1
0
Fork 0
mirror of https://github.com/denoland/deno.git synced 2024-11-22 15:06:54 -05:00
denoland-deno/cli/util/text_encoding.rs

174 lines
4.6 KiB
Rust
Raw Normal View History

// Copyright 2018-2023 the Deno authors. All rights reserved. MIT license.
2020-09-05 20:34:02 -04:00
2023-10-26 12:39:04 -04:00
use base64::prelude::BASE64_STANDARD;
use base64::Engine;
use deno_core::ModuleCode;
use encoding_rs::*;
use std::borrow::Cow;
use std::io::Error;
use std::io::ErrorKind;
pub const BOM_CHAR: char = '\u{FEFF}';
/// Attempts to detect the character encoding of the provided bytes.
///
/// Supports UTF-8, UTF-16 Little Endian and UTF-16 Big Endian.
pub fn detect_charset(bytes: &'_ [u8]) -> &'static str {
const UTF16_LE_BOM: &[u8] = b"\xFF\xFE";
const UTF16_BE_BOM: &[u8] = b"\xFE\xFF";
if bytes.starts_with(UTF16_LE_BOM) {
"utf-16le"
} else if bytes.starts_with(UTF16_BE_BOM) {
"utf-16be"
} else {
// Assume everything else is utf-8
"utf-8"
}
}
/// Attempts to convert the provided bytes to a UTF-8 string.
///
/// Supports all encodings supported by the encoding_rs crate, which includes
/// all encodings specified in the WHATWG Encoding Standard, and only those
/// encodings (see: <https://encoding.spec.whatwg.org/>).
pub fn convert_to_utf8<'a>(
bytes: &'a [u8],
charset: &'_ str,
) -> Result<Cow<'a, str>, Error> {
match Encoding::for_label(charset.as_bytes()) {
Some(encoding) => encoding
.decode_without_bom_handling_and_without_replacement(bytes)
.ok_or_else(|| ErrorKind::InvalidData.into()),
None => Err(Error::new(
ErrorKind::InvalidInput,
format!("Unsupported charset: {charset}"),
)),
}
}
/// Strips the byte order mark from the provided text if it exists.
pub fn strip_bom(text: &str) -> &str {
if text.starts_with(BOM_CHAR) {
&text[BOM_CHAR.len_utf8()..]
} else {
text
}
}
static SOURCE_MAP_PREFIX: &[u8] =
b"//# sourceMappingURL=data:application/json;base64,";
pub fn source_map_from_code(code: &ModuleCode) -> Option<Vec<u8>> {
let bytes = code.as_bytes();
let last_line = bytes.rsplit(|u| *u == b'\n').next()?;
if last_line.starts_with(SOURCE_MAP_PREFIX) {
let input = last_line.split_at(SOURCE_MAP_PREFIX.len()).1;
2023-10-26 12:39:04 -04:00
let decoded_map = BASE64_STANDARD
.decode(input)
.expect("Unable to decode source map from emitted file.");
Some(decoded_map)
} else {
None
}
}
/// Truncate the source code before the source map.
pub fn code_without_source_map(mut code: ModuleCode) -> ModuleCode {
let bytes = code.as_bytes();
for i in (0..bytes.len()).rev() {
if bytes[i] == b'\n' {
if bytes[i + 1..].starts_with(SOURCE_MAP_PREFIX) {
code.truncate(i + 1);
}
return code;
}
}
code
}
#[cfg(test)]
mod tests {
use super::*;
fn test_detection(test_data: &[u8], expected_charset: &str) {
let detected_charset = detect_charset(test_data);
assert_eq!(
expected_charset.to_lowercase(),
detected_charset.to_lowercase()
);
}
#[test]
fn test_detection_utf8_no_bom() {
let test_data = "Hello UTF-8 it is \u{23F0} for Deno!"
.to_owned()
.into_bytes();
test_detection(&test_data, "utf-8");
}
#[test]
fn test_detection_utf16_little_endian() {
let test_data = b"\xFF\xFEHello UTF-16LE".to_owned().to_vec();
test_detection(&test_data, "utf-16le");
}
#[test]
fn test_detection_utf16_big_endian() {
let test_data = b"\xFE\xFFHello UTF-16BE".to_owned().to_vec();
test_detection(&test_data, "utf-16be");
}
#[test]
fn test_decoding_unsupported_charset() {
let test_data = Vec::new();
let result = convert_to_utf8(&test_data, "utf-32le");
assert!(result.is_err());
let err = result.expect_err("Err expected");
assert!(err.kind() == ErrorKind::InvalidInput);
}
#[test]
fn test_decoding_invalid_utf8() {
let test_data = b"\xFE\xFE\xFF\xFF".to_vec();
let result = convert_to_utf8(&test_data, "utf-8");
assert!(result.is_err());
let err = result.expect_err("Err expected");
assert!(err.kind() == ErrorKind::InvalidData);
}
#[test]
fn test_source_without_source_map() {
run_test("", "");
run_test("\n", "\n");
run_test("\r\n", "\r\n");
run_test("a", "a");
run_test("a\n", "a\n");
run_test("a\r\n", "a\r\n");
run_test("a\r\nb", "a\r\nb");
run_test("a\nb\n", "a\nb\n");
run_test("a\r\nb\r\n", "a\r\nb\r\n");
run_test(
"test\n//# sourceMappingURL=data:application/json;base64,test",
"test\n",
);
run_test(
"test\r\n//# sourceMappingURL=data:application/json;base64,test",
"test\r\n",
);
run_test(
"\n//# sourceMappingURL=data:application/json;base64,test",
"\n",
);
fn run_test(input: &'static str, output: &'static str) {
assert_eq!(
code_without_source_map(ModuleCode::from_static(input))
.as_str()
.to_owned(),
output
);
}
}
}