[Truffle] Shim regexps with modifiers.

jruby · Jan 3, 2016 · 1150491 · b1tfury · Jan 3, 2016 · chrisseaton
1 parent 81c658a
commit 1150491
Showing 3 changed files with 38 additions and 5 deletions.
diff --git a/spec/truffle/tags/language/regexp/modifiers_tags.txt b/spec/truffle/tags/language/regexp/modifiers_tags.txt
diff --git a/truffle/src/main/java/org/jruby/truffle/nodes/core/RegexpNodes.java b/truffle/src/main/java/org/jruby/truffle/nodes/core/RegexpNodes.java
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2013, 2015 Oracle and/or its affiliates. All rights reserved. This
+ * Copyright (c) 2013, 2016 Oracle and/or its affiliates. All rights reserved. This
  * code is released under a tri EPL/GPL/LGPL license. You can use it,
  * redistribute it and/or modify it under the terms of the:
  *
@@ -198,8 +198,42 @@ private static void setLocalVariable(Frame frame, String name, Object value) {
         }
     }
 
+    public static ByteList shimModifiers(ByteList bytes) {
+        // Joni doesn't support (?u) etc but we can shim some common cases
+
+        String bytesString = bytes.toString();
+
+        if (bytesString.startsWith("(?u)") || bytesString.startsWith("(?d)") || bytesString.startsWith("(?a)")) {
+            final char modifier = (char) bytes.get(2);
+            bytesString = bytesString.substring(4);
+
+            switch (modifier) {
+                case 'u': {
+                    bytesString = bytesString.replace("\\w", "[[:alpha:]]");
+                } break;
+
+                case 'd': {
+
+                } break;
+
+                case 'a': {
+                    bytesString = bytesString.replace("[[:alpha:]]", "[a-zA-Z]");
+                } break;
+
+                default:
+                    throw new UnsupportedOperationException();
+            }
+
+            bytes = ByteList.create(bytesString);
+        }
+
+        return bytes;
+    }
+
     @TruffleBoundary
     public static Regex compile(Node currentNode, RubyContext context, ByteList bytes, RegexpOptions options) {
+        bytes = shimModifiers(bytes);
+
         try {
             /*
                     // This isn't quite right - we shouldn't be looking up by name, we need a real reference to this constants

diff --git a/truffle/src/main/java/org/jruby/truffle/nodes/rubinius/RegexpPrimitiveNodes.java b/truffle/src/main/java/org/jruby/truffle/nodes/rubinius/RegexpPrimitiveNodes.java
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2015 Oracle and/or its affiliates. All rights reserved. This
+ * Copyright (c) 2015, 2016 Oracle and/or its affiliates. All rights reserved. This
  * code is released under a tri EPL/GPL/LGPL license. You can use it,
  * redistribute it and/or modify it under the terms of the:
  *
@@ -138,8 +138,8 @@ public Object searchRegion(DynamicObject regexp, DynamicObject string, int start
             final ByteList stringBl = StringOperations.getByteList(string);
             final ByteList bl = Layouts.REGEXP.getSource(regexp);
             final Encoding enc = RegexpNodes.checkEncoding(regexp, StringOperations.getCodeRangeable(string), true);
-            final ByteList preprocessed = RegexpSupport.preprocess(getContext().getRuntime(), bl, enc, new Encoding[]{null}, RegexpSupport.ErrorMode.RAISE);
-
+            ByteList preprocessed = RegexpSupport.preprocess(getContext().getRuntime(), bl, enc, new Encoding[]{null}, RegexpSupport.ErrorMode.RAISE);
+            preprocessed = RegexpNodes.shimModifiers(preprocessed);
             final Regex r = new Regex(preprocessed.getUnsafeBytes(), preprocessed.getBegin(), preprocessed.getBegin() + preprocessed.getRealSize(), Layouts.REGEXP.getRegex(regexp).getOptions(), RegexpNodes.checkEncoding(regexp, StringOperations.getCodeRangeable(string), true));
             final Matcher matcher = r.matcher(stringBl.getUnsafeBytes(), stringBl.begin(), stringBl.begin() + stringBl.realSize());