
(define (you-said text)
  (para "You said: " text))

(display (you-said "Hi!"))
@print{} <p>You said: Hi!</p>
@end example

This is a perfectly valid implementation, provided that the incoming
text does not contain the special HTML characters @samp{<}, @samp{>}, or
@samp{&}.  But this provision of a restricted character set is not
reflected anywhere in the program itself: we must @emph{assume} that the
programmer understands this, and performs the check elsewhere.

Unfortunately, the short history of the practice of programming does not
bear out this assumption.  A @dfn{cross-site scripting} (@acronym{XSS})
vulnerability is just such a common error in which unfiltered user input
is allowed into the output.  A user could submit a crafted comment to
your web site which results in visitors running malicious Javascript,
within the security context of your domain:

@example
(display (you-said "<script src=\"http://bad.com/nasty.js\" />"))
@print{} <p>You said: <script src="http://bad.com/nasty.js" /></p>
@end example

The fundamental problem here is that both user data and the program
template are represented using strings.  This identity means that types
can't help the programmer to make a distinction between these two, so
they get confused.

There are a number of possible solutions, but perhaps the best is to
treat HTML not as strings, but as native s-expressions: as SXML.  The
basic idea is that HTML is either text, represented by a string, or an
element, represented as a tagged list.  So @samp{foo} becomes
@samp{"foo"}, and @samp{<b>foo</b>} becomes @samp{(b "foo")}.
Attributes, if present, go in a tagged list headed by @samp{@@}, like
@samp{(img (@@ (src "http://example.com/foo.png")))}.  @xref{sxml
simple}, for more information.

The good thing about SXML is that HTML elements cannot be confused with
text.  Let's make a new definition of @code{para}:

@example
(define (para . contents)
  `(p ,@@contents))

(use-modules (sxml simple))
(sxml->xml (you-said "Hi!"))
@print{} <p>You said: Hi!</p>

(sxml->xml (you-said "<i>Rats, foiled again!</i>"))
@print{} <p>You said: &lt;i&gt;Rats, foiled again!&lt;/i&gt;</p>
@end example

So we see in the second example that HTML elements cannot be unwittingly
introduced into the output.  However it is now perfectly acceptable to
pass SXML to @code{you-said}; in fact, that is the big advantage of SXML
over everything-as-a-string.

@example
(sxml->xml (you-said (you-said "<Hi!>")))
@print{} <p>You said: <p>You said: &lt;Hi!&gt;</p></p>
@end example

The SXML types allow procedures to @emph{compose}.  The types make
manifest which parts are HTML elements, and which are text.  So you
needn't worry about escaping user input; the type transition back to a
string handles that for you.  @acronym{XSS} vulnerabilities are a thing
of the past.

Well.  That's all very nice and opinionated and such, but how do I use
the thing?  Read on!

@node URIs
@subsection Universal Resource Identifiers

Guile provides a standard data type for Universal Resource Identifiers
(URIs), as defined in RFC 3986.

The generic URI syntax is as follows:

@example
URI := scheme ":" ["//" [userinfo "@@"] host [":" port]] path \
       [ "?" query ] [ "#" fragment ]
@end example

For example, in the URI, @indicateurl{http://www.gnu.org/help/}, the
scheme is @code{http}, the host is @code{www.gnu.org}, the path is
@code{/help/}, and there is no userinfo, port, query, or path.  All URIs
have a scheme and a path (though the path might be empty).  Some URIs
have a host, and some of those have ports and userinfo.  Any URI might
have a query part or a fragment.

Userinfo is something of an abstraction, as some legacy URI schemes
allowed userinfo of the form @code{@var{username}:@var{passwd}}.  But
since passwords do not belong in URIs, the RFC does not want to condone
this practice, so it calls anything before the @code{@@} sign
@dfn{userinfo}.

Properly speaking, a fragment is not part of a URI.  For example, when a
web browser follows a link to @indicateurl{http://example.com/#foo}, it
sends a request for @indicateurl{http://example.com/}, then looks in the
resulting page for the fragment identified @code{foo} reference.  A
fragment identifies a part of a resource, not the resource itself.  But
it is useful to have a fragment field in the URI record itself, so we
hope you will forgive the inconsistency.

@example
(use-modules (web uri))
@end example

The following procedures can be found in the @code{(web uri)}
module. Load it into your Guile, using a form like the above, to have
access to them.

@deffn {Scheme Procedure} build-uri scheme [#:userinfo=@code{#f}] [#:host=@code{#f}] @
       [#:port=@code{#f}] [#:path=@code{""}] [#:query=@code{#f}] @
       [#:fragment=@code{#f}] [#:validate?=@code{#t}]
Construct a URI object.  @var{scheme} should be a symbol, and the rest
of the fields are either strings or @code{#f}.  If @var{validate?} is
true, also run some consistency checks to make sure that the constructed
URI is valid.
@end deffn

@deffn {Scheme Procedure} uri? x
@deffnx {Scheme Procedure} uri-scheme uri
@deffnx {Scheme Procedure} uri-userinfo uri
@deffnx {Scheme Procedure} uri-host uri
@deffnx {Scheme Procedure} uri-port uri
@deffnx {Scheme Procedure} uri-path uri
@deffnx {Scheme Procedure} uri-query uri
@deffnx {Scheme Procedure} uri-fragment uri
A predicate and field accessors for the URI record type.  The URI scheme
will be a symbol, and the rest either strings or @code{#f} if not
present.
@end deffn

@deffn {Scheme Procedure} string->uri string
Parse @var{string} into a URI object.  Return @code{#f} if the string
could not be parsed.
@end deffn

@deffn {Scheme Procedure} uri->string uri
Serialize @var{uri} to a string.  If the URI has a port that is the
default port for its scheme, the port is not included in the
serialization.
@end deffn

@deffn {Scheme Procedure} declare-default-port! scheme port
Declare a default port for the given URI scheme.
@end deffn

@deffn {Scheme Procedure} uri-decode str [#:encoding=@code{"utf-8"}]
Percent-decode the given @var{str}, according to @var{encoding}, which
should be the name of a character encoding.

Note that this function should not generally be applied to a full URI
string. For paths, use split-and-decode-uri-path instead. For query
strings, split the query on @code{&} and @code{=} boundaries, and decode
the components separately.

Note also that percent-encoded strings encode @emph{bytes}, not
characters.  There is no guarantee that a given byte sequence is a valid
string encoding. Therefore this routine may signal an error if the
decoded bytes are not valid for the given encoding. Pass @code{#f} for
@var{encoding} if you want decoded bytes as a bytevector directly.
@xref{Ports, @code{set-port-encoding!}}, for more information on
character encodings.

Returns a string of the decoded characters, or a bytevector if
@var{encoding} was @code{#f}.
@end deffn

Fixme: clarify return type. indicate default values. type of
unescaped-chars.

@deffn {Scheme Procedure} uri-encode str [#:encoding=@code{"utf-8"}] [#:unescaped-chars]
Percent-encode any character not in the character set,
@var{unescaped-chars}.

