| Title: | Public Key Infrastucture for R Based on the X.509 Standard |
|---|---|
| Description: | Public Key Infrastucture functions such as verifying certificates, RSA encription and signing which can be used to build PKI infrastructure and perform cryptographic tasks. |
| Authors: | Simon Urbanek <[email protected]> |
| Maintainer: | Simon Urbanek <[email protected]> |
| License: | GPL-2 | GPL-3 | file LICENSE |
| Version: | 0.1-14 |
| Built: | 2024-11-13 06:15:31 UTC |
| Source: | CRAN |
ASN1.decode decodes ASN.1 binary format into
raw format chunks tagged with class types.
ASN1.encode converts structured objects into
ASN.1 binary format.
ASN1.item creates an item - basic object in
structures that can be encoded using ASN1.encode.
ASN1.type extracts the class type from an
ASN.1 item
ASN1.decode(what) ASN1.encode(what) ASN1.item(what, type) ASN1.type(what)ASN1.decode(what) ASN1.encode(what) ASN1.item(what, type) ASN1.type(what)
what |
object to decode/encode/query |
type |
class type of the item (integer value) |
This is a suite of low-level tools to deal with ASN.1 (Abstract Syntax
Notation One) binary formats DER, BER and CER. The tools were written
specifically to handle the various DER-encoded key structures so it
provides only a subset of the ASN.1 specification. They are used
internally by the PKI poackage.
ASN1.decode decodes the binary representation (as raw vector)
into individual items. Sequences are convered into lists, all other
objects are retained in their binary form and tagged with the integer
class type - which can be obtained using ASN1.type function.
ASN1.encode expects item (or a list of items) either created
using ASN1.decode or ASN1.item and converts them into
DER binary format.
The result of ASN1.encode(ASN1.decode(x)) will be x if
x was in DER format.
ASN1.decode returns either one item or a list.
ASN1.encode returns a raw vector in DER format.
ASN1.type returns an integer class type
ASN1.item returns an ASN.1 item object
ASN1.encode uses a fixed buffer for encoding which currently
limits the total size of the resulting structure to 1MB.
Only definite length forms are supported. The validity of individual items is not checked.
Simon Urbanek
# generate a small key key <- PKI.genRSAkey(bits = 512L) # extract private and public parts in DER format prv <- PKI.save.key(key, format="DER") pub <- PKI.save.key(key, private=FALSE, format="DER") # parse the public key x <- ASN1.decode(pub) x # the second element is the actual key # as a bit string that's itself in DER # two integers - modulus and exponent # Note that this is in fact the pure PKCS#1 key format ASN1.decode(x[[2]]) # encoding it back should yield the same representation since it is DER stopifnot(identical(ASN1.encode(x), as.raw(pub)))# generate a small key key <- PKI.genRSAkey(bits = 512L) # extract private and public parts in DER format prv <- PKI.save.key(key, format="DER") pub <- PKI.save.key(key, private=FALSE, format="DER") # parse the public key x <- ASN1.decode(pub) x # the second element is the actual key # as a bit string that's itself in DER # two integers - modulus and exponent # Note that this is in fact the pure PKCS#1 key format ASN1.decode(x[[2]]) # encoding it back should yield the same representation since it is DER stopifnot(identical(ASN1.encode(x), as.raw(pub)))
as.BIGNUMint encodes integer in BIGNUM format as raw vector as
used by ASN.1 format.
as.BIGNUMint(what, scalar = TRUE)as.BIGNUMint(what, scalar = TRUE)
what |
representation of an integer or a vector
thereof. Currently supported formats include |
scalar |
if |
The BIGNUM representation as used in ASN.1 is a big-endian encoding of
variable length stored in a raw vector. Negative numbers are stored in
two-complement's encoding, but are currently unsupported by
as.BIGNUMint.
Raw vector in BIGNUM integer representation.
