From 0648e8725ee0b2d292a5065fe886b11cd4a6da3c Mon Sep 17 00:00:00 2001
From: Stefano Mazzocchi <stefano@apache.org>
Date: Tue, 1 Jun 2010 08:54:17 +0000
Subject: [PATCH] adding regexp group capturing GEL function

git-svn-id: http://google-refine.googlecode.com/svn/trunk@932 7d457c2a-affb-35e4-300a-418c747d4874
---
 .../expr/functions/strings/Match.java         | 55 +++++++++++++++++++
 .../gel/ControlFunctionRegistry.java          |  2 +
 2 files changed, 57 insertions(+)
 create mode 100644 main/src/com/metaweb/gridworks/expr/functions/strings/Match.java

diff --git a/main/src/com/metaweb/gridworks/expr/functions/strings/Match.java b/main/src/com/metaweb/gridworks/expr/functions/strings/Match.java
new file mode 100644
index 000000000..1c2d2460d
--- /dev/null
+++ b/main/src/com/metaweb/gridworks/expr/functions/strings/Match.java
@@ -0,0 +1,55 @@
+package com.metaweb.gridworks.expr.functions.strings;
+
+import java.util.Properties;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+import org.json.JSONException;
+import org.json.JSONWriter;
+
+import com.metaweb.gridworks.expr.EvalError;
+import com.metaweb.gridworks.gel.ControlFunctionRegistry;
+import com.metaweb.gridworks.gel.Function;
+
+public class Match implements Function {
+
+    public Object call(Properties bindings, Object[] args) {
+        if (args.length == 2) {
+            Object s = args[0];
+            Object p = args[1];
+            
+            if (s != null && p != null && (p instanceof String || p instanceof Pattern)) {
+                
+                Pattern pattern = (p instanceof String) ? Pattern.compile((String) p) : (Pattern) p;
+
+                Matcher matcher = pattern.matcher(s.toString());
+                
+                if (matcher.matches()) {
+                    int count = matcher.groupCount();
+    
+                    String[] groups = new String[count];
+                    for (int i = 0; i < count; i++) {
+                        groups[i] = matcher.group(i + 1);
+                    }
+    
+                    return groups;
+                } else {
+                    return null;
+                }
+            }
+            
+            return null;
+        }
+        return new EvalError(ControlFunctionRegistry.getFunctionName(this) + " expects a string or a regexp");
+    }
+    
+    public void write(JSONWriter writer, Properties options)
+        throws JSONException {
+    
+        writer.object();
+        writer.key("description"); writer.value("Returns an array of the groups matching the given regular expression");
+        writer.key("params"); writer.value("string or regexp");
+        writer.key("returns"); writer.value("array of strings");
+        writer.endObject();
+    }
+}
diff --git a/main/src/com/metaweb/gridworks/gel/ControlFunctionRegistry.java b/main/src/com/metaweb/gridworks/gel/ControlFunctionRegistry.java
index cc8b18ab8..e1f80c461 100644
--- a/main/src/com/metaweb/gridworks/gel/ControlFunctionRegistry.java
+++ b/main/src/com/metaweb/gridworks/gel/ControlFunctionRegistry.java
@@ -42,6 +42,7 @@ import com.metaweb.gridworks.expr.functions.strings.Fingerprint;
 import com.metaweb.gridworks.expr.functions.strings.IndexOf;
 import com.metaweb.gridworks.expr.functions.strings.LastIndexOf;
 import com.metaweb.gridworks.expr.functions.strings.MD5;
+import com.metaweb.gridworks.expr.functions.strings.Match;
 import com.metaweb.gridworks.expr.functions.strings.NGram;
 import com.metaweb.gridworks.expr.functions.strings.NGramFingerprint;
 import com.metaweb.gridworks.expr.functions.strings.Partition;
@@ -149,6 +150,7 @@ public class ControlFunctionRegistry {
         registerFunction("reinterpret", new Reinterpret());
         registerFunction("jsonize", new Jsonize());
         registerFunction("ngram", new NGram());
+        registerFunction("match", new Match());
         
         registerFunction("indexOf", new IndexOf());
         registerFunction("lastIndexOf", new LastIndexOf());