INTERNET-DRAFT Ken A L Coar
draft-coar-cgi-v11-01.{html,txt} IBM Corporation
D.R.T. Robinson
ESI
8 December, 1998
The WWW Common Gateway Interface
Version 1.1
Status of this Memo
This document is an Internet-Draft. Internet-Drafts are
working documents of the Internet Engineering Task Force
(IETF), its areas, and its working groups. Note that other
groups may also distribute working documents as
Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of
six months and may be updated, replaced, or obsoleted by
other documents at any time. It is inappropriate to use
Internet-Drafts as reference material or to cite them other
than as "work in progress."
To view the entire list of current Internet-Drafts, please
check the "1id-abstracts.txt" listing contained in the
Internet-Drafts Shadow Directories on ftp.is.co.za
(Africa), ftp.nordu.net (Northern Europe), ftp.nis.garr.it
(Southern Europe), munnari.oz.au (Pacific Rim),
ftp.ietf.org (US East Coast), or ftp.isi.edu (US West
Coast).
Abstract
The Common Gateway Interface (CGI) is a simple interface for
running external programs, software or gateways under an
information server in a platform-independent manner.
Currently, the supported information servers are HTTP servers.
The interface has been in use by the World-Wide Web since
1993. This specification defines the "current practice"
parameters of the 'CGI/1.1' interface developed and documented
at the U.S. National Centre for Supercomputing Applications
[NCSA-CGI]. This document also defines the use of the CGI/1.1
interface on the Unix and AmigaDOS(tm) systems.
Discussion of this draft occurs on the CGI-WG mailing list;
see the project Web page at
for details on the
mailing list and the status of the project.
Table of Contents
1 Introduction..............................................2
1.1 Purpose................................................2
1.2 Requirements...........................................2
1.3 Specifications.........................................2
Coar, et al. INTERNET-DRAFT [Page 1]
CGI/1.1 Expires: 14 May, 1999
1.4 Terminology............................................3
2 Notational Conventions and Generic Grammar................3
2.1 Augmented BNF..........................................3
2.2 Basic Rules............................................4
3 Protocol Parameters.......................................5
3.1 URL Encoding...........................................5
3.2 The Script URI.........................................5
4 Invoking the Script.......................................6
5 The CGI Script Command Line...............................6
6 Data Input to the CGI Script..............................6
6.1 Request Metadata (Meta-Variables)......................7
6.1.1 AUTH_TYPE...........................................7
6.1.2 CONTENT_LENGTH......................................8
6.1.3 CONTENT_TYPE........................................8
6.1.4 GATEWAY_INTERFACE...................................9
6.1.5 HTTP_*..............................................9
6.1.6 PATH_INFO...........................................10
6.1.7 PATH_TRANSLATED.....................................10
6.1.8 QUERY_STRING........................................11
6.1.9 REMOTE_ADDR.........................................11
6.1.10 REMOTE_HOST........................................12
6.1.11 REMOTE_IDENT.......................................12
6.1.12 REMOTE_USER........................................12
6.1.13 REQUEST_METHOD.....................................12
6.1.14 SCRIPT_NAME........................................13
6.1.15 SERVER_NAME........................................13
6.1.16 SERVER_PORT........................................13
6.1.17 SERVER_PROTOCOL....................................14
6.1.18 SERVER_SOFTWARE....................................14
6.2 Request Content-Bodies................................14
7 Data Output from the CGI Script...........................15
7.1 Non-Parsed Header Output...............................15
7.2 Parsed Header Output...................................15
7.2.1 CGI header fields...................................16
7.2.1.1 Content-Type.....................................16
7.2.1.2 Location.........................................16
7.2.1.3 Status...........................................17
7.2.1.4 Extension header fields..........................17
7.2.2 HTTP header fields..................................17
8 Server Implementation.....................................17
8.1 Requirements for Servers...............................18
8.2 Recommendations for Servers............................18
8.3 Summary of Meta-Variables..............................19
9 Script Implementation.....................................20
9.1 Requirements for Scripts...............................20
9.2 Recommendations for Scripts............................20
10 System Specifications....................................21
10.1 AmigaDOS..............................................21
10.2 Unix..................................................21
11 Security Considerations..................................22
11.1 Safe Methods..........................................22
11.2 HTTP Header Fields Containing Sensitive Information...22
11.3 Script Interference with the Server...................22
12 Acknowledgments..........................................23
Coar, et al. INTERNET-DRAFT [Page 2]
CGI/1.1 Expires: 14 May, 1999
13 References...............................................23
14 Authors' Addresses.......................................24
1. Introduction
1.1. Purpose
Together the HTTP [3],[8] server and the CGI script are
responsible for servicing a client request by sending back
responses. The client request comprises a Universal Resource
Identifier (URI) [1], a request method, and various ancillary
information about the request provided by the transport
mechanism.
The CGI defines the abstract parameters, known as
meta-variables, which describe the client's request. Together
with a concrete programmer interface this specifies a
platform-independent interface between the script and the HTTP
server.
1.2. Requirements
This specification uses the same words as RFC 1123 [5] to
define the significance of each particular requirement. These
are:
MUST
This word or the adjective 'required' means that the
item is an absolute requirement of the specification.
SHOULD
This word or the adjective 'recommended' means that
there may exist valid reasons in particular
circumstances to ignore this item, but the full
implications should be understood and the case
carefully weighed before choosing a different course.
MAY
This word or the adjective 'optional' means that this
item is truly optional. One vendor may choose to
include the item because a particular marketplace
requires it or because it enhances the product, for
example; another vendor may omit the same item.
An implementation is not compliant if it fails to satisfy one
or more of the 'must' requirements for the protocols it
implements. An implementation that satisfies all of the 'must'
and all of the 'should' requirements for its features is said
to be 'unconditionally compliant'; one that satisfies all of
the 'must' requirements but not all of the 'should'
requirements for its features is said to be 'conditionally
compliant.'