Unless the input is of class "bigz" then 32-bit platforms only
support integers up to 32-bit, 64-bit platforms up to 53-bit (when
real vectors are used).
Simon Urbanek
as.BIGNUMint(65537)as.BIGNUMint(65537)
Object Identifiers (OIDs) are entities defined by international standards (ITU-T, ISO, IEC) used to identify objects. In the PKI context they are used for example to identify encyrption algorithms. Each root (first integer - see below) denotes the standards body governing the allocations.
OIDs consist of a hierarchy of integers with each component having a
meaning in the hierarchy. For example, the OID of the DER encoding is
defined in the ITU-T X.680 standard as
joint-iso-itu-t(2) asn1(1) ber-derived(2) distinguished-encoding(1)
where the text before each integer describes its meaning in that
context and the integer is the encoding of that meaning. So the OID
itself would be in character form "2.1.2.1" (also called the dot
notation introduced by IETF) and in R integer form
c(2, 1, 2, 1). Internally, OIDs are represented in their ASN.1
encoding as raw vectors which is the way they are typically used
in files or communication payload.
The following functions are used to operate on OIDs.
oid creates an OID.
Coercion methods as.integer and as.character convert the
OID into its numeric and textural form respectively. as.oid is
a generic for convering objects into OIDs and is implemented for at
least the above cases.
is.oid returns TRUE if the object is an OID.
oid(x) as.oid(x, ...) ## Default S3 method: as.oid(x, ...) is.oid(x) ## S3 method for class 'oid' Ops(e1, e2) ## S3 method for class 'oid' print(x, ...) ## S3 method for class 'oid' as.character(x, ...) ## S3 method for class 'oid' as.integer(x, ...)oid(x) as.oid(x, ...) ## Default S3 method: as.oid(x, ...) is.oid(x) ## S3 method for class 'oid' Ops(e1, e2) ## S3 method for class 'oid' print(x, ...) ## S3 method for class 'oid' as.character(x, ...) ## S3 method for class 'oid' as.integer(x, ...)
x |
object to covert/create/check |
e1 |
left-hand side argument for binary operators |
e2 |
right-hand side arguemnt for binary operators |
... |
further arguments (currently unused) |
The only allowed oparators on OIDs are == and != which
return TRUE or FALSE.
The oid(x) constructor (and also the as.oid default
method) support following types: scalar string (expected to be in
dot-notation), integer vector, numeric vector (it is coerced to integer
vector implicitly), raw vector (must be ASN.1 encoding of the OID).
The S3 class of OID objects is "oid". It consists of a raw
vector repesenting the ASN.1 encoded OID (without the type
specifier). An additional attribute "type" is set to 6L
for compatiblity with ASN1.encode.
Simon Urbanek
## RSA algorithm OID: ## iso(1) member-body(2) us(840) rsadsi(113549) ## pkcs(1) pkcs-1(1) rsaEncryption(1) o <- oid("1.2.840.113549.1.1.1") as.raw(o) as.integer(o) as.character(o) as.oid(as.integer(o)) == o is.oid(o) (a <- ASN1.encode(o)) as.oid(ASN1.decode(a)) == o## RSA algorithm OID: ## iso(1) member-body(2) us(840) rsadsi(113549) ## pkcs(1) pkcs-1(1) rsaEncryption(1) o <- oid("1.2.840.113549.1.1.1") as.raw(o) as.integer(o) as.character(o) as.oid(as.integer(o)) == o is.oid(o) (a <- ASN1.encode(o)) as.oid(ASN1.decode(a)) == o
PKI.encrypt encrypts a raw vector
PKI.decrypt decrypts a raw vector
PKI.encrypt(what, key, cipher = NULL, iv = NULL) PKI.decrypt(what, key, cipher = NULL, iv = NULL)PKI.encrypt(what, key, cipher = NULL, iv = NULL) PKI.decrypt(what, key, cipher = NULL, iv = NULL)
what |
raw vector to encrypt/decrypt. It must not exceed the key size minus padding |
key |
key to use for encryption/decryption |
cipher |
cipher to use for encryption/decryption |
iv |
initialization vector for ciphers that use it (e.g.,
CBC). |
Raw vector (encrypted/decrypted)
The cipher is optional for key objects that already contain the cipher information such as RSA keys (in fact it is ignored in that case).
