Merge pull request #16 from zeroae/f/disk-lru-cache

Add DiskLRU Based cache for Gate responses

Author: sodre

build.gradle

@@ -16,6 +16,7 @@ dependencies {
     implementation 'uk.ac.gate.plugins:format-json:8.7'
     implementation 'com.amazonaws:aws-lambda-java-core:1.2.1'
     implementation 'com.amazonaws:aws-lambda-java-events:2.2.9'
+    implementation 'com.jakewharton:disklrucache:2.0.2'
 
     runtimeOnly 'com.amazonaws:aws-lambda-java-log4j2:1.2.0'
 

src/main/java/co/zeroae/gate/App.java

@@ -7,17 +7,21 @@
 
 import gate.*;
 import gate.corpora.export.GATEJsonExporter;
-import gate.creole.ExecutionException;
 import gate.creole.ResourceInstantiationException;
 import gate.util.GateException;
 import gate.util.persistence.PersistenceManager;
+
+import com.jakewharton.disklrucache.DiskLruCache;
+
 import org.apache.log4j.LogManager;
 import org.apache.log4j.Logger;
+import org.codehaus.httpcache4j.util.Hex;
 
-import java.io.ByteArrayOutputStream;
-import java.io.File;
-import java.io.IOException;
+import javax.xml.stream.XMLStreamException;
+import java.io.*;
 import java.net.URL;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.Objects;
@@ -36,38 +40,94 @@ public class App implements RequestHandler<APIGatewayProxyRequestEvent, APIGatew
         }
     }
 
+    private static final String GATE_APP_NAME = System.getenv("GATE_APP_NAME");
+    private static final String CACHE_DIR = System.getenv().getOrDefault(
+            "CACHE_DIR_PREFIX", "/tmp/lru/" + GATE_APP_NAME );
+    private static final double CACHE_DIR_USAGE = .9;
+
     private static final Logger logger = LogManager.getLogger(App.class);
     private static final CorpusController application = loadApplication();
     private static final GATEJsonExporter gateJsonExporter = new GATEJsonExporter();
 
+
+    private static final DiskLruCache cache = initializeCache();
+
     public APIGatewayProxyResponseEvent handleRequest(APIGatewayProxyRequestEvent input, final Context context) {
         final String responseType = input.getHeaders().getOrDefault("Accept", "application/xml");
         final APIGatewayProxyResponseEvent response = new APIGatewayProxyResponseEvent()
                 .withHeaders(new HashMap<>());
         response.getHeaders().put("Content-Type", "text/plain");
 
         try {
-            final Document doc = Factory.newDocument(input.getBody());
-            final Corpus corpus = application.getCorpus();
-            corpus.add(doc);
+            final String bodyDigest = computeMessageDigest(input.getBody());
+            response.getHeaders().put("x-zae-gate-cache", "HIT");
+            final Document doc = cacheComputeIfNull(
+                    bodyDigest,
+                    () -> {
+                        final Document rv = Factory.newDocument(input.getBody());
+                        final Corpus corpus = application.getCorpus();
+                        corpus.add(rv);
+                        try {
+                            application.execute();
+                        } finally {
+                            corpus.clear();
+                        }
+                        response.getHeaders().put("x-zae-gate-cache", "MISS");
+                        return rv;
+                    }
+            );
             try {
-                application.execute();
                 response.getHeaders().put("Content-Type", responseType);
                 return response.withBody(export(doc, responseType)).withStatusCode(200);
             } finally {
-                corpus.clear();
                 Factory.deleteResource(doc);
             }
-        } catch (ExecutionException e) {
+        } catch (GateException e) {
             logger.error(e);
             return response.withBody(e.getMessage()).withStatusCode(400);
         } catch (IOException e) {
             logger.error(e);
             return response.withBody(e.getMessage()).withStatusCode(406);
-        } catch (ResourceInstantiationException e) {
+        }
+    }
+
+    private Document cacheComputeIfNull(String key, TextProcessor processor) throws GateException {
+        try {
+            final DiskLruCache.Snapshot snapshot = cache.get(key);
+            if (snapshot == null) {
+                final Document doc = processor.process();
+                try {
+                    DiskLruCache.Editor editor = cache.edit(key);
+                    editor.set(0, doc.toXml());
+                    editor.commit();
+                } catch (IOException e) {
+                    logger.warn(e);
+                }
+                return doc;
+            } else {
+                try {
+                    return Utils.xmlToDocument(new InputStreamReader(snapshot.getInputStream(0)));
+                } catch (ResourceInstantiationException | XMLStreamException e) {
+                    logger.warn(e);
+                    cache.remove(key);
+                    return processor.process();
+                }
+            }
+        } catch (IOException e) {
             logger.warn(e);
-            return response.withBody(e.getMessage()).withStatusCode(400);
+            return processor.process();
+        }
+    }
+
+    private String computeMessageDigest(String text) {
+        final String sha256;
+        try {
+            final MessageDigest md = MessageDigest.getInstance("SHA-256");
+            sha256 = Hex.encode(md.digest(text.getBytes()));
+        } catch (NoSuchAlgorithmException e) {
+            throw new RuntimeException(e);
         }
+        return sha256;
     }
 