The default character set includes alphanumerics from ASCII, as well as
the special characters @samp{-}, @samp{.}, @samp{_}, and @samp{~}.  Any
other character will be percent-encoded, by writing out the character to
a bytevector within the given @var{encoding}, then encoding each byte as
@code{%@var{HH}}, where @var{HH} is the hexadecimal representation of
the byte.
@end deffn

@deffn {Scheme Procedure} split-and-decode-uri-path path
Split @var{path} into its components, and decode each component,
removing empty components.

For example, @code{"/foo/bar%20baz/"} decodes to the two-element list,
@code{("foo" "bar baz")}.
@end deffn

@deffn {Scheme Procedure} encode-and-join-uri-path parts
URI-encode each element of @var{parts}, which should be a list of
strings, and join the parts together with @code{/} as a delimiter.

For example, the list @code{("scrambled eggs" "biscuits&gravy")} encodes
as @code{"scrambled%20eggs/biscuits%26gravy"}.
@end deffn

@node HTTP
@subsection The Hyper-Text Transfer Protocol

The initial motivation for including web functionality in Guile, rather
than rely on an external package, was to establish a standard base on
which people can share code.  To that end, we continue the focus on data
types by providing a number of low-level parsers and unparsers for
elements of the HTTP protocol.

If you are want to skip the low-level details for now and move on to web
pages, @pxref{Web Client}, and @pxref{Web Server}.  Otherwise, load the
HTTP module, and read on.

@example
(use-modules (web http))
@end example

The focus of the @code{(web http)} module is to parse and unparse
standard HTTP headers, representing them to Guile as native data
structures.  For example, a @code{Date:} header will be represented as a
SRFI-19 date record (@pxref{SRFI-19}), rather than as a string.

Guile tries to follow RFCs fairly strictly---the road to perdition being
paved with compatibility hacks---though some allowances are made for
not-too-divergent texts.

Header names are represented as lower-case symbols.

@deffn {Scheme Procedure} string->header name
Parse @var{name} to a symbolic header name.
@end deffn

@deffn {Scheme Procedure} header->string sym
Return the string form for the header named @var{sym}.
@end deffn

For example:

@example
(string->header "Content-Length")
@result{} content-length
(header->string 'content-length)
@result{} "Content-Length"

(string->header "FOO")
@result{} foo
(header->string 'foo)
@result{} "Foo"
@end example

Guile keeps a registry of known headers, their string names, and some
parsing and serialization procedures.  If a header is unknown, its
string name is simply its symbol name in title-case.

@deffn {Scheme Procedure} known-header? sym
Return @code{#t} iff @var{sym} is a known header, with associated
parsers and serialization procedures.
@end deffn

@deffn {Scheme Procedure} header-parser sym
Return the value parser for headers named @var{sym}.  The result is a
procedure that takes one argument, a string, and returns the parsed
value.  If the header isn't known to Guile, a default parser is returned
that passes through the string unchanged.
@end deffn

@deffn {Scheme Procedure} header-validator sym
Return a predicate which returns @code{#t} if the given value is valid
for headers named @var{sym}.  The default validator for unknown headers
is @code{string?}.
@end deffn

@deffn {Scheme Procedure} header-writer sym
Return a procedure that writes values for headers named @var{sym} to a
port.  The resulting procedure takes two arguments: a value and a port.
The default writer is @code{display}.
@end deffn

For more on the set of headers that Guile knows about out of the box,
@pxref{HTTP Headers}.  To add your own, use the @code{declare-header!}
procedure:

@deffn {Scheme Procedure} declare-header! name parser validator writer [#:multiple?=@code{#f}]
Declare a parser, validator, and writer for a given header.
@end deffn

For example, let's say you are running a web server behind some sort of
proxy, and your proxy adds an @code{X-Client-Address} header, indicating
the IPv4 address of the original client.  You would like for the HTTP
request record to parse out this header to a Scheme value, instead of
leaving it as a string.  You could register this header with Guile's
HTTP stack like this:

@example
(declare-header! "X-Client-Address"
  (lambda (str)
    (inet-aton str))
  (lambda (ip)
    (and (integer? ip) (exact? ip) (<= 0 ip #xffffffff)))
  (lambda (ip port)
    (display (inet-ntoa ip) port)))
@end example

@deffn {Scheme Procedure} valid-header? sym val
Return a true value iff @var{val} is a valid Scheme value for the header
with name @var{sym}.
@end deffn

Now that we have a generic interface for reading and writing headers, we
do just that.

@deffn {Scheme Procedure} read-header port
Read one HTTP header from @var{port}. Return two values: the header
name and the parsed Scheme value. May raise an exception if the header
was known but the value was invalid.

Returns the end-of-file object for both values if the end of the message
body was reached (i.e., a blank line).
@end deffn

@deffn {Scheme Procedure} parse-header name val
Parse @var{val}, a string, with the parser for the header named
@var{name}.  Returns the parsed value.
@end deffn

@deffn {Scheme Procedure} write-header name val port
Write the given header name and value to @var{port}, using the writer
from @code{header-writer}.
@end deffn

@deffn {Scheme Procedure} read-headers port
Read the headers of an HTTP message from @var{port}, returning the
headers as an ordered alist.
@end deffn

@deffn {Scheme Procedure} write-headers headers port
Write the given header alist to @var{port}. Doesn't write the final
@samp{\r\n}, as the user might want to add another header.
@end deffn

The @code{(web http)} module also has some utility procedures to read
and write request and response lines.

@deffn {Scheme Procedure} parse-http-method str [start] [end]
Parse an HTTP method from @var{str}. The result is an upper-case symbol,
like @code{GET}.
@end deffn

@deffn {Scheme Procedure} parse-http-version str [start] [end]
Parse an HTTP version from @var{str}, returning it as a major-minor
pair. For example, @code{HTTP/1.1} parses as the pair of integers,
@code{(1 . 1)}.
@end deffn

@deffn {Scheme Procedure} parse-request-uri str [start] [end]
Parse a URI from an HTTP request line. Note that URIs in requests do not
have to have a scheme or host name. The result is a URI object.
@end deffn

@deffn {Scheme Procedure} read-request-line port
Read the first line of an HTTP request from @var{port}, returning three
values: the method, the URI, and the version.
@end deffn

@deffn {Scheme Procedure} write-request-line method uri version port
Write the first line of an HTTP request to @var{port}.
@end deffn

@deffn {Scheme Procedure} read-response-line port
Read the first line of an HTTP response from @var{port}, returning three
values: the HTTP version, the response code, and the "reason phrase".
@end deffn

@deffn {Scheme Procedure} write-response-line version code reason-phrase port
Write the first line of an HTTP response to @var{port}.
@end deffn


@node HTTP Headers
@subsection HTTP Headers

In addition to defining the infrastructure to parse headers, the
@code{(web http)} module defines specific parsers and unparsers for all
headers defined in the HTTP/1.1 standard.