1.3. Specifications
Coar, et al. INTERNET-DRAFT [Page 3]
CGI/1.1 Expires: 14 May, 1999
Not all of the functions and features of the CGI are defined
in the main part of this specification. The following phrases
are used to describe the features which are not specified:
system defined
The feature may differ between systems, but must be the
same for different implementations using the same
system. A system will usually identify a class of
operating-systems. Some systems are defined in section
12 of this document. New systems may be defined by new
specifications without revision of this document.
implementation defined
The behaviour of the feature may vary from
implementation to implementation, but a particular
implementation must document its behaviour.
1.4. Terminology
This specification uses many terms defined in the HTTP/1.1
specification [8]; however, the following terms are used here
in a sense which may not accord with their definitions in that
document, or with their common meaning.
meta-variable
A named parameter that carries information from the
server to the script. It is not necessarily a variable
in the operating-system's environment, although that is
the most common implementation.
script
The software which is invoked by the server via this
interface. It need not be a standalone program, but
could be a dynamically-loaded or shared library, or
even a subroutine in the server.
server
The application program which invokes the script in
order to service requests.
2. Notational Conventions and Generic Grammar
2.1. Augmented BNF
All of the mechanisms specified in this document are described
in both prose and an augmented Backus-Naur Form (BNF) similar
to that used by RFC 822 [6]. This augmented BNF contains the
following constructs:
name = definition
The definition by the equal character ("="). Whitespace
is only significant in that continuation lines of a
definition are indented.
Coar, et al. INTERNET-DRAFT [Page 4]
CGI/1.1 Expires: 14 May, 1999
"literal"
Quotation marks (") surround literal text, except for a
literal quotation mark, which is surrounded by
angle-brackets ("<" and ">"). Unless stated otherwise,
the text is case-sensitive.
rule1 | rule2
Alternative rules are separated by a vertical bar
("|").
(rule1 rule2 rule3)
Elements enclosed in parentheses are treated as a
single element.
*rule
A rule preceded by an asterisk ("*") may have zero or
more occurrences. A rule preceded by an integer
followed by an asterisk must occur at least the
specified number of times.
[rule]
A element enclosed in square brackets ("[" and "]") is
optional.
2.2. Basic Rules
The following rules are used throughout this specification to
describe basic parsing constructs.
alpha = lowalpha | hialpha
lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h"
| "i" | "j" | "k" | "l" | "m" | "n" | "o" | "p"
| "q" | "r" | "s" | "t" | "u" | "v" | "w" | "x"
| "y" | "z"
hialpha = "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H"
| "I" | "J" | "K" | "L" | "M" | "N" | "O" | "P"
| "Q" | "R" | "S" | "T" | "U" | "V" | "W" | "X"
| "Y" | "Z"
hex = digit | "A" | "B" | "C" | "D" | "E" | "F" | "a"
| "b" | "c" | "d" | "e" | "f"
digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7"
| "8" | "9"
OCTET =
CHAR =
CTL =
SP =
HT =
NL =
LWSP = SP | HT | NL
tspecial = "(" | ")" | "@" | "," | ";" | ":" | "\" | <">
| "/" | "[" | "]" | "?" | "<" | ">" | "{" | "}"
| SP | HT
token = 1*
Coar, et al. INTERNET-DRAFT [Page 5]
CGI/1.1 Expires: 14 May, 1999
quoted-string = ( <"> *qdtext <"> ) | ( "<" *qatext ">")
qdtext = and CTLs but including LWSP>
qatext = " and CTLs but
including LWSP>
Note that newline (NL) need not be a single character, but can
be a character sequence.
3. Protocol Parameters
3.1. URL Encoding
Some variables and constructs used here are described as being
'URL-encoded'. This encoding is described in section 2 of RFC
2396 [4].
An alternate "shortcut" encoding for representing the space
character exists and is in common use. Scripts MUST be
prepared to recognise both '+' and '%20' as an encoded space
in a URL-encoded value.
Note that some unsafe characters may have different semantics
if they are encoded. The definition of which characters are
unsafe depends on the context. For example, the following two
URLs do not necessarily refer to the same resource:
http://somehost.com/somedir%2Fvalue
http://somehost.com/somedir/value
See section 2 of RFC 2396 [4] for authoritative treatment of
this issue.
3.2. The Script URI
The 'Script-URI' is defined as the URI of the resource
identified by the meta-variables. Often, this URI will be the
same as the URI requested by the client (the 'Client-URI');
however, it need not be. Instead, it could be a URI invented
by the server, and so it can only be used in the context of
the server and its CGI interface.
The Script-URI has the syntax of generic-RL as defined in
section 2.1 of RFC 1808 [7], with the exception that object
parameters and fragment identifiers are not permitted:
:///?
The various components of the script URI are defined by some
of the meta-variables (see section 4 below);
script-uri = protocol "://" SERVER_NAME ":" SERVER_PORT enc-script
enc-path-info "?" QUERY_STRING
where 'protocol' is obtained from SERVER_PROTOCOL,
Coar, et al. INTERNET-DRAFT [Page 6]
CGI/1.1 Expires: 14 May, 1999
'enc-script' is a URL-encoded version of SCRIPT_NAME and
'enc-path-info' is a URL-encoded version of PATH_INFO. See
section 4.6 for more information about the PATH_INFO
meta-variable.
Note that the scheme and the protocol are not identical; for
instance, a resource accessed via an SSL mechanism may have a
Client-URI with a scheme of "https" rather than "http".
CGI/1.1 provides no means for the script to reconstruct this,
and therefore the Script-URI includes the base protocol used.
