Skip to content

Commit 45b128b

Browse files
committed
Add whitespace \xa0
1 parent 3f768a5 commit 45b128b

File tree

2 files changed

+8
-3
lines changed

2 files changed

+8
-3
lines changed

src/main/kotlin/ink/meodinger/htmlparser/HParser.kt

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -67,7 +67,7 @@ fun parse(htmlText: String): HPage {
6767
builder.append(tokenStream.next().value).append(tokenStream.next().value)
6868
}
6969

70-
// I won't add more ill-format fix like href=http://xxx.yyy anymore.
70+
// I won't add more ill-format fix like href=http://xxx.yyy.
7171

7272
builder.toString()
7373
}

src/main/kotlin/ink/meodinger/htmlparser/internal/TokenStream.kt

Lines changed: 7 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -21,6 +21,11 @@ class TokenStream(private val stringStream: StringStream) {
2121
* Token
2222
*/
2323
data class Token(val type: TokenType, val value: String) {
24+
25+
companion object {
26+
val EOF: Token = Token(TokenType.EOF, "")
27+
}
28+
2429
fun isEOF(): Boolean = type == TokenType.EOF
2530
fun isText(): Boolean = type == TokenType.TEXT
2631
fun isSymbol(): Boolean = type == TokenType.SYMBOL
@@ -45,7 +50,7 @@ class TokenStream(private val stringStream: StringStream) {
4550
addAll('A'..'Z')
4651
addAll('0'..'9')
4752
}.toCharArray()
48-
private val Whitespaces: CharArray = charArrayOf(' ', '\n', '\t', '\r', '\uFEFF')
53+
private val Whitespaces: CharArray = charArrayOf(' ', '\n', '\t', '\r', '\uFEFF', '\u3000', '\u00A0')
4954
private val Symbols: CharArray = charArrayOf('=', '<', '>', '/')
5055

5156
private fun isCommentStart(char: Char): Boolean = CommentHeads.contains(char)
@@ -135,7 +140,7 @@ class TokenStream(private val stringStream: StringStream) {
135140

136141
private fun takeNext(): Token {
137142
readWhile(Companion::isWhitespace)
138-
if (eof()) return Token(TokenType.EOF, "")
143+
if (eof()) return Token.EOF
139144

140145
val char = stringStream.peek()
141146

0 commit comments

Comments
 (0)