Copyright © 2002 The Internet Society & W3C® (MIT, INRIA, Keio), All Rights Reserved. W3C liability, trademark, docum ent use and software licensing rules apply.
This document specifies XML digital signature processing rules and syntax. XML Signatures provide integrity, message authentication, and/or signer authentication services for data of any type, whether located within the XML that includes the signature or elsewhere.
This document has been reviewed by W3C Members and other interested parties and has been endorsed by the Director as a W3C Recommendation. It is a stable document and may be used as reference material or cited as a normative reference from another document. W3C's role in making the Recommendation is to draw attention to the specification and to promote its widespread deployment. This enhances the functionality and interoperability of the Web.
This specification was produced by the IETF/W3C XML Signature Working Group (W3C Activity Statement) which believes the specification is sufficient for the creation of independent interoperable implementations; the Interoperability Report shows at least 10 implementations with at least two interoperable implementations over every feature.
Patent disclosures relevant to this specification may be found on the Working Group's patent disclosure page, in conformance with W3C policy, and the IETF Page of Intellectual Property Rights Notices, in conformance with IETF policy.
Please report errors in this document to w3c-ietf-xmldsig@w3.org (archive).
The list of known errors in this specification is available at http://www.w3.org/2001/10/xmldsig-errata.
The English version of this specification is the only normative version. Information about translations of this document (if any) is available http://www.w3.org/Signature/2002/02/xmldsig-translations
A list of current W3C Technical Reports can be found at http://www.w3.org/TR/.
This document specifies XML syntax and processing rules for creating and representing digital signatures. XML Signatures can be applied to any digital content (data object), including XML. An XML Signature may be applied to the content of one or more resources. Enveloped or enveloping signatures are over data within the same XML document as the signature; detached signatures are over data external to the signature element. More specifically, this specification defines an XML signature element type and an XML signature application; conformance requirements for each are specified by way of schema definitions and prose respectively. This specification also includes other useful types that identify methods for referencing collections of resources, algorithms, and keying and management information.
The XML Signature is a method of associating a key with referenced data (octets); it does not normatively specify how keys are associated with persons or institutions, nor the meaning of the data being referenced and signed. Consequently, while this specification is an important component of secure XML applications, it itself is not sufficient to address all application security/trust concerns, particularly with respect to using signed XML (or other data formats) as a basis of human-to-human communication and agreement. Such an application must specify additional key, algorithm, processing and rendering requirements. For further information, please see Security Considerations (section 8).
For readability, brevity, and historic reasons this document uses the term "signature" to generally refer to digital authentication values of all types. Obviously, the term is also strictly used to refer to authentication values that are based on public keys and that provide signer authentication. When specifically discussing authentication values based on symmetric secret key codes we use the terms authenticators or authentication codes. (See Check the Security Model, section 8.3.)
This specification provides an XML Schema [XML-schema] and DTD [XML]. The schema definition is normative.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this specification are to be interpreted as described in RFC2119 [KEYWORDS]:
"they MUST only be used where it is actually required for interoperation or to limit behavior which has potential for causing harm (e.g., limiting retransmissions)"
Consequently, we use these capitalized key words to unambiguously specify requirements over protocol and application features and behavior that affect the interoperability and security of implementations. These key words are not used (capitalized) to describe XML grammar; schema definitions unambiguously describe such requirements and we wish to reserve the prominence of these terms for the natural language descriptions of protocols and features. For instance, an XML attribute might be described as being "optional." Compliance with the Namespaces in XML specification [XML-ns] is described as "REQUIRED."
The design philosophy and requirements of this specification are addressed in the XML-Signature Requirements document [XML-Signature-RD].
No provision is made for an explicit version number in this syntax. If a future version is needed, it will use a different namespace. The XML namespace [XML-ns] URI that MUST be used by implementations of this (dated) specification is:
xmlns="http://www.w3.org/2000/09/xmldsig#"
This namespace is also used as the prefix for algorithm identifiers used by this specification. While applications MUST support XML and XML namespaces, the use of internal entities [XML] or our "dsig" XML namespace prefix and defaulting/scoping conventions are OPTIONAL; we use these facilities to provide compact and readable examples.
This specification uses Uniform Resource Identifiers [URI] to identify resources, algorithms, and semantics. The URI in the namespace declaration above is also used as a prefix for URIs under the control of this specification. For resources not under the control of this specification, we use the designated Uniform Resource Names [URN] or Uniform Resource Locators [URL] defined by its normative external specification. If an external specification has not allocated itself a Uniform Resource Identifier we allocate an identifier under our own namespace. For instance:
SignatureProperties
is identified and defined by this
specification's namespace
Finally, in order to provide for terse namespace declarations we sometimes use XML internal entities [XML] within URIs. For instance:
<?xml version='1.0'?> <!DOCTYPE Signature SYSTEM "xmldsig-core-schema.dtd" [ <!ENTITY dsig "http://www.w3.org/2000/09/xmldsig#"> ]> <Signature xmlns="&dsig;" Id="MyFirstSignature"> <SignedInfo> ...
The contributions of the following Working Group members to this specification are gratefully acknowledged:
As are the Last Call comments from the following:
This section provides an overview and examples of XML digital signature syntax. The specific processing is given in Processing Rules (section 3). The formal syntax is found in Core Signature Syntax (section 4) and Additional Signature Syntax (section 5).
In this section, an informal representation and examples are used to describe the structure of the XML signature syntax. This representation and examples may omit attributes, details and potential features that are fully explained later.
XML Signatures are applied to arbitrary digital content (data objects) via an indirection. Data
objects are digested, the resulting value is placed in an element (with other
information) and that element is then digested and cryptographically signed. XML
digital signatures are represented by the Signature
element which
has the following structure (where "?" denotes zero or one occurrence; "+"
denotes one or more occurrences; and "*" denotes zero or more occurrences):
<Signature ID?> <SignedInfo> <CanonicalizationMethod/> <SignatureMethod/> (<Reference URI? > (<Transforms>)? <DigestMethod> <DigestValue> </Reference>)+ </SignedInfo> <SignatureValue> (<KeyInfo>)? (<Object ID?>)* </Signature>
Signatures are related to data
objects via URIs [URI]. Within an XML document,
signatures are related to local data objects via fragment identifiers. Such local
data can be included within an enveloping signature or can enclose an enveloped signature. Detached signatures are over
external network resources or local data objects that reside within the same XML
document as sibling elements; in this case, the signature is neither enveloping
(signature is parent) nor enveloped (signature is child). Since a
Signature
element (and its Id
attribute value/name) may
co-exist or be combined with other elements (and their IDs) within a single XML
document, care should be taken in choosing names such that there are no
subsequent collisions that violate the ID uniqueness validity constraint [XML].
Signature
, SignedInfo
, Methods
, and
Reference
)s
The following example is a detached signature of the content of the HTML4 in XML specification.
[s01] <Signature Id="MyFirstSignature" xmlns="http://www.w3.org/2000/09/xmldsig#"> [s02] <SignedInfo> [s03] <CanonicalizationMethod Algorithm="http://www.w3.org/TR/2001/REC-xml-c14n-20010315"/> [s04] <SignatureMethod Algorithm="http://www.w3.org/2000/09/xmldsig#dsa-sha1"/> [s05] <Reference URI="http://www.w3.org/TR/2000/REC-xhtml1-20000126/"> [s06] <Transforms> [s07] <Transform Algorithm="http://www.w3.org/TR/2001/REC-xml-c14n-20010315"/> [s08] </Transforms> [s09] <DigestMethod Algorithm="http://www.w3.org/2000/09/xmldsig#sha1"/> [s10] <DigestValue>j6lwx3rvEPO0vKtMup4NbeVu8nk=</DigestValue> [s11] </Reference> [s12] </SignedInfo> [s13] <SignatureValue>MC0CFFrVLtRlk=...</SignatureValue> [s14] <KeyInfo> [s15a] <KeyValue> [s15b] <DSAKeyValue> [s15c] <P>...</P><Q>...</Q><G>...</G><Y>...</Y> [s15d] </DSAKeyValue> [s15e] </KeyValue> [s16] </KeyInfo> [s17] </Signature>
[s02-12]
The required SignedInfo
element is the
information that is actually signed. Core validation of SignedInfo
consists of two
mandatory processes: validation of the signature over SignedInfo
and
validation of each
Reference
digest within SignedInfo
. Note that the
algorithms used in calculating the SignatureValue
are also included
in the signed information while the SignatureValue
element is
outside SignedInfo
.
[s03]
The CanonicalizationMethod
is the algorithm that
is used to canonicalize the SignedInfo
element before it is digested
as part of the signature operation. Note that this example, and all examples in
this specification, are not in canonical form.
[s04]
The SignatureMethod
is the algorithm that is used
to convert the canonicalized SignedInfo
into the
SignatureValue
. It is a combination of a digest algorithm and a key
dependent algorithm and possibly other algorithms such as padding, for example
RSA-SHA1. The algorithm names are signed to resist attacks based on substituting
a weaker algorithm. To promote application interoperability we specify a set of
signature algorithms that MUST be implemented, though their use is at the
discretion of the signature creator. We specify additional algorithms as
RECOMMENDED or OPTIONAL for implementation; the design also permits arbitrary
user specified algorithms.
[s05-11]
Each Reference
element includes the digest
method and resulting digest value calculated over the identified data object. It
also may include transformations that produced the input to the digest operation.
A data object is signed by computing its digest value and a signature over that
value. The signature is later checked via reference and signature validation.
[s14-16]
KeyInfo
indicates the key to be used to
validate the signature. Possible forms for identification include certificates,
key names, and key agreement algorithms and information -- we define only a few.
KeyInfo
is optional for two reasons. First, the signer may not wish
to reveal key information to all document processing parties. Second, the
information may be known within the application's context and need not be
represented explicitly. Since KeyInfo
is outside of
SignedInfo
, if the signer wishes to bind the keying information to
the signature, a Reference
can easily identify and include the
KeyInfo
as part of the signature.
Reference
[s05] <Reference URI="http://www.w3.org/TR/2000/REC-xhtml1-20000126/"> [s06] <Transforms> [s07] <Transform Algorithm="http://www.w3.org/TR/2001/REC-xml-c14n-20010315"/> [s08] </Transforms> [s09] <DigestMethod Algorithm="http://www.w3.org/2000/09/xmldsig#sha1"/> [s10] <DigestValue>j6lwx3rvEPO0vKtMup4NbeVu8nk=</DigestValue> [s11] </Reference>
[s05]
The optional URI
attribute of
Reference
identifies the data object to be signed. This attribute
may be omitted on at most one Reference
in a Signature
.
(This limitation is imposed in order to ensure that references and objects may be
matched unambiguously.)
[s05-08]
This identification, along with the transforms, is a
description provided by the signer on how they obtained the signed data object in
the form it was digested (i.e. the digested content). The verifier may obtain the
digested content in another method so long as the digest verifies. In particular,
the verifier may obtain the content from a different location such as a local
store than that specified in the URI
.
[s06-08] Transforms
is an optional ordered list of processing steps
that were applied to the resource's content before it was digested. Transforms
can include operations such as canonicalization, encoding/decoding (including
compression/inflation), XSLT, XPath, XML schema validation, or XInclude. XPath
transforms permit the signer to derive an XML document that omits portions of the
source document. Consequently those excluded portions can change without
affecting signature validity. For example, if the resource being signed encloses
the signature itself, such a transform must be used to exclude the signature
value from its own computation. If no Transforms
element is present,
the resource's content is digested directly. While the Working Group has
specified mandatory (and optional) canonicalization and decoding algorithms, user
specified transforms are permitted.