Supported symmetric ciphers are AES-128, AES-256 and BF
(blowfish). Each cipher can be used in CBC (default), ECB or OFB
modes which are specified as suffix, so "aes256ofb" would
specify AES-256 in OFB mode. Case and non-alphanumeric characters are
ignored, so the same could be specified as "AES-256-OFB".
PKCS padding is used to fill up to the block size. Analogously, PKCS
padding is expected when decoding.
Note that the payload for RSA encryption should be very small since it must fit into the key size including padding. For example, 1024-bit key can only encrypt 87 bytes, while 2048-bit key can encrypt 215 bytes. Therefore a typical use is to use RSA to transfer a symmeric key to the peer and subsequently use symmetric ciphers like AES for encryption of larger amounts of data.
Simon Urbanek
key <- PKI.genRSAkey(2048) x <- charToRaw("Hello, world!") e <- PKI.encrypt(x, key) y <- PKI.decrypt(e, key) stopifnot(identical(x, y)) print(rawToChar(y)) ## AES symmetric - use SHA256 to support arbitrarily long key strings key <- PKI.digest(charToRaw("hello"), "SHA256") ae <- PKI.encrypt(x, key, "aes256") ae ad <- PKI.decrypt(ae, key, "aes256") stopifnot(identical(x, ad))key <- PKI.genRSAkey(2048) x <- charToRaw("Hello, world!") e <- PKI.encrypt(x, key) y <- PKI.decrypt(e, key) stopifnot(identical(x, y)) print(rawToChar(y)) ## AES symmetric - use SHA256 to support arbitrarily long key strings key <- PKI.digest(charToRaw("hello"), "SHA256") ae <- PKI.encrypt(x, key, "aes256") ae ad <- PKI.decrypt(ae, key, "aes256") stopifnot(identical(x, ad))
PKI.digest computes digsest sum based on the hash function
specified
PKI.digest(what, hash = c("SHA1", "SHA256", "MD5"))PKI.digest(what, hash = c("SHA1", "SHA256", "MD5"))
what |
raw vector of bytes to digest |
hash |
type of the hash function. Note that "MD5" should not be used for cryptographic purposes as it is not secure |
Raw vector containg the hash
Simon Urbanek
PKI.digest(as.raw(1:10))PKI.digest(as.raw(1:10))
PKI.genpass generates n cryptographically strong
pseudo-random password by using a given set of allowed characters.
PKI.genpass(n=15, set=c(alphanum, ".", "/"), block=5, sep="-")PKI.genpass(n=15, set=c(alphanum, ".", "/"), block=5, sep="-")
n |
positive integer, number of random elements in the password |
set |
character vector, set of characters to use in the
password, ideally its length should be a power of 2 and must be at
most 256. Internal variable |
block |
non-negative integer, number of character blocks in the password or 0 if no separated blocks are desired. |
sep |
string, separator between blocks (only used if 0 <
|
PKI.genpass generates a password based on a set of allowable
characters by subsetting the set with bytes generated using
PKI.random.
If block is >0 and <n then blocks of block
characters are separated by the separator string sep. This is
typically used to guarantee at least one special character in the
password. The default results in a 90-bit random password of the form
XXXXX-XXXXX-XXXXX.
String, generated password.
This is just a utility front-end to PKI.random(n) to
subset set modulo its length. If the set does not have a length
which is a power of 2 then a warning is issued and the leading
elements are more likely to be used, reducing the password strength.
Simon Urbanek
PKI.genpass()PKI.genpass()
PKI.info returns information about the engine which is
providing the PKI functionality.
