JoBoEVD/UglifyJS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Improve multi-line comment parsing

Browse Source 
* Make sure comments are skipped correctly with surrogates
* Fix regression in multiline comments with nlb
This commit is contained in:
Anthony Van de Gejuchte
2016-07-04 14:30:27 +02:00
parent 0db7caf13b
commit ff7f6139ba
3 changed files with 29 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
lib/parse.js
View File

@@ -148,6 +148,21 @@ function get_full_char_code(str, pos) {
return str.charCodeAt(pos);
}
function get_full_char_length(str) {
var surrogates = 0;
for (var i = 0; i < str.length; i++) {
if (str.charCodeAt(i) >= 0xd800 && str.charCodeAt(i) <= 0xdbff) {
if (str.charCodeAt(i + 1) >= 0xdc00 && str.charCodeAt(i + 1) <= 0xdfff) {
surrogates++;
i++;
}
}
}
return str.length - surrogates;
}
function from_char_code(code) {
// Based on https://github.com/mathiasbynens/String.fromCodePoint/blob/master/fromcodepoint.js
if (code > 0xFFFF) {
@@ -493,8 +508,9 @@ function tokenizer($TEXT, filename, html5_comments, shebang) {
var i = find("*/", true);
var text = S.text.substring(S.pos, i).replace(/\r\n|\r|\u2028|\u2029/g, '\n');
// update stream position
forward(text.length /* doesn't count \r\n as 2 char while S.pos - i does */ + 2);
forward(get_full_char_length(text) /* text length doesn't count \r\n as 2 char while S.pos - i does */ + 2);
S.comments_before.push(token("comment2", text, true));
S.newline_before = S.newline_before || text.indexOf("\n") >= 0;
S.regex_allowed = regex_allowed;
return next_token;
});