[s09-10] DigestMethod
is the algorithm applied to the data after
Transforms
is applied (if specified) to yield the
DigestValue
. The signing of the DigestValue
is what
binds a resources content to the signer's key.
Object
and SignatureProperty
)
This specification does not address mechanisms for making statements or
assertions. Instead, this document defines what it means for something to be
signed by an XML Signature (integrity, message authentication, and/or signer authentication).
Applications that wish to represent other semantics must rely upon other
technologies, such as [XML, RDF].
For instance, an application might use a foo:assuredby
attribute
within its own markup to reference a Signature
element.
Consequently, it's the application that must understand and know how to make
trust decisions given the validity of the signature and the meaning of
assuredby
syntax. We also define a SignatureProperties
element type for the inclusion of assertions about the signature itself (e.g.,
signature semantics, the time of signing or the serial number of hardware used in
cryptographic processes). Such assertions may be signed by including a
Reference
for the SignatureProperties
in
SignedInfo
. While the signing application should be very careful
about what it signs (it should understand what is in the
SignatureProperty
) a receiving application has no obligation to
understand that semantic (though its parent trust engine may wish to). Any
content about the signature generation may be located within the
SignatureProperty
element. The mandatory Target
attribute references the Signature
element to which the property
applies.
Consider the preceding example with an additional reference to a local
Object
that includes a SignatureProperty
element. (Such
a signature would not only be detached [p02]
but enveloping
[p03]
.)
[ ] <Signature Id="MySecondSignature" ...> [p01] <SignedInfo> [ ] ... [p02] <Reference URI="http://www.w3.org/TR/xml-stylesheet/"> [ ] ... [p03] <Reference URI="#AMadeUpTimeStamp" [p04] Type="http://www.w3.org/2000/09/xmldsig#SignatureProperties"> [p05] <DigestMethod Algorithm="http://www.w3.org/2000/09/xmldsig#sha1"/> [p06] <DigestValue>k3453rvEPO0vKtMup4NbeVu8nk=</DigestValue> [p07] </Reference> [p08] </SignedInfo> [p09] ... [p10] <Object> [p11] <SignatureProperties> [p12] <SignatureProperty Id="AMadeUpTimeStamp" Target="#MySecondSignature"> [p13] <timestamp xmlns="http://www.ietf.org/rfcXXXX.txt"> [p14] <date>19990908</date> [p15] <time>14:34:34:34</time> [p16] </timestamp> [p17] </SignatureProperty> [p18] </SignatureProperties> [p19] </Object> [p20]</Signature>
[p04]
The optional Type
attribute of
Reference
provides information about the resource identified by the
URI
. In particular, it can indicate that it is an
Object
, SignatureProperty
, or Manifest
element. This can be used by applications to initiate special processing of some
Reference
elements. References to an XML data element within an
Object
element SHOULD identify the actual element pointed to. Where
the element content is not XML (perhaps it is binary or encoded data) the
reference should identify the Object
and the Reference
Type
, if given, SHOULD indicate Object
. Note that
Type
is advisory and no action based on it or checking of its
correctness is required by core behavior.
[p10]
Object
is an optional element for including data
objects within the signature element or elsewhere. The Object
can be
optionally typed and/or encoded.
[p11-18]
Signature properties, such as time of signing, can be
optionally signed by identifying them from within a Reference
.
(These properties are traditionally called signature "attributes" although that
term has no relationship to the XML term "attribute".)
Object
and Manifest
)
The Manifest
element is provided to meet additional requirements not
directly addressed by the mandatory parts of this specification. Two requirements
and the way the Manifest
satisfies them follow.
First, applications frequently need to efficiently sign multiple data objects
even where the signature operation itself is an expensive public key signature.
This requirement can be met by including multiple Reference
elements
within SignedInfo
since the inclusion of each digest secures the
data digested. However, some applications may not want the core validation behavior
associated with this approach because it requires every Reference
within SignedInfo
to undergo reference validation -- the DigestValue
elements are checked. These applications may wish to reserve reference validation
decision logic to themselves. For example, an application might receive a signature valid
SignedInfo
element that includes three Reference
elements. If a single Reference
fails (the identified data object
when digested does not yield the specified DigestValue
) the
signature would fail core
validation. However, the application may wish to treat the signature over the
two valid Reference
elements as valid or take different actions
depending on which fails. To accomplish this, SignedInfo
would
reference a Manifest
element that contains one or more
Reference
elements (with the same structure as those in
SignedInfo
). Then, reference validation of the Manifest
is under application control.
Second, consider an application where many signatures (using different keys) are
applied to a large number of documents. An inefficient solution is to have a
separate signature (per key) repeatedly applied to a large
SignedInfo
element (with many Reference
s); this is
wasteful and redundant. A more efficient solution is to include many references
in a single Manifest
that is then referenced from multiple
Signature
elements.
The example below includes a Reference
that signs a
Manifest
found within the Object
element.
[ ] ... [m01] <Reference URI="#MyFirstManifest" [m02] Type="http://www.w3.org/2000/09/xmldsig#Manifest"> [m03] <DigestMethod Algorithm="http://www.w3.org/2000/09/xmldsig#sha1"/> [m04] <DigestValue>345x3rvEPO0vKtMup4NbeVu8nk=</DigestValue> [m05] </Reference> [ ] ... [m06] <Object> [m07] <Manifest Id="MyFirstManifest"> [m08] <Reference> [m09] ... [m10] </Reference> [m11] <Reference> [m12] ... [m13] </Reference> [m14] </Manifest> [m15] </Object>
The sections below describe the operations to be performed as part of signature generation and validation.
The REQUIRED steps include the generation of Reference
elements and
the SignatureValue
over SignedInfo
.
For each data object being signed:
Transforms
, as determined by the application, to the
data object.
Reference
element, including the (optional)
identification of the data object, any (optional) transform elements, the
digest algorithm and the DigestValue
. (Note, it is the canonical
form of these references that are signed in 3.1.2 and validated in 3.2.1 .)
SignedInfo
element with SignatureMethod
,
CanonicalizationMethod
and Reference
(s).
SignatureValue
over
SignedInfo
based on algorithms specified in
SignedInfo
.
Signature
element that includes
SignedInfo
, Object
(s) (if desired, encoding may be
different than that used for signing), KeyInfo
(if required), and
SignatureValue
.
Note, if the Signature
includes same-document references, [XML] or [XML-schema]
validation of the document might introduce changes that break the signature.
Consequently, applications should be careful to consistently process the
document or refrain from using external contributions (e.g., defaults and
entities).
The REQUIRED steps of core
validation include (1) reference validation, the verification of the digest
contained in each Reference
in SignedInfo
, and (2) the
cryptographic signature
validation of the signature calculated over SignedInfo
.
Note, there may be valid signatures that some signature applications are unable to validate. Reasons for this include failure to implement optional parts of this specification, inability or unwillingness to execute specified algorithms, or inability or unwillingness to dereference specified URIs (some URI schemes may cause undesirable side effects), etc.
Comparison of values in reference and signature validation are over the numeric (e.g., integer) or decoded octet sequence of the value. Different implementations may produce different encoded digest and signature values when processing the same resources because of variances in their encoding, such as accidental white space. But if one uses numeric or octet comparison (choose one) on both the stated and computed values these problems are eliminated.
SignedInfo
element based on the
CanonicalizationMethod
in SignedInfo
.
Reference
in SignedInfo
:
URI
and execute
Transforms
provided by the signer in the
Reference
element, or it may obtain the content through other
means such as a local cache.)
DigestMethod
specified in its Reference
specification.
DigestValue
in the
SignedInfo
Reference
; if there is any mismatch,
validation fails.
Note, SignedInfo
is canonicalized in step 1. The application must
ensure that the CanonicalizationMethod has no dangerous side affects, such as
rewriting URIs, (see CanonicalizationMethod
(section 4.3)) and that it Sees What is Signed, which is
the canonical form.
KeyInfo
or from an external source.
SignatureMethod
using the
CanonicalizationMethod
and use the result (and previously
obtained KeyInfo
) to confirm the SignatureValue
over
the SignedInfo
element.
Note, KeyInfo
(or some transformed
version thereof) may be signed via a Reference
element.
Transformation and validation of this reference (3.2.1) is orthogonal to
Signature Validation which uses the KeyInfo
as parsed.
Additionally, the SignatureMethod
URI may have been altered by the
canonicalization of SignedInfo
(e.g., absolutization of relative
URIs) and it is the canonical form that MUST be used. However, the required
canonicalization [XML-C14N] of this specification
does not change URIs.
The general structure of an XML signature is described in Signature Overview (section 2). This section provides detailed syntax of the core signature features. Features described in this section are mandatory to implement unless otherwise indicated. The syntax is defined via DTDs and [XML-Schema] with the following XML preamble, declaration, and internal entity.
Schema Definition: <?xml version="1.0" encoding="utf-8"?> <!DOCTYPE schema PUBLIC "-//W3C//DTD XMLSchema 200102//EN" "http://www.w3.org/2001/XMLSchema.dtd" [ <!ATTLIST schema xmlns:ds CDATA #FIXED "http://www.w3.org/2000/09/xmldsig#"> <!ENTITY dsig 'http://www.w3.org/2000/09/xmldsig#'> <!ENTITY % p ''> <!ENTITY % s ''> ]> <schema xmlns="http://www.w3.org/2001/XMLSchema" xmlns:ds="http://www.w3.org/2000/09/xmldsig#" targetNamespace="http://www.w3.org/2000/09/xmldsig#" version="0.1" elementFormDefault="qualified">
DTD: <!-- The following entity declarations enable external/flexible content in the Signature content model. #PCDATA emulates schema:string; when combined with element types it emulates schema mixed="true". %foo.ANY permits the user to include their own element types from other namespaces, for example: <!ENTITY % KeyValue.ANY '| ecds:ECDSAKeyValue'> ... <!ELEMENT ecds:ECDSAKeyValue (#PCDATA) > --> <!ENTITY % Object.ANY ''> <!ENTITY % Method.ANY ''> <!ENTITY % Transform.ANY ''> <!ENTITY % SignatureProperty.ANY ''> <!ENTITY % KeyInfo.ANY ''> <!ENTITY % KeyValue.ANY ''> <!ENTITY % PGPData.ANY ''> <!ENTITY % X509Data.ANY ''> <!ENTITY % SPKIData.ANY ''>
This specification defines the ds:CryptoBinary
simple type for
representing arbitrary-length integers (e.g. "bignums") in XML as octet strings.
The integer value is first converted to a "big endian" bitstring. The bitstring
is then padded with leading zero bits so that the total number of bits == 0 mod 8
(so that there are an integral number of octets). If the bitstring contains
entire leading octets that are zero, these are removed (so the high-order octet
is always non-zero). This octet string is then base64 [MIME] encoded. (The conversion from integer to octet string
is equivalent to IEEE 1363's I2OSP [1363] with minimal
length).
This type is used by "bignum" values such as RSAKeyValue
and
DSAKeyValue
. If a value can be of type base64Binary
or
ds:CryptoBinary
they are defined as base64Binary
.
For example, if the signature algorithm is RSA or DSA then
SignatureValue
represents a bignum and could be
ds:CryptoBinary
. However, if HMAC-SHA1 is the signature algorithm
then SignatureValue
could have leading zero octets that must be
preserved. Thus SignatureValue
is generically defined as of type
base64Binary
.
Schema Definition: <simpleType name="CryptoBinary"> <restriction base="base64Binary"> </restriction> </simpleType>
Signature
element
The Signature
element is the root element of an XML Signature.