PKI.info()PKI.info()
Named list:
engine |
string, name of the engine, currently either
|
version |
numeric, version of the engine as a real number in the
form |
description |
string, description of the engine, its version and any futher information that the engine may provide |
This function should be treated as informational only. The return value is subject to change, mainly we may extend it to possibly supply information on available ciphers etc.
Older versions of OpenSSL did not provide functional API to retrieve version inforation, so versions < 1.1 may not reflect the true version, but rather the values from the headers at compile time which may not be the same as the loaded library at run-time.
Simon Urbanek
str(PKI.info())str(PKI.info())
PKI.random generates n cryptographically strong
pseudo-random bytes.
PKI.random(n)PKI.random(n)
n |
non-negative integer, number of bytes to generate |
PKI.random is the preferred way to generate cryptographically
strong random content that can be used as keys, seeds etc. Not to be
confused with random number generators in R, it is entirely separate
for cryptographics purposes.
Raw vector of n cryptographically strong pseudo-random bytes.
Simon Urbanek
PKI.random(10)PKI.random(10)
PKI.sign signs content using RSA with the specified hash function
PKI.verify verifies a signature of RSA-signed content
PKI.sign(what, key, hash = c("SHA1", "SHA256", "MD5"), digest) PKI.verify(what, signature, key, hash = c("SHA1", "SHA256", "MD5"), digest)PKI.sign(what, key, hash = c("SHA1", "SHA256", "MD5"), digest) PKI.verify(what, signature, key, hash = c("SHA1", "SHA256", "MD5"), digest)
what |
raw vector: content to sign |
key |
RSA private key to use for signing; RSA public key or certificate to use for verification. |
hash |
hash function to use. "MD5" should not be used unless absolutely needed for compatibility as it is less secure. |
digest |
raw vector: it is possible to supply the digest of
the content directly instead of specifying |
signature |
raw vector: signature |
Objects are signed by computing a hash function digest (typically
using SHA1 hash function) and then signing the digest with a
RSA key. Verification is done by computing the digest and then
comparing the signature to the digest. Private key is needed for
signing whereas public key is needed for verification.
Both functions call PKI.digest on what if
digest is not specified.
PKI.sign signature (raw vector)
PKI.verify logical: TRUE if the digest and signature
match, FALSE otherwise
Simon Urbanek
PKI.pubkey, PKI.genRSAkey,
PKI.digest
key <- PKI.genRSAkey(2048) x <- charToRaw("My message to sign") sig <- PKI.sign(x, key) stopifnot(PKI.verify(x, sig, key))key <- PKI.genRSAkey(2048) x <- charToRaw("My message to sign") sig <- PKI.sign(x, key) stopifnot(PKI.verify(x, sig, key))
PKI.sign.tar appends a signature to a tar file
PKI.verify.tar verifies the signature in a tar file
PKI.sign.tar(tarfile, key, certificate, output = tarfile) PKI.verify.tar(tarfile, key, silent = FALSE, enforce.cert = FALSE)PKI.sign.tar(tarfile, key, certificate, output = tarfile) PKI.verify.tar(tarfile, key, silent = FALSE, enforce.cert = FALSE)
tarfile |
string, file name of the file to sign |
key |
|
certificate |
optional, certificate to embed in the signature
with the public key matching |
output |
file name, connection or raw vector determining how to store the signed tar file |
silent |
if |
enforce.cert |
if |
PKI.tar.sign adds extra entry .signature with the
signature based on the contents of the tarfile. Note that
any existing signatures are retained. key is a mandatory
private key used to sign the content. certificate is optional
but if present, it will be embedded in the signature.
The tarfile can be in compressed form (gzip, bzip2 or xz) in
which case it is decompressed internally before the signature is
applied. If output is a file name then the same compression is
applied to the output, otherwise the output is uncompressed.