For example, if you receive a header named @samp{Accept-Language} with a
value @samp{en, es;q=0.8}, Guile parses it as a quality list (defined
below):

@example
(parse-header 'accept-language "en, es;q=0.8")
@result{} ((1000 . "en") (800 . "es"))
@end example

The format of the value for @samp{Accept-Language} headers is defined
below, along with all other headers defined in the HTTP standard.  (If
the header were unknown, the value would have been returned as a
string.)

For brevity, the header definitions below are given in the form,
@var{Type} @code{@var{name}}, indicating that values for the header
@code{@var{name}} will be of the given @var{Type}.  Since Guile
internally treats header names in lower case, in this document we give
types title-cased names.  A short description of the each header's
purpose and an example follow.

For full details on the meanings of all of these headers, see the HTTP
1.1 standard, RFC 2616.

@subsubsection HTTP Header Types

Here we define the types that are used below, when defining headers.

@deftp {HTTP Header Type} Date
A SRFI-19 date.
@end deftp

@deftp {HTTP Header Type} KVList
A list whose elements are keys or key-value pairs.  Keys are parsed to
symbols.  Values are strings by default.  Non-string values are the
exception, and are mentioned explicitly below, as appropriate.
@end deftp

@deftp {HTTP Header Type} SList
A list of strings.
@end deftp

@deftp {HTTP Header Type} Quality
An exact integer between 0 and 1000.  Qualities are used to express
preference, given multiple options.  An option with a quality of 870,
for example, is preferred over an option with quality 500.

(Qualities are written out over the wire as numbers between 0.0 and
1.0, but since the standard only allows three digits after the decimal,
it's equivalent to integers between 0 and 1000, so that's what Guile
uses.)
@end deftp

@deftp {HTTP Header Type} QList
A quality list: a list of pairs, the car of which is a quality, and the
cdr a string.  Used to express a list of options, along with their
qualities.
@end deftp

@deftp {HTTP Header Type} ETag
An entity tag, represented as a pair.  The car of the pair is an opaque
string, and the cdr is @code{#t} if the entity tag is a ``strong'' entity
tag, and @code{#f} otherwise.
@end deftp

@subsubsection General Headers

General HTTP headers may be present in any HTTP message.

@deftypevr {HTTP Header} KVList cache-control
A key-value list of cache-control directives.  See RFC 2616, for more
details.

If present, parameters to @code{max-age}, @code{max-stale},
@code{min-fresh}, and @code{s-maxage} are all parsed as non-negative
integers.

If present, parameters to @code{private} and @code{no-cache} are parsed
as lists of header names, as symbols.

@example
(parse-header 'cache-control "no-cache,no-store"
@result{} (no-cache no-store)
(parse-header 'cache-control "no-cache=\"Authorization,Date\",no-store"
@result{} ((no-cache . (authorization date)) no-store)
(parse-header 'cache-control "no-cache=\"Authorization,Date\",max-age=10"
@result{} ((no-cache . (authorization date)) (max-age . 10))
@end example
@end deftypevr

@deftypevr {HTTP Header} List connection
A list of header names that apply only to this HTTP connection, as
symbols.  Additionally, the symbol @samp{close} may be present, to
indicate that the server should close the connection after responding to
the request.
@example
(parse-header 'connection "close")
@result{} (close)
@end example
@end deftypevr

@deftypevr {HTTP Header} Date date
The date that a given HTTP message was originated.
@example
(parse-header 'date "Tue, 15 Nov 1994 08:12:31 GMT")
@result{} #<date ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} KVList pragma
A key-value list of implementation-specific directives.
@example
(parse-header 'pragma "no-cache, broccoli=tasty")
@result{} (no-cache (broccoli . "tasty"))
@end example
@end deftypevr

@deftypevr {HTTP Header} List trailer
A list of header names which will appear after the message body, instead
of with the message headers.
@example
(parse-header 'trailer "ETag")
@result{} (etag)
@end example
@end deftypevr

@deftypevr {HTTP Header} List transfer-encoding
A list of transfer codings, expressed as key-value lists.  The only
transfer coding defined by the specification is @code{chunked}.
@example
(parse-header 'transfer-encoding "chunked")
@result{} ((chunked))
@end example
@end deftypevr

@deftypevr {HTTP Header} List upgrade
A list of strings, indicating additional protocols that a server could use
in response to a request.
@example
(parse-header 'upgrade "WebSocket")
@result{} ("WebSocket")
@end example
@end deftypevr

FIXME: parse out more fully?
@deftypevr {HTTP Header} List via
A list of strings, indicating the protocol versions and hosts of
intermediate servers and proxies.  There may be multiple @code{via}
headers in one message.
@example
(parse-header 'via "1.0 venus, 1.1 mars")
@result{} ("1.0 venus" "1.1 mars")
@end example
@end deftypevr

@deftypevr {HTTP Header} List warning
A list of warnings given by a server or intermediate proxy.  Each
warning is a itself a list of four elements: a code, as an exact integer
between 0 and 1000, a host as a string, the warning text as a string,
and either @code{#f} or a SRFI-19 date.

There may be multiple @code{warning} headers in one message.
@example
(parse-header 'warning "123 foo \"core breach imminent\"")
@result{} ((123 "foo" "core-breach imminent" #f))
@end example
@end deftypevr


@subsubsection Entity Headers

Entity headers may be present in any HTTP message, and refer to the
resource referenced in the HTTP request or response.

@deftypevr {HTTP Header} List allow
A list of allowed methods on a given resource, as symbols.
@example
(parse-header 'allow "GET, HEAD")
@result{} (GET HEAD)
@end example
@end deftypevr

@deftypevr {HTTP Header} List content-encoding
A list of content codings, as symbols.
@example
(parse-header 'content-encoding "gzip")
@result{} (GET HEAD)
@end example
@end deftypevr

@deftypevr {HTTP Header} List content-language
The languages that a resource is in, as strings.
@example
(parse-header 'content-language "en")
@result{} ("en")
@end example
@end deftypevr

@deftypevr {HTTP Header} UInt content-length
The number of bytes in a resource, as an exact, non-negative integer.
@example
(parse-header 'content-length "300")
@result{} 300
@end example
@end deftypevr

@deftypevr {HTTP Header} URI content-location
The canonical URI for a resource, in the case that it is also accessible
from a different URI.
@example
(parse-header 'content-location "http://example.com/foo")
@result{} #<<uri> ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} String content-md5
The MD5 digest of a resource.
@example
(parse-header 'content-md5 "ffaea1a79810785575e29e2bd45e2fa5")
@result{} "ffaea1a79810785575e29e2bd45e2fa5"
@end example
@end deftypevr

@deftypevr {HTTP Header} List content-range
A range specification, as a list of three elements: the symbol
@code{bytes}, either the symbol @code{*} or a pair of integers,
indicating the byte rage, and either @code{*} or an integer, for the
instance length.  Used to indicate that a response only includes part of
a resource.
@example
(parse-header 'content-range "bytes 10-20/*")
@result{} (bytes (10 . 20) *)
@end example
@end deftypevr

