regex: java.util.regex companion code (groups, lookarounds, replaceAll lambdas, ReDoS and timeouts)

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01KqJyCidz3ZgRyHABv2GVJh

regex/output/01-basics.txt

@@ -0,0 +1,23 @@
+matches()    = true
+group(0)     = 2026-03-14 ERROR disk /dev/sda1 is 97% full
+group(2)     = ERROR
+level        = ERROR
+start(msg)   = 17, end(msg) = 43
+namedGroups  = {date=1, level=2, msg=3}
+
+matches()    on "order-42 shipped" = false
+lookingAt()  on "order-42 shipped" = false
+find()       on "order-42 shipped" = true
+
+results() over "a1 b22 c333":
+  a1 -> group(1)=1
+  b22 -> group(1)=22
+  c333 -> group(1)=333
+
+no flags            : 0
+CASE_INSENSITIVE    : 0
+MULTILINE           : 1
+MULTILINE|CASE_INS. : 3
+embedded (?im)      : 3
+'.' vs newline      : 0 / DOTALL 1
+COMMENTS            : 1

regex/output/02-lookarounds.txt

@@ -0,0 +1,12 @@
+\d+(?=px)                    -> "120", "48"
+foo(?!bar)                   -> "foo", "foo"
+(?<=\$)\d+(?:\.\d\d)?        -> "19.99", "23"
+(?<![\d$.])\d+               -> "3"
+policy abcdefgh   -> false
+policy Abcdefg1   -> true
+policy Ab1        -> false
+policy ABCDEFG1x  -> true
+1,234,567
+compiled  (?<=a+)b
+compiled  (?<=a{1,9})b
+rejected  (?<=(?:ab)+)c : Look-behind group does not have an obvious maximum length

regex/output/03-replace.txt

@@ -0,0 +1,9 @@
+price: 5 dollars, fee: 12 dollars
+price: USD 5, fee: USD 12
+replaceAll("$")  -> IllegalArgumentException: Illegal group reference: group index is missing
+price: 5 $, fee: 12 $
+price: 10 USD, fee: 24 USD
+unquoted lambda  -> IndexOutOfBoundsException: No group 5
+hello ankur, you owe $5
+hi ${nobody}
+rEgUlAr ExprEssIOns

regex/output/04-blowup.txt

@@ -0,0 +1,28 @@
+cap per run: 2000 ms; input = n letters 'a' followed by '!' (no match possible)
+
+pattern (a+)+b
+  n=20     0 ms  matches=false
+  n=36     0 ms  matches=false
+  n=500    1 ms  matches=false
+  n=1000   23 ms  matches=false
+  n=2000   63 ms  matches=false
+  n=4000   161 ms  matches=false
+
+pattern (a+)+\1?b
+  n=20     27 ms  matches=false
+  n=24     690 ms  matches=false
+  n=26     ABORTED after 2036 ms (cap)
+
+pattern (?:(?:a+)+)+b
+  n=16     3 ms  matches=false
+  n=18     13 ms  matches=false
+  n=20     101 ms  matches=false
+  n=22     896 ms  matches=false
+  n=24     ABORTED after 2000 ms (cap)
+
+pattern a*a*a*a*b
+  n=50     2 ms  matches=false
+  n=100    69 ms  matches=false
+  n=150    285 ms  matches=false
+  n=200    824 ms  matches=false
+

regex/output/05-fixes.txt

@@ -0,0 +1,12 @@
+input: 40 x 'a' + '!'   budget 1000 ms
+vulnerable   (?:(?:a+)+)+b    ABORTED after 1019 ms
+rewritten    a+b              0 ms  matches=false
+possessive   (?:a++)++b       7 ms  matches=false
+atomic       (?>(?:a+)+)b     0 ms  matches=false
+
+find() a+b        n=5000    139 ms  found=false
+find() a+b        n=10000   489 ms  found=false
+find() a+b        n=20000  1906 ms  found=false
+find() (?<!a)a+b  n=5000      6 ms  found=false
+find() (?<!a)a+b  n=10000     6 ms  found=false
+find() (?<!a)a+b  n=20000    15 ms  found=false

regex/output/06-limits.txt

@@ -0,0 +1,6 @@
+(?:a|b)*c  length 1000 -> true
+(?:a|b)*c  length 5000 -> StackOverflowError
+(?:a|b)*c  length 20000 -> StackOverflowError
+(?:a|b)*c  length 100000 -> StackOverflowError
+[ab]*c     length 100000 -> true
+(?:a|b)*+c length 100000 -> true

regex/output/07-jdk-source.txt

@@ -0,0 +1,22 @@
+$ java -version
+openjdk version "25.0.4.1" 2026-08-18 LTS
+OpenJDK Runtime Environment Temurin-25.0.4.1+1 (build 25.0.4.1+1-LTS)
+OpenJDK 64-Bit Server VM Temurin-25.0.4.1+1 (build 25.0.4.1+1-LTS, mixed mode, sharing)
+
+$ unzip -p lib/src.zip java.base/java/util/regex/Pattern.java | sed -n "/Optimize the greedy Loop/,/^        }$/p"
+        // Optimize the greedy Loop to prevent exponential backtracking, IF there
+        // is no group ref in this pattern. With a non-negative localTCNCount value,
+        // the greedy type Loop, Curly will skip the backtracking for any starting
+        // position "i" that failed in the past.
+        if (!hasGroupRef) {
+            for (Node node : topClosureNodes) {
+                if (node instanceof Loop) {
+                    // non-deterministic-greedy-group
+                    ((Loop)node).posIndex = localTCNCount++;
+                }
+            }
+        }
+
+$ grep -ci timeout Pattern.java Matcher.java
+Pattern.java: 0
+Matcher.java: 0

regex/output/08-tests.txt

@@ -0,0 +1,4 @@
+-------------------------------------------------------------------------------
+Test set: com.ankurm.regex.RegexClaimsTest
+-------------------------------------------------------------------------------
+Tests run: 11, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 1.131 s -- in com.ankurm.regex.RegexClaimsTest