Skip to content

Base64 XML variants #16586

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 2 commits into from
Closed

Base64 XML variants #16586

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a137777
serialize/base64: Add XML-specific encodings
gioele Aug 18, 2014
74a70b3
serialize/base64: Make from_base64 config-aware
gioele Aug 18, 2014
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
106 changes: 79 additions & 27 deletions src/libserialize/base64.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,11 @@ pub enum CharacterSet {
/// The standard character set (uses `+` and `/`)
Standard,
/// The URL safe character set (uses `-` and `_`)
UrlSafe
UrlSafe,
/// The XML/Name character set (uses `_` and `:`)
XmlName,
/// The XML/Nmtoken character set (uses `.` and `-`)
XmlNmtoken,
}

/// Contains configuration parameters for `to_base64`.
Expand All @@ -44,6 +48,14 @@ pub static URL_SAFE: Config =
pub static MIME: Config =
Config {char_set: Standard, pad: true, line_length: Some(76)};

/// Configuration for base64 encoding acceptable in XML element names
pub static XML_NAME: Config =
Config {char_set: XmlName, pad: false, line_length: None};

/// Configuration for base64 encoding acceptable in XML NMTOKENs
pub static XML_NMTOKEN: Config =
Config {char_set: XmlNmtoken, pad: false, line_length: None};

static STANDARD_CHARS: &'static[u8] = b"ABCDEFGHIJKLMNOPQRSTUVWXYZ\
abcdefghijklmnopqrstuvwxyz\
0123456789+/";
Expand All @@ -52,6 +64,14 @@ static URLSAFE_CHARS: &'static[u8] = b"ABCDEFGHIJKLMNOPQRSTUVWXYZ\
abcdefghijklmnopqrstuvwxyz\
0123456789-_";

static XML_NAME_CHARS: &'static[u8] = b"ABCDEFGHIJKLMNOPQRSTUVWXYZ\
abcdefghijklmnopqrstuvwxyz\
0123456789_:";

static XML_NMTOKEN_CHARS: &'static[u8] = b"ABCDEFGHIJKLMNOPQRSTUVWXYZ\
abcdefghijklmnopqrstuvwxyz\
0123456789.-";

/// A trait for converting a value to base64 encoding.
pub trait ToBase64 {
/// Converts the value of `self` to a base64 value following the specified
Expand All @@ -78,7 +98,9 @@ impl<'a> ToBase64 for &'a [u8] {
fn to_base64(&self, config: Config) -> String {
let bytes = match config.char_set {
Standard => STANDARD_CHARS,
UrlSafe => URLSAFE_CHARS
UrlSafe => URLSAFE_CHARS,
XmlNmtoken => XML_NMTOKEN_CHARS,
XmlName => XML_NAME_CHARS,
};

let mut v = Vec::new();
Expand Down Expand Up @@ -157,7 +179,7 @@ impl<'a> ToBase64 for &'a [u8] {
pub trait FromBase64 {
/// Converts the value of `self`, interpreted as base64 encoded data, into
/// an owned vector of bytes, returning the vector.
fn from_base64(&self) -> Result<Vec<u8>, FromBase64Error>;
fn from_base64(&self, config: Config) -> Result<Vec<u8>, FromBase64Error>;
}