@deftypevr {HTTP Header} List content-type
The MIME type of a resource, as a symbol, along with any parameters.
@example
(parse-header 'content-length "text/plain")
@result{} (text/plain)
(parse-header 'content-length "text/plain;charset=utf-8")
@result{} (text/plain (charset . "utf-8"))
@end example
Note that the @code{charset} parameter is something is a misnomer, and
the HTTP specification admits this.  It specifies the @emph{encoding} of
the characters, not the character set.
@end deftypevr

@deftypevr {HTTP Header} Date expires
The date/time after which the resource given in a response is considered
stale.
@example
(parse-header 'expires "Tue, 15 Nov 1994 08:12:31 GMT")
@result{} #<date ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} Date last-modified
The date/time on which the resource given in a response was last
modified.
@example
(parse-header 'expires "Tue, 15 Nov 1994 08:12:31 GMT")
@result{} #<date ...>
@end example
@end deftypevr


@subsubsection Request Headers

Request headers may only appear in an HTTP request, not in a response.

@deftypevr {HTTP Header} List accept
A list of preferred media types for a response.  Each element of the
list is itself a list, in the same format as @code{content-type}.  
@example
(parse-header 'accept "text/html,text/plain;charset=utf-8")
@result{} ((text/html) (text/plain (charset . "utf-8")))
@end example
Preference is expressed with quality values:
@example
(parse-header 'accept "text/html;q=0.8,text/plain;q=0.6")
@result{} ((text/html (q . 800)) (text/plain (q . 600)))
@end example
@end deftypevr

@deftypevr {HTTP Header} QList accept-charset
A quality list of acceptable charsets.  Note again that what HTTP calls
a ``charset'' is what Guile calls a ``character encoding''.
@example
(parse-header 'accept-charset "iso-8859-5, unicode-1-1;q=0.8")
@result{} ((1000 . "iso-8859-5") (800 . "unicode-1-1"))
@end example
@end deftypevr

@deftypevr {HTTP Header} QList accept-encoding
A quality list of acceptable content codings.
@example
(parse-header 'accept-encoding "gzip,identity=0.8")
@result{} ((1000 . "gzip") (800 . "identity"))
@end example
@end deftypevr

@deftypevr {HTTP Header} QList accept-language
A quality list of acceptable languages.
@example
(parse-header 'accept-language "cn,en=0.75")
@result{} ((1000 . "cn") (750 . "en"))
@end example
@end deftypevr

@deftypevr {HTTP Header} Pair authorization
Authorization credentials.  The car of the pair indicates the
authentication scheme, like @code{basic}.  For basic authentication, the
cdr of the pair will be the base64-encoded @samp{@var{user}:@var{pass}}
string.  For other authentication schemes, like @code{digest}, the cdr
will be a key-value list of credentials.
@example
(parse-header 'authorization "Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
@result{} (basic . "QWxhZGRpbjpvcGVuIHNlc2FtZQ==")
@end example
@end deftypevr

@deftypevr {HTTP Header} List expect
A list of expectations that a client has of a server.  The expectations
are key-value lists.
@example
(parse-header 'expect "100-continue")
@result{} ((100-continue))
@end example
@end deftypevr

@deftypevr {HTTP Header} String from
The email address of a user making an HTTP request.
@example
(parse-header 'from "bob@@example.com")
@result{} "bob@@example.com"
@end example
@end deftypevr

@deftypevr {HTTP Header} Pair host
The host for the resource being requested, as a hostname-port pair.  If
no port is given, the port is @code{#f}.
@example
(parse-header 'host "gnu.org:80")
@result{} ("gnu.org" . 80)
(parse-header 'host "gnu.org")
@result{} ("gnu.org" . #f)
@end example
@end deftypevr

@deftypevr {HTTP Header} *|List if-match
A set of etags, indicating that the request should proceed if and only
if the etag of the resource is in that set.  Either the symbol @code{*},
indicating any etag, or a list of entity tags.
@example
(parse-header 'if-match "*")
@result{} *
(parse-header 'if-match "asdfadf")
@result{} (("asdfadf" . #t))
(parse-header 'if-match W/"asdfadf")
@result{} (("asdfadf" . #f))
@end example
@end deftypevr

@deftypevr {HTTP Header} Date if-modified-since
Indicates that a response should proceed if and only if the resource has
been modified since the given date.
@example
(parse-header 'if-modified-since "Tue, 15 Nov 1994 08:12:31 GMT")
@result{} #<date ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} *|List if-none-match
A set of etags, indicating that the request should proceed if and only
if the etag of the resource is not in the set.  Either the symbol
@code{*}, indicating any etag, or a list of entity tags.
@example
(parse-header 'if-none-match "*")
@result{} *
@end example
@end deftypevr

@deftypevr {HTTP Header} ETag|Date if-range
Indicates that the range request should proceed if and only if the
resource matches a modification date or an etag.  Either an entity tag,
or a SRFI-19 date.
@example
(parse-header 'if-range "\"original-etag\"")
@result{} ("original-etag" . #t)
@end example
@end deftypevr

@deftypevr {HTTP Header} Date if-unmodified-since
Indicates that a response should proceed if and only if the resource has
not been modified since the given date.
@example
(parse-header 'if-not-modified-since "Tue, 15 Nov 1994 08:12:31 GMT")
@result{} #<date ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} UInt max-forwards
The maximum number of proxy or gateway hops that a request should be
subject to.
@example
(parse-header 'max-forwards "10")
@result{} 10
@end example
@end deftypevr

@deftypevr {HTTP Header} Pair proxy-authorization
Authorization credentials for a proxy connection.  See the documentation
for @code{authorization} above for more information on the format.
@example
(parse-header 'proxy-authorization "Digest foo=bar,baz=qux"
@result{} (digest (foo . "bar") (baz . "qux"))
@end example
@end deftypevr

@deftypevr {HTTP Header} Pair range
A range request, indicating that the client wants only part of a
resource.  The car of the pair is the symbol @code{bytes}, and the cdr
is a list of pairs. Each element of the cdr indicates a range; the car
is the first byte position and the cdr is the last byte position, as
integers, or @code{#f} if not given.
@example
(parse-header 'range "bytes=10-30,50-")
@result{} (bytes (10 . 30) (50 . #f))
@end example
@end deftypevr

@deftypevr {HTTP Header} URI referer
The URI of the resource that referred the user to this resource.  The
name of the header is a misspelling, but we are stuck with it.
@example
(parse-header 'referer "http://www.gnu.org/")
@result{} #<uri ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} List te
A list of transfer codings, expressed as key-value lists.  A common
transfer coding is @code{trailers}.
@example
(parse-header 'te "trailers")
@result{} ((trailers))
@end example
@end deftypevr