Implementation MUST generate laxly
schema valid [XML-schema]
Signature
elements as specified by the following schema:
Schema Definition: <element name="Signature" type="ds:SignatureType"/> <complexType name="SignatureType"> <sequence> <element ref="ds:SignedInfo"/> <element ref="ds:SignatureValue"/> <element ref="ds:KeyInfo" minOccurs="0"/> <element ref="ds:Object" minOccurs="0" maxOccurs="unbounded"/> </sequence> <attribute name="Id" type="ID" use="optional"/> </complexType>
DTD: <!ELEMENT Signature (SignedInfo, SignatureValue, KeyInfo?, Object*) > <!ATTLIST Signature xmlns CDATA #FIXED 'http://www.w3.org/2000/09/xmldsig#' Id ID #IMPLIED >
SignatureValue
Element
The SignatureValue
element contains the actual value of the digital
signature; it is always encoded using base64 [MIME].
While we identify two SignatureMethod
algorithms, one mandatory and
one optional to implement, user specified algorithms may be used as well.
Schema Definition: <element name="SignatureValue" type="ds:SignatureValueType"/> <complexType name="SignatureValueType"> <simpleContent> <extension base="base64Binary"> <attribute name="Id" type="ID" use="optional"/> </extension> </simpleContent> </complexType>
DTD: <!ELEMENT SignatureValue (#PCDATA) > <!ATTLIST SignatureValue Id ID #IMPLIED>
SignedInfo
Element
The structure of SignedInfo
includes the canonicalization algorithm,
a signature algorithm, and one or more references. The SignedInfo
element may contain an optional ID attribute that will allow it to be referenced
by other signatures and objects.
SignedInfo
does not include explicit signature or digest properties
(such as calculation time, cryptographic device serial number, etc.). If an
application needs to associate properties with the signature or digest, it may
include such information in a SignatureProperties
element within an
Object
element.
Schema Definition: <element name="SignedInfo" type="ds:SignedInfoType"/> <complexType name="SignedInfoType"> <sequence> <element ref="ds:CanonicalizationMethod"/> <element ref="ds:SignatureMethod"/> <element ref="ds:Reference" maxOccurs="unbounded"/> </sequence> <attribute name="Id" type="ID" use="optional"/> </complexType>
DTD: <!ELEMENT SignedInfo (CanonicalizationMethod, SignatureMethod, Reference+) > <!ATTLIST SignedInfo Id ID #IMPLIED
CanonicalizationMethod
Element
CanonicalizationMethod
is a required element that specifies the
canonicalization algorithm applied to the SignedInfo
element prior
to performing signature calculations. This element uses the general structure for
algorithms described in Algorithm Identifiers and
Implementation Requirements (section 6.1). Implementations MUST support the
REQUIRED canonicalization algorithms.
Alternatives to the REQUIRED canonicalization algorithms (section 6.5), such as Canonical XML with Comments (section 6.5.1) or a minimal canonicalization (such as CRLF and charset normalization), may be explicitly specified but are NOT REQUIRED. Consequently, their use may not interoperate with other applications that do not support the specified algorithm (see XML Canonicalization and Syntax Constraint Considerations, section 7). Security issues may also arise in the treatment of entity processing and comments if non-XML aware canonicalization algorithms are not properly constrained (see section 8.2: Only What is "Seen" Should be Signed).
The way in which the SignedInfo
element is presented to the
canonicalization method is dependent on that method. The following applies to
algorithms which process XML as nodes or characters:
SignedInfo
and currently indicating the
SignedInfo
, its descendants, and the attribute and namespace nodes
of SignedInfo
and its descendant elements.
We recommend applications that implement a text-based instead of XML-based canonicalization -- such as resource constrained apps -- generate canonicalized XML as their output serialization so as to mitigate interoperability and security concerns. For instance, such an implementation SHOULD (at least) generate standalone XML instances [XML].
NOTE: The signature application must
exercise great care in accepting and executing an arbitrary
CanonicalizationMethod
. For example, the canonicalization method
could rewrite the URIs of the Reference
s being validated. Or, the
method could massively transform SignedInfo
so that validation would
always succeed (i.e., converting it to a trivial signature with a known key over
trivial data). Since CanonicalizationMethod
is inside
SignedInfo
, in the resulting canonical form it could erase itself
from SignedInfo
or modify the SignedInfo
element so
that it appears that a different canonicalization function was used! Thus a
Signature
which appears to authenticate the desired data with the
desired key, DigestMethod
, and SignatureMethod
, can be
meaningless if a capricious CanonicalizationMethod
is used.
Schema Definition: <element name="CanonicalizationMethod" type="ds:CanonicalizationMethodType"/> <complexType name="CanonicalizationMethodType" mixed="true"> <sequence> <any namespace="##any" minOccurs="0" maxOccurs="unbounded"/> <!-- (0,unbounded) elements from (1,1) namespace --> </sequence> <attribute name="Algorithm" type="anyURI" use="required"/> </complexType>
DTD: <!ELEMENT CanonicalizationMethod (#PCDATA %Method.ANY;)* > <!ATTLIST CanonicalizationMethod Algorithm CDATA #REQUIRED >
SignatureMethod
Element
SignatureMethod
is a required element that specifies the algorithm
used for signature generation and validation. This algorithm identifies all
cryptographic functions involved in the signature operation (e.g. hashing, public
key algorithms, MACs, padding, etc.). This element uses the general structure
here for algorithms described in section 6.1: Algorithm
Identifiers and Implementation Requirements. While there is a single
identifier, that identifier may specify a format containing multiple distinct
signature values.
Schema Definition: <element name="SignatureMethod" type="ds:SignatureMethodType"/> <complexType name="SignatureMethodType" mixed="true"> <sequence> <element name="HMACOutputLength" minOccurs="0" type="ds:HMACOutputLengthType"/> <any namespace="##other" minOccurs="0" maxOccurs="unbounded"/> <!-- (0,unbounded) elements from (1,1) external namespace --> </sequence> <attribute name="Algorithm" type="anyURI" use="required"/> </complexType>
DTD: <!ELEMENT SignatureMethod (#PCDATA|HMACOutputLength %Method.ANY;)* > <!ATTLIST SignatureMethod Algorithm CDATA #REQUIRED >
Reference
Element
Reference
is an element that may occur one or more times. It
specifies a digest algorithm and digest value, and optionally an identifier of
the object being signed, the type of the object, and/or a list of transforms to
be applied prior to digesting. The identification (URI) and transforms describe
how the digested content (i.e., the input to the digest method) was created. The
Type
attribute facilitates the processing of referenced data. For
example, while this specification makes no requirements over external data, an
application may wish to signal that the referent is a Manifest
. An
optional ID attribute permits a Reference
to be referenced from
elsewhere.
Schema Definition: <element name="Reference" type="ds:ReferenceType"/> <complexType name="ReferenceType"> <sequence> <element ref="ds:Transforms" minOccurs="0"/> <element ref="ds:DigestMethod"/> <element ref="ds:DigestValue"/> </sequence> <attribute name="Id" type="ID" use="optional"/> <attribute name="URI" type="anyURI" use="optional"/> <attribute name="Type" type="anyURI" use="optional"/> </complexType>
DTD: <!ELEMENT Reference (Transforms?, DigestMethod, DigestValue) > <!ATTLIST Reference Id ID #IMPLIED URI CDATA #IMPLIED Type CDATA #IMPLIED>
URI
Attribute
The URI
attribute identifies a data object using a URI-Reference, as
specified by RFC2396 [URI]. The set of allowed characters
for URI
attributes is the same as for XML, namely [Unicode]. However, some Unicode characters are
disallowed from URI references including all non-ASCII characters and the
excluded characters listed in RFC2396 [URI, section 2.4].
However, the number sign (#), percent sign (%), and square bracket characters
re-allowed in RFC 2732 [URI-Literal] are
permitted. Disallowed characters must be escaped as follows:
XML signature applications MUST be able to parse URI syntax. We RECOMMEND they be able to dereference URIs in the HTTP scheme. Dereferencing a URI in the HTTP scheme MUST comply with the Status Code Definitions of [HTTP] (e.g., 302, 305 and 307 redirects are followed to obtain the entity-body of a 200 status code response). Applications should also be cognizant of the fact that protocol parameter and state information, (such as HTTP cookies, HTML device profiles or content negotiation), may affect the content yielded by dereferencing a URI.
If a resource is identified by more than one URI, the most specific should be used (e.g. http://www.w3.org/2000/06/interop-pressrelease.html.en instead of http://www.w3.org/2000/06/interop-pressrelease). (See the Reference Validation (section 3.2.1) for a further information on reference processing.)
If the URI
attribute is omitted altogether, the receiving
application is expected to know the identity of the object. For example, a
lightweight data protocol might omit this attribute given the identity of the
object is part of the application context. This attribute may be omitted from at
most one Reference
in any particular SignedInfo
, or
Manifest
.
The optional Type attribute contains information about the type of object being signed. This is represented as a URI. For example:
Type="http://www.w3.org/2000/09/xmldsig#Object"
Type="http://www.w3.org/2000/09/xmldsig#Manifest"
The Type attribute applies to the item being pointed at, not its contents. For
example, a reference that identifies an Object
element containing a
SignatureProperties
element is still of type #Object
.
The type attribute is advisory. No validation of the type information is required
by this specification.
Note: XPath is RECOMMENDED. Signature applications need not conform to [XPath] specification in order to conform to this specification. However, the XPath data model, definitions (e.g., node-sets) and syntax is used within this document in order to describe functionality for those that want to process XML-as-XML (instead of octets) as part of signature generation. For those that want to use these features, a conformant [XPath] implementation is one way to implement these features, but it is not required. Such applications could use a sufficiently functional replacement to a node-set and implement only those XPath expression behaviors REQUIRED by this specification. However, for simplicity we generally will use XPath terminology without including this qualification on every point. Requirements over "XPath node-sets" can include a node-set functional equivalent. Requirements over XPath processing can include application behaviors that are equivalent to the corresponding XPath behavior.
The data-type of the result of URI dereferencing or subsequent Transforms is either an octet stream or an XPath node-set.
The Transforms
specified in this document are defined with respect
to the input they require. The following is the default signature application
behavior:
Users may specify alternative transforms that override these defaults in
transitions between transforms that expect different inputs. The final octet
stream contains the data octets being secured. The digest algorithm specified by
DigestMethod
is then applied to these data octets, resulting in the
DigestValue
.
Unless the URI-Reference is a 'same-document' reference as defined in [URI, Section 4.2], the result of dereferencing the URI-Reference MUST be an octet stream. In particular, an XML document identified by URI is not parsed by the signature application unless the URI is a same-document reference or unless a transform that requires XML parsing is applied. (See Transforms (section 4.3.3.1).)
When a fragment is preceded by an absolute or relative URI in the URI-Reference,
the meaning of the fragment is defined by the resource's MIME type. Even for XML
documents, URI dereferencing (including the fragment processing) might be done
for the signature application by a proxy. Therefore, reference validation might
fail if fragment processing is not performed in a standard way (as defined in the
following section for same-document references). Consequently, we RECOMMEND that
the URI
attribute not include fragment identifiers and that
such processing be specified as an additional XPath
Transform.
When a fragment is not preceded by a URI in the URI-Reference, XML signature applications MUST support the null URI and barename XPointer. We RECOMMEND support for the same-document XPointers '#xpointer(/)' and '#xpointer(id('ID'))' if the application also intends to support any canonicalization that preserves comments. (Otherwise URI="#foo" will automatically remove comments before the canonicalization can even be invoked.) All other support for XPointers is OPTIONAL, especially all support for barename and other XPointers in external resources since the application may not have control over how the fragment is generated (leading to interoperability problems and validation failures).