     /**
@@ -97,7 +157,7 @@ private String export(Document doc, String responseType) throws IOException {
 
     private static CorpusController loadApplication() {
         try {
-            final String gappResourcePah = System.getenv("GATE_APP_NAME") + "/application.xgapp";
+            final String gappResourcePah = GATE_APP_NAME + "/application.xgapp";
             final URL gappUrl = App.class.getClassLoader().getResource(gappResourcePah);
             final File gappFile = new File(Objects.requireNonNull(gappUrl).getFile());
             final CorpusController rv =
@@ -109,4 +169,26 @@ private static CorpusController loadApplication() {
             throw new RuntimeException(e);
         }
     }
+
+    private static DiskLruCache initializeCache() {
+        File cacheDir = new File(CACHE_DIR);
+        if (!cacheDir.exists() && !cacheDir.mkdirs()) {
+            throw new RuntimeException("Unable to create cache directory '" + cacheDir.getName() + "'.");
+        }
+        for (File file: Objects.requireNonNull(cacheDir.listFiles())) file.delete();
+        try {
+            long usableSpace = (long) (cacheDir.getUsableSpace()*CACHE_DIR_USAGE);
+            return DiskLruCache.open(cacheDir,
+                    1,
+                    1,
+                    usableSpace);
+        } catch (IOException e) {
+            throw new RuntimeException(e);
+        }
+    }
+
+    private interface TextProcessor {
+        Document process() throws GateException;
+    }
+
 }

src/main/java/co/zeroae/gate/Utils.java

@@ -0,0 +1,32 @@
+package co.zeroae.gate;
+
+import gate.Document;
+import gate.Factory;
+import gate.creole.ResourceInstantiationException;
+
+import javax.xml.stream.XMLInputFactory;
+import javax.xml.stream.XMLStreamException;
+import javax.xml.stream.XMLStreamReader;
+import java.io.InputStream;
+import java.io.InputStreamReader;
+import java.io.Reader;
+
+public class Utils {
+    /**
+     *
+     * @param gateXMLReader
+     * @return
+     * @throws ResourceInstantiationException if Document
+     * @throws XMLStreamException
+     */
+    static Document xmlToDocument(Reader gateXMLReader) throws ResourceInstantiationException, XMLStreamException {
+        final Document doc = Factory.newDocument("");
+        XMLStreamReader reader;
+        reader = XMLInputFactory.newFactory().createXMLStreamReader(gateXMLReader);
+        do {
+            reader.next();
+        } while(reader.getEventType() != XMLStreamReader.START_ELEMENT);
+        gate.corpora.DocumentStaxUtils.readGateXmlDocument(reader, doc);
+        return doc;
+    }
+}

src/test/java/co/zeroae/gate/AppTest.java

@@ -5,13 +5,10 @@
 import com.fasterxml.jackson.core.JsonFactory;
 import com.fasterxml.jackson.core.JsonParser;
 import gate.Document;
-import gate.Factory;
 import org.junit.Before;
 import org.junit.BeforeClass;
 import org.junit.Test;
 
-import javax.xml.stream.XMLInputFactory;
-import javax.xml.stream.XMLStreamReader;
 import java.io.StringReader;
 import java.util.HashMap;
 
@@ -27,7 +24,6 @@ public static void setUpClass() throws Exception {
                 .execute(App::new);
     }
 
-
     private static App app = null;
     private static TestContext context = new TestContext();
 
@@ -66,14 +62,7 @@ public void testGateXMLToDocument() throws Exception {
         final String resultBody = result.getBody();
         assertNotNull(resultBody);
 
-        Document doc = Factory.newDocument("");
-        XMLStreamReader reader = XMLInputFactory.newFactory().createXMLStreamReader(
-                new StringReader(resultBody)
-        );
-        do {
-            reader.next();
-        } while(reader.getEventType() != XMLStreamReader.START_ELEMENT);
-        gate.corpora.DocumentStaxUtils.readGateXmlDocument(reader, doc);
+        Document doc = Utils.xmlToDocument(new StringReader(resultBody));
         assertEquals(doc.getContent().toString(), content);
     }
 
@@ -102,4 +91,20 @@ public void testApplicationJsonResponse() throws Exception {
             parser.nextToken();
         }
     }
+
+    @Test
+    public void testCache() {
+        // Create the Input
+        input_headers.put("Content-Type", "text/plain");
+        input.withBody("This is a test. My name is LambdaTestFunction, and I just watched the T.V. show Wanda Vision.");
+
+        // Invoke the App
+        final APIGatewayProxyResponseEvent result = app.handleRequest(input, context);
+        assertEquals(result.getStatusCode().intValue(), 200);
+        assertEquals(result.getHeaders().get("x-zae-gate-cache"), "MISS");
+
+        final APIGatewayProxyResponseEvent cachedResult = app.handleRequest(input, context);
+        assertEquals(cachedResult.getStatusCode().intValue(), 200);
+        assertEquals(cachedResult.getHeaders().get("x-zae-gate-cache"), "HIT");
+    }
 }