@deftypevr {HTTP Header} String user-agent
A string indicating the user agent making the request.  The
specification defines a structured format for this header, but it is
widely disregarded, so Guile does not attempt to parse strictly.
@example
(parse-header 'user-agent "Mozilla/5.0")
@result{} "Mozilla/5.0"
@end example
@end deftypevr


@subsubsection Response Headers

@deftypevr {HTTP Header} List accept-ranges
A list of range units that the server supports, as symbols.
@example
(parse-header 'accept-ranges "bytes")
@result{} (bytes)
@end example
@end deftypevr

@deftypevr {HTTP Header} UInt age
The age of a cached response, in seconds.
@example
(parse-header 'age "3600")
@result{} 3600
@end example
@end deftypevr

@deftypevr {HTTP Header} ETag etag
The entity-tag of the resource.
@example
(parse-header 'etag "\"foo\"")
@result{} ("foo" . #t)
@end example
@end deftypevr

@deftypevr {HTTP Header} URI location
A URI on which a request may be completed.  Used in combination with a
redirecting status code to perform client-side redirection.
@example
(parse-header 'location "http://example.com/other")
@result{} #<uri ...>
@end example
@end deftypevr

@deftypevr {HTTP Header} List proxy-authenticate
A list of challenges to a proxy, indicating the need for authentication.
@example
(parse-header 'proxy-authenticate "Basic realm=\"foo\"")
@result{} ((basic (realm . "foo")))
@end example
@end deftypevr

@deftypevr {HTTP Header} UInt|Date retry-after
Used in combination with a server-busy status code, like 503, to
indicate that a client should retry later.  Either a number of seconds,
or a date.
@example
(parse-header 'retry-after "60")
@result{} 60
@end example
@end deftypevr

@deftypevr {HTTP Header} String server
A string identifying the server.
@example
(parse-header 'server "My first web server")
@result{} "My first web server"
@end example
@end deftypevr

@deftypevr {HTTP Header} *|List vary
A set of request headers that were used in computing this response.
Used to indicate that server-side content negotiation was performed, for
example in response to the @code{accept-language} header.  Can also be
the symbol @code{*}, indicating that all headers were considered.
@example
(parse-header 'vary "Accept-Language, Accept")
@result{} (accept-language accept)
@end example
@end deftypevr

@deftypevr {HTTP Header} List www-authenticate
A list of challenges to a user, indicating the need for authentication.
@example
(parse-header 'www-authenticate "Basic realm=\"foo\"")
@result{} ((basic (realm . "foo")))
@end example
@end deftypevr


@node Requests
@subsection HTTP Requests

@example
(use-modules (web request))
@end example

The request module contains a data type for HTTP requests.  

@subsubsection An Important Note on Character Sets

HTTP requests consist of two parts: the request proper, consisting of a
request line and a set of headers, and (optionally) a body.  The body
might have a binary content-type, and even in the textual case its
length is specified in bytes, not characters.

Therefore, HTTP is a fundamentally binary protocol.  However the request
line and headers are specified to be in a subset of ASCII, so they can
be treated as text, provided that the port's encoding is set to an
ASCII-compatible one-byte-per-character encoding.  ISO-8859-1 (latin-1)
is just such an encoding, and happens to be very efficient for Guile.

So what Guile does when reading requests from the wire, or writing them
out, is to set the port's encoding to latin-1, and treating the request
headers as text.

The request body is another issue.  For binary data, the data is
probably in a bytevector, so we use the R6RS binary output procedures to
write out the binary payload.  Textual data usually has to be written
out to some character encoding, usually UTF-8, and then the resulting
bytevector is written out to the port.

In summary, Guile reads and writes HTTP over latin-1 sockets, without
any loss of generality.

@subsubsection Request API

@deffn {Scheme Procedure} request? 
@deffnx {Scheme Procedure} request-method 
@deffnx {Scheme Procedure} request-uri 
@deffnx {Scheme Procedure} request-version 
@deffnx {Scheme Procedure} request-headers 
@deffnx {Scheme Procedure} request-meta 
@deffnx {Scheme Procedure} request-port 
A predicate and field accessors for the request type.  The fields are as
follows:
@table @code
@item method
The HTTP method, for example, @code{GET}.
@item uri
The URI as a URI record.
@item version
The HTTP version pair, like @code{(1 . 1)}.
@item headers
The request headers, as an alist of parsed values.
@item meta
An arbitrary alist of other data, for example information returned in
the @code{sockaddr} from @code{accept} (@pxref{Network Sockets and
Communication}).
@item port
The port on which to read or write a request body, if any.
@end table
@end deffn

@deffn {Scheme Procedure} read-request port [meta='()]
Read an HTTP request from @var{port}, optionally attaching the given
metadata, @var{meta}.

As a side effect, sets the encoding on @var{port} to ISO-8859-1
(latin-1), so that reading one character reads one byte. See the
discussion of character sets above, for more information.

Note that the body is not part of the request.  Once you have read a
request, you may read the body separately, and likewise for writing
requests.
@end deffn

@deffn {Scheme Procedure} build-request uri [#:method='GET] [#:version='(1 . 1)] [#:headers='()] [#:port=#f] [#:meta='()] [#:validate-headers?=#t]
Construct an HTTP request object. If @var{validate-headers?} is true,
the headers are each run through their respective validators.
@end deffn

@deffn {Scheme Procedure} write-request r port
Write the given HTTP request to @var{port}.

Return a new request, whose @code{request-port} will continue writing
on @var{port}, perhaps using some transfer encoding.
@end deffn

@deffn {Scheme Procedure} read-request-body r
Reads the request body from @var{r}, as a bytevector.  Return @code{#f}
if there was no request body.
@end deffn

@deffn {Scheme Procedure} write-request-body r bv
Write @var{bv}, a bytevector, to the port corresponding to the HTTP
request @var{r}.
@end deffn

The various headers that are typically associated with HTTP requests may
be accessed with these dedicated accessors.  @xref{HTTP Headers}, for
more information on the format of parsed headers.