The following examples demonstrate what the URI attribute identifies and how it is dereferenced:
URI="http://example.com/bar.xml"
URI="http://example.com/bar.xml#chapter1"
URI=""
URI="#chapter1"
Dereferencing a same-document reference MUST result in an XPath node-set suitable
for use by Canonical XML [XML-C14N]. Specifically,
dereferencing a null URI (URI=""
) MUST result in an XPath node-set
that includes every non-comment node of the XML document containing the
URI
attribute. In a fragment URI, the characters after the number
sign ('#') character conform to the XPointer syntax [Xptr]. When processing an XPointer, the application MUST
behave as if the root node of the XML document containing the URI
attribute were used to initialize the XPointer evaluation context. The
application MUST behave as if the result of XPointer processing were a node-set
derived from the resultant location-set as follows:
The second to last replacement is necessary because XPointer typically indicates a subtree of an XML document's parse tree using just the element node at the root of the subtree, whereas Canonical XML treats a node-set as a set of nodes in which absence of descendant nodes results in absence of their representative text from the canonical form.
The last step is performed for null URIs, barename XPointers and child sequence
XPointers. It's necessary because when [XML-C14N] is
passed a node-set, it processes the node-set as is: with or without comments.
Only when it's called with an octet stream does it invoke its own XPath
expressions (default or without comments). Therefore to retain the default
behavior of stripping comments when passed a node-set, they are removed in the
last step if the URI is not a full XPointer. To retain comments while selecting
an element by an identifier ID, use the following full XPointer:
URI='#xpointer(id('ID'))'
. To retain comments while selecting the
entire document, use the following full XPointer:
URI='#xpointer(/)'
. This XPointer contains a simple XPath expression
that includes the root node, which the second to last step above replaces with
all nodes of the parse tree (all descendants, plus all attributes, plus all
namespaces nodes).
Transforms
Element
The optional Transforms
element contains an ordered list of
Transform
elements; these describe how the signer obtained the data
object that was digested. The output of each Transform
serves as
input to the next Transform
. The input to the first
Transform
is the result of dereferencing the URI
attribute of the Reference
element. The output from the last
Transform
is the input for the DigestMethod
algorithm.
When transforms are applied the signer is not signing the native (original)
document but the resulting (transformed) document. (See Only What is Signed is Secure (section 8.1).)
Each Transform
consists of an Algorithm
attribute and
content parameters, if any, appropriate for the given algorithm. The
Algorithm
attribute value specifies the name of the algorithm to be
performed, and the Transform
content provides additional data to
govern the algorithm's processing of the transform input. (See Algorithm Identifiers and Implementation Requirements
(section 6).)
As described in The Reference Processing Model (section 4.3.3.2), some transforms take an XPath node-set as input, while others require an octet stream. If the actual input matches the input needs of the transform, then the transform operates on the unaltered input. If the transform input requirement differs from the format of the actual input, then the input must be converted.
Some Transform
s may require explicit MIME type, charset (IANA
registered "character set"), or other such information concerning the data they
are receiving from an earlier Transform
or the source data, although
no Transform
algorithm specified in this document needs such
explicit information. Such data characteristics are provided as parameters to the
Transform
algorithm and should be described in the specification for
the algorithm.
Examples of transforms include but are not limited to base64 decoding [MIME], canonicalization [XML-C14N], XPath filtering [XPath], and XSLT [XSLT]. The
generic definition of the Transform
element also allows
application-specific transform algorithms. For example, the transform could be a
decompression routine given by a Java class appearing as a base64 encoded
parameter to a Java Transform
algorithm. However, applications
should refrain from using application-specific transforms if they wish their
signatures to be verifiable outside of their application domain. Transform Algorithms (section 6.6) defines the list of
standard transformations.
Schema Definition: <element name="Transforms" type="ds:TransformsType"/> <complexType name="TransformsType"> <sequence> <element ref="ds:Transform" maxOccurs="unbounded"/> </sequence> </complexType> <element name="Transform" type="ds:TransformType"/> <complexType name="TransformType" mixed="true"> <choice minOccurs="0" maxOccurs="unbounded"> <any namespace="##other" processContents="lax"/> <!-- (1,1) elements from (0,unbounded) namespaces --> <element name="XPath" type="string"/> </choice> <attribute name="Algorithm" type="anyURI" use="required"/> </complexType>
DTD: <!ELEMENT Transforms (Transform+)> <!ELEMENT Transform (#PCDATA|XPath %Transform.ANY;)* > <!ATTLIST Transform Algorithm CDATA #REQUIRED > <!ELEMENT XPath (#PCDATA) >
DigestMethod
Element
DigestMethod
is a required element that identifies the digest
algorithm to be applied to the signed object. This element uses the general
structure here for algorithms specified in Algorithm
Identifiers and Implementation Requirements (section 6.1).
If the result of the URI dereference and application of Transforms is an XPath node-set (or sufficiently functional replacement implemented by the application) then it must be converted as described in the Reference Processing Model (section 4.3.3.2). If the result of URI dereference and application of transforms is an octet stream, then no conversion occurs (comments might be present if the Canonical XML with Comments was specified in the Transforms). The digest algorithm is applied to the data octets of the resulting octet stream.
Schema Definition: <element name="DigestMethod" type="ds:DigestMethodType"/> <complexType name="DigestMethodType" mixed="true"> <sequence> <any namespace="##other" processContents="lax" minOccurs="0" maxOccurs="unbounded"/> </sequence> <attribute name="Algorithm" type="anyURI" use="required"/> </complexType>
DTD: <!ELEMENT DigestMethod (#PCDATA %Method.ANY;)* > <!ATTLIST DigestMethod Algorithm CDATA #REQUIRED >
DigestValue
Element
DigestValue is an element that contains the encoded value of the digest. The digest is always encoded using base64 [MIME].
Schema Definition: <element name="DigestValue" type="ds:DigestValueType"/> <simpleType name="DigestValueType"> <restriction base="base64Binary"/> </simpleType>
DTD:
<!ELEMENT DigestValue (#PCDATA) >
<!-- base64 encoded digest value -->
KeyInfo
Element
KeyInfo
is an optional element that enables the recipient(s) to
obtain the key needed to validate the signature. KeyInfo
may
contain keys, names, certificates and other public key management information,
such as in-band key distribution or key agreement data. This specification
defines a few simple types but applications may extend those types or all
together replace them with their own key identification and exchange semantics
using the XML namespace facility. [XML-ns] However,
questions of trust of such key information (e.g., its authenticity or
strength) are out of scope of this specification and left to the application.
If KeyInfo
is omitted, the recipient is expected to be able to
identify the key based on application context. Multiple declarations within
KeyInfo
refer to the same key. While applications may define and use
any mechanism they choose through inclusion of elements from a different
namespace, compliant versions MUST implement KeyValue
(section 4.4.2) and SHOULD
implement RetrievalMethod
(section 4.4.3).
The schema/DTD specifications of many of KeyInfo
's children (e.g.,
PGPData
, SPKIData
, X509Data
) permit their
content to be extended/complemented with elements from another namespace. This
may be done only if it is safe to ignore these extension elements while claiming
support for the types defined in this specification. Otherwise, external
elements, including alternative structures to those defined by this
specification, MUST be a child of KeyInfo
. For example, should a
complete XML-PGP standard be defined, its root element MUST be a child of
KeyInfo
. (Of course, new structures from external namespaces can
incorporate elements from the &dsig;
namespace via features of
the type definition language. For instance, they can create a DTD that mixes
their own and dsig qualified elements, or a schema that permits, includes,
imports, or derives new types based on &dsig;
elements.)
The following list summarizes the KeyInfo
types that are allocated
an identifier in the &dsig;
namespace; these can be used within
the RetrievalMethod
Type
attribute to describe a remote
KeyInfo
structure.
In addition to the types above for which we define an XML structure, we specify one additional type to indicate a binary (ASN.1 DER) X.509 Certificate.
Schema Definition: <element name="KeyInfo" type="ds:KeyInfoType"/> <complexType name="KeyInfoType" mixed="true"> <choice maxOccurs="unbounded"> <element ref="ds:KeyName"/> <element ref="ds:KeyValue"/> <element ref="ds:RetrievalMethod"/> <element ref="ds:X509Data"/> <element ref="ds:PGPData"/> <element ref="ds:SPKIData"/> <element ref="ds:MgmtData"/> <any processContents="lax" namespace="##other"/> <!-- (1,1) elements from (0,unbounded) namespaces --> </choice> <attribute name="Id" type="ID" use="optional"/> </complexType>
DTD: <!ELEMENT KeyInfo (#PCDATA|KeyName|KeyValue|RetrievalMethod| X509Data|PGPData|SPKIData|MgmtData %KeyInfo.ANY;)* > <!ATTLIST KeyInfo Id ID #IMPLIED >
KeyName
Element
The KeyName
element contains a string value (in which white space is
significant) which may be used by the signer to communicate a key identifier to
the recipient. Typically, KeyName
contains an identifier related to
the key pair used to sign the message, but it may contain other protocol-related
information that indirectly identifies a key pair. (Common uses of
KeyName
include simple string names for keys, a key index, a
distinguished name (DN), an email address, etc.)
Schema Definition: <element name="KeyName" type="string"/>
DTD: <!ELEMENT KeyName (#PCDATA) >
KeyValue
Element
The KeyValue
element contains a single public key that may be useful
in validating the signature. Structured formats for defining DSA (REQUIRED) and
RSA (RECOMMENDED) public keys are defined in Signature Algorithms (section 6.4). The
KeyValue
element may include externally defined public keys values
represented as PCDATA or element types from an external namespace.
Schema Definition: <element name="KeyValue" type="ds:KeyValueType"/> <complexType name="KeyValueType" mixed="true"> <choice> <element ref="ds:DSAKeyValue"/> <element ref="ds:RSAKeyValue"/> <any namespace="##other" processContents="lax"/> </choice> </complexType>
DTD: <!ELEMENT KeyValue (#PCDATA|DSAKeyValue|RSAKeyValue %KeyValue.ANY;)* >
DSAKeyValue
Element
Type="http://www.w3.org/2000/09/xmldsig#DSAKeyValue"
(this can be used within a RetrievalMethod
or
Reference
element to identify the referent's type)
DSA keys and the DSA signature algorithm are specified in [DSS]. DSA public key values can have the following fields:
P
Q
G
Y
J
seed
pgenCounter
Parameter J is available for inclusion solely for efficiency as it is
calculatable from P and Q. Parameters seed and pgenCounter are used in the DSA
prime number generation algorithm specified in [DSS]. As such, they are optional
but must either both be present or both be absent. This prime generation
algorithm is designed to provide assurance that a weak prime is not being used
and it yields a P and Q value. Parameters P, Q, and G can be public and common to
a group of users. They might be known from application context. As such, they are
optional but P and Q must either both appear or both be absent. If all of
P
, Q
, seed
, and pgenCounter
are present, implementations are not required to check if they are consistent and
are free to use either P
and Q
or seed
and
pgenCounter
. All parameters are encoded as base64 [MIME] values.
Arbitrary-length integers (e.g. "bignums" such as RSA moduli) are represented in
XML as octet strings as defined by the ds:CryptoBinary
type.