PKI.verify.tar retrieves the last .signature entry from
the tar file (if tarfile is a file name then the same
compression auto-detection is applied as above) and
verifies the signature against either the supplied (public) key
or against the key or certificate stored in the signature. The result
is TRUE or FALSE except when enforce.cert is
set. In that case the result is the certificate contained in the
signature if the validation succeeded (and thus it can be further
verified against a chain of trust), otherwise FALSE.
The signature format is ASN.1 DER encoded as follows:
SEQ(signature BITSTRING, subjectPublicKeyInfo, Certificate[opt])
The subjectPublicKeyInfo can be NULL in which case the
certificate must be present (in X.509 DER format).
The signature is appended as tar entry named
.signature. However, terminating blocks are not removed from
the file, so the signature is placed after the EOF blocks and thus
doesn't affect extraction.
Simon Urbanek
raw2hex convers a raw vector into hexadecimal representation
raw2hex(what, sep, upper = FALSE)raw2hex(what, sep, upper = FALSE)
what |
raw vector |
sep |
optional separator string |
upper |
logical, if |
If sep is omitted or NULL then the resulting character
vector will have as many elements as the raw vector. Otherwise the
elements are concatenated using the specified separator into one
character string. This is much more efficient than using
paste(raw2hex(x), collapse=sep), but has the same effect.
Character vector with the hexadecimal representation of the raw vector.
Simon Urbanek
raw2hex(PKI.digest(raw(), "SHA1"), "") raw2hex(PKI.digest(raw(), "MD5"), ":") ## this is jsut a performance comparison and a test that ## raw2hex can handle long strings x <- as.raw(runif(1e5) * 255.9) system.time(h1 <- raw2hex(x, " ")) system.time(h2 <- paste(raw2hex(x), collapse=" ")) stopifnot(identical(h1, h2))raw2hex(PKI.digest(raw(), "SHA1"), "") raw2hex(PKI.digest(raw(), "MD5"), ":") ## this is jsut a performance comparison and a test that ## raw2hex can handle long strings x <- as.raw(runif(1e5) * 255.9) system.time(h1 <- raw2hex(x, " ")) system.time(h2 <- paste(raw2hex(x), collapse=" ")) stopifnot(identical(h1, h2))
PKI.load.key loads an RSA key in PKCS#1/8 PEM or DER format.
PKI.save.key creates a PEM or DER representation of a RSA key.
PKI.genRSAkey generates RSA public/private key pair.
PKI.mkRSApubkey creates a RSA public key with the supplied
modulus and exponent.
PKI.load.OpenSSH.pubkey loads public key in OpenSSH format
(as used in .ssh/authorized_keys file)
PKI.load.key(what, format = c("PEM", "DER"), private, file, password="") PKI.save.key(key, format = c("PEM", "DER"), private, target) PKI.genRSAkey(bits = 2048L) PKI.mkRSApubkey(modulus, exponent=65537L, format = c("DER", "PEM", "key")) PKI.load.OpenSSH.pubkey(what, first=TRUE, format = c("DER", "PEM", "key"))PKI.load.key(what, format = c("PEM", "DER"), private, file, password="") PKI.save.key(key, format = c("PEM", "DER"), private, target) PKI.genRSAkey(bits = 2048L) PKI.mkRSApubkey(modulus, exponent=65537L, format = c("DER", "PEM", "key")) PKI.load.OpenSSH.pubkey(what, first=TRUE, format = c("DER", "PEM", "key"))
what |
string, raw vector or connection to load the key from |
key |
RSA key object |
format |
format - PEM is ASCII (essentially base64-encoded DER with header/footer), DER is binary and key means an acutal key object |
private |
logical, whether to use the private key ( |
file |
filename to load the key from - |
password |
string, used only if |
target |
optional connection or a file name to store the result in. If missing, the result is just returned form the function as either a character vector (PEM) or a raw vector (DER). |
bits |
size of the generated key in bits. Must be |
modulus |
modulus either as a raw vector (see
|
exponent |
exponent either as a raw vector (see
|
first |
logical, if |
PKI.load.key: private or public key object
PKI.save.key: raw vector (DER format) or character vector (PEM
format).