@deffn {Scheme Procedure} request-accept request [default='()]
@deffnx {Scheme Procedure} request-accept-charset request [default='()]
@deffnx {Scheme Procedure} request-accept-encoding request [default='()]
@deffnx {Scheme Procedure} request-accept-language request [default='()]
@deffnx {Scheme Procedure} request-allow request [default='()]
@deffnx {Scheme Procedure} request-authorization request [default=#f]
@deffnx {Scheme Procedure} request-cache-control request [default='()]
@deffnx {Scheme Procedure} request-connection request [default='()]
@deffnx {Scheme Procedure} request-content-encoding request [default='()]
@deffnx {Scheme Procedure} request-content-language request [default='()]
@deffnx {Scheme Procedure} request-content-length request [default=#f]
@deffnx {Scheme Procedure} request-content-location request [default=#f]
@deffnx {Scheme Procedure} request-content-md5 request [default=#f]
@deffnx {Scheme Procedure} request-content-range request [default=#f]
@deffnx {Scheme Procedure} request-content-type request [default=#f]
@deffnx {Scheme Procedure} request-date request [default=#f]
@deffnx {Scheme Procedure} request-expect request [default='()]
@deffnx {Scheme Procedure} request-expires request [default=#f]
@deffnx {Scheme Procedure} request-from request [default=#f]
@deffnx {Scheme Procedure} request-host request [default=#f]
@deffnx {Scheme Procedure} request-if-match request [default=#f]
@deffnx {Scheme Procedure} request-if-modified-since request [default=#f]
@deffnx {Scheme Procedure} request-if-none-match request [default=#f]
@deffnx {Scheme Procedure} request-if-range request [default=#f]
@deffnx {Scheme Procedure} request-if-unmodified-since request [default=#f]
@deffnx {Scheme Procedure} request-last-modified request [default=#f]
@deffnx {Scheme Procedure} request-max-forwards request [default=#f]
@deffnx {Scheme Procedure} request-pragma request [default='()]
@deffnx {Scheme Procedure} request-proxy-authorization request [default=#f]
@deffnx {Scheme Procedure} request-range request [default=#f]
@deffnx {Scheme Procedure} request-referer request [default=#f]
@deffnx {Scheme Procedure} request-te request [default=#f]
@deffnx {Scheme Procedure} request-trailer request [default='()]
@deffnx {Scheme Procedure} request-transfer-encoding request [default='()]
@deffnx {Scheme Procedure} request-upgrade request [default='()]
@deffnx {Scheme Procedure} request-user-agent request [default=#f]
@deffnx {Scheme Procedure} request-via request [default='()]
@deffnx {Scheme Procedure} request-warning request [default='()]
Return the given request header, or @var{default} if none was present.
@end deffn

@deffn {Scheme Procedure} request-absolute-uri r [default-host=#f] [default-port=#f]
A helper routine to determine the absolute URI of a request, using the
@code{host} header and the default host and port.
@end deffn


@node Responses
@subsection HTTP Responses

@example
(use-modules (web response))
@end example

As with requests (@pxref{Requests}), Guile offers a data type for HTTP
responses.  Again, the body is represented separately from the request.

@deffn {Scheme Procedure} response? 
@deffnx {Scheme Procedure} response-version 
@deffnx {Scheme Procedure} response-code 
@deffnx {Scheme Procedure} response-reason-phrase response
@deffnx {Scheme Procedure} response-headers 
@deffnx {Scheme Procedure} response-port 
A predicate and field accessors for the response type.  The fields are as
follows:
@table @code
@item version
The HTTP version pair, like @code{(1 . 1)}.
@item code
The HTTP response code, like @code{200}.
@item reason-phrase
The reason phrase, or the standard reason phrase for the response's
code.
@item headers
The response headers, as an alist of parsed values.
@item port
The port on which to read or write a response body, if any.
@end table
@end deffn

@deffn {Scheme Procedure} read-response port
Read an HTTP response from @var{port}.

As a side effect, sets the encoding on @var{port} to ISO-8859-1
(latin-1), so that reading one character reads one byte. See the
discussion of character sets in @ref{Responses}, for more information.
@end deffn

@deffn {Scheme Procedure} build-response [#:version='(1 . 1)] [#:code=200] [#:reason-phrase=#f] [#:headers='()] [#:port=#f] [#:validate-headers?=#t]
Construct an HTTP response object. If @var{validate-headers?} is true,
the headers are each run through their respective validators.
@end deffn

@deffn {Scheme Procedure} adapt-response-version response version
Adapt the given response to a different HTTP version.  Return a new HTTP
response.

The idea is that many applications might just build a response for the
default HTTP version, and this method could handle a number of
programmatic transformations to respond to older HTTP versions (0.9 and
1.0). But currently this function is a bit heavy-handed, just updating
the version field.
@end deffn

@deffn {Scheme Procedure} write-response r port
Write the given HTTP response to @var{port}.

Return a new response, whose @code{response-port} will continue writing
on @var{port}, perhaps using some transfer encoding.
@end deffn

@deffn {Scheme Procedure} response-must-not-include-body? r
Some responses, like those with status code 304, are specified as never
having bodies.  This predicate returns @code{#t} for those responses.

Note also, though, that responses to @code{HEAD} requests must also not
have a body.
@end deffn

@deffn {Scheme Procedure} read-response-body r
Read the response body from @var{r}, as a bytevector.  Returns @code{#f}
if there was no response body.
@end deffn

@deffn {Scheme Procedure} write-response-body r bv
Write @var{bv}, a bytevector, to the port corresponding to the HTTP
response @var{r}.
@end deffn

As with requests, the various headers that are typically associated with
HTTP responses may be accessed with these dedicated accessors.
@xref{HTTP Headers}, for more information on the format of parsed
headers.

@deffn {Scheme Procedure} response-accept-ranges response [default=#f]
@deffnx {Scheme Procedure} response-age response [default='()]
@deffnx {Scheme Procedure} response-allow response [default='()]
@deffnx {Scheme Procedure} response-cache-control response [default='()]
@deffnx {Scheme Procedure} response-connection response [default='()]
@deffnx {Scheme Procedure} response-content-encoding response [default='()]
@deffnx {Scheme Procedure} response-content-language response [default='()]
@deffnx {Scheme Procedure} response-content-length response [default=#f]
@deffnx {Scheme Procedure} response-content-location response [default=#f]
@deffnx {Scheme Procedure} response-content-md5 response [default=#f]
@deffnx {Scheme Procedure} response-content-range response [default=#f]
@deffnx {Scheme Procedure} response-content-type response [default=#f]
@deffnx {Scheme Procedure} response-date response [default=#f]
@deffnx {Scheme Procedure} response-etag response [default=#f]
@deffnx {Scheme Procedure} response-expires response [default=#f]
@deffnx {Scheme Procedure} response-last-modified response [default=#f]
@deffnx {Scheme Procedure} response-location response [default=#f]
@deffnx {Scheme Procedure} response-pragma response [default='()]
@deffnx {Scheme Procedure} response-proxy-authenticate response [default=#f]
@deffnx {Scheme Procedure} response-retry-after response [default=#f]
@deffnx {Scheme Procedure} response-server response [default=#f]
@deffnx {Scheme Procedure} response-trailer response [default='()]
@deffnx {Scheme Procedure} response-transfer-encoding response [default='()]
@deffnx {Scheme Procedure} response-upgrade response [default='()]
@deffnx {Scheme Procedure} response-vary response [default='()]
@deffnx {Scheme Procedure} response-via response [default='()]
@deffnx {Scheme Procedure} response-warning response [default='()]
@deffnx {Scheme Procedure} response-www-authenticate response [default=#f]
Return the given response header, or @var{default} if none was present.
@end deffn


@node Web Client
@subsection Web Client

@code{(web client)} provides a simple, synchronous HTTP client, built on
the lower-level HTTP, request, and response modules.