Schema Definition:
<element name="DSAKeyValue" type="ds:DSAKeyValueType"/>
<complexType name="DSAKeyValueType">
<sequence>
<sequence minOccurs="0">
<element name="P" type="ds:CryptoBinary"/>
<element name="Q" type="ds:CryptoBinary"/>
</sequence>
<element name="G" type="ds:CryptoBinary" minOccurs="0"/>
<element name="Y" type="ds:CryptoBinary"/>
<element name="J" type="ds:CryptoBinary" minOccurs="0"/>
<sequence minOccurs="0">
<element name="Seed" type="ds:CryptoBinary"/>
<element name="PgenCounter" type="ds:CryptoBinary"/>
</sequence>
</sequence>
</complexType>
DTD Definition:
<!ELEMENT DSAKeyValue ((P, Q)?, G?, Y, J?, (Seed, PgenCounter)?) >
<!ELEMENT P (#PCDATA) >
<!ELEMENT Q (#PCDATA) >
<!ELEMENT G (#PCDATA) >
<!ELEMENT Y (#PCDATA) >
<!ELEMENT J (#PCDATA) >
<!ELEMENT Seed (#PCDATA) >
<!ELEMENT PgenCounter (#PCDATA) >
RSAKeyValue
Element
Type="http://www.w3.org/2000/09/xmldsig#RSAKeyValue"
(this can be used within a RetrievalMethod
or
Reference
element to identify the referent's type)
RSA key values have two fields: Modulus and Exponent.
<RSAKeyValue> <Modulus>xA7SEU+e0yQH5rm9kbCDN9o3aPIo7HbP7tX6WOocLZAtNfyxSZDU16ksL6W jubafOqNEpcwR3RdFsT7bCqnXPBe5ELh5u4VEy19MzxkXRgrMvavzyBpVRgBUwUlV 5foK5hhmbktQhyNdy/6LpQRhDUDsTvK+g9Ucj47es9AQJ3U= </Modulus> <Exponent>AQAB</Exponent> </RSAKeyValue>
Arbitrary-length integers (e.g. "bignums" such as RSA moduli) are represented in
XML as octet strings as defined by the ds:CryptoBinary
type.
Schema Definition:
<element name="RSAKeyValue" type="ds:RSAKeyValueType"/>
<complexType name="RSAKeyValueType">
<sequence>
<element name="Modulus" type="ds:CryptoBinary"/>
<element name="Exponent" type="ds:CryptoBinary"/>
</sequence>
</complexType>
DTD Definition:
<!ELEMENT RSAKeyValue (Modulus, Exponent) >
<!ELEMENT Modulus (#PCDATA) >
<!ELEMENT Exponent (#PCDATA) >
RetrievalMethod
Element
A RetrievalMethod
element within KeyInfo
is used to
convey a reference to KeyInfo
information that is stored at another
location. For example, several signatures in a document might use a key verified
by an X.509v3 certificate chain appearing once in the document or remotely
outside the document; each signature's KeyInfo
can reference this
chain using a single RetrievalMethod
element instead of including
the entire chain with a sequence of X509Certificate
elements.
RetrievalMethod
uses the same syntax and dereferencing behavior as
Reference
's URI (section 4.3.3.1) and The Reference Processing Model (section
4.3.3.2) except that there is no DigestMethod
or
DigestValue
child elements and presence of the URI is mandatory.
Type
is an optional identifier for the type of data to be retrieved.
The result of dereferencing a RetrievalMethod
Reference
for all KeyInfo
types defined by this specification
(section 4.4) with a corresponding XML structure is an XML element or document
with that element as the root. The rawX509Certificate
KeyInfo
(for which there is no XML structure) returns a binary X509
certificate.
Schema Definition <element name="RetrievalMethod" type="ds:RetrievalMethodType"/> <complexType name="RetrievalMethodType"> <sequence> <element ref="ds:Transforms" minOccurs="0"/> </sequence> <attribute name="URI" type="anyURI"/> <attribute name="Type" type="anyURI" use="optional"/> </complexType>
DTD <!ELEMENT RetrievalMethod (Transforms?) > <!ATTLIST RetrievalMethod URI CDATA #REQUIRED Type CDATA #IMPLIED >
X509Data
Element
Type="http://www.w3.org/2000/09/xmldsig#X509Data
"RetrievalMethod
or
Reference
element to identify the referent's type)
An X509Data
element within KeyInfo
contains one or more
identifiers of keys or X509 certificates (or certificates' identifiers or a
revocation list). The content of X509Data
is:
X509IssuerSerial
element, which contains an X.509 issuer
distinguished name/serial number pair that SHOULD be compliant with RFC2253
[LDAP-DN],
X509SubjectName
element, which contains an X.509 subject
distinguished name that SHOULD be compliant with RFC2253 [LDAP-DN],
X509SKI
element, which contains the base64 encoded plain
(i.e. non-DER-encoded) value of a X509 V.3 SubjectKeyIdentifier extension.
X509Certificate
element, which contains a base64-encoded
[X509v3] certificate, and
X509CRL
element, which contains a base64-encoded
certificate revocation list (CRL) [X509v3].
Any X509IssuerSerial
, X509SKI
, and
X509SubjectName
elements that appear MUST refer to the certificate
or certificates containing the validation key. All such elements that refer to a
particular individual certificate MUST be grouped inside a single
X509Data
element and if the certificate to which they refer appears,
it MUST also be in that X509Data
element.
Any X509IssuerSerial
, X509SKI
, and
X509SubjectName
elements that relate to the same key but different
certificates MUST be grouped within a single KeyInfo
but MAY occur
in multiple X509Data
elements.
All certificates appearing in an X509Data
element MUST relate to the
validation key by either containing it or being part of a certification chain
that terminates in a certificate containing the validation key.
No ordering is implied by the above constraints. The comments in the following instance demonstrate these constraints:
<KeyInfo>
<X509Data> <!-- two pointers to certificate-A -->
<X509IssuerSerial>
<X509IssuerName>CN=TAMURA Kent, OU=TRL, O=IBM,
L=Yamato-shi, ST=Kanagawa, C=JP</X509IssuerName>
<X509SerialNumber>12345678</X509SerialNumber>
</X509IssuerSerial>
<X509SKI>31d97bd7</X509SKI>
</X509Data>
<X509Data><!-- single pointer to certificate-B -->
<X509SubjectName>Subject of Certificate B</X509SubjectName>
</X509Data>
<X509Data> <!-- certificate chain -->
<!--Signer cert, issuer CN=arbolCA,OU=FVT,O=IBM,C=US, serial 4-->
<X509Certificate>MIICXTCCA..</X509Certificate>
<!-- Intermediate cert subject CN=arbolCA,OU=FVT,O=IBM,C=US
issuer CN=tootiseCA,OU=FVT,O=Bridgepoint,C=US -->
<X509Certificate>MIICPzCCA...</X509Certificate>
<!-- Root cert subject CN=tootiseCA,OU=FVT,O=Bridgepoint,C=US -->
<X509Certificate>MIICSTCCA...</X509Certificate>
</X509Data>
</KeyInfo>
Note, there is no direct provision for a PKCS#7 encoded "bag" of certificates or
CRLs. However, a set of certificates and CRLs can occur within an
X509Data
element and multiple X509Data
elements can
occur in a KeyInfo
. Whenever multiple certificates occur in an
X509Data
element, at least one such certificate must contain the
public key which verifies the signature.
Also, strings in DNames
(X509IssuerSerial
,X509SubjectName
, and
KeyName
if approriate) should be encoded as follows:
Schema Definition <element name="X509Data" type="ds:X509DataType"/> <complexType name="X509DataType"> <sequence maxOccurs="unbounded"> <choice> <element name="X509IssuerSerial" type="ds:X509IssuerSerialType"/> <element name="X509SKI" type="base64Binary"/> <element name="X509SubjectName" type="string"/> <element name="X509Certificate" type="base64Binary"/> <element name="X509CRL" type="base64Binary"/> <any namespace="##other" processContents="lax"/> </choice> </sequence> </complexType> <complexType name="X509IssuerSerialType"> <sequence> <element name="X509IssuerName" type="string"/> <element name="X509SerialNumber" type="integer"/> </sequence> </complexType>
DTD <!ELEMENT X509Data ((X509IssuerSerial | X509SKI | X509SubjectName | X509Certificate | X509CRL)+ %X509.ANY;)> <!ELEMENT X509IssuerSerial (X509IssuerName, X509SerialNumber) > <!ELEMENT X509IssuerName (#PCDATA) > <!ELEMENT X509SubjectName (#PCDATA) > <!ELEMENT X509SerialNumber (#PCDATA) > <!ELEMENT X509SKI (#PCDATA) > <!ELEMENT X509Certificate (#PCDATA) > <!ELEMENT X509CRL (#PCDATA) > <!-- Note, this DTD and schema permitX509Data
to be empty; this is precluded by the text inKeyInfo
Element (section 4.4) which states that at least one element from the dsig namespace should be present in the PGP, SPKI, and X509 structures. This is easily expressed for the other key types, but not for X509Data because of its rich structure. -->
PGPData
Element
Type="http://www.w3.org/2000/09/xmldsig#PGPData
"RetrievalMethod
or
Reference
element to identify the referent's type)
The PGPData
element within KeyInfo
is used to convey
information related to PGP public key pairs and signatures on such keys. The
PGPKeyID
's value is a base64Binary sequence containing a standard
PGP public key identifier as defined in [PGP, section
11.2]. The PGPKeyPacket
contains a base64-encoded Key Material
Packet as defined in [PGP, section 5.5]. These children
element types can be complemented/extended by siblings from an external namespace
within PGPData
, or PGPData
can be replaced all together
with an alternative PGP XML structure as a child of KeyInfo
.
PGPData
must contain one PGPKeyID
and/or one
PGPKeyPacket
and 0 or more elements from an external namespace.
Schema Definition: <element name="PGPData" type="ds:PGPDataType"/> <complexType name="PGPDataType"> <choice> <sequence> <element name="PGPKeyID" type="base64Binary"/> <element name="PGPKeyPacket" type="base64Binary" minOccurs="0"/> <any namespace="##other" processContents="lax" minOccurs="0" maxOccurs="unbounded"/> </sequence> <sequence> <element name="PGPKeyPacket" type="base64Binary"/> <any namespace="##other" processContents="lax" minOccurs="0" maxOccurs="unbounded"/> </sequence> </choice> </complexType>
DTD: <!ELEMENT PGPData ((PGPKeyID, PGPKeyPacket?) | (PGPKeyPacket) %PGPData.ANY;) > <!ELEMENT PGPKeyPacket (#PCDATA) > <!ELEMENT PGPKeyID (#PCDATA) >
SPKIData
Element
Type="http://www.w3.org/2000/09/xmldsig#SPKIData
"RetrievalMethod
or
Reference
element to identify the referent's type)
The SPKIData
element within KeyInfo
is used to convey
information related to SPKI public key pairs, certificates and other SPKI data.
SPKISexp
is the base64 encoding of a SPKI canonical S-expression.
SPKIData
must have at least one SPKISexp
;
SPKISexp
can be complemented/extended by siblings from an external
namespace within SPKIData
, or SPKIData
can be entirely
replaced with an alternative SPKI XML structure as a child of
KeyInfo
.
Schema Definition: <element name="SPKIData" type="ds:SPKIDataType"/> <complexType name="SPKIDataType"> <sequence maxOccurs="unbounded"> <element name="SPKISexp" type="base64Binary"/> <any namespace="##other" processContents="lax" minOccurs="0"/> </sequence> </complexType>
DTD: <!ELEMENT SPKIData (SPKISexp %SPKIData.ANY;) > <!ELEMENT SPKISexp (#PCDATA) >
MgmtData
Element
Type="http://www.w3.org/2000/09/xmldsig#MgmtData
"RetrievalMethod
or
Reference
element to identify the referent's type)
The MgmtData
element within KeyInfo
is a string value
used to convey in-band key distribution or agreement data. For example, DH key
exchange, RSA key encryption, etc. Use of this element is NOT RECOMMENDED. It
provides a syntactic hook where in-band key distribution or agreement data can be
placed. However, superior interoperable child elements of KeyInfo
for the transmission of encrypted keys and for key agreement are being specified
by the W3C XML Encryption Working Group and they should be used instead of
MgmtData
.