PKI.genRSAkey: private + public key object
PKI.mkRSApubkey, PKI.load.OpenSSH.pubkey: raw vector
(DER format) or character vector (PEM format) or a "public.key"
object.
The output format for private keys in PEM is PKCS#1, but for public keys it is X.509 SubjectPublicKeyInfo (certificate public key). This is consistent with OpenSSL RSA command line tool which uses the same convention.
PKI.load.key can auto-detect the contained format based on
the header if 'PEM' format is used. In that case it supports PKCS#1
(naked RSA key), PKCS#8 (wrapped key with identifier - for public
keys X.509 SubjectPublicKeyInfo) and encrypted private key in
PKCS#8 (password must be passed to decrypt). 'DER' format provides no
way to define the type so 'private' cannot be 'NA' and only the
default format (PKCS#1 for private keys and X.509
SubjectPublicKeyInfo for public keys) is supported.
The OpenSSH format is one line beginning with "ssh-rsa ".
SSH2 PEM public keys (rfc4716) are supported in PKI.load.key
and the binary payload is the same as the OpenSSH, only with
different wrapping.
Simon Urbanek
PKI.encrypt, PKI.decrypt, PKI.pubkey
# generate 2048-bit RSA key key <- PKI.genRSAkey(bits = 2048L) # extract private and public parts as PEM priv.pem <- PKI.save.key(key) pub.pem <- PKI.save.key(key, private=FALSE) # load back the public key separately pub.k <- PKI.load.key(pub.pem) # encrypt with the public key x <- PKI.encrypt(charToRaw("Hello, world!"), pub.k) # decrypt with private key rawToChar(PKI.decrypt(x, key)) # compute SHA1 hash (fingerprint) of the public key PKI.digest(PKI.save.key(key, "DER", private=FALSE)) # convert OpenSSH public key to PEM format # (the example is split into multiple lines just # so it is readable in the documentation, in reality you can # simply use the full line from is_rsa.pub without gsub) PKI.load.OpenSSH.pubkey(gsub("\n","", "ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAuvOXqfZ3pJeWeqyQOIXZwmg M1RBqPUmVx3XgntpA+YtOZjKfuoJSpg3LhBuI/wXx8L2QZXNFibvX4qX2qoYsb Hvkz2uonA3F7HRhCR/BJURR5nT135znVqALZo328v86HDsVWYR2/JzY1X8GI2R 2iKUMGXF0hVuRphdwLB735CU= foo@mycomputer"), format="PEM")# generate 2048-bit RSA key key <- PKI.genRSAkey(bits = 2048L) # extract private and public parts as PEM priv.pem <- PKI.save.key(key) pub.pem <- PKI.save.key(key, private=FALSE) # load back the public key separately pub.k <- PKI.load.key(pub.pem) # encrypt with the public key x <- PKI.encrypt(charToRaw("Hello, world!"), pub.k) # decrypt with private key rawToChar(PKI.decrypt(x, key)) # compute SHA1 hash (fingerprint) of the public key PKI.digest(PKI.save.key(key, "DER", private=FALSE)) # convert OpenSSH public key to PEM format # (the example is split into multiple lines just # so it is readable in the documentation, in reality you can # simply use the full line from is_rsa.pub without gsub) PKI.load.OpenSSH.pubkey(gsub("\n","", "ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAuvOXqfZ3pJeWeqyQOIXZwmg M1RBqPUmVx3XgntpA+YtOZjKfuoJSpg3LhBuI/wXx8L2QZXNFibvX4qX2qoYsb Hvkz2uonA3F7HRhCR/BJURR5nT135znVqALZo328v86HDsVWYR2/JzY1X8GI2R 2iKUMGXF0hVuRphdwLB735CU= foo@mycomputer"), format="PEM")
PKI.load.cert creates a certificate object from a string,
connection or file.
