mirror of
https://github.com/internetarchive/heritrix3.git
synced 2026-09-26 15:46:31 +00:00
Allow spaces in JavaScript urls, but only if they have a known good file extension
This commit is contained in:
@@ -67,7 +67,7 @@ public class ExtractorJS extends ContentExtractor {
|
||||
// (areas between paired ' or " characters, possibly backslash-quoted
|
||||
// on the ends, but not in the middle)
|
||||
protected static final String JAVASCRIPT_STRING_EXTRACTOR =
|
||||
"(\\\\{0,8}+(?:['\"]|u002[27]))([^\\s'\"]{1,"+UURI.MAX_URL_LENGTH+"})(?:\\1)";
|
||||
"(\\\\{0,8}+(?:['\"]|u002[27]))([^'\"]{0,"+UURI.MAX_URL_LENGTH+"})(?:\\1)";
|
||||
|
||||
// GROUPS:
|
||||
// (G1) ' or " with optional leading backslashes
|
||||
|
||||
Reference in New Issue
Block a user