svn commit: r1059877 - in /tika/trunk/tika-core/src/main/java/org/apache/tika: fork/ForkServer.java parser/ParseContext.java

[ju...@apache.org]

Author: jukka
Date: Mon Jan 17 11:17:51 2011
New Revision: 1059877

URL: http://svn.apache.org/viewvc?rev=1059877&view=rev
Log:
TIKA-416: Out-of-process text extraction

Make ParseContext more lightweight for easier serialization.
Some ForkServer javadocs.

Modified:
    tika/trunk/tika-core/src/main/java/org/apache/tika/fork/ForkServer.java
    tika/trunk/tika-core/src/main/java/org/apache/tika/parser/ParseContext.java

Modified: tika/trunk/tika-core/src/main/java/org/apache/tika/fork/ForkServer.java
URL: http://svn.apache.org/viewvc/tika/trunk/tika-core/src/main/java/org/apache/tika/fork/ForkServer.java?rev=1059877&r1=1059876&r2=1059877&view=diff
==============================================================================
--- tika/trunk/tika-core/src/main/java/org/apache/tika/fork/ForkServer.java (original)
+++ tika/trunk/tika-core/src/main/java/org/apache/tika/fork/ForkServer.java Mon Jan 17 11:17:51 2011
@@ -43,9 +43,18 @@ class ForkServer extends ClassLoader {
 
     public static final byte FIND_RESOURCES = 3;
 
+    /**
+     * Starts a forked server process.
+     * 
+     * @param args command line arguments, ignored
+     * @throws Exception if the server could not be started
+     */
     public static void main(String[] args) throws Exception {
         ForkServer server =
             new ForkServer(System.in, System.out);
+
+        // Set the server instance as the context class loader
+        // to make classes from the parent process available
         Thread.currentThread().setContextClassLoader(server);
 
         // Redirect standard input and output streams to prevent
@@ -53,6 +62,7 @@ class ForkServer extends ClassLoader {
         System.setIn(new ByteArrayInputStream(new byte[0]));
         System.setOut(System.err);
 
+        // Start processing request
         server.run();
     }
 
@@ -62,6 +72,14 @@ class ForkServer extends ClassLoader {
 
     private int count = 0;
 
+    /**
+     * Sets up a forked server instance using the given stdin/out
+     * communication channel.
+     *
+     * @param input input stream for reading from the parent process
+     * @param output output stream for writing to the parent process
+     * @throws IOException if the server instance could not be created
+     */
     public ForkServer(InputStream input, OutputStream output)
             throws IOException {
         this.input = new DataInputStream(input);

Modified: tika/trunk/tika-core/src/main/java/org/apache/tika/parser/ParseContext.java
URL: http://svn.apache.org/viewvc/tika/trunk/tika-core/src/main/java/org/apache/tika/parser/ParseContext.java?rev=1059877&r1=1059876&r2=1059877&view=diff
==============================================================================
--- tika/trunk/tika-core/src/main/java/org/apache/tika/parser/ParseContext.java (original)
+++ tika/trunk/tika-core/src/main/java/org/apache/tika/parser/ParseContext.java Mon Jan 17 11:17:51 2011
@@ -38,23 +38,48 @@ import org.xml.sax.SAXNotSupportedExcept
  */
 public class ParseContext implements Serializable {
 
-    /**
-     * Serial version UID.
-     */
+    /** Serial version UID. */
     private static final long serialVersionUID = -5921436862145826534L;
 
-    private final Map<Class<?>, Object> context =
-        new HashMap<Class<?>, Object>();
+    /** Map of objects in this context */
+    private final Map<String, Object> context = new HashMap<String, Object>();
  
+    /**
+     * Adds the given value to the context as an implementation of the given
+     * interface.
+     *
+     * @param key the interface implemented by the given value
+     * @param value the value to be added, or <code>null</code> to remove
+     */
     public <T> void set(Class<T> key, T value) {
-        context.put(key, value);
+        if (value != null) {
+            context.put(key.getName(), value);
+        } else {
+            context.remove(key.getName());
+        }
     }
 
+    /**
+     * Returns the object in this context that implements the given interface.
+     *
+     * @param key the interface implemented by the requested object
+     * @return the object that implements the given interface,
+     *         or <code>null</code> if not found
+     */
     @SuppressWarnings("unchecked")
     public <T> T get(Class<T> key) {
-        return (T) context.get(key);
+        return (T) context.get(key.getName());
     }
 
+    /**
+     * Returns the object in this context that implements the given interface,
+     * or the given default value if such an object is not found.
+     *
+     * @param key the interface implemented by the requested object
+     * @param defaultValue value to return if the requested object is not found
+     * @return the object that implements the given interface,
+     *         or the given default value if not found
+     */
     public <T> T get(Class<T> key, T defaultValue) {
         T value = get(key);
         if (value != null) {