PKI.verifyCA verifies a certificate against a given chain of
trust.
PKI.pubkey extracts public key from a certificate.
PKI.get.subject extracts the subject name from the certificate.
PKI.get.cert.info decodes information from the certificate.
PKI.load.cert(what, format = c("PEM", "DER"), file) PKI.verifyCA(certificate, ca, default = FALSE, partial = FALSE) PKI.pubkey(certificate) PKI.get.subject(certificate) PKI.get.cert.info(certificate)PKI.load.cert(what, format = c("PEM", "DER"), file) PKI.verifyCA(certificate, ca, default = FALSE, partial = FALSE) PKI.pubkey(certificate) PKI.get.subject(certificate) PKI.get.cert.info(certificate)
what |
string, raw vector or connection to load the certificate from |
format |
format used to encode the certificate |
file |
filename to load the certificate from - |
certificate |
a certificate object (as returned by
|
ca |
a certificate object of the Certificate Authority (CA) or a list of such objects if a chain of certificates is involved |
default |
logical, if |
partial |
logical, if |
PKI.verifyCA is used to verify the validity of a certificate
by following a chain of trust. In the most simple case the
certificate was issued by a certificate authority (CA) directly,
which has a self-signed certificate. This is typically the case when
you (or your organization) have created your own CA for internal use.
In that case you only need to supply that CA's certificate to
ca and that's it. It is also possible that your self-signed
CA issued an intermediate certificate - if that is the case then pass
a list of both certificates (order doesn't matter) to ca.
Another use case is that you have a certificate which has been issued
by publicly trusted CA - this is commonly the case with SSL
certificates used by web servers. In that case, the chain doesn't
end with an internal self-signed certificate, but instead it will
end with a publicly known root CA. OpenSSL manages a list of such
trusted CAs and you can check against them with
default=TRUE. However, in most cases your certificate won't
be issued directly by a root CA, but by an intermetiate authority so
you have to pass the intermediate certificate(s) in the ca
argument.
Finally, it is sometimes possible that the default list of trusted
certificates does not include the root CA that you need. If that is
the case, and you still want to trust that chain, you can set
partial=TRUE and then PKI.verifyCA will trust the
certificates provided in ca unconditinally, even if they
don't lead to a trusted root or are not self-signed. Note, however,
that this is the least secure option and you should only use it if
the certificates are supplied by you and not the user. If you want
to support user-supplied intermediate certificates then you can use
PKI.verifyCA first to verify the integrity of the
user-supplied chain with partial=TRUE and then verify just
the intermediate certificate against your trusted certificate. That
way you won't trust the intermediate certificate inadvertently.
PKI.load.cert: a certificate object
PKI.verifyCA: TRUE is the certificate can be trusted,
FALSE otherwise
PKI.pubkey: public key object
PKI.get.subject: string containing the subject information in
one-line RFC2253 format but in UTF8 encoding instead of MBS escapes.
NOTE: this is experimantal, we may choose to parse the contents and
return it in native R form as a named vector instead.
Simon Urbanek
(ca <- PKI.load.cert(file=system.file("certs", "RForge-ca.crt", package="PKI"))) (my.cert <- PKI.load.cert(readLines(system.file("certs", "demo.crt", package="PKI")))) PKI.verifyCA(my.cert, ca) PKI.pubkey(my.cert) PKI.get.subject(my.cert) PKI.get.cert.info(my.cert)(ca <- PKI.load.cert(file=system.file("certs", "RForge-ca.crt", package="PKI"))) (my.cert <- PKI.load.cert(readLines(system.file("certs", "demo.crt", package="PKI")))) PKI.verifyCA(my.cert, ca) PKI.pubkey(my.cert) PKI.get.subject(my.cert) PKI.get.cert.info(my.cert)