[Docs] [txt|pdf] [Tracker] [Email] [Diff1] [Diff2] [Nits]

Versions: 00 01 02 03 04 RFC 3875

   INTERNET-DRAFT                                           Ken A L Coar
   draft-coar-cgi-v11-02.{html,txt}                      IBM Corporation
                                                         D.R.T. Robinson
                                                           2 April, 1999

                  The WWW Common Gateway Interface
                            Version 1.1

Status of this Memo

     This document is an Internet-Draft and is in full
     conformance with all provisions of Section 10 of RFC2026.

     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."

     The list of current Internet-Drafts can be accessed at

     The list of Internet-Draft Shadow Directories can be
     accessed at <http://www.ietf.org/shadow.html>.


   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
   <URL:http://Web.Golux.Com/coar/cgi/> for details on the
   mailing list and the status of the project.

Table of Contents

  1 Introduction..............................................3
   1.1 Purpose................................................3
   1.2 Requirements...........................................3
   1.3 Specifications.........................................4

Coar, et al.                 INTERNET-DRAFT                     [Page 1]

                                 CGI/1.1       Expires: 14 October, 1999

   1.4 Terminology............................................4
  2 Notational Conventions and Generic Grammar................4
   2.1 Augmented BNF..........................................4
   2.2 Basic Rules............................................5
  3 Protocol Parameters.......................................6
   3.1 URL Encoding...........................................6
   3.2 The Script URI.........................................6
  4 Invoking the Script.......................................7
  5 The CGI Script Command Line...............................7
  6 Data Input to the CGI Script..............................7
   6.1 Request Metadata (Metavariables).......................8
    6.1.1 AUTH_TYPE...........................................8
    6.1.2 CONTENT_LENGTH......................................9
    6.1.3 CONTENT_TYPE........................................9
    6.1.4 GATEWAY_INTERFACE...................................10
    6.1.5 HTTP_*..............................................10
    6.1.6 PATH_INFO...........................................11
    6.1.7 PATH_TRANSLATED.....................................11
    6.1.8 QUERY_STRING........................................12
    6.1.9 REMOTE_ADDR.........................................12
    6.1.10 REMOTE_HOST........................................13
    6.1.11 REMOTE_IDENT.......................................13
    6.1.12 REMOTE_USER........................................13
    6.1.13 REQUEST_METHOD.....................................13
    6.1.14 SCRIPT_NAME........................................14
    6.1.15 SERVER_NAME........................................14
    6.1.16 SERVER_PORT........................................14
    6.1.17 SERVER_PROTOCOL....................................15
    6.1.18 SERVER_SOFTWARE....................................15
    6.2 Request Content-Bodies................................15
  7 Data Output from the CGI Script...........................16
   7.1 Non-Parsed Header Output...............................16
   7.2 Parsed Header Output...................................16
    7.2.1 CGI header fields...................................17 Content-Type.....................................17 Location.........................................17 Status...........................................18 Extension header fields..........................18
    7.2.2 HTTP header fields..................................18
  8 Server Implementation.....................................19
   8.1 Requirements for Servers...............................19
    8.1.1 Script-URI..........................................19
    8.1.2 Request Content-body Handling.......................19
    8.1.3 Required Metavariables..............................20
   8.2 Recommendations for Servers............................20
   8.3 Summary of Metavariables...............................21
  9 Script Implementation.....................................22
   9.1 Requirements for Scripts...............................22
   9.2 Recommendations for Scripts............................22
  10 System Specifications....................................23
   10.1 AmigaDOS..............................................23
   10.2 Unix..................................................23
  11 Security Considerations..................................24
   11.1 Safe Methods..........................................24

Coar, et al.                 INTERNET-DRAFT                     [Page 2]

                                 CGI/1.1       Expires: 14 October, 1999

   11.2 HTTP Header Fields Containing Sensitive Information...24
   11.3 Script Interference with the Server...................24
  12 Acknowledgments..........................................25
  13 References...............................................25
  14 Authors' Addresses.......................................26

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

   The CGI defines the abstract parameters, known as
   metavariables, which describe the client's request. Together
   with a concrete programmer interface this specifies a
   platform-independent interface between the script and the HTTP

1.2. Requirements

   This specification uses the same words as RFC 1123 [5] to
   define the significance of each particular requirement. These

          This word or the adjective 'required' means that the
          item is an absolute requirement of the specification.