Schema Definition: <element name="MgmtData" type="string"/>
DTD: <!ELEMENT MgmtData (#PCDATA)>
Object
Element
Type="http://www.w3.org/2000/09/xmldsig#Object"
(this can be used within a Reference
element to identify
the referent's type)
Object
is an optional element that may occur one or more times. When
present, this element may contain any data. The Object
element may
include optional MIME type, ID, and encoding attributes.
The Object
's Encoding
attributed may be used to provide
a URI that identifies the method by which the object is encoded (e.g., a binary
file).
The MimeType
attribute is an optional attribute which describes the
data within the Object
(independent of its encoding). This is a
string with values defined by [MIME]. For example, if the
Object
contains base64 encoded PNG, the Encoding
may be
specified as 'base64' and the MimeType
as 'image/png'. This
attribute is purely advisory; no validation of the MimeType
information is required by this specification. Applications which require
normative type and encoding information for signature validation should specify
Transforms
with well defined resulting
types and/or encodings.
The Object
's Id
is commonly referenced from a
Reference
in SignedInfo
, or Manifest
. This
element is typically used for enveloping signatures where the object being signed is to be
included in the signature element. The digest is calculated over the entire
Object
element including start and end tags.
Note, if the application wishes to exclude the <Object>
tags
from the digest calculation the Reference
must identify the actual
data object (easy for XML documents) or a transform must be used to remove the
Object
tags (likely where the data object is non-XML). Exclusion of
the object tags may be desired for cases where one wants the signature to remain
valid if the data object is moved from inside a signature to outside the
signature (or vice versa), or where the content of the Object
is an
encoding of an original binary document and it is desired to extract and decode
so as to sign the original bitwise representation.
Schema Definition: <element name="Object" type="ds:ObjectType"/> <complexType name="ObjectType" mixed="true"> <sequence minOccurs="0" maxOccurs="unbounded"> <any namespace="##any" processContents="lax"/> </sequence> <attribute name="Id" type="ID" use="optional"/> <attribute name="MimeType" type="string" use="optional"/> <attribute name="Encoding" type="anyURI" use="optional"/> </complexType>
DTD: <!ELEMENT Object (#PCDATA|Signature|SignatureProperties|Manifest %Object.ANY;)* > <!ATTLIST Object Id ID #IMPLIED MimeType CDATA #IMPLIED Encoding CDATA #IMPLIED >
This section describes the optional to implement Manifest
and
SignatureProperties
elements and describes the handling of XML
processing instructions and comments. With respect to the elements
Manifest
and SignatureProperties
this section specifies
syntax and little behavior -- it is left to the application. These elements can
appear anywhere the parent's content model permits; the Signature
content model only permits them within Object
.
Manifest
Element
Type="http://www.w3.org/2000/09/xmldsig#Manifest"
(this can be used within a Reference
element to identify
the referent's type)
The Manifest
element provides a list of Reference
s. The
difference from the list in SignedInfo
is that it is application
defined which, if any, of the digests are actually checked against the objects
referenced and what to do if the object is inaccessible or the digest compare
fails. If a Manifest
is pointed to from SignedInfo
, the
digest over the Manifest
itself will be checked by the core
signature validation behavior. The digests within such a Manifest
are checked at the application's discretion. If a Manifest
is
referenced from another Manifest
, even the overall digest of this
two level deep Manifest
might not be checked.
Schema Definition: <element name="Manifest" type="ds:ManifestType"/> <complexType name="ManifestType"> <sequence> <element ref="ds:Reference" maxOccurs="unbounded"/> </sequence> <attribute name="Id" type="ID" use="optional"/> </complexType>
DTD: <!ELEMENT Manifest (Reference+) > <!ATTLIST Manifest Id ID #IMPLIED >
SignatureProperties
Element
Type="http://www.w3.org/2000/09/xmldsig#SignatureProperties"
(this can be used within a Reference
element to identify
the referent's type)
Additional information items concerning the generation of the signature(s) can be
placed in a SignatureProperty
element (i.e., date/time stamp or the
serial number of cryptographic hardware used in signature generation).
Schema Definition: <element name="SignatureProperties" type="ds:SignaturePropertiesType"/> <complexType name="SignaturePropertiesType"> <sequence> <element ref="ds:SignatureProperty" maxOccurs="unbounded"/> </sequence> <attribute name="Id" type="ID" use="optional"/> </complexType> <element name="SignatureProperty" type="ds:SignaturePropertyType"/> <complexType name="SignaturePropertyType" mixed="true"> <choice maxOccurs="unbounded"> <any namespace="##other" processContents="lax"/> <!-- (1,1) elements from (1,unbounded) namespaces --> </choice> <attribute name="Target" type="anyURI" use="required"/> <attribute name="Id" type="ID" use="optional"/> </complexType>
DTD: <!ELEMENT SignatureProperties (SignatureProperty+) > <!ATTLIST SignatureProperties Id ID #IMPLIED > <!ELEMENT SignatureProperty (#PCDATA %SignatureProperty.ANY;)* > <!ATTLIST SignatureProperty Target CDATA #REQUIRED Id ID #IMPLIED >
No XML processing instructions (PIs) are used by this specification.
Note that PIs placed inside SignedInfo
by an application will be
signed unless the CanonicalizationMethod
algorithm discards them.
(This is true for any signed XML content.) All of the
CanonicalizationMethod
s identified within this specification retain
PIs. When a PI is part of content that is signed (e.g., within
SignedInfo
or referenced XML documents) any change to the PI will
obviously result in a signature failure.
XML comments are not used by this specification.
Note that unless CanonicalizationMethod
removes comments within
SignedInfo
or any other referenced XML (which [XML-C14N] does), they will be signed. Consequently, if
they are retained, a change to the comment will cause a signature failure.
Similarly, the XML signature over any XML data will be sensitive to comment
changes unless a comment-ignoring canonicalization/transform method, such as the
Canonical XML [XML-C14N], is specified.
This section identifies algorithms used with the XML digital signature
specification. Entries contain the identifier to be used in
Signature
elements, a reference to the formal specification, and
definitions, where applicable, for the representation of keys and the results of
cryptographic operations.
Algorithms are identified by URIs that appear as an attribute to the element that
identifies the algorithms' role (DigestMethod
,
Transform
, SignatureMethod
, or
CanonicalizationMethod
). All algorithms used herein take parameters
but in many cases the parameters are implicit. For example, a
SignatureMethod
is implicitly given two parameters: the keying info
and the output of CanonicalizationMethod
. Explicit additional
parameters to an algorithm appear as content elements within the algorithm role
element. Such parameter elements have a descriptive element name, which is
frequently algorithm specific, and MUST be in the XML Signature namespace or an
algorithm specific namespace.
This specification defines a set of algorithms, their URIs, and requirements for implementation. Requirements are specified over implementation, not over requirements for signature use. Furthermore, the mechanism is extensible; alternative algorithms may be used by signature applications.
* The Enveloped Signature transform removes the Signature
element
from the calculation of the signature when the signature is within the content
that it is being signed. This MAY be implemented via the RECOMMENDED XPath
specification specified in 6.6.4: Enveloped
Signature Transform; it MUST have the same effect as that specified by the XPath Transform.
Only one digest algorithm is defined herein. However, it is expected that one or more additional strong digest algorithms will be developed in connection with the US Advanced Encryption Standard effort. Use of MD5 [MD5] is NOT RECOMMENDED because recent advances in cryptanalysis have cast doubt on its strength.
The SHA-1 algorithm [SHA-1] takes no explicit parameters. An example of an SHA-1 DigestAlg element is:
<DigestMethod Algorithm="
http://www.w3.org/2000/09/xmldsig#sha1"/>
A SHA-1 digest is a 160-bit string. The content of the DigestValue element shall be the base64 encoding of this bit string viewed as a 20-octet octet stream. For example, the DigestValue element for the message digest:
A9993E36 4706816A BA3E2571 7850C26C 9CD0D89D
from Appendix A of the SHA-1 standard would be:
<DigestValue>qZk+NkcGgWq6PiVxeFDCbJzQ2J0=</DigestValue>
MAC algorithms take two implicit parameters, their keying material determined
from KeyInfo
and the octet stream output by
CanonicalizationMethod
. MACs and signature algorithms are
syntactically identical but a MAC implies a shared secret key.
The HMAC algorithm (RFC2104 [HMAC]) takes the truncation length in bits as a parameter;
if the parameter is not specified then all the bits of the hash are output. An
example of an HMAC SignatureMethod
element:
<SignatureMethod Algorithm="http://www.w3.org/2000/09/xmldsig#hmac-sha1"> <HMACOutputLength>128</HMACOutputLength> </SignatureMethod>
The output of the HMAC algorithm is ultimately the output (possibly truncated) of the chosen digest algorithm. This value shall be base64 encoded in the same straightforward fashion as the output of the digest algorithms. Example: the SignatureValue element for the HMAC-SHA1 digest
9294727A 3638BB1C 13F48EF8 158BFC9D
from the test vectors in [HMAC] would be
<SignatureValue>kpRyejY4uxwT9I74FYv8nQ==</SignatureValue>
Schema Definition: <simpleType name="HMACOutputLengthType"> <restriction base="integer"/> </simpleType>
DTD: <!ELEMENT HMACOutputLength (#PCDATA)>
Signature algorithms take two implicit parameters, their keying material
determined from KeyInfo
and the octet stream output by
CanonicalizationMethod
. Signature and MAC algorithms are
syntactically identical but a signature implies public key cryptography.
The DSA algorithm [DSS] takes no explicit parameters. An
example of a DSA SignatureMethod
element is:
<SignatureMethod Algorithm="http://www.w3.org/2000/09/xmldsig#dsa-sha1"/>
The output of the DSA algorithm consists of a pair of integers usually referred
by the pair (r, s). The signature value consists of the base64 encoding of the
concatenation of two octet-streams that respectively result from the
octet-encoding of the values r and s in that order. Integer to octet-stream
conversion must be done according to the I2OSP operation defined in the RFC 2437 [PKCS1] specification with a l
parameter equal
to 20. For example, the SignatureValue element for a DSA signature
(r
, s
) with values specified in hexadecimal:
r = 8BAC1AB6 6410435C B7181F95 B16AB97C 92B341C0
s = 41E2345F 1F56DF24 58F426D1 55B4BA2D B6DCD8C8
from the example in Appendix 5 of the DSS standard would be
<SignatureValue>
i6watmQQQ1y3GB+VsWq5fJKzQcBB4jRfH1bfJFj0JtFVtLotttzYyA==</SignatureValue>
The expression "RSA algorithm" as used in this draft refers to the RSASSA-PKCS1-v1_5 algorithm described in RFC 2437 [PKCS1]. The RSA algorithm takes no explicit parameters. An example of an RSA SignatureMethod element is:
<SignatureMethod Algorithm="http://www.w3.org/2000/09/xmldsig#rsa-sha1"/>
The SignatureValue
content for an RSA signature is the base64 [MIME] encoding of the octet string computed as per RFC 2437 [PKCS1, section 8.1.1: Signature generation for the
RSASSA-PKCS1-v1_5 signature scheme]. As specified in the EMSA-PKCS1-V1_5-ENCODE
function RFC 2437 [PKCS1, section 9.2.1], the value input to the signature
function MUST contain a pre-pended algorithm object identifier for the hash
function, but the availability of an ASN.1 parser and recognition of OIDs is not
required of a signature verifier. The PKCS#1 v1.5 representation appears as:
CRYPT (PAD (ASN.1 (OID, DIGEST (data))))
Note that the padded ASN.1 will be of the following form:
01 | FF* | 00 | prefix | hash
where "|" is concatenation, "01", "FF", and "00" are fixed octets of the corresponding hexadecimal value, "hash" is the SHA1 digest of the data, and "prefix" is the ASN.1 BER SHA1 algorithm designator prefix required in PKCS1 [RFC 2437], that is,
hex 30 21 30 09 06 05 2B 0E 03 02 1A 05 00 04 14
This prefix is included to make it easier to use standard cryptographic libraries. The FF octet MUST be repeated the maximum number of times such that the value of the quantity being CRYPTed is one octet shorter than the RSA modulus.