@deffn {Scheme Procedure} open-socket-for-uri uri
@end deffn

@deffn {Scheme Procedure} http-get uri [#:port=(open-socket-for-uri uri)] [#:version='(1 . 1)] [#:keep-alive?=#f] [#:extra-headers='()] [#:decode-body?=#t]
Connect to the server corresponding to @var{uri} and ask for the
resource, using the @code{GET} method.  If you already have a port open,
pass it as @var{port}.  The port will be closed at the end of the
request unless @var{keep-alive?} is true.  Any extra headers in the
alist @var{extra-headers} will be added to the request.

If @var{decode-body?} is true, as is the default, the body of the
response will be decoded to string, if it is a textual content-type.
Otherwise it will be returned as a bytevector.
@end deffn

@code{http-get} is useful for making one-off requests to web sites.  If
you are writing a web spider or some other client that needs to handle a
number of requests in parallel, it's better to build an event-driven URL
fetcher, similar in structure to the web server (@pxref{Web Server}).

Another option, good but not as performant, would be to use threads,
possibly via par-map or futures.

More helper procedures for the other common HTTP verbs would be a good
addition to this module.  Send your code to
@email{guile-user@@gnu.org}.


@node Web Server
@subsection Web Server

@code{(web server)} is a generic web server interface, along with a main
loop implementation for web servers controlled by Guile.

@example
(use-modules (web server))
@end example

The lowest layer is the @code{<server-impl>} object, which defines a set
of hooks to open a server, read a request from a client, write a
response to a client, and close a server.  These hooks -- @code{open},
@code{read}, @code{write}, and @code{close}, respectively -- are bound
together in a @code{<server-impl>} object.  Procedures in this module take a
@code{<server-impl>} object, if needed.

A @code{<server-impl>} may also be looked up by name.  If you pass the
@code{http} symbol to @code{run-server}, Guile looks for a variable
named @code{http} in the @code{(web server http)} module, which should
be bound to a @code{<server-impl>} object.  Such a binding is made by
instantiation of the @code{define-server-impl} syntax.  In this way the
run-server loop can automatically load other backends if available.

The life cycle of a server goes as follows:

@enumerate
@item
The @code{open} hook is called, to open the server. @code{open} takes 0 or
more arguments, depending on the backend, and returns an opaque
server socket object, or signals an error.

@item
The @code{read} hook is called, to read a request from a new client.
The @code{read} hook takes one argument, the server socket.  It should
return three values: an opaque client socket, the request, and the
request body. The request should be a @code{<request>} object, from
@code{(web request)}.  The body should be a string or a bytevector, or
@code{#f} if there is no body.

If the read failed, the @code{read} hook may return #f for the client
socket, request, and body.

@item
A user-provided handler procedure is called, with the request and body
as its arguments.  The handler should return two values: the response,
as a @code{<response>} record from @code{(web response)}, and the
response body as bytevector, or @code{#f} if not present.

The respose and response body are run through @code{sanitize-response},
documented below.  This allows the handler writer to take some
convenient shortcuts: for example, instead of a @code{<response>}, the
handler can simply return an alist of headers, in which case a default
response object is constructed with those headers.  Instead of a
bytevector for the body, the handler can return a string, which will be
serialized into an appropriate encoding; or it can return a procedure,
which will be called on a port to write out the data.  See the
@code{sanitize-response} documentation, for more.

@item
The @code{write} hook is called with three arguments: the client
socket, the response, and the body.  The @code{write} hook returns no
values.

@item
At this point the request handling is complete. For a loop, we
loop back and try to read a new request.

@item
If the user interrupts the loop, the @code{close} hook is called on
the server socket.
@end enumerate

A user may define a server implementation with the following form:

@deffn {Scheme Procedure} define-server-impl name open read write close
Make a @code{<server-impl>} object with the hooks @var{open},
@var{read}, @var{write}, and @var{close}, and bind it to the symbol
@var{name} in the current module.
@end deffn

@deffn {Scheme Procedure} lookup-server-impl impl
Look up a server implementation. If @var{impl} is a server
implementation already, it is returned directly. If it is a symbol, the
binding named @var{impl} in the @code{(web server @var{impl})} module is
looked up. Otherwise an error is signaled.

Currently a server implementation is a somewhat opaque type, useful only
for passing to other procedures in this module, like @code{read-client}.
@end deffn

The @code{(web server)} module defines a number of routines that use
@code{<server-impl>} objects to implement parts of a web server.  Given
that we don't expose the accessors for the various fields of a
@code{<server-impl>}, indeed these routines are the only procedures with
any access to the impl objects.

@deffn {Scheme Procedure} open-server impl open-params
Open a server for the given implementation.  Return one value, the new
server object. The implementation's @code{open} procedure is applied to
@var{open-params}, which should be a list.
@end deffn

@deffn {Scheme Procedure} read-client impl server
Read a new client from @var{server}, by applying the implementation's
@code{read} procedure to the server.  If successful, return three
values: an object corresponding to the client, a request object, and the
request body. If any exception occurs, return @code{#f} for all three
values.
@end deffn

@deffn {Scheme Procedure} handle-request handler request body state
Handle a given request, returning the response and body.

The response and response body are produced by calling the given
@var{handler} with @var{request} and @var{body} as arguments.

The elements of @var{state} are also passed to @var{handler} as
arguments, and may be returned as additional values. The new
@var{state}, collected from the @var{handler}'s return values, is then
returned as a list. The idea is that a server loop receives a handler
from the user, along with whatever state values the user is interested
in, allowing the user's handler to explicitly manage its state.
@end deffn

@deffn {Scheme Procedure} sanitize-response request response body
"Sanitize" the given response and body, making them appropriate for the
given request.

As a convenience to web handler authors, @var{response} may be given as
an alist of headers, in which case it is used to construct a default
response. Ensures that the response version corresponds to the request
version. If @var{body} is a string, encodes the string to a bytevector,
in an encoding appropriate for @var{response}. Adds a
@code{content-length} and @code{content-type} header, as necessary.

If @var{body} is a procedure, it is called with a port as an argument,
and the output collected as a bytevector. In the future we might try to
instead use a compressing, chunk-encoded port, and call this procedure
later, in the write-client procedure. Authors are advised not to rely on
the procedure being called at any particular time.
@end deffn

@deffn {Scheme Procedure} write-client impl server client response body
Write an HTTP response and body to @var{client}. If the server and
client support persistent connections, it is the implementation's
responsibility to keep track of the client thereafter, presumably by
attaching it to the @var{server} argument somehow.
@end deffn

@deffn {Scheme Procedure} close-server impl server
Release resources allocated by a previous invocation of
@code{open-server}.
@end deffn

Given the procedures above, it is a small matter to make a web server:

@deffn {Scheme Procedure} serve-one-client handler impl server state
Read one request from @var{server}, call @var{handler} on the request
and body, and write the response to the client.  Return the new state
produced by the handler procedure.
@end deffn

@deffn {Scheme Procedure} run-server handler [impl='http] [open-params='()] . state
Run Guile's built-in web server.

@var{handler} should be a procedure that takes two or more arguments,
the HTTP request and request body, and returns two or more values, the
response and response body.