          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.

          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

Coar, et al.                 INTERNET-DRAFT                     [Page 3]

                                 CGI/1.1       Expires: 14 October, 1999


1.3. Specifications

   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.

          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.

          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.

          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

Coar, et al.                 INTERNET-DRAFT                     [Page 4]

                                 CGI/1.1       Expires: 14 October, 1999

          The definition by the equal character ("="). Whitespace
          is only significant in that continuation lines of a
          definition are indented.

          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.

          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.

          An element enclosed in square brackets ("[" and "]") is

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         = <any 8-bit byte>
    CHAR          = <any character>
    CTL           = <any control character>
    SP            = <space character>
    HT            = <horizontal tab character>
    NL            = <newline>
    LWSP          = SP | HT | NL
    tspecial      = "(" | ")" | "@" | "," | ";" | ":" | "\" | <">

Coar, et al.                 INTERNET-DRAFT                     [Page 5]

                                 CGI/1.1       Expires: 14 October, 1999

                    | "/" | "[" | "]" | "?" | "<" | ">" | "{" | "}"
                    | SP | HT
    token         = 1*<any CHAR except CTLs or tspecials>
    quoted-string = ( <"> *qdtext <"> ) | ( "<" *qatext ">")
    qdtext        = <any CHAR except <"> and CTLs but including LWSP>
    qatext        = <any CHAR except "<", ">" 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:


   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 metavariables. 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 metavariables (see section 4 below);

Coar, et al.                 INTERNET-DRAFT                     [Page 6]

                                 CGI/1.1       Expires: 14 October, 1999

    script-uri = protocol "://" SERVER_NAME ":" SERVER_PORT enc-script
                 enc-path-info "?" QUERY_STRING

   where 'protocol' is obtained from SERVER_PROTOCOL,
   '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

   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

Coar, et al.                 INTERNET-DRAFT                     [Page 7]

                                 CGI/1.1       Expires: 14 October, 1999

   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.

6.1. Request Metadata (Metavariables)

   Each CGI server implementation MUST define a mechanism to pass
   data about the request from the server to the script. The
   metavariables containing these data are accessed by the script
   in a system defined manner. In all cases, a missing
   metavariable is equivalent to a zero-length (NULL) value, and
   vice versa. The representation of the characters in the
   metavariables is system defined.

   Case is not significant in the metavariable 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.

   metavariable values MUST be considered case-sensitive except
   as noted otherwise.

   The canonical variables defined by this specification are:


6.1.1. AUTH_TYPE

   This variable is specific to requests made via the HTTP

   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"

Coar, et al.                 INTERNET-DRAFT                     [Page 8]

                                 CGI/1.1       Expires: 14 October, 1999

   header field. therwise it is set to NULL.

    AUTH_TYPE   = "" | auth-scheme
    auth-scheme = "Basic" | token

   HTTP access authentication schemes are described in section 11
   of the HTTP/1.1 specification [8]. The auth-scheme is not

   Servers SHOULD provide this metavariable to scripts if the
   request header included an "Authorization" field.


   This metavariable 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 metavariable 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 metavariable to scripts if the
   request was accompanied by a content-body entity.


   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].



   There is no default value for this variable. If and only if it
   is unset, then the script MAY attempt to determine the media

Coar, et al.                 INTERNET-DRAFT                     [Page 9]

                                 CGI/1.1       Expires: 14 October, 1999

   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 for more
   information about returning error status values.

   Servers MUST provide this metavariable 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
   metavariable when communicating the request information to the


   This metavariable 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

   Servers MUST provide this metavariable 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 metavariable (see section 6.1.17).

   metavariables 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 metavariable 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 metavariable. 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

Coar, et al.                 INTERNET-DRAFT                    [Page 10]

                                 CGI/1.1       Expires: 14 October, 1999

   of the data (for example, the character set) to be appropriate
   for a CGI metavariable.

   Servers are not required to create metavariables for all the
   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

6.1.6. PATH_INFO

   The PATH_INFO metavariable 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     = <any CHAR except "/">

   The PATH_INFO string is the trailing part of the <path>
   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.


   PATH_TRANSLATED is derived by taking any path-info component
   of the request 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.


   For a request such as the following:


Coar, et al.                 INTERNET-DRAFT                    [Page 11]

                                 CGI/1.1       Expires: 14 October, 1999

   the PATH_INFO component would be decoded, and the result
   parsed as though it were a request for the following:


   This would then be translated to a filesystem location,
   perhaps something like this:


   This resulting filesystem path is the value of

   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 metavariable to scripts.


   A URL-encoded string; the <query> 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.


   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

Coar, et al.                 INTERNET-DRAFT                    [Page 12]

                                 CGI/1.1       Expires: 14 October, 1999

    hostnumber   = ipv4-address | ipv6-address
    ipv4-address = digits "." digits "." digits "." digits
    ipv6-address = hexbit16 ":" hexbit16 ":" hexbit16 ":" hexbit16 ":"
                   hexbit16 ":" hexbit16 ":" hexbit16 ":" hexbit16
    digits       = 1*digit
    hexbit16     = 1*hex

   Servers MUST supply this value to scripts.


   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.


   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.


   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.


   If the request required authentication using the "Basic"
   mechanism (i.e., the AUTH_TYPE metavariable is set to
   "Basic"), then the value of the REMOTE_USER metavariable is
   set to the user-ID supplied. In all other cases the value of
   this metavariable is undefined.

    REMOTE_USER = "" | userid | *OCTET
    userid      = token

   This variable is specific to requests made via the HTTP

   Servers SHOULD provide this metavariable to scripts.


   The REQUEST_METHOD metavariable is set to the method with

Coar, et al.                 INTERNET-DRAFT                    [Page 13]

                                 CGI/1.1       Expires: 14 October, 1999

   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].

    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 metavariable to scripts.


   The SCRIPT_NAME metavariable 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 <path>
   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

   Servers MUST provide this metavariable to scripts.


   The SERVER_NAME metavariable is set to the name of the server,
   as derived from the <host> part of the script-URI (see section

    SERVER_NAME = hostname | hostnumber

   Servers MUST provide this metavariable to scripts.


   The SERVER_PORT metavariable is set to the port on which the
   request was received, as used in the <port> part of the

    SERVER_PORT = 1*digit

   If the <port> portion of the script-URI is blank, the actual
   port number upon which the request was received MUST be

Coar, et al.                 INTERNET-DRAFT                    [Page 14]

                                 CGI/1.1       Expires: 14 October, 1999


   Servers MUST provide this metavariable to scripts.


   The SERVER_PROTOCOL metavariable iss 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 <scheme> 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 metavariable to scripts.


   The SERVER_SOFTWARE metavariable is set to the name and
   version of the information server software answering the
   request (and running the gateway).


   Servers MUST provide this metavariable 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

Coar, et al.                 INTERNET-DRAFT                    [Page 15]

                                 CGI/1.1       Expires: 14 October, 1999

   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
   directly from the client, with minimal buffering. For all
   scripts the data will be as supplied by the client.

   Section 8.1.2 describes the requirements of servers with
   regard to requests that include content-bodies.

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

Coar, et al.                 INTERNET-DRAFT                    [Page 16]

                                 CGI/1.1       Expires: 14 October, 1999

   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,
   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*<any CHAR, excluding CTLs, SP and ":">
    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. 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. Location

   This is used to specify to the server that the script is
   returning a reference to a document rather than an actual

    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

Coar, et al.                 INTERNET-DRAFT                    [Page 17]

                                 CGI/1.1       Expires: 14 October, 1999

    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
   within the server's URI space (i.e., omitting the scheme and
   network-related fields) and optional query-string. If an
   absolute URI is returned by the script, then the server MUST
   generate a '302 redirect' HTTP response message unless the
   script has supplied an explicit Status response header field,
   and if no entity body is supplied by the script, the server
   MUST produce one. If the Location value is a path, then the
   server MUST 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. Status

   The "Status" header field is used to indicate to the server
   what status code the server MUST use in the response message.

    Status        = "Status" ":" digit digit digit SP reason-phrase NL
    reason-phrase = *<CHAR, excluding CTLs, NL>

   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

   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. 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

Coar, et al.                 INTERNET-DRAFT                    [Page 18]

                                 CGI/1.1       Expires: 14 October, 1999

   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

   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

   In order to be considered CGI/1.1-compliant, a server must
   meet certain basic criteria and provide certain minimal
   functionality. The details of these requirements are described
   in the following sections.

8.1.1. Script-URI

   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
   metavariables. 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.

8.1.2. Request Content-body Handling

   Because CGI/1.1 requires that the CONTENT_LENGTH metavariable
   be provided when a script is activated (see section 8.3),
   servers MUST take care to correctly process requests that
   include a content-body but do not specify a Content-Length
   field in the request header (such as a request with a
   Transfer-Encoding of "chunked").

   Requests that require special treatment meet all of the
   following criteria:

    1. The resource identified by the request-URI is a CGI/1.1
       script, and
    2. The request header does not include a Content-Length

Coar, et al.                 INTERNET-DRAFT                    [Page 19]

                                 CGI/1.1       Expires: 14 October, 1999

       field, and
    3. The request includes a content-body, and
    4. The request header includes a Transfer-Encoding field that
       specifies a method (such as "chunked") which either does
       not require or prohibits a Content-Length field.

   If an HTTP/1.1 server determines that a request meets all of
   the above criteria, it MUST process the request in one of the
   two following ways:

     * The server may read (buffer) the entire request in order
       to determine the correct value of the CONTENT_LENGTH
       metavariable, and supply the buffered information to the
       script as though the content-body had not been
       transfer-encoded. The server MAY abort the request with a
       '413 Request Entity Too Large' status and close the
       connection if the size of the content-body is greater than
       the server is prepared to buffer. Or,

     * The server may reject the request with a '411 Length
       Required' status.

8.1.3. Required Metavariables

   Servers MUST provide scripts with certain information and
   metavariables as described in section 8.3.

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 metavariable 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

Coar, et al.                 INTERNET-DRAFT                    [Page 20]

                                 CGI/1.1       Expires: 14 October, 1999

       particular whether ASCII NUL is permitted;
    2. define any restrictions on allowed path segments, in
       particular whether non-terminal NULL segments are
    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

   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 MetaVariables

   Servers MUST provide the following metavariables to scripts.
   See the individual descriptions for exceptions and semantics.

    CONTENT_LENGTH (section 6.1.2)
    CONTENT_TYPE (section 6.1.3)
    GATEWAY_INTERFACE (section 6.1.4)
    PATH_INFO (section 6.1.6)
    QUERY_STRING (section 6.1.8)
    REMOTE_ADDR (section 6.1.9)
    REQUEST_METHOD (section 6.1.13)
    SCRIPT_NAME (section 6.1.14)
    SERVER_NAME (section 6.1.15)
    SERVER_PORT (section 6.1.16)
    SERVER_PROTOCOL (section 6.1.17)
    SERVER_SOFTWARE (section 6.1.18)

   Servers SHOULD define the following metavariables for scripts.
   See the individual descriptions for exceptions and semantics.