4. Invoking the Script
The script is invoked in a system defined manner. Unless
specified otherwise, this will be by treating the file
containing the script as an executable program, and running it
as a child process of the server.
5. The CGI Script Command Line
Some systems support a method for supplying an array of
strings to the CGI script. This is only used in the case of an
'indexed' query. This is identified by a "GET" or "HEAD" HTTP
request with a URL query string not containing any unencoded
"=" characters. For such a request, servers SHOULD parse the
search string into words, using the following rules:
search-string = search-word *( "+" search-word )
search-word = 1*schar
schar = xunreserved | escape | xreserved
xunreserved = alpha | digit | xsafe | extra
xsafe = "$" | "-" | "_" | "."
xreserved = ";" | "/" | "?" | ":" | "@" | "&"
After parsing, each word is URL-decoded, optionally encoded in
a system defined manner, and then the argument list is set to
the list of words.
If the server cannot create any part of the argument list,
then the server SHOULD NOT generate any command line
information. For example, the number of arguments may be
greater than operating system or server limitations permit, or
one of the words may not be representable as an argument.
Scripts SHOULD check to see if the QUERY_STRING value contains
an unencoded "=" character, and SHOULD NOT use the command
line arguments if it does.
6. Data Input to the CGI Script
Information about a request comes from two different sources:
the request header, and any associated content-body. Servers
MUST make portions of this information available to scripts.
Coar, et al. INTERNET-DRAFT [Page 7]
CGI/1.1 Expires: 14 May, 1999
6.1. Request Metadata (Meta-Variables)
Each CGI server implementation MUST define a mechanism to pass
data about the request from the server to the script. The
meta-variables containing these data are accessed by the
script in a system defined manner. In all cases, a missing
meta-variable is equivalent to a zero-length (NULL) value, and
vice versa. The representation of the characters in the
meta-variables is system defined.
Case is not significant in the meta-variable names, in that
there cannot be two different variables whose names differ in
case only. Here they are shown using a canonical
representation of capitals plus underscore ("_"). The actual
representation of the names is system defined; for a
particular system the representation MAY be defined
differently than this.
Meta-variable values MUST be considered case-sensitive except
as noted otherwise.
The canonical variables defined by this specification are:
AUTH_TYPE
CONTENT_LENGTH
CONTENT_TYPE
GATEWAY_INTERFACE
HTTP_*
PATH_INFO
PATH_TRANSLATED
QUERY_STRING
REMOTE_ADDR
REMOTE_HOST
REMOTE_IDENT
REMOTE_USER
REQUEST_METHOD
SCRIPT_NAME
SERVER_NAME
SERVER_PORT
SERVER_PROTOCOL
SERVER_SOFTWARE
6.1.1. AUTH_TYPE
This variable is specific to requests made via the HTTP
scheme.
If the script-URI required access authentication for external
access, then the server SHOULD set the value of this variable
from the 'auth-scheme' token in the request's "Authorization"
header field. therwise it is set to NULL.
AUTH_TYPE = "" | auth-scheme
auth-scheme = "Basic" | token
Coar, et al. INTERNET-DRAFT [Page 8]
CGI/1.1 Expires: 14 May, 1999
HTTP access authentication schemes are described in section 11
of the HTTP/1.1 specification [8]. The auth-scheme is not
case-sensitive.
Servers SHOULD provide this meta-variable to scripts if the
request header included an "Authorization" field.
6.1.2. CONTENT_LENGTH
This meta-variable is set to the size of the content-body
entity attached to the request, if any, in decimal number of
octets. If no data are attached, then this meta-variable is
either NULL or not defined. The syntax is the same as for the
HTTP "Content-Length" header field (section 14.14, HTTP/1.1
specification [8]).
CONTENT_LENGTH = "" | 1*digit
Servers MUST provide this meta-variable to scripts if the
request was accompanied by a content-body entity.
6.1.3. CONTENT_TYPE
If the request includes a content-body, CONTENT_TYPE is set
tothe Internet Media Type [9] of the attached entity if the
type was provided via a "Content-type" field in the request
header, or if the server can determine it in the absence of a
supplied "Content-type" field. The syntax is the same as for
the HTTP "Content-Type" header field.
CONTENT_TYPE = "" | media-type
media-type = type "/" subtype *( ";" parameter)
type = token
subtype = token
parameter = attribute "=" value
attribute = token
value = token | quoted-string
The type, subtype, and parameter attribute names are not
case-sensitive. Parameter values MAY be case sensitive. Media
types and their use in HTTP are described in section 3.7 of
the HTTP/1.1 specification [8].
Example:
application/x-www-form-urlencoded
There is no default value for this variable. If and only if it
is unset, then the script MAY attempt to determine the media
type from the data received. If the type remains unknown, then
the script MAY choose to either assume a content-type of
application/octet-stream or reject the request with a 415
("Unsupported Media Type") error. See section 7.2.1.3 for more
Coar, et al. INTERNET-DRAFT [Page 9]
CGI/1.1 Expires: 14 May, 1999
information about returning error status values.
Servers MUST provide this meta-variable to scripts if a
"Content-Type" field was present in the original request
header. If the server receives a request with an attached
entity but no "Content-Type" header field, it MAY attempt to
determine the correct datatype, or it MAY omit this
meta-variable when communicating the request information to
the script.
6.1.4. GATEWAY_INTERFACE
This meta-variable is set to the dialect of CGI being used by
the server to communicate with the script. Syntax:
GATEWAY_INTERFACE = "CGI" "/" major "." minor
major = 1*digit
minor = 1*digit
Note that the major and minor numbers are treated as separate
integers and hence each may be more than a single digit. Thus
CGI/2.4 is a lower version than CGI/2.13 which in turn is
lower than CGI/12.3. Leading zeros in either the major or the
minor number MUST be ignored by scripts and SHOULD NOT be
generated by servers.
