Code cleanup (#96)
diff --git a/lib/src/char_encodings.dart b/lib/src/char_encodings.dart
deleted file mode 100644
index 95755d0..0000000
--- a/lib/src/char_encodings.dart
+++ /dev/null
@@ -1,40 +0,0 @@
-import 'dart:convert' show ascii, utf8;
-
-// TODO(jmesserly): this function is conspicuously absent from dart:utf.
-/// Returns true if the [bytes] starts with a UTF-8 byte order mark.
-/// Since UTF-8 doesn't have byte order, it's somewhat of a misnomer, but it is
-/// used in HTML to detect the UTF-
-bool hasUtf8Bom(List<int> bytes, [int offset = 0, int length]) {
- int end = length != null ? offset + length : bytes.length;
- return (offset + 3) <= end &&
- bytes[offset] == 0xEF &&
- bytes[offset + 1] == 0xBB &&
- bytes[offset + 2] == 0xBF;
-}
-
-// TODO(jmesserly): it's unfortunate that this has to be one-shot on the entire
-// file, but dart:utf does not expose stream-based decoders yet.
-/// Decodes the [bytes] with the provided [encoding] and returns an iterable for
-/// the codepoints. Supports the major unicode encodings as well as ascii and
-/// and windows-1252 encodings.
-Iterable<int> decodeBytes(String encoding, List<int> bytes) {
- switch (encoding) {
- case 'ascii':
- return ascii.decode(bytes).runes;
-
- case 'utf-8':
- // NOTE: To match the behavior of the other decode functions, we eat the
- // UTF-8 BOM here. This is the default behavior of `utf8.decode`.
- return utf8.decode(bytes).runes;
-
- default:
- throw ArgumentError('Encoding $encoding not supported');
- }
-}
-
-// TODO(jmesserly): use dart:utf once http://dartbug.com/6476 is fixed.
-/// Returns the code points for the [input]. This works like [String.charCodes]
-/// but it decodes UTF-16 surrogate pairs.
-List<int> toCodepoints(String input) {
- return input.runes.toList();
-}
diff --git a/lib/src/html_input_stream.dart b/lib/src/html_input_stream.dart
index 84d2f13..42b1741 100644
--- a/lib/src/html_input_stream.dart
+++ b/lib/src/html_input_stream.dart
@@ -1,8 +1,8 @@
import 'dart:collection';
+import 'dart:convert' show ascii, utf8;
import 'package:source_span/source_span.dart';
-import 'char_encodings.dart';
import 'constants.dart';
import 'encoding_parser.dart';
import 'utils.dart';
@@ -66,7 +66,7 @@
this.sourceUrl])
: charEncodingName = codecName(encoding) {
if (source is String) {
- _rawChars = toCodepoints(source);
+ _rawChars = source.runes.toList();
charEncodingName = 'utf-8';
charEncodingCertain = true;
} else if (source is List<int>) {
@@ -92,7 +92,7 @@
_chars = <int>[];
if (_rawChars == null) {
- _rawChars = decodeBytes(charEncodingName, _rawBytes);
+ _rawChars = _decodeBytes(charEncodingName, _rawBytes);
}
bool skipNewline = false;
@@ -177,7 +177,7 @@
/// encoding otherwise return null.
String detectBOM() {
// Try detecting the BOM using bytes from the string
- if (hasUtf8Bom(_rawBytes)) {
+ if (_hasUtf8Bom(_rawBytes)) {
return 'utf-8';
}
return null;
@@ -292,3 +292,32 @@
var canonicalName = encoding.replaceAll(asciiPunctuation, '').toLowerCase();
return encodings[canonicalName];
}
+
+/// Returns true if the [bytes] starts with a UTF-8 byte order mark.
+/// Since UTF-8 doesn't have byte order, it's somewhat of a misnomer, but it is
+/// used in HTML to detect the UTF-
+bool _hasUtf8Bom(List<int> bytes, [int offset = 0, int length]) {
+ int end = length != null ? offset + length : bytes.length;
+ return (offset + 3) <= end &&
+ bytes[offset] == 0xEF &&
+ bytes[offset + 1] == 0xBB &&
+ bytes[offset + 2] == 0xBF;
+}
+
+/// Decodes the [bytes] with the provided [encoding] and returns an iterable for
+/// the codepoints. Supports the major unicode encodings as well as ascii and
+/// and windows-1252 encodings.
+Iterable<int> _decodeBytes(String encoding, List<int> bytes) {
+ switch (encoding) {
+ case 'ascii':
+ return ascii.decode(bytes).runes;
+
+ case 'utf-8':
+ // NOTE: To match the behavior of the other decode functions, we eat the
+ // UTF-8 BOM here. This is the default behavior of `utf8.decode`.
+ return utf8.decode(bytes).runes;
+
+ default:
+ throw ArgumentError('Encoding $encoding not supported');
+ }
+}