The resulting base64 [MIME] string is the value of the child text node of the SignatureValue element, e.g.
<SignatureValue> IWijxQjUrcXBYoCei4QxjWo9Kg8D3p9tlWoT4t0/gyTE96639In0FZFY2/rvP+/bMJ01EArmKZsR5VW3rwoPxw= </SignatureValue>
If canonicalization is performed over octets, the canonicalization algorithms take two implicit parameters: the content and its charset. The charset is derived according to the rules of the transport protocols and media types (e.g, RFC2376 [XML-MT] defines the media types for XML). This information is necessary to correctly sign and verify documents and often requires careful server side configuration.
Various canonicalization algorithms require conversion to [UTF-8].The two algorithms below understand at least [UTF-8] and [UTF-16] as input encodings. We RECOMMEND that externally specified algorithms do the same. Knowledge of other encodings is OPTIONAL.
Various canonicalization algorithms transcode from a non-Unicode encoding to Unicode. The two algorithms below perform text normalization during transcoding [NFC, NFC-Corrigendum]. We RECOMMEND that externally specified canonicalization algorithms do the same. (Note, there can be ambiguities in converting existing charsets to Unicode, for an example see the XML Japanese Profile [XML-Japanese] Note.)
An example of an XML canonicalization element is:
<CanonicalizationMethod Algorithm="
http://www.w3.org/TR/2001/REC-xml-c14n-20010315"/>
The normative specification of Canonical XML is [XML-C14N]. The algorithm is capable of taking as input either an octet stream or an XPath node-set (or sufficiently functional alternative). The algorithm produces an octet stream as output. Canonical XML is easily parameterized (via an additional URI) to omit or retain comments.
Transform
Algorithms
A Transform
algorithm has a single implicit parameter: an octet
stream from the Reference
or the output of an earlier
Transform
.
Application developers are strongly encouraged to support all transforms listed in this section as RECOMMENDED unless the application environment has resource constraints that would make such support impractical. Compliance with this recommendation will maximize application interoperability and libraries should be available to enable support of these transforms in applications without extensive development.
Any canonicalization algorithm that can be used for
CanonicalizationMethod
(such as those in Canonicalization Algorithms (section 6.5)) can be used as
a Transform
.
The normative specification for base64 decoding transforms is [MIME]. The base64 Transform
element has no
content. The input is decoded by the algorithms. This transform is useful if an
application needs to sign the raw data associated with the encoded content of an
element.
This transform requires an octet stream for input. If an XPath node-set (or
sufficiently functional alternative) is given as input, then it is converted to
an octet stream by performing operations logically equivalent to 1) applying an
XPath transform with expression self::text()
, then 2) taking the
string-value of the node-set. Thus, if an XML element is identified by a barename
XPointer in the Reference
URI, and its content consists solely of
base64 encoded character data, then this transform automatically strips away the
start and end tags of the identified element and any of its descendant elements
as well as any descendant comments and processing instructions. The output of
this transform is an octet stream.
The normative specification for XPath expression evaluation is [XPath]. The XPath expression to be evaluated appears as the
character content of a transform parameter child element named
XPath
.
The input required by this transform is an XPath node-set. Note that if the actual input is an XPath node-set resulting from a null URI or barename XPointer dereference, then comment nodes will have been omitted. If the actual input is an octet stream, then the application MUST convert the octet stream to an XPath node-set suitable for use by Canonical XML with Comments. (A subsequent application of the REQUIRED Canonical XML algorithm would strip away these comments.) In other words, the input node-set should be equivalent to the one that would be created by the following process:
(//. | //@* | //namespace::*)
The evaluation of this expression includes all of the document's nodes (including comments) in the node-set representing the octet stream.
The transform output is also an XPath node-set. The XPath expression appearing in
the XPath
parameter is evaluated once for each node in the input
node-set. The result is converted to a boolean. If the boolean is true, then the
node is included in the output node-set. If the boolean is false, then the node
is omitted from the output node-set.
Note: Even if the input node-set has had comments removed, the
comment nodes still exist in the underlying parse tree and can separate text
nodes. For example, the markup <e>Hello, <!-- comment
-->world!</e>
contains two text nodes. Therefore, the expression
self::text()[string()="Hello, world!"]
would fail. Should this
problem arise in the application, it can be solved by either canonicalizing the
document before the XPath transform to physically remove the comments or by
matching the node based on the parent element's string value (e.g. by using the
expression self::text()[string(parent::e)="Hello, world!"]
).
The primary purpose of this transform is to ensure that only specifically defined changes to the input XML document are permitted after the signature is affixed. This is done by omitting precisely those nodes that are allowed to change once the signature is affixed, and including all other input nodes in the output. It is the responsibility of the XPath expression author to include all nodes whose change could affect the interpretation of the transform output in the application context.
An important scenario would be a document requiring two enveloped signatures. Each signature must omit itself from its own digest calculations, but it is also necessary to exclude the second signature element from the digest calculations of the first signature so that adding the second signature does not break the first signature.
The XPath transform establishes the following evaluation context for each node of the input node-set:
As a result of the context node setting, the XPath expressions appearing in this
transform will be quite similar to those used in used in [XSLT], except that the size and position are always 1 to
reflect the fact that the transform is automatically visiting every node (in
XSLT, one recursively calls the command apply-templates
to visit the
nodes of the input tree).
The function here()
is defined as follows:
The here function returns a node-set containing the attribute or processing instruction node or the parent element of the text node that directly bears the XPath expression. This expression results in an error if the containing XPath expression does not appear in the same XML document against which the XPath expression is being evaluated.
As an example, consider creating an enveloped signature (a Signature
element that is a descendant of an element being signed). Although the signed
content should not be changed after signing, the elements within the
Signature
element are changing (e.g. the digest value must be put
inside the DigestValue
and the SignatureValue
must be
subsequently calculated). One way to prevent these changes from invalidating the
digest value in DigestValue
is to add an XPath
Transform
that omits all Signature
elements and their
descendants. For example,
<Document> ... <Signature xmlns="http://www.w3.org/2000/09/xmldsig#"> <SignedInfo> ... <Reference URI=""> <Transforms> <Transform Algorithm="http://www.w3.org/TR/1999/REC-xpath-19991116"> <XPath xmlns:dsig="&dsig;"> not(ancestor-or-self::dsig:Signature) </XPath> </Transform> </Transforms> <DigestMethod Algorithm="http://www.w3.org/2000/09/xmldsig#sha1"/> <DigestValue></DigestValue> </Reference> </SignedInfo> <SignatureValue></SignatureValue> </Signature> ... </Document>
Due to the null Reference
URI in this example, the XPath transform
input node-set contains all nodes in the entire parse tree starting at the root
node (except the comment nodes). For each node in this node-set, the node is
included in the output node-set except if the node or one of its ancestors has a
tag of Signature
that is in the namespace given by the replacement
text for the entity &dsig;
.
A more elegant solution uses the here function to omit only the
Signature
containing the XPath Transform, thus allowing enveloped
signatures to sign other signatures. In the example above, use the
XPath
element:
<XPath xmlns:dsig="&dsig;"> count(ancestor-or-self::dsig:Signature | here()/ancestor::dsig:Signature[1]) > count(ancestor-or-self::dsig:Signature)</XPath>
Since the XPath equality operator converts node sets to string values before
comparison, we must instead use the XPath union operator (|). For each node of
the document, the predicate expression is true if and only if the node-set
containing the node and its Signature
element ancestors does not
include the enveloped Signature
element containing the XPath
expression (the union does not produce a larger set if the enveloped
Signature
element is in the node-set given by
ancestor-or-self::Signature
).
An enveloped signature transform T removes the whole
Signature
element containing T from the
digest calculation of the Reference
element containing
T. The entire string of characters used by an XML
processor to match the Signature
with the XML production
element
is removed. The output of the transform is equivalent to the
output that would result from replacing T with an XPath
transform containing the following XPath
parameter element:
<XPath xmlns:dsig="&dsig;"> count(ancestor-or-self::dsig:Signature | here()/ancestor::dsig:Signature[1]) > count(ancestor-or-self::dsig:Signature)</XPath>
The input and output requirements of this transform are identical to those of the XPath transform, but may only be applied to a node-set from its parent XML document. Note that it is not necessary to use an XPath expression evaluator to create this transform. However, this transform MUST produce output in exactly the same manner as the XPath transform parameterized by the XPath expression above.
The normative specification for XSL Transformations is [XSLT]. Specification of a namespace-qualified stylesheet
element, which MUST be the sole child of the Transform
element,
indicates that the specified style sheet should be used. Whether this
instantiates in-line processing of local XSLT declarations within the resource is
determined by the XSLT processing model; the ordered application of multiple
stylesheet may require multiple Transforms
. No special provision is
made for the identification of a remote stylesheet at a given URI because it can
be communicated via an
xsl:include
or
xsl:import
within the stylesheet
child of the
Transform
.
This transform requires an octet stream as input. If the actual input is an XPath node-set, then the signature application should attempt to convert it to octets (apply Canonical XML]) as described in the Reference Processing Model (section 4.3.3.2).
The output of this transform is an octet stream. The processing rules for the XSL
style sheet or transform element are stated in the XSLT specification [XSLT]. We RECOMMEND that XSLT transform authors use an
output method of xml
for XML and HTML. As XSLT implementations do
not produce consistent serializations of their output, we further RECOMMEND
inserting a transform after the XSLT transform to canonicalize the output. These
steps will help to ensure interoperability of the resulting signatures among
applications that support the XSLT transform. Note that if the output is actually
HTML, then the result of these steps is logically equivalent [XHTML].
Digital signatures only work if the verification calculations are performed on exactly the same bits as the signing calculations. If the surface representation of the signed data can change between signing and verification, then some way to standardize the changeable aspect must be used before signing and verification. For example, even for simple ASCII text there are at least three widely used line ending sequences. If it is possible for signed text to be modified from one line ending convention to another between the time of signing and signature verification, then the line endings need to be canonicalized to a standard form before signing and verification or the signatures will break.
XML is subject to surface representation changes and to processing which discards some surface information. For this reason, XML digital signatures have a provision for indicating canonicalization methods in the signature so that a verifier can use the same canonicalization as the signer.
Throughout this specification we distinguish between the canonicalization of a
Signature
element and other signed XML data objects. It is possible
for an isolated XML document to be treated as if it were binary data so that no
changes can occur. In that case, the digest of the document will not change and
it need not be canonicalized if it is signed and verified as such. However, XML
that is read and processed using standard XML parsing and processing techniques
is frequently changed such that some of its surface representation information is
lost or modified. In particular, this will occur in many cases for the
Signature
and enclosed SignedInfo
elements since they,
and possibly an encompassing XML document, will be processed as XML.