This document defines the 1.1 version of the CGI interface
("CGI/1.1").
Servers MUST provide this meta-variable to scripts.
6.1.5. HTTP_*
These variables are specific to requests made via the HTTP
scheme. Interpretation of these variables depends on the value
of theSERVER_PROTOCOL meta-variable (see section 6.1.17).
Meta-variables with names beginning with "HTTP_" contain
values from the request header, if the scheme used was HTTP.
Each HTTP header field name is converted to upper case, has
all occurrences of "-" replaced with "_", and has "HTTP_"
prepended to form the meta-variable name. The header data MAY
be presented as sent by the client, or MAY be rewritten in
ways which do not change its semantics. If multiple header
fields with the same field-name are received then the server
MUST rewrite them as though they had been received as a single
header field having the same semantics before being
represented in a meta-variable. Similarly, a header field that
is received on more than one line MUST be merged into a single
line. The server MUST, if necessary, change the representation
of the data (for example, the character set) to be appropriate
for a CGI meta-variable.
Servers are not required to create meta-variables for all the
Coar, et al. INTERNET-DRAFT [Page 10]
CGI/1.1 Expires: 14 May, 1999
request header fields that they receive. In particular, they
MAY decline to make available any header fields carrying
authentication information, such as "Authorization"; and they
MAY header fields the values of which are available to the
script via other metavariables, such as "Content-Length" and
"Content-Type".
6.1.6. PATH_INFO
The PATH_INFO meta-variable specifies a path to be interpreted
by the CGI script. It identifies the source or sub-resource to
be returned by the CGI script, and it is derived from the
portion of the URI path following the script name but
preceding any query data. The syntax and semantics are similar
to a decoded HTTP URL 'path' token (defined in RFC 2396 [4]),
with the exception that a PATH_INFO of "/" represents a single
void path segment.
PATH_INFO = "" | ( "/" path )
path = segment *( "/" segment )
segment = *pchar
pchar =
The PATH_INFO string is the trailing part of the
component of the script-URI (see section 3.2) that follows the
SCRIPT_NAME portion of the path.
Servers MAY impose their own restrictions and limitations on
what values they will accept for PATH_INFO, and MAY reject or
edit any values they considers objectionable before passing
them to the script.
Servers MUST make this URI component available to CGI scripts.
The PATH_INFO quantity is case-sensitive, and the server MUST
preserve the case of the PATH_INFO element of the URI when
making it available to scripts.
6.1.7. PATH_TRANSLATED
PATH_TRANSLATED is derived by taking any path-info component
of the requiest URI (see section 6.1.6), decoding it (see
section 3.1), parsing it as a URI in its own right, and
performing any virtual-to-physical filesystem translation.
PATH_TRANSLATED = *CHAR
For a request such as the following:
http://somehost.com/cgi-bin/somescript/this%2eis%2epath%2einfo
the PATH_INFO component would be decoded, and the result
parsed as though it were a request for the following:
http://somehost.com/this.is.the.path.info
Coar, et al. INTERNET-DRAFT [Page 11]
CGI/1.1 Expires: 14 May, 1999
This would then be translated to a filesystem location,
perhaps something like this:
/usr/local/www/htdocs/this.is.the.path.info
This resulting filesystem path is the value of
PATH_TRANSLATED.
The value of PATH_TRANSLATED may or may not map to a valid
filesystem location or file. Servers MUST preserve the case of
the path-info segment if and only if the underlying filesystem
supports case-sensitive names. If the filesystem is only
case-aware, case-preserving, or case-blind with regard to
filenames, servers are not required to preserve the case of
the original segment through the translation.
The filesystem translation algorithm the server uses to derive
PATH_TRANSLATED is implementation defined; CGI scripts which
use this variable may suffer limited portability.
Servers SHOULD provide this meta-variable to scripts.
6.1.8. QUERY_STRING
A URL-encoded string; the part of the script-URI. (See
section 3.2.)
QUERY_STRING = query-string
query-string = *qchar
qchar = unreserved | escape | reserved
unreserved = alpha | digit | safe | extra
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "="
safe = "$" | "-" | "_" | "." | "+"
extra = "!" | "*" | "'" | "(" | ")" | ","
escape = "%" hex hex
The URL syntax for a query string is described in section 3 of
RFC 2396 [4].
Servers MUST supply this value to scripts. The QUERY_STRING
value is case-sensitive.
6.1.9. REMOTE_ADDR
The IP address of the agent sending the request to the server.
This is not necessarily that of the client (such as if the
request came through a proxy).
REMOTE_ADDR = hostnumber
hostnumber = ipv4-address | ipv6-address
ipv4-address = digits "." digits "." digits "." digits
ipv6-address = hexbit16 ":" hexbit16 ":" hexbit16 ":" hexbit16 ":"
Coar, et al. INTERNET-DRAFT [Page 12]
CGI/1.1 Expires: 14 May, 1999
hexbit16 ":" hexbit16 ":" hexbit16 ":" hexbit16
digits = 1*digit
hexbit16 = 1*hex
Servers MUST supply this value to scripts.
6.1.10. REMOTE_HOST
The fully qualified domain name of the agent sending the
request to the server, if available, otherwise NULL. (See
section 6.1.9.) Fully qualified domain names take the form as
described in section 3.5 of RFC 1034 [10] and section 2.1 of
RFC 1123 [5]. Domain names are not case sensitive.
Servers SHOULD provide this information to scripts.
6.1.11. REMOTE_IDENT
The identity information reported about the connection by a
RFC 1413 [11] request to the remote agent, if available.
Servers MAY choose not to support this feature, or not to
request the data for efficiency reasons.
REMOTE_IDENT = *CHAR
The data returned may be used for authentication purposes, but
the level of trust reposed in them should be minimal.