For examples, skip ahead to the next section, @ref{Web Examples}.

The response and body will be run through @code{sanitize-response}
before sending back to the client.

Additional arguments to @var{handler} are taken from @var{state}.
Additional return values are accumulated into a new @var{state}, which
will be used for subsequent requests. In this way a handler can
explicitly manage its state.
@end deffn

The default web server implementation is @code{http}, which binds to a
socket, listening for request on that port.

@deffn {HTTP Implementation} http [#:host=#f] [#:family=AF_INET] [#:addr=INADDR_LOOPBACK] [#:port 8080] [#:socket]
The default HTTP implementation.  We document it as a function with
keyword arguments, because that is precisely the way that it is -- all
of the @var{open-params} to @code{run-server} get passed to the
implementation's open function.

@example
;; The defaults: localhost:8080
(run-server handler)
;; Same thing
(run-server handler 'http '())
;; On a different port
(run-server handler 'http '(#:port 8081))
;; IPv6
(run-server handler 'http '(#:family AF_INET6 #:port 8081))
;; Custom socket
(run-server handler 'http `(#:socket ,(sudo-make-me-a-socket)))
@end example
@end deffn

@node Web Examples
@subsection Web Examples

Well, enough about the tedious internals.  Let's make a web application!

@subsubsection Hello, World!

The first program we have to write, of course, is ``Hello, World!''.
This means that we have to implement a web handler that does what we
want.

Now we define a handler, a function of two arguments and two return
values:

@example
(define (handler request request-body)
  (values @var{response} @var{response-body}))
@end example

In this first example, we take advantage of a short-cut, returning an
alist of headers instead of a proper response object. The response body
is our payload:

@example
(define (hello-world-handler request request-body)
  (values '((content-type . (text/plain)))
          "Hello World!"))
@end example

Now let's test it, by running a server with this handler. Load up the
web server module if you haven't yet done so, and run a server with this
handler:

@example
(use-modules (web server))
(run-server hello-world-handler)
@end example

By default, the web server listens for requests on
@code{localhost:8080}.  Visit that address in your web browser to
test.  If you see the string, @code{Hello World!}, sweet!

@subsubsection Inspecting the Request

The Hello World program above is a general greeter, responding to all
URIs.  To make a more exclusive greeter, we need to inspect the request
object, and conditionally produce different results.  So let's load up
the request, response, and URI modules, and do just that.

@example
(use-modules (web server)) ; you probably did this already
(use-modules (web request)
             (web response)
             (web uri))

(define (request-path-components request)
  (split-and-decode-uri-path (uri-path (request-uri request))))

(define (hello-hacker-handler request body)
  (if (equal? (request-path-components request)
              '("hacker"))
      (values '((content-type . (text/plain)))
              "Hello hacker!")
      (not-found request)))

(run-server hello-hacker-handler)
@end example

Here we see that we have defined a helper to return the components of
the URI path as a list of strings, and used that to check for a request
to @code{/hacker/}. Then the success case is just as before -- visit
@code{http://localhost:8080/hacker/} in your browser to check.

You should always match against URI path components as decoded by
@code{split-and-decode-uri-path}. The above example will work for
@code{/hacker/}, @code{//hacker///}, and @code{/h%61ck%65r}.

But we forgot to define @code{not-found}!  If you are pasting these
examples into a REPL, accessing any other URI in your web browser will
drop your Guile console into the debugger:

@example
<unnamed port>:38:7: In procedure module-lookup:
<unnamed port>:38:7: Unbound variable: not-found

Entering a new prompt.  Type `,bt' for a backtrace or `,q' to continue.
scheme@@(guile-user) [1]> 
@end example

So let's define the function, right there in the debugger.  As you
probably know, we'll want to return a 404 response.

@example
;; Paste this in your REPL
(define (not-found request)
  (values (build-response #:code 404)
          (string-append "Resource not found: "
                         (uri->string (request-uri request)))))

;; Now paste this to let the web server keep going:
,continue
@end example

Now if you access @code{http://localhost/foo/}, you get this error
message.  (Note that some popular web browsers won't show
server-generated 404 messages, showing their own instead, unless the 404
message body is long enough.)

@subsubsection Higher-Level Interfaces

The web handler interface is a common baseline that all kinds of Guile
web applications can use.  You will usually want to build something on
top of it, however, especially when producing HTML.  Here is a simple
example that builds up HTML output using SXML (@pxref{sxml simple}).

First, load up the modules:

@example
(use-modules (web server)
             (web request)
             (web response)
             (sxml simple))
@end example

Now we define a simple templating function that takes a list of HTML
body elements, as SXML, and puts them in our super template:

@example
(define (templatize title body)
  `(html (head (title ,title))
         (body ,@@body)))
@end example

For example, the simplest Hello HTML can be produced like this:

@example
(sxml->xml (templatize "Hello!" '((b "Hi!"))))
@print{}
<html><head><title>Hello!</title></head><body><b>Hi!</b></body></html>
@end example

Much better to work with Scheme data types than to work with HTML as
strings. Now we define a little response helper:

@example
(define* (respond #:optional body #:key
                  (status 200)
                  (title "Hello hello!")
                  (doctype "<!DOCTYPE html>\n")
                  (content-type-params '((charset . "utf-8")))
                  (content-type 'text/html)
                  (extra-headers '())
                  (sxml (and body (templatize title body))))
  (values (build-response
           #:code status
           #:headers `((content-type
                        . (,content-type ,@@content-type-params))
                       ,@@extra-headers))
          (lambda (port)
            (if sxml
                (begin
                  (if doctype (display doctype port))
                  (sxml->xml sxml port))))))
@end example

Here we see the power of keyword arguments with default initializers. By
the time the arguments are fully parsed, the @code{sxml} local variable
will hold the templated SXML, ready for sending out to the client.

Also, instead of returning the body as a string, @code{respond} gives a
procedure, which will be called by the web server to write out the
response to the client.

Now, a simple example using this responder, which lays out the incoming
headers in an HTML table.

@example
(define (debug-page request body)
  (respond
   `((h1 "hello world!")
     (table
      (tr (th "header") (th "value"))
      ,@@(map (lambda (pair)
               `(tr (td (tt ,(with-output-to-string
                               (lambda () (display (car pair))))))
                    (td (tt ,(with-output-to-string
                               (lambda ()
                                 (write (cdr pair))))))))
             (request-headers request))))))

(run-server debug-page)
@end example

Now if you visit any local address in your web browser, we actually see
some HTML, finally.

@subsubsection Conclusion

Well, this is about as far as Guile's built-in web support goes, for
now.  There are many ways to make a web application, but hopefully by
standardizing the most fundamental data types, users will be able to
choose the approach that suits them best, while also being able to
switch between implementations of the server.  This is a relatively new
part of Guile, so if you have feedback, let us know, and we can take it
into account.  Happy hacking on the web!

@c Local Variables:
@c TeX-master: "guile.texi"
@c End: