INTERNET-DRAFT Ken A L Coar
draft-coar-cgi-v11-00.{html,txt} The Apache Group
D.R.T. Robinson
ESI
28 May, 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.............................................3
1.3 Specifications...........................................3
1.4 Terminology..............................................4
2 Notational Conventions and Generic Grammar..................4
Coar, et al. INTERNET-DRAFT [Page 1]
CGI/1.1 Expires: 29 November, 1998
2.1 Augmented BNF............................................4
2.2 Basic Rules..............................................5
3 Protocol Parameters.........................................5
3.1 URL Encoding.............................................5
3.2 The Script URI...........................................6
4 Request Metadata (Meta-Variables)...........................6
4.1 AUTH_TYPE................................................7
4.2 CONTENT_LENGTH...........................................7
4.3 CONTENT_TYPE.............................................7
4.4 GATEWAY_INTERFACE........................................8
4.5 HTTP_*...................................................8
4.6 PATH_INFO................................................9
4.7 PATH_TRANSLATED..........................................9
4.8 QUERY_STRING............................................10
4.9 REMOTE_ADDR.............................................10
4.10 REMOTE_HOST............................................10
4.11 REMOTE_IDENT...........................................10
4.12 REMOTE_USER............................................11
4.13 REQUEST_METHOD.........................................11
4.14 SCRIPT_NAME............................................11
4.15 SERVER_NAME............................................11
4.16 SERVER_PORT............................................12
4.17 SERVER_PROTOCOL........................................12
4.18 SERVER_SOFTWARE........................................12
5 Invoking the Script........................................12
6 The CGI Script Command Line................................12
7 Data Input to the CGI Script...............................13
8 Data Output from the CGI Script............................13
8.1 Non-Parsed Header Output................................13
8.2 Parsed Header Output....................................14
9 Requirements for Servers...................................16
10 Recommendations for Scripts...............................17
11 System Specifications.....................................17
11.1 AmigaDOS...............................................17
11.2 Unix...................................................17
12 Security Considerations...................................18
12.1 Safe Methods...........................................18
12.2 HTTP Header Fields Containing Sensitive Information....18
12.3 Script Interference with the Server....................18
13 Acknowledgments...........................................19
14 References................................................19
15 Authors' Addresses........................................20
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.
Coar, et al. INTERNET-DRAFT [Page 2]
CGI/1.1 Expires: 29 November, 1998
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
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.
Coar, et al. INTERNET-DRAFT [Page 3]
CGI/1.1 Expires: 29 November, 1998
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.
"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.
Coar, et al. INTERNET-DRAFT [Page 4]
CGI/1.1 Expires: 29 November, 1998
[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"
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*
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 2.2 of RFC 1738
[4]. In a URL encoded string an escape sequence consists of a percent
character ("%") followed by two hexadecimal digits, where the two
hexadecimal digits form an octet. An escape sequence represents the
graphic character which has the octet as its code within the US-ASCII
[12] coded character set, if it exists. If no such graphic character
exists, then the escape sequence represents the octet value itself.
Coar, et al. INTERNET-DRAFT [Page 5]
CGI/1.1 Expires: 29 November, 1998
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.
Note that some unsafe characters may have different semantics if they
are encoded. The definition of which characters are unsafe depends on
the context. See section 2.2 of RFC 1738 [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 below);
script-uri = protocol "://" SERVER_NAME ":" SERVER_PORT enc-script
enc-path-info "?" QUERY_STRING
where 'protocol' is found 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.
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". There is no
way in CGI/1.1 for the script to reconstruct this, and therefore the
Script-URI includes the base protocol used.
4. Request Metadata (Meta-Variables)
Each CGI 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
Coar, et al. INTERNET-DRAFT [Page 6]
CGI/1.1 Expires: 29 November, 1998
defined; for a particular system the representation MAY be defined
differently than this.
The variables 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
4.1. AUTH_TYPE
This variable is specific to requests made with HTTP.
If the script URI would require access authentication for external
access, then this variable is found from the 'auth-scheme' token in
the request, otherwise 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 case-sensitive.
4.2. CONTENT_LENGTH
The size of the entity attached to the request, if any, in decimal
number of octets. If no data are attached, then NULL. The syntax is
the same as the HTTP Content-Length header field (section 14.14,
HTTP/1.1 specification [8]).
CONTENT_LENGTH = "" | 1*digit
4.3. CONTENT_TYPE
The Internet Media Type [9] of the attached entity. The syntax is the
same as the HTTP Content-Type header field.
Coar, et al. INTERNET-DRAFT [Page 7]
CGI/1.1 Expires: 29 November, 1998
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 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 either a 406 ("Not Acceptable") or 415
("Unsupported Media Type") error.
4.4. GATEWAY_INTERFACE
The version of the CGI specification to which this server complies.
Syntax:
GATEWAY_INTERFACE = "CGI" "/" 1*digit "." 1*digit
Note that the major and minor numbers are treated as separate
integers and hence each may be incremented higher 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 MUST be ignored by scripts and
SHOULD NOT be generated by servers.
This document defines the 1.1 version of the CGI interface.
4.5. HTTP_*
These variables are specific to requests made with HTTP.
Interpretation of these variables depends on the value of
SERVER_PROTOCOL.
Meta-variables with names beginning with "HTTP_" contain header data
read from the client, if the protocol used was HTTP. The HTTP header
field name is converted to upper case, has all occurrences of "-"
replaced with "_" and has "HTTP_" prepended to give 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 they MUST be
Coar, et al. INTERNET-DRAFT [Page 8]
CGI/1.1 Expires: 29 November, 1998
rewritten 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.
The server is not required to create meta-variables for all the
header fields that it receives. In particular, it MAY remove any
header fields carrying authentication information, such as
"Authorization"; and it MAY remove header fields whose value is
available to the script via other variables, such as "Content-Length"
and "Content-Type".
4.6. PATH_INFO
A path to be interpreted by the CGI script. It identifies the source
or sub-resource to be returned by the CGI script. The syntax and
semantics are similar to a decoded HTTP URL 'hpath' token (defined in
RFC 1738 [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 that follows the SCRIPT_NAME part of the path.
4.7. PATH_TRANSLATED
The OS path to the file that the server would attempt to access were
the client to request the absolute URL containing the path PATH_INFO.
I.e., for a request of
protocol "://" SERVER_NAME ":" SERVER_PORT enc-path-info
where 'enc-path-info' is a URL-encoded version of PATH_INFO. If
PATH_INFO is NULL then PATH_TRANSLATED is set to NULL.
PATH_TRANSLATED = *CHAR
PATH_TRANSLATED need not be supported by the server. The server may
choose to set PATH_TRANSLATED to NULL for reasons of security, or
because the path would not be interpretable by a CGI script; such as
the object it represented was internal to the server and not visible
in the file-system; or for any other reason.
The algorithm the server uses to derive PATH_TRANSLATED is obviously
implementation defined; CGI scripts which use this variable may
Coar, et al. INTERNET-DRAFT [Page 9]
CGI/1.1 Expires: 29 November, 1998
suffer limited portability.
4.8. QUERY_STRING
A URL-encoded search string; the part of the script URI.
QUERY_STRING = query-string
query-string = *qchar
qchar = unreserved | escape | reserved
unreserved = alpha | digit | safe | extra
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "="
safe = "$" | "-" | "_" | "." | "+"
extra = "!" | "*" | "'" | "(" | ")" | ","
escape = "%" hex hex
hex = digit | "A" | "B" | "C" | "D" | "E" | "F" | "a"
| "b" | "c" | "d" | "e" | "f"
The URL syntax for a search string is described in RFC 1738 [4].
4.9. REMOTE_ADDR
The IP address of the agent sending the request to the server. This
is not necessarily that of the client.
REMOTE_ADDR = hostnumber
hostnumber = digits "." digits "." digits "." digits
digits = 1*digit
4.10. REMOTE_HOST
The fully qualified domain name of the agent sending the request to
the server, if available, otherwise NULL. Not necessarily that of the
client. 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]; a
sequence of domain labels separated by ".", each domain label
starting and ending with an alphanumerical character and possibly
also containing "-" characters. The rightmost domain label will never
start with a digit. Domain names are not case sensitive.
REMOTE_HOST = "" | hostname
hostname = *( domainlabel ".") toplabel
domainlabel = alphadigit [ *alphahypdigit alphadigit ]
toplabel = alpha [ *alphahypdigit alphadigit ]
alphahypdigit = alphadigit | "-"
alphadigit = alpha | digit
4.11. REMOTE_IDENT
The identity information reported about the connection by a RFC 1413
[11] request to the remote agent, if available. The server MAY choose
not to support this feature, or not to request the data for
Coar, et al. INTERNET-DRAFT [Page 10]
CGI/1.1 Expires: 29 November, 1998
efficiency reasons.
REMOTE_IDENT = *CHAR
The data returned are not appropriate for use as authentication
information.
4.12. REMOTE_USER
This variable is specific to requests made with HTTP.
If AUTH_TYPE is "Basic", then the user-ID sent by the client. If
AUTH_TYPE is NULL, then NULL, otherwise undefined.
REMOTE_USER = "" | userid | *OCTET
userid = token
4.13. REQUEST_METHOD
This variable is specific to requests made with HTTP.
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].
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.
4.14. SCRIPT_NAME
A URL path that could identify the CGI script (rather than the
particular CGI output). The syntax and semantics are identical to a
decoded HTTP URL 'hpath' token [4].
SCRIPT_NAME = "" | ( "/" [ path ] )
The leading "/" is not part of the path. It is optional if the path
is NULL.
The SCRIPT_NAME string is some leading part of the component
of the script URI derived in some implementation defined manner.
4.15. SERVER_NAME
Coar, et al. INTERNET-DRAFT [Page 11]
CGI/1.1 Expires: 29 November, 1998
The name for this server, as used in the part of the script
URI. Thus either a fully qualified domain name, or an IP address.
SERVER_NAME = hostname | hostnumber
4.16. SERVER_PORT
The port on which this request was received, as used in the
part of the script URI.
SERVER_PORT = 1*digit
4.17. SERVER_PROTOCOL
The name and revision of the information protocol with which this
request arrived. This is not necessarily the same as the protocol
version used by the server in its response.
SERVER_PROTOCOL = HTTP-Version | extension-version
HTTP-Version = "HTTP" "/" 1*digit "." 1*digit
extension-version = protocol "/" 1*digit "." 1*digit
protocol = 1*( alpha | digit | "+" | "-" | "." )
'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. By convention, 'protocol' is in upper case.
4.18. SERVER_SOFTWARE
The name and version of the information server software answering the
request (and running the gateway).
SERVER_SOFTWARE = *CHAR
5. Invoking the Script
This 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.
6. 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
search string not containing any unencoded "=" characters. For such a
request, the server SHOULD parse the search string into words, using
the rules:
Coar, et al. INTERNET-DRAFT [Page 12]
CGI/1.1 Expires: 29 November, 1998
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.
7. Data Input to the CGI Script
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.
There MUST be at least CONTENT_LENGTH bytes available for the script
to read if CONTENT_LENGTH is not NULL. The script is not obliged to
read the data, but it MUST NOT attempt to read more than
CONTENT_LENGTH bytes, even if more data are available.
For non-parsed header (NPH) scripts (see below), the server 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. 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 the script can give; non-parsed
header (NPH) output, and parsed header output. A server is only
required to support the latter; distinguishing between the two types
of output (or scripts) is implementation defined.
8.1. Non-Parsed Header Output
The script MUST return a complete HTTP response message, as described
in Section 6 of the HTTP specifications [3],[8]. The script MUST use
the SERVER_PROTOCOL variable to determine the appropriate format for
a response.
Coar, et al. INTERNET-DRAFT [Page 13]
CGI/1.1 Expires: 29 November, 1998
The server SHOULD attempt to ensure that the script output is sent
directly to the client, with minimal internal and no
transport-visible buffering.
8.2. Parsed Header Output
The script returns a CGI response message as follows:
CGI-Response = *( CGI-Header | HTTP-Header ) NL [ Entity-Body ]
CGI-Header = Content-type
| Location
| Status
| extension-header
The response comprises a header and a body, separated by a blank
line. The header fields are either CGI header fields to be
interpreted by the server, or HTTP headers to be included in the
response returned to the client if the request method is HTTP. At
least one CGI-Header MUST be supplied, but no CGI header field can be
repeated with the same field-name. If a body is supplied, then a
Content-type header field is required, otherwise the script MUST send
a Location or Status header field. If a Location CGI-header field is
returned, then the script MUST NOT supply any HTTP-Headers.
All header fields occurring in a CGI-Response MUST be specified one
per line; CGI/1.1 makes no provision for continuation lines.
The CGI header fields have the generic syntax:
generic-header = 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.
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-Header rather than a CGI-header
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
Coar, et al. INTERNET-DRAFT [Page 14]
CGI/1.1 Expires: 29 November, 1998
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
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
protocol "://" SERVER_NAME ":" SERVER_PORT rel-URL-abs-path
The location header field MUST only be sent if the
REQUEST_METHOD is HEAD or GET.
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"
Coar, et al. INTERNET-DRAFT [Page 15]
CGI/1.1 Expires: 29 November, 1998
in the header data returned to the server.
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]). 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.
9. Requirements for Servers
Servers MUST support the standard mechanism (described below) which
allows the script author 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 value for SCRIPT_NAME is governed by the server configuration and
the location of the script in the OS file-system. Given this, any
access to the partial URL
SCRIPT_NAME extra-path ? query-information
where extra-path is either NULL or begins with a "/" and satisfies
any other server requirements, will cause the CGI script to be
executed with PATH_INFO set to the decoded extra-path, and
QUERY_STRING set to query-information (not decoded).
Servers MAY reject with error 404 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 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
Coar, et al. INTERNET-DRAFT [Page 16]
CGI/1.1 Expires: 29 November, 1998
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).
10. Recommendations for Scripts
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 reject the request with
404 Not Found if PATH_INFO is not NULL.
If the output of a form is being processed, check that CONTENT_TYPE
is "application/x-www-form-urlencoded" [2].
If parsing PATH_INFO, PATH_TRANSLATED or SCRIPT_NAME then 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.
It is very unlikely that any other use could be made of these.
As it is impossible for the script to determine the client URI that
initiated this 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, the script SHOULD try to send the CGI
header fields as soon as possible, and preferably before any HTTP
header fields. This may help reduce the server's memory requirements.
11. System Specifications
11.1. 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.
11.2. Unix
Coar, et al. INTERNET-DRAFT [Page 17]
CGI/1.1 Expires: 29 November, 1998
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.
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.
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.
12. Security Considerations
12.1. 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.
12.2. 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.
12.3. 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 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
Coar, et al. INTERNET-DRAFT [Page 18]
CGI/1.1 Expires: 29 November, 1998
SHOULD be taken to protect the core memory of the server, or to
ensure that untrusted code cannot be executed.
13. 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.
14. 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., Masinter, L. and McCahill, M., Editors,
'Uniform Resource Locators (URL)', RFC 1738, CERN, Xerox
Corporation, University of Minnesota, December 1994.
[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
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
Coar, et al. INTERNET-DRAFT [Page 19]
CGI/1.1 Expires: 29 November, 1998
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.
15. Authors' Addresses
Ken A L Coar
MeepZor Consulting
26B Bay Ridge Drive
Nashua, NH 03062
U.S.A.
Tel: +1 (603) 891.2243
Fax: not available
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 20]