Servers MAY supply this information to scripts if the RFC1413
[11] lookup is performed.
6.1.12. REMOTE_USER
If the request required authentication using the "Basic"
mechanism (i.e., the AUTH_TYPE meta-variable is set to
"Basic"), then the value of the REMOTE_USER meta-variable is
set to the user-ID supplied. In all other cases the value of
this meta-variable is undefined.
REMOTE_USER = "" | userid | *OCTET
userid = token
This variable is specific to requests made via the HTTP
protocol.
Servers SHOULD provide this meta-variable to scripts.
6.1.13. REQUEST_METHOD
The REQUEST_METHOD meta-variable is set to the method with
which the request was made, as described in section 5.1.1 of
the HTTP/1.0 specification [3] and section 5.1.1 of the
HTTP/1.1 specification [8].
Coar, et al. INTERNET-DRAFT [Page 13]
CGI/1.1 Expires: 14 May, 1999
REQUEST_METHOD = http-method
http-method = "GET" | "HEAD" | "POST" | "PUT" | "DELETE"
| extension-method
extension-method = token
The method is case sensitive. Note that of the new methods
defined by the HTTP/1.1 specification [8], OPTIONS and TRACE
are not appropriate for the CGI/1.1 environment.
This variable is specific to requests made with HTTP.
Servers MUST provide this meta-variable to scripts.
6.1.14. SCRIPT_NAME
The SCRIPT_NAME meta-variable is set to a URL path that could
identify the CGI script (rather than the script's output). The
syntax and semantics are identical to a decoded HTTP URL
'path' token (see RFC 2396 [4]).
SCRIPT_NAME = "" | ( "/" [ path ] )
The SCRIPT_NAME string is some leading part of the
component of the script-URI derived in some implementation
defined manner. No PATH_INFO or QUERY_STRING segments (see
sections 6.1.6 and 6.1.8) are included in the SCRIPT_NAME
value.
Servers MUST provide this meta-variable to scripts.
6.1.15. SERVER_NAME
The SERVER_NAME meta-variable is set to the name of the
server, as derived from the part of the script-URI (see
section 3.2).
SERVER_NAME = hostname | hostnumber
Servers MUST provide this meta-variable to scripts.
6.1.16. SERVER_PORT
The SERVER_PORT meta-variable is set to the port on which the
request was received, as used in the part of the
script-URI.
SERVER_PORT = 1*digit
If the portion of the script-URI is blank, the actual
port number upon which the request was received MUST be
supplied.
Servers MUST provide this meta-variable to scripts.
Coar, et al. INTERNET-DRAFT [Page 14]
CGI/1.1 Expires: 14 May, 1999
6.1.17. SERVER_PROTOCOL
The SERVER_PROTOCOL meta-variable is set to the name and
revision of the information protocol with which thLe request
arrived. This is not necessarily the same as the protocol
version used by the server in its response to the client.
SERVER_PROTOCOL = HTTP-Version | extension-version
| extension-token
HTTP-Version = "HTTP" "/" 1*digit "." 1*digit
extension-version = protocol "/" 1*digit "." 1*digit
protocol = 1*( alpha | digit | "+" | "-" | "." )
extension-token = token
'protocol' is a version of the part of the
script-URI, but is not identical to it. For example, the
scheme of a request may be "https" while the protocol remains
"http". The protocol is not case sensitive, but by convention,
'protocol' is in upper case.
A well-known extension token value is "INCLUDED", which
signals that the current document is being included as part of
a composite document, rather than being the direct target of
the client request.
Servers MUST provide this meta-variable to scripts.
6.1.18. SERVER_SOFTWARE
The SERVER_SOFTWARE meta-variable is set to the name and
version of the information server software answering the
request (and running the gateway).
SERVER_SOFTWARE = *CHAR
Servers MUST provide this meta-variable to scripts.
6.2. Request Content-Bodies
As there may be a data entity attached to the request, there
MUST be a system defined method for the script to read these
data. Unless defined otherwise, this will be via the 'standard
input' file descriptor.
If the CONTENT_LENGTH value (see section 6.1.2) is non-NULL,
the server MUST supply at least that many bytes to scripts on
the standard input stream. Scripts are not obliged to read the
data. Servers MAY signal an EOF condition after CONTENT_LENGTH
bytes have been read, but are are not obligated to do so.
Therefore, scripts MUST NOT attempt to read more than
CONTENT_LENGTH bytes, even if more data are available.
For non-parsed header (NPH) scripts (see section 7.1 below),
servers SHOULD attempt to ensure that the script input comes
Coar, et al. INTERNET-DRAFT [Page 15]
CGI/1.1 Expires: 14 May, 1999
directly from the client, with minimal buffering. For all
scripts the data will be as supplied by the client.
7. Data Output from the CGI Script
There MUST be a system defined method for the script to send
data back to the server or client; a script MUST always return
some data. Unless defined otherwise, this will be via the
'standard output' file descriptor.
There are two forms of output that scripts can supply to
servers: non-parsed header (NPH) output, and parsed header
output. Servers MUST support parsed header output and MAY
support NPH output. The method of distinguishing between the
two types of output (or scripts) is implementation defined.
Servers MAY implement a timeout period within which data must
be received from scripts. If a server implementation defines
such a timeout and receives no data from a script within the
timeout period, the server MAY terminate the script process
and SHOULD abort the client request with a '500 Internal
Server Error' response.
7.1. Non-Parsed Header Output
Scripts using the NPH output form MUST return a complete HTTP
response message, as described in Section 6 of the HTTP
specifications [3],[8]. NPH scripts MUST use the
SERVER_PROTOCOL variable to determine the appropriate format
for a response.
Servers SHOULD attempt to ensure that the script output is
sent directly to the client, with minimal internal and no
transport-visible buffering.