    AUTH_TYPE (section 6.1.1)
    REMOTE_HOST (section 6.1.10)

   In addition, servers SHOULD provide metavariables for all
   fields present in the HTTP request header, with the exception
   of those involved with access control. Servers MAY at their
   discretion provide metavariables for access control fields.

   Servers MAY define the following metavariables. See the
   individual descriptions for exceptions and semantics.

    PATH_TRANSLATED (section 6.1.7)

Coar, et al.                 INTERNET-DRAFT                    [Page 21]

                                 CGI/1.1       Expires: 14 October, 1999

    REMOTE_IDENT (section 6.1.11)
    REMOTE_USER (section 6.1.12)

   Servers may at their discretion define additional
   implementation-specific extension metavariables provided their
   names do not conflict with defined header field names.
   Implementation-specific metavariable names SHOULD be prefixed
   with "X_" (e.g., "X_DBA") to avoid the potential for such

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 metavariable to determine the appropriate
   format. (See section 7.1.)

   Scripts MUST check the value of the REQUEST_METHOD
   metavariable in order to provide an appropriate response. (See
   section 6.1.13.)

   Scripts MUST be prepared to handled URL-encoded values in
   metavariables. 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 metavariable 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

Coar, et al.                 INTERNET-DRAFT                    [Page 22]

                                 CGI/1.1       Expires: 14 October, 1999

   not intend processing the PATH_INFO data, then it SHOULD
   reject the request with '404 Not Found' if PATH_INFO is not

   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.

   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
   "<BASE>" 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

Coar, et al.                 INTERNET-DRAFT                    [Page 23]

                                 CGI/1.1       Expires: 14 October, 1999

   providing request metadata to CGI scripts.

   For Unix compatible operating systems, the following are

   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 metavariable 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
   metavariable without careful consideration. This also applies
   to the Proxy-Authorization header field and the corresponding

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

Coar, et al.                 INTERNET-DRAFT                    [Page 24]

                                 CGI/1.1       Expires: 14 October, 1999

   process. It SHOULD therefore be ensured that the script cannot
   interfere with the server process, its configuration, or

   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

          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.
          Berners-Lee, T. and Connolly, D., 'Hypertext Markup
          Language - 2.0', RFC 1866, MIT/W3C, November 1995.
          Berners-Lee, T., Fielding, R. T. and Frystyk, H.,
          'Hypertext Transfer Protocol -- HTTP/1.0', RFC 1945,
          MIT/LCS, UC Irvine, May 1996.
          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.
          Braden, R., Editor, 'Requirements for Internet Hosts --
          Application and Support', STD 3, RFC 1123, IETF,
          October 1989.
          Crocker, D.H., 'Standard for the Format of ARPA
          Internet Text Messages', STD 11, RFC 822, University of
          Delaware, August 1982.
          Fielding, R., 'Relative Uniform Resource Locators', RFC
          1808, UC Irvine, June 1995.

Coar, et al.                 INTERNET-DRAFT                    [Page 25]

                                 CGI/1.1       Expires: 14 October, 1999

          Fielding, R., Gettys, J., Mogul, J., Frystyk, H. and
          Berners-Lee, T., 'Hypertext Transfer Protocol --
          HTTP/1.1', RFC 2068, UC Irvine, DEC, MIT/LCS, January
          Freed, N. and Borenstein N., 'Multipurpose Internet
          Mail Extensions (MIME) Part Two: Media Types', RFC
          2046, Innosoft, First Virtual, November 1996.
          Mockapetris, P., 'Domain Names - Concepts and
          Facilities', STD 13, RFC 1034, ISI, November 1987.
          St. Johns, M., 'Identification Protocol', RFC 1431, US
          Department of Defense, February 1993.
          '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
    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
    Tel: +44 (1223) 566926
    Fax: +44 (1223) 506288
    Email: drtr@esi.co.uk

Coar, et al.                 INTERNET-DRAFT                    [Page 26]

Html markup produced by rfcmarkup 1.129c, available from https://tools.ietf.org/tools/rfcmarkup/