[LUCENE-10501] StackOverflow when RegExp encounters a very large string

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 9.1
• Component/s: core/queryparser
• Lucene Fields: New
• Assignee: Unassigned
• Reporter: Kartik Ganesh
• Created: 05/Apr/22 21:46
• Updated: 28/Aug/22 16:38

Description

When RegExp encounters a very large string, it hits a Stack Overflow exception when parsing it.

Simple program to repro:

$ ls
RegExpTest.java       lucene-core-9.1.0.jar
$ cat RegExpTest.java
class RegExpTest {
    public static void main(String[] args) {
        StringBuilder strBuilder = new StringBuilder();
        for (int i = 0; i < 50000; i++) {
            strBuilder.append("a");
        }
        try {
            new org.apache.lucene.util.automaton.RegExp(strBuilder.toString());
        } catch (StackOverflowError e) {
            System.out.println("Stack overflow");
            System.exit(-1);
        }
        System.out.println("Success");
    }
}
$ javac -cp './lucene-core-9.1.0.jar:.' RegExpTest.java
$ java -cp './lucene-core-9.1.0.jar:.' RegExpTest
Stack overflow
$ java -Xss1G -cp './lucene-core-9.1.0.jar:.' RegExpTest
Success

Based on https://issues.apache.org/jira/browse/LUCENE-6156 , this appears to be due to the recursive parsing implementation.