/// Errors that can occur when decoding a base64 encoded string
Expand Down Expand Up @@ -197,7 +219,7 @@ impl<'a> FromBase64 for &'a str {
* fn main () {
* let hello_str = b"Hello, World".to_base64(STANDARD);
* println!("base64 output: {}", hello_str);
* let res = hello_str.as_slice().from_base64();
* let res = hello_str.as_slice().from_base64(STANDARD);
* if res.is_ok() {
* let opt_bytes = String::from_utf8(res.unwrap());
* if opt_bytes.is_ok() {
Expand All @@ -208,17 +230,26 @@ impl<'a> FromBase64 for &'a str {
* ```
*/
#[inline]
fn from_base64(&self) -> Result<Vec<u8>, FromBase64Error> {
self.as_bytes().from_base64()
fn from_base64(&self, config: Config) -> Result<Vec<u8>, FromBase64Error> {
self.as_bytes().from_base64(config)
}
}

impl<'a> FromBase64 for &'a [u8] {
fn from_base64(&self) -> Result<Vec<u8>, FromBase64Error> {
fn from_base64(&self, config: Config) -> Result<Vec<u8>, FromBase64Error> {
let mut r = Vec::new();
let mut buf: u32 = 0;
let mut modulus = 0i;

let chars = match config.char_set {
Standard => STANDARD_CHARS,
UrlSafe => URLSAFE_CHARS,
XmlNmtoken => XML_NMTOKEN_CHARS,
XmlName => XML_NAME_CHARS,
};

let (c62, c63) = (chars[62], chars[63]);

let mut it = self.iter().enumerate();
for (idx, &byte) in it {
let val = byte as u32;
Expand All @@ -227,8 +258,8 @@ impl<'a> FromBase64 for &'a [u8] {
b'A'..b'Z' => buf |= val - 0x41,
b'a'..b'z' => buf |= val - 0x47,
b'0'..b'9' => buf |= val + 0x04,
b'+' | b'-' => buf |= 0x3E,
b'/' | b'_' => buf |= 0x3F,
b if (b == c62) => buf |= 0x3E,
b if (b == c63) => buf |= 0x3F,
b'\r' | b'\n' => continue,
b'=' => break,
_ => return Err(InvalidBase64Byte(self[idx], idx)),
Expand Down Expand Up @@ -271,7 +302,7 @@ impl<'a> FromBase64 for &'a [u8] {
mod tests {
extern crate test;
use self::test::Bencher;
use base64::{Config, FromBase64, ToBase64, STANDARD, URL_SAFE};
use base64::{Config, FromBase64, ToBase64, STANDARD, URL_SAFE, XML_NAME, XML_NMTOKEN};

#[test]
fn test_to_base64_basic() {
Expand Down Expand Up @@ -306,44 +337,66 @@ mod tests {
assert_eq!([251, 255].to_base64(STANDARD), "+/8=".to_string());
}

#[test]
fn test_to_base64_xml_token() {
assert_eq!([251, 255].to_base64(XML_NMTOKEN), ".-8".to_string());
}

#[test]
fn test_to_base64_xml_name() {
assert_eq!([251, 255].to_base64(XML_NAME), "_:8".to_string());
}

#[test]
fn test_from_base64_basic() {
assert_eq!("".from_base64().unwrap().as_slice(), "".as_bytes());
assert_eq!("Zg==".from_base64().unwrap().as_slice(), "f".as_bytes());
assert_eq!("Zm8=".from_base64().unwrap().as_slice(), "fo".as_bytes());
assert_eq!("Zm9v".from_base64().unwrap().as_slice(), "foo".as_bytes());
assert_eq!("Zm9vYg==".from_base64().unwrap().as_slice(), "foob".as_bytes());
assert_eq!("Zm9vYmE=".from_base64().unwrap().as_slice(), "fooba".as_bytes());
assert_eq!("Zm9vYmFy".from_base64().unwrap().as_slice(), "foobar".as_bytes());
assert_eq!("".from_base64(STANDARD).unwrap().as_slice(), "".as_bytes());
assert_eq!("Zg==".from_base64(STANDARD).unwrap().as_slice(), "f".as_bytes());
assert_eq!("Zm8=".from_base64(STANDARD).unwrap().as_slice(), "fo".as_bytes());
assert_eq!("Zm9v".from_base64(STANDARD).unwrap().as_slice(), "foo".as_bytes());
assert_eq!("Zm9vYg==".from_base64(STANDARD).unwrap().as_slice(), "foob".as_bytes());
assert_eq!("Zm9vYmE=".from_base64(STANDARD).unwrap().as_slice(), "fooba".as_bytes());
assert_eq!("Zm9vYmFy".from_base64(STANDARD).unwrap().as_slice(), "foobar".as_bytes());
}

#[test]
fn test_from_base64_bytes() {
assert_eq!(b"Zm9vYmFy".from_base64().unwrap().as_slice(), "foobar".as_bytes());
assert_eq!(b"Zm9vYmFy".from_base64(STANDARD).unwrap().as_slice(), "foobar".as_bytes());
}

#[test]
fn test_from_base64_newlines() {
assert_eq!("Zm9v\r\nYmFy".from_base64().unwrap().as_slice(),
assert_eq!("Zm9v\r\nYmFy".from_base64(STANDARD).unwrap().as_slice(),
"foobar".as_bytes());
assert_eq!("Zm9vYg==\r\n".from_base64().unwrap().as_slice(),
assert_eq!("Zm9vYg==\r\n".from_base64(STANDARD).unwrap().as_slice(),
"foob".as_bytes());
}

#[test]
fn test_from_base64_urlsafe() {
assert_eq!("-_8".from_base64().unwrap(), "+/8=".from_base64().unwrap());
assert_eq!("-_8".from_base64(URL_SAFE).unwrap(), "+/8=".from_base64(STANDARD).unwrap());
}

#[test]
fn test_from_base64_xml_name() {
let plain_text = "+/8=".from_base64(STANDARD).unwrap();
assert_eq!("_:8".from_base64(XML_NAME).unwrap(), plain_text);
}

#[test]
fn test_from_base64_xml_nmtoken() {
let plain_text = "+/8=".from_base64(STANDARD).unwrap();
assert_eq!(".-8".from_base64(XML_NMTOKEN).unwrap(), plain_text);
}

#[test]
fn test_from_base64_invalid_char() {
assert!("Zm$=".from_base64().is_err())
assert!("Zg==$".from_base64().is_err());
assert!("Zm$=".from_base64(STANDARD).is_err())
assert!("Zg==$".from_base64(STANDARD).is_err());
}

#[test]
fn test_from_base64_invalid_padding() {
assert!("Z===".from_base64().is_err());
assert!("Z===".from_base64(STANDARD).is_err());
}

#[test]
Expand All @@ -356,7 +409,7 @@ mod tests {
assert_eq!(v.as_slice()
.to_base64(STANDARD)
.as_slice()
.from_base64()
.from_base64(STANDARD)
.unwrap()
.as_slice(),
v.as_slice());
Expand All @@ -379,9 +432,8 @@ mod tests {
ウヰノオクヤマ ケフコエテ アサキユメミシ ヱヒモセスン";
let sb = s.as_bytes().to_base64(STANDARD);
b.iter(|| {
sb.as_slice().from_base64().unwrap();
sb.as_slice().from_base64(STANDARD).unwrap();
});
b.bytes = sb.len() as u64;
}

}