7.2. Parsed Header Output
Scripts using the parsed header output form MUST supply a CGI
response message to the server as follows:
CGI-Response = *( CGI-field | HTTP-Field ) NL [ Entity-Body ]
CGI-Field = Content-type
| Location
| Status
| extension-header
The response comprises a header and a body, separated by a
blank line. The body may be NULL. The header fields are either
CGI header fields to be interpreted by the server, or HTTP
header fields to be included in the response returned to the
client if the request method is HTTP. At least one CGI-Field
MUST be supplied, but no CGI field name may be used more than
once in a response. If a body is supplied, then a
"Content-type" header field MUST be supplied by the script,
Coar, et al. INTERNET-DRAFT [Page 16]
CGI/1.1 Expires: 14 May, 1999
otherwise the script MUST send a "Location" or "Status" header
field. If a Location CGI-field is returned, then the script
MUST NOT supply any HTTP-Fields.
All header fields occurring in a CGI-Response MUST be
specified one per line; CGI/1.1 makes no provision for
continuation lines.
7.2.1. CGI header fields
The CGI header fields have the generic syntax:
generic-field = field-name ":" [ field-value ] NL
field-name = 1*
field-value = *( field-content | LWSP )
field-content = *( token | tspecial | quoted-string )
The field-name is not case sensitive; a NULL field value is
equivalent to the header field not being sent.
7.2.1.1. Content-Type
The Internet Media Type [9] of the entity body, which is to be
sent unmodified to the client.
Content-Type = "Content-Type" ":" media-type NL
This is actually an HTTP-Field rather than a CGI-field, but it
is listed here because of its importance in the CGI dialogue
as a member of the "one of these is required" set of header
fields.
7.2.1.2. Location
This is used to specify to the server that the script is
returning a reference to a document rather than an actual
document.
Location = "Location" ":"
( fragment-URI | rel-URL-abs-path ) NL
fragment-URI = URI [ # fragmentid ]
URI = scheme ":" *qchar
fragmentid = *qchar
rel-URL-abs-path = "/" [ hpath ] [ "?" query-string ]
hpath = fpsegment *( "/" psegment )
fpsegment = 1*hchar
psegment = *hchar
hchar = alpha | digit | safe | extra
| ":" | "@" | "& | "="
The Location value is either an absolute URI with optional
fragment, as defined in RFC 1630 [1], or an absolute path and
optional query-string. If an absolute URI is returned by the
script, then the server will generate a '302 redirect' HTTP
Coar, et al. INTERNET-DRAFT [Page 17]
CGI/1.1 Expires: 14 May, 1999
response message, and if no entity body is supplied by the
script, then the server will produce one. If the Location
value is a path, then the server will generate the response
that it would have produced in response to a request
containing the URL
scheme "://" SERVER_NAME ":" SERVER_PORT rel-URL-abs-path
Note: If the request was accompanied by a content-body (such
as for a POST request), the content-body will be lost if the
script redirects the request with a Location field.
7.2.1.3. Status
The "Status" header field is used to indicate to the server
what status code the server MUST use in the response message.
It SHOULD NOT be sent if the script returns a "Location"
header field.
Status = "Status" ":" digit digit digit SP reason-phrase NL
reason-phrase = *
The valid status codes are listed in section 6.1.1 of the
HTTP/1.0 specifications [3]. If the SERVER_PROTOCOL is
"HTTP/1.1", then the status codes defined in the HTTP/1.1
specification [8] may be used. If the script does not return a
"Status" header field, then "200 OK" SHOULD be assumed by the
server.
If a script is being used to handle a particular error or
condition encountered by the server, such as a '404 Not Found'
error, the script SHOULD use the "Status" CGI header field to
propagate the error condition back to the client. E.g., in the
example mentioned it SHOULD include a "Status: 404 Not Found"
in the header data returned to the server.
7.2.1.4. Extension header fields
Scripts MAY include in their CGI response header additional
fields not defined in this or the HTTP specification. These
are called "extension" fields, and have the syntax of a
generic-field as defined in section 7.2.1. The name of an
extension field MUST NOT conflict with a field name defined in
this or any other specification; extension field names SHOULD
begin with "X-CGI-" to ensure uniqueness.
7.2.2. HTTP header fields
The script MAY return any other header fields defined by the
specification for the SERVER_PROTOCOL (HTTP/1.0 [3] or
HTTP/1.1 [8]). Servers MUST resolve conflicts beteen CGI
header and HTTP header formats or names (see section 8).
8. Server Implementation
Coar, et al. INTERNET-DRAFT [Page 18]
CGI/1.1 Expires: 14 May, 1999
This section defines the requirements that must be met by HTTP
servers in order to provide a coherent and correct CGI/1.1
environment in which scripts may function. It is intended
primarily for server implementors, but it is useful for script
authors to be familiar with the information as well.
8.1. Requirements for Servers
Servers MUST support the standard mechanism (described below)
which allows script authors to determine what URL to use in
documents which reference the script; specifically, what URL
to use in order to achieve particular settings of the
meta-variables. This mechanism is as follows:
The server MUST translate the header data from the CGI header
field syntax to the HTTP header field syntax if these differ.
For example, the character sequence for newline (such as
Unix's ASCII NL) used by CGI scripts may not be the same as
that used by HTTP (ASCII CR followed by LF). The server MUST
also resolve any conflicts between header fields returned by
the script and header fields that it would otherwise send
itself. The fields affected and the resolution method used
SHOULD be documented as part of the server implementation.
8.2. Recommendations for Servers
Servers SHOULD provide the "query" component of the script-URI
as command-line arguments to scripts if it does not contain
any unencoded '=' characters and the command-line arguments
can be generated in an unambiguous manner. (See section 5.)
Servers SHOULD set the AUTH_TYPE meta-variable to the value of
the 'auth-scheme' token of the "Authorization" field if it was
supplied as part of the request header. (See section 6.1.1.)
