Author: sylvain
Date: Sat Mar 12 05:04:38 2005
New Revision: 157253
URL: http://svn.apache.org/viewcvs?view=rev&rev=157253
Log:
Add support for flowscript file-specific encoding using a first-line comment
Modified:
cocoon/trunk/src/java/org/apache/cocoon/components/flow/javascript/fom/FOM_JavaScriptInterpreter.java
cocoon/trunk/status.xml
Modified:
cocoon/trunk/src/java/org/apache/cocoon/components/flow/javascript/fom/FOM_JavaScriptInterpreter.java
URL:
http://svn.apache.org/viewcvs/cocoon/trunk/src/java/org/apache/cocoon/components/flow/javascript/fom/FOM_JavaScriptInterpreter.java?view=diff&r1=157252&r2=157253
==============================================================================
---
cocoon/trunk/src/java/org/apache/cocoon/components/flow/javascript/fom/FOM_JavaScriptInterpreter.java
(original)
+++
cocoon/trunk/src/java/org/apache/cocoon/components/flow/javascript/fom/FOM_JavaScriptInterpreter.java
Sat Mar 12 05:04:38 2005
@@ -40,6 +40,9 @@
import org.apache.commons.jxpath.JXPathIntrospector;
import org.apache.commons.jxpath.ri.JXPathContextReferenceImpl;
import org.apache.excalibur.source.Source;
+import org.apache.regexp.RE;
+import org.apache.regexp.RECompiler;
+import org.apache.regexp.REProgram;
import org.mozilla.javascript.Context;
import org.mozilla.javascript.EcmaError;
import org.mozilla.javascript.EvaluatorException;
@@ -62,9 +65,11 @@
import java.awt.Dimension;
import java.awt.Toolkit;
import java.io.BufferedReader;
+import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.OutputStream;
+import java.io.PushbackInputStream;
import java.io.Reader;
import java.util.ArrayList;
import java.util.List;
@@ -499,14 +504,42 @@
protected Script compileScript(Context cx, Scriptable scope, Source src)
throws Exception {
- InputStream is = src.getInputStream();
+ PushbackInputStream is = new PushbackInputStream(src.getInputStream(),
ENCODING_BUF_SIZE);
try {
- Reader reader = new BufferedReader(new InputStreamReader(is));
+ String encoding = findEncoding(is);
+ Reader reader = encoding == null ? new InputStreamReader(is) : new
InputStreamReader(is, encoding);
+ reader = new BufferedReader(reader);
Script compiledScript = cx.compileReader(scope, reader,
src.getURI(), 1, null);
return compiledScript;
} finally {
is.close();
+ }
+ }
+
+ // A charset name can be up to 40 characters taken from the printable
characters of US-ASCII
+ // (see http://www.iana.org/assignments/character-sets). So reading 100
bytes should be more than enough.
+ private final static int ENCODING_BUF_SIZE = 100;
+ // Match 'encoding = xxxx' on the first line
+ REProgram encodingRE = new
RECompiler().compile("^.*encoding\\s*=\\s*([^\\s]*)");
+
+ /**
+ * Find the encoding of the stream, or null if not specified
+ */
+ String findEncoding(PushbackInputStream is) throws IOException {
+ // Read some bytes
+ byte[] buffer = new byte[ENCODING_BUF_SIZE];
+ int len = is.read(buffer, 0, buffer.length);
+ // and push them back
+ is.unread(buffer, 0, len);
+
+ // Interpret them as an ASCII string
+ String str = new String(buffer, 0, len, "ASCII");
+ RE re = new RE(encodingRE);
+ if (re.match(str)) {
+ return re.getParen(1);
+ } else {
+ return null;
}
}
Modified: cocoon/trunk/status.xml
URL:
http://svn.apache.org/viewcvs/cocoon/trunk/status.xml?view=diff&r1=157252&r2=157253
==============================================================================
--- cocoon/trunk/status.xml (original)
+++ cocoon/trunk/status.xml Sat Mar 12 05:04:38 2005
@@ -429,6 +429,11 @@
</action>
</release>
<release version="2.1.7" date="TBD">
+ <action dev="SW" type="add">
+ A flowscript's file encoding can now be specified by a comment on the
script's
+ very first line, in the form "<code>// encoding = xxxx</code>".
Otherwise, the
+ default platform encoding is used.
+ </action>
<action dev="BD" type="fix">
Tour block: fix "shapes" sample using if/else in flowscript
instead of switch.