Similarly, these considerations apply to Manifest
,
Object
, and SignatureProperties
elements if those
elements have been digested, their DigestValue
is to be checked, and
they are being processed as XML.
The kinds of changes in XML that may need to be canonicalized can be divided into four categories. There are those related to the basic [XML], as described in 7.1 below. There are those related to [DOM], [SAX], or similar processing as described in 7.2 below. Third, there is the possibility of coded character set conversion, such as between UTF-8 and UTF-16, both of which all [XML] compliant processors are required to support, which is described in the paragraph immediately below. And, fourth, there are changes that related to namespace declaration and XML namespace attribute context as described in 7.3 below.
Any canonicalization algorithm should yield output in a specific fixed coded
character set. All canonicalization algorithms
identified in this document use UTF-8 (without a byte order mark (BOM)) and do
not provide character normalization. We RECOMMEND that signature applications
create XML content (Signature
elements and their
descendents/content) in Normalization Form C [NFC, NFC-Corrigendum] and check that any XML being
consumed is in that form as well; (if not, signatures may consequently fail to
validate). Additionally, none of these algorithms provide data type
normalization. Applications that normalize data types in varying formats (e.g.,
(true, false) or (1,0)) may not be able to validate each other's signatures.
XML 1.0 [XML] defines an interface where a conformant application reading XML is given certain information from that XML and not other information. In particular,
Note that items (2), (4), and (5.3) depend on the presence of a schema, DTD or
similar declarations. The Signature
element type is laxly
schema valid [XML-schema], consequently
external XML or even XML within the same document as the signature may be (only)
well-formed or from another namespace (where permitted by the signature schema);
the noted items may not be present. Thus, a signature with such content will only
be verifiable by other signature applications if the following syntax constraints
are observed when generating any signed material including the
SignedInfo
element:
In addition to the canonicalization and syntax constraints discussed above, many XML applications use the Document Object Model [DOM] or the Simple API for XML [SAX]. DOM maps XML into a tree structure of nodes and typically assumes it will be used on an entire document with subsequent processing being done on this tree. SAX converts XML into a series of events such as a start tag, content, etc. In either case, many surface characteristics such as the ordering of attributes and insignificant white space within start/end tags is lost. In addition, namespace declarations are mapped over the nodes to which they apply, losing the namespace prefixes in the source text and, in most cases, losing where namespace declarations appeared in the original instance.
If an XML Signature is to be produced or verified on a system using the DOM or SAX processing, a canonical method is needed to serialize the relevant part of a DOM tree or sequence of SAX events. XML canonicalization specifications, such as [XML-C14N], are based only on information which is preserved by DOM and SAX. For an XML Signature to be verifiable by an implementation using DOM or SAX, not only must the XML 1.0 syntax constraints given in the previous section be followed but an appropriate XML canonicalization MUST be specified so that the verifier can re-serialize DOM/SAX mediated input into the same octet stream that was signed.
In [XPath] and consequently the Canonical XML data model an element has namespace nodes that correspond to those declarations within the element and its ancestors:
"Note: An element E has namespace nodes that represent its namespace declarations as well as any namespace declarations made by its ancestors that have not been overridden in E's declarations, the default namespace if it is non-empty, and the declaration of the prefix
xml
." [XML-C14N]
When serializing a Signature
element or signed XML data that's the
child of other elements using these data models, that Signature
element and its children, may contain namespace declarations from its ancestor
context. In addition, the Canonical XML and Canonical XML with Comments
algorithms import all xml namespace attributes (such as xml:lang
)
from the nearest ancestor in which they are declared to the apex node of
canonicalized XML unless they are already declared at that node. This may
frustrate the intent of the signer to create a signature in one context which
remains valid in another. For example, given a signature which is a child of
B
and a grandchild of A
:
<A xmlns:n1="&foo;"> <B xmlns:n2="&bar;"> <Signature xmlns="&dsig;"> ... <Reference URI="#signme"/> ... </Signature> <C ID="signme" xmlns="&baz;"/> </B> </A>
when either the element B
or the signed element C
is
moved into a [SOAP] envelope for transport:
<SOAP:Envelope xmlns:SOAP="http://schemas.xmlsoap.org/soap/envelope/"> ... <SOAP:Body> <B xmlns:n2="&bar;"> <Signature xmlns="&dsig;"> ... </Signature> <C ID="signme" xmlns="&baz;"/> </B> </SOAP:Body> </SOAP:Envelope>
The canonical form of the signature in this context will contain new namespace
declarations from the SOAP:Envelope
context, invalidating the
signature. Also, the canonical form will lack namespace declarations it may have
originally had from element A
's context, also invalidating the
signature. To avoid these problems, the application may:
The XML Signature specification provides a very flexible digital signature mechanism. Implementors must give consideration to their application threat models and to the following factors.
A requirement of this specification is to permit signatures to "apply to a
part or totality of a XML document." (See [XML-Signature-RD, section 3.1.3].) The
Transforms
mechanism meets this requirement by permitting one to
sign data derived from processing the content of the identified resource. For
instance, applications that wish to sign a form, but permit users to enter
limited field data without invalidating a previous signature on the form might
use [XPath] to exclude those portions the user needs to
change. Transforms
may be arbitrarily specified and may include
encoding transforms, canonicalization instructions or even XSLT transformations.
Three cautions are raised with respect to this feature in the following sections.
Note, core validation behavior does not confirm that the signed data was obtained by applying each step of the indicated transforms. (Though it does check that the digest of the resulting content matches that specified in the signature.) For example, some applications may be satisfied with verifying an XML signature over a cached copy of already transformed data. Other applications might require that content be freshly dereferenced and transformed.
First, obviously, signatures over a transformed document do not secure any information discarded by transforms: only what is signed is secure.
Note that the use of Canonical XML [XML-C14N]
ensures that all internal entities and XML namespaces are expanded within the
content being signed. All entities are replaced with their definitions and the
canonical form explicitly represents the namespace that an element would
otherwise inherit. Applications that do not canonicalize XML content (especially
the SignedInfo
element) SHOULD NOT use internal entities and SHOULD
represent the namespace explicitly within the content being signed since they can
not rely upon canonicalization to do this for them. Also, users concerned with
the integrity of the element type definitions associated with the XML instance
being signed may wish to sign those definitions as well (i.e., the schema, DTD,
or natural language description associated with the namespace/identifier).
Second, an envelope containing signed information is not secured by the signature. For instance, when an encrypted envelope contains a signature, the signature does not protect the authenticity or integrity of unsigned envelope headers nor its ciphertext form, it only secures the plaintext actually signed.
Additionally, the signature secures any information introduced by the transform: only what is "seen" (that which is represented to the user via visual, auditory or other media) should be signed. If signing is intended to convey the judgment or consent of a user (an automated mechanism or person), then it is normally necessary to secure as exactly as practical the information that was presented to that user. Note that this can be accomplished by literally signing what was presented, such as the screen images shown a user. However, this may result in data which is difficult for subsequent software to manipulate. Instead, one can sign the data along with whatever filters, style sheets, client profile or other information that affects its presentation.
Just as a user should only sign what he or she "sees," persons and automated
mechanism that trust the validity of a transformed document on the basis of a
valid signature should operate over the data that was transformed (including
canonicalization) and signed, not the original pre-transformed data. This
recommendation applies to transforms specified within the signature as well as
those included as part of the document itself. For instance, if an XML document
includes an embedded
style sheet [XSLT] it is the transformed document
that should be represented to the user and signed. To meet this recommendation
where a document references an external style sheet, the content of that external
resource should also be signed as via a signature Reference
otherwise the content of that external content might change which alters the
resulting document without invalidating the signature.
Some applications might operate over the original or intermediary data but should be extremely careful about potential weaknesses introduced between the original and transformed data. This is a trust decision about the character and meaning of the transforms that an application needs to make with caution. Consider a canonicalization algorithm that normalizes character case (lower to upper) or character composition ('e and accent' to 'accented-e'). An adversary could introduce changes that are normalized and consequently inconsequential to signature validity but material to a DOM processor. For instance, by changing the case of a character one might influence the result of an XPath selection. A serious risk is introduced if that change is normalized for signature validation but the processor operates over the original data and returns a different result than intended.
As a result:
This specification uses public key signatures and keyed hash authentication codes. These have substantially different security models. Furthermore, it permits user specified algorithms which may have other models.
With public key signatures, any number of parties can hold the public key and verify signatures while only the parties with the private key can create signatures. The number of holders of the private key should be minimized and preferably be one. Confidence by verifiers in the public key they are using and its binding to the entity or capabilities represented by the corresponding private key is an important issue, usually addressed by certificate or online authority systems.
Keyed hash authentication codes, based on secret keys, are typically much more efficient in terms of the computational effort required but have the characteristic that all verifiers need to have possession of the same key as the signer. Thus any verifier can forge signatures.
This specification permits user provided signature algorithms and keying information designators. Such user provided algorithms may have different security models. For example, methods involving biometrics usually depend on a physical characteristic of the authorized user that can not be changed the way public or secret keys can be and may have other security model differences.
The strength of a particular signature depends on all links in the security chain. This includes the signature and digest algorithms used, the strength of the key generation [RANDOM] and the size of the key, the security of key and certificate authentication and distribution mechanisms, certificate chain validation policy, protection of cryptographic processing from hostile observation and tampering, etc.
Care must be exercised by applications in executing the various algorithms that may be specified in an XML signature and in the processing of any "executable content" that might be provided to such algorithms as parameters, such as XSLT transforms. The algorithms specified in this document will usually be implemented via a trusted library but even there perverse parameters might cause unacceptable processing or memory demand. Even more care may be warranted with application defined algorithms.
The security of an overall system will also depend on the security and integrity of its operating procedures, its personnel, and on the administrative enforcement of those procedures. All the factors listed in this section are important to the overall security of a system; however, most are beyond the scope of this specification.
schemaLocation
to aid
automated schema fetching and validation.
Object
designates a specific XML element. Occasionally we refer to
a data object as a document or as a resource's content. The term element content
is used to describe the data between XML start and end tags [XML]. The term XML document is used to describe
data objects which conform to the XML specification [XML].
Object
element
is merely one type of digital data (or document) that can be signed via a
Reference
.
Signature
element type and its children
(including SignatureValue
) and the specified algorithms.
Signature
element,
and can be identified via a URI
or transform. Consequently, the
signature is "detached" from the content it signs. This definition typically
applies to separate data objects, but it also includes the instance where the
Signature
and data object reside within the same XML document but
are sibling elements.
Object
element of
the signature itself. The Object
(or its content) is identified
via a Reference
(via a URI
fragment identifier or
transform).
SignatureValue
.
SignedInfo
reference validation.
Reference
, matches its specified DigestValue
.
SignatureValue
matches the result of processing
SignedInfo
with CanonicalizationMethod
and
SignatureMethod
as specified in Core
Validation (section 3.2).
Donald E. Eastlake 3rd
Motorola, 20 Forbes Boulevard
Mansfield, MA 02048 USA
Phone: 1-508-261-5434
Email: Donald.Eastlake@motorola.com
Joseph M. Reagle Jr., W3C
Massachusetts Institute of Technology
Laboratory for Computer Science
NE43-350, 545 Technology Square
Cambridge, MA 02139
Phone: + 1.617.258.7621
Email: reagle@w3.org
David Solo
Citigroup
909 Third Ave, 16th Floor
NY, NY 10043 USA
Phone +1-212-559-2900
Email: dsolo@alum.mit.edu