Where applicable, servers SHOULD set the current working
directory to the directory in which the script is located
before invoking it.
Servers MAY reject with error '404 Not Found' any requests
that would result in an encoded "/" being decoded into
PATH_INFO or SCRIPT_NAME, as this might represent a loss of
information to the script.
Although the server and the CGI script need not be consistent
in their handling of URL paths (client URLs and the PATH_INFO
data, respectively), server authors may wish to impose
consistency. So the server implementation SHOULD define its
behaviour for the following cases:
1. define any restrictions on allowed characters, in
particular whether ASCII NUL is permitted;
2. define any restrictions on allowed path segments, in
Coar, et al. INTERNET-DRAFT [Page 19]
CGI/1.1 Expires: 14 May, 1999
particular whether non-terminal NULL segments are
permitted;
3. define the behaviour for "." or ".." path segments; i.e.,
whether they are prohibited, treated as ordinary path
segments or interpreted in accordance with the relative
URL specification [7];
4. define any limits of the implementation, including limits
on path or search string lengths, and limits on the volume
of header data the server will parse.
Servers MAY generate the script URI in any way from the client
URI, or from any other data (but the behaviour SHOULD be
documented).
For non-parsed header (NPH) scripts (see section 7.1), servers
SHOULD attempt to ensure that the script input comes directly
from the client, with minimal buffering. For all scripts the
data will be as supplied by the client.
8.3. Summary of Meta-Variables
Servers MUST provide the following meta-variables to scripts.
See the individual descriptions for exceptions and semantics.
CONTENT_LENGTH
CONTENT_TYPE
GATEWAY_INTERFACE
PATH_INFO
QUERY_STRING
REQUEST_METHOD
REMOTE_ADDR
SCRIPT_NAME
SERVER_NAME
SERVER_PORT
SERVER_PROTOCOL
SERVER_SOFTWARE
Servers SHOULD define the following meta-variables for
scripts. See the individual descriptions for exceptions and
semantics.
AUTH_TYPE
REMOTE_HOST
In addition, servers SHOULD provide meta-variables for all
fields present in the HTTP request header, with the exception
of those involved with access control. Servers MAY at their
discretion provide meta-variables for access control fields.
Servers MAY define the following meta-variables. See the
individual descriptions for exceptions and semantics.
PATH_TRANSLATED
REMOTE_IDENT
Coar, et al. INTERNET-DRAFT [Page 20]
CGI/1.1 Expires: 14 May, 1999
REMOTE_USER
Servers may at their discretion define additional
implementation-specific extension meta-variables provided
their names do not conflict with defined header field names.
Implementation-specific meta-variable names SHOULD be prefixed
with "X_" (e.g., "X_DBA") to avoid the potential for such
conflicts.
9. Script Implementation
This section defines the requirements and recommendations for
scripts that are intended to function in a CGI/1.1
environment. It is intended primarily as a reference for
script authors, but server implementors should be familiar
with these issues as well.
9.1. Requirements for Scripts
Scripts using the parsed-header method to communicate with
servers MUST supply a response header to the server. (See
section 7.)
Scripts using the NPH method to communicate with servers MUST
provide complete HTTP responses, and MUST use the value of the
SERVER_PROTOCOL meta-variable to determine the appropriate
format. (See section 7.1.)
Scripts MUST check the value of the REQUEST_METHOD
meta-variable in order to provide an appropriate response.
(See section 6.1.13.)
Scripts MUST be prepared to handled URL-encoded values in
meta-variables. In addition, they MUST recognise both "+" and
"%20" in URL-encoded quantities as representing the space
character. (See section 3.1.)
Scripts MUST ignore leading zeros in the major and minor
version numbers in the GATEWAY_INTERFACE meta-variable value.
(See section 6.1.4.)
When processing requests that include a content-body, scripts
MUST NOT read more than CONTENT_LENGTH bytes from the input
stream. (See sections 6.1.2 and 6.2.)
9.2. Recommendations for Scripts
Servers may interrupt or terminate script execution at any
time and without warning, so scripts SHOULD be prepared to
deal with abnormal termination.
Scripts SHOULD reject unexpected methods (such as DELETE,
etc.) with error '405 Method Not Allowed'. If the script does
not intend processing the PATH_INFO data, then it SHOULD
Coar, et al. INTERNET-DRAFT [Page 21]
CGI/1.1 Expires: 14 May, 1999
reject the request with '404 Not Found' if PATH_INFO is not
NULL.
If a script is processing the output of a form , it SHOULD
verify that the CONTENT_TYPE is
"application/x-www-form-urlencoded" [2] or whatever other
media type is expected.
Scripts parsing PATH_INFO, PATH_TRANSLATED, or SCRIPT_NAME
SHOULD be careful of void path segments ("//") and special
path segments ("." and ".."). They SHOULD either be removed
from the path before use in OS system calls, or the request
SHOULD be rejected with '404 Not Found'.
As it is impossible for scripts to determine the client URI
that initiated a request without knowledge of the specific
server in use, the script SHOULD NOT return "text/html"
documents containing relative URL links without including a
"" tag in the document.
When returning header fields, scripts SHOULD try to send the
CGI header fields (see section 7.2) as soon as possible, and
preferably before any HTTP header fields. This may help reduce
the server's memory requirements.
10. System Specifications
10.1. AmigaDOS
The implementation of the CGI on an AmigaDOS operating system
platform SHOULD use environment variables as the mechanism of
providing request metadata to CGI scripts.
Environment variables
These are accessed by the DOS library routine GetVar.
The flags argument SHOULD be 0. Case is ignored, but
upper case is recommended for compatibility with
case-sensitive systems.
The current working directory
The current working directory for the script is set to
the directory containing the script.
Character set
The US-ASCII character set is used for the definition
of environment variable names and header field names;
the newline (NL) sequence is LF; servers SHOULD also
accept CR LF as a newline.
10.2. Unix
The implementation of the CGI on a UNIX operating system
platform SHOULD use environment variables as the mechanism of
providing request metadata to CGI scripts.
Coar, et al. INTERNET-DRAFT [Page 22]
CGI/1.1 Expires: 14 May, 1999
For Unix compatible operating systems, the following are
defined:
Environment variables
These are accessed by the C library routine getenv.
The command line
This is accessed using the the argc and argv arguments
to main(). The words have any characters which are
'active' in the Bourne shell escaped with a backslash.
If the value of the QUERY_STRING meta-variable contains
an unencoded equals-sign '=', then the command line
SHOULD NOT be used by the script.
The current working directory
The current working directory for the script SHOULD be
set to the directory containing the script.
Character set
The US-ASCII character set is used for the definition
of environment variable names and header field names;
the newline (NL) sequence is LF; servers SHOULD also
accept CR LF as a newline.
11. Security Considerations
11.1. Safe Methods
As discussed in the security considerations of the HTTP
specifications [3],[8], the convention has been established
that the GET and HEAD methods should be 'safe'; they should
cause no side-effects and only have the significance of
resource retrieval.
11.2. HTTP Header Fields Containing Sensitive Information
Some HTTP header fields may carry sensitive information which
the server SHOULD NOT pass on to the script unless explicitly
configured to do so. For example, if the server protects the
script using the "Basic" authentication scheme, then the
client will send an "Authorization" header field containing a
username and password. If the server, rather than the script,
validates this information then the password SHOULD NOT be
passed on to the script via the HTTP_AUTHORIZATION
meta-variable without careful consideration. This also applies
to the Proxy-Authorization header field and the corresponding
HTTP_PROXY_AUTHORIZATION meta-variable.
11.3. Script Interference with the Server
The most common implementation of CGI invokes the script as a
child process using the same user and group as the server
process. It SHOULD therefore be ensured that the script cannot
Coar, et al. INTERNET-DRAFT [Page 23]
CGI/1.1 Expires: 14 May, 1999
interfere with the server process, its configuration, or
documents.
If the script is executed by calling a function linked in to
the server software (either at compile-time or run-time) then
precautions SHOULD be taken to protect the core memory of the
server, or to ensure that untrusted code cannot be executed.
12. Acknowledgements
This work is based on a draft published in 1997 by David R.
Robinson in 1997, which in turn was based on the original CGI
interface that arose out of discussions on the www-talk
mailing list. In particular, Rob McCool, John Franks, Ari
Luotonen, George Phillips and Tony Sanders deserve special
recognition for their efforts in defining and implementing the
early versions of this interface.
This document has also greatly benefited from the comments and
suggestions made by Chris Adie, Dave Kristol, Mike Meyer,
David Morris, and Harald Alvestrand.
13. References
[1]
Berners-Lee, T., 'Universal Resource Identifiers in
WWW: A Unifying Syntax for the Expression of Names and
Addresses of Objects on the Network as used in the
World-Wide Web', RFC 1630, CERN, June 1994.
[2]
Berners-Lee, T. and Connolly, D., 'Hypertext Markup
Language - 2.0', RFC 1866, MIT/W3C, November 1995.
[3]
Berners-Lee, T., Fielding, R. T. and Frystyk, H.,
'Hypertext Transfer Protocol -- HTTP/1.0', RFC 1945,
MIT/LCS, UC Irvine, May 1996.
[4]
Berners-Lee, T., Fielding, R., and Masinter, L.,
Editors, 'Uniform Resource Identifiers (URI): Generic
Syntax', RFC 2396, MIT, U.C. Irvine, Xerox Corporation,
August 1996.
[5]
Braden, R., Editor, 'Requirements for Internet Hosts --
Application and Support', STD 3, RFC 1123, IETF,
October 1989.
[6]
Crocker, D.H., 'Standard for the Format of ARPA
Internet Text Messages', STD 11, RFC 822, University of
Delaware, August 1982.
[7]
Fielding, R., 'Relative Uniform Resource Locators', RFC
1808, UC Irving, June 1995.
[8]
Fielding, R., Gettys, J., Mogul, J., Frystyk, H. and
Coar, et al. INTERNET-DRAFT [Page 24]
CGI/1.1 Expires: 14 May, 1999
Berners-Lee, T., 'Hypertext Transfer Protocol --
HTTP/1.1', RFC 2068, UC Irving, DEC, MIT/LCS, January
1997.
[9]
Freed, N. and Borenstein N., 'Multipurpose Internet
Mail Extensions (MIME) Part Two: Media Types', RFC
2046, Innosoft, First Virtual, November 1996.
[10]
Mockapetris, P., 'Domain Names - Concepts and
Facilities', STD 13, RFC 1034, ISI, November 1987.
[11]
St. Johns, M., 'Identification Protocol', RFC 1431, US
Department of Defense, February 1993.
[12]
'Coded Character Set -- 7-bit American Standard Code
for Information Interchange', ANSI X3.4-1986.
14. Authors' Addresses
Ken A L Coar
MeepZor Consulting
7824 Mayfaire Crest Lane, Suite 202
Raleigh, NC 27615-4875
U.S.A.
Tel: +1 (919) 254.4237
Fax: +1 (919) 254.5250
Email: Ken.Coar@Golux.Com
David Robinson
Electronic Share Information Ltd
Mount Pleasant House
2 Mount Pleasant
Huntingdon Road
Cambridge CB3 0RN
UK
Tel: +44 (1223) 566926
Fax: +44 (1223) 506288
Email: drtr@esi.co.uk
Coar, et al. INTERNET-DRAFT [Page 25]