Author: nick Date: Thu Aug 20 10:02:19 2015 New Revision: 1696749 URL: http://svn.apache.org/r1696749 Log: TIKA-1710 patch from Yaniv Kunda - Use Commons IO instead of the Tika Core IO copies, and java.nio.charset.StandardCharsets
Modified: tika/trunk/tika-batch/pom.xml tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcess.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcessDriverCLI.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/Interrupter.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/BasicTikaFSConsumer.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/FSBatchProcessCLI.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/RecursiveParserWrapperFSConsumer.java tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/strawman/StrawManTikaAppDriver.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/CommandLineParserBuilderTest.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchDriverTest.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchProcessTest.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/FSBatchTestBase.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/HandlerBuilderTest.java tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/StringStreamGobbler.java Modified: tika/trunk/tika-batch/pom.xml URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/pom.xml?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/pom.xml (original) +++ tika/trunk/tika-batch/pom.xml Thu Aug 20 10:02:19 2015 @@ -67,6 +67,11 @@ <version>${cli.version}</version> </dependency> <dependency> + <groupId>commons-io</groupId> + <artifactId>commons-io</artifactId> + <version>${commons.io.version}</version> + </dependency> + <dependency> <groupId>org.apache.tika</groupId> <artifactId>tika-core</artifactId> <version>${project.version}</version> @@ -85,12 +90,6 @@ <artifactId>junit</artifactId> <scope>test</scope> </dependency> - <dependency> - <groupId>commons-io</groupId> - <artifactId>commons-io</artifactId> - <scope>test</scope> - <version>${commons.io.version}</version> - </dependency> </dependencies> <build> Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcess.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcess.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcess.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcess.java Thu Aug 20 10:02:19 2015 @@ -31,10 +31,10 @@ import java.util.concurrent.Executors; import java.util.concurrent.Future; import java.util.concurrent.TimeUnit; -import org.apache.tika.io.IOUtils; import org.slf4j.Logger; import org.slf4j.LoggerFactory; +import static java.nio.charset.StandardCharsets.UTF_8; /** * This is the main processor class for a single process. @@ -134,7 +134,7 @@ public class BatchProcess implements Cal //System.err should be redirected to System.out PrintStream sysErr = System.err; try { - outputStreamWriter = new PrintStream(sysErr, true, IOUtils.UTF_8.toString()); + outputStreamWriter = new PrintStream(sysErr, true, UTF_8.toString()); } catch (IOException e) { throw new RuntimeException("Can't redirect streams"); } Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcessDriverCLI.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcessDriverCLI.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcessDriverCLI.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/BatchProcessDriverCLI.java Thu Aug 20 10:02:19 2015 @@ -29,10 +29,12 @@ import java.util.ArrayList; import java.util.List; import java.util.Locale; -import org.apache.tika.io.IOUtils; +import org.apache.commons.io.IOUtils; import org.slf4j.Logger; import org.slf4j.LoggerFactory; +import static java.nio.charset.StandardCharsets.UTF_8; + public class BatchProcessDriverCLI { /** @@ -285,7 +287,7 @@ public class BatchProcessDriverCLI { private BufferedReader reader; private InterruptWatcher(InputStream is) { - reader = new BufferedReader(new InputStreamReader(is, IOUtils.UTF_8)); + reader = new BufferedReader(new InputStreamReader(is, UTF_8)); } @Override @@ -310,7 +312,7 @@ public class BatchProcessDriverCLI { private final Writer writer; private InterruptWriter(OutputStream os) { - this.writer = new OutputStreamWriter(os, IOUtils.UTF_8); + this.writer = new OutputStreamWriter(os, UTF_8); } @Override @@ -337,8 +339,7 @@ public class BatchProcessDriverCLI { protected boolean running = true; private StreamGobbler(InputStream is) { - this.reader = new BufferedReader(new InputStreamReader(new BufferedInputStream(is), - IOUtils.UTF_8)); + this.reader = new BufferedReader(new InputStreamReader(new BufferedInputStream(is), UTF_8)); } @Override Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/Interrupter.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/Interrupter.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/Interrupter.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/Interrupter.java Thu Aug 20 10:02:19 2015 @@ -22,10 +22,11 @@ import java.io.IOException; import java.io.InputStreamReader; import java.util.concurrent.Callable; -import org.apache.tika.io.IOUtils; import org.slf4j.Logger; import org.slf4j.LoggerFactory; +import static java.nio.charset.StandardCharsets.UTF_8; + /** * Class that waits for input on System.in. If the user enters a keystroke on @@ -39,7 +40,7 @@ public class Interrupter implements Call private Logger logger = LoggerFactory.getLogger(Interrupter.class); public IFileProcessorFutureResult call(){ try{ - BufferedReader reader = new BufferedReader(new InputStreamReader(System.in, IOUtils.UTF_8)); + BufferedReader reader = new BufferedReader(new InputStreamReader(System.in, UTF_8)); while (true){ if (reader.ready()){ reader.readLine(); Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/BasicTikaFSConsumer.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/BasicTikaFSConsumer.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/BasicTikaFSConsumer.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/BasicTikaFSConsumer.java Thu Aug 20 10:02:19 2015 @@ -22,16 +22,18 @@ import java.io.OutputStream; import java.io.UnsupportedEncodingException; import java.util.concurrent.ArrayBlockingQueue; +import org.apache.commons.io.IOUtils; import org.apache.tika.batch.FileResource; import org.apache.tika.batch.OutputStreamFactory; import org.apache.tika.batch.ParserFactory; import org.apache.tika.config.TikaConfig; -import org.apache.tika.io.IOUtils; import org.apache.tika.parser.ParseContext; import org.apache.tika.parser.Parser; import org.apache.tika.sax.ContentHandlerFactory; import org.xml.sax.ContentHandler; +import static java.nio.charset.StandardCharsets.UTF_8; + /** * Basic FileResourceConsumer that reads files from an input * directory and writes content to the output directory. @@ -47,7 +49,7 @@ public class BasicTikaFSConsumer extends private final ContentHandlerFactory contentHandlerFactory; private final OutputStreamFactory fsOSFactory; private final TikaConfig config; - private String outputEncoding = IOUtils.UTF_8.toString(); + private String outputEncoding = UTF_8.toString(); public BasicTikaFSConsumer(ArrayBlockingQueue<FileResource> queue, Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/FSBatchProcessCLI.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/FSBatchProcessCLI.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/FSBatchProcessCLI.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/FSBatchProcessCLI.java Thu Aug 20 10:02:19 2015 @@ -32,12 +32,12 @@ import org.apache.commons.cli.GnuParser; import org.apache.commons.cli.HelpFormatter; import org.apache.commons.cli.Option; import org.apache.commons.cli.Options; +import org.apache.commons.io.IOUtils; import org.apache.tika.batch.BatchProcess; import org.apache.tika.batch.BatchProcessDriverCLI; import org.apache.tika.batch.ParallelFileProcessingResult; import org.apache.tika.batch.builders.BatchProcessBuilder; import org.apache.tika.batch.builders.CommandLineParserBuilder; -import org.apache.tika.io.IOUtils; import org.apache.tika.io.TikaInputStream; import org.slf4j.Logger; import org.slf4j.LoggerFactory; Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/RecursiveParserWrapperFSConsumer.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/RecursiveParserWrapperFSConsumer.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/RecursiveParserWrapperFSConsumer.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/RecursiveParserWrapperFSConsumer.java Thu Aug 20 10:02:19 2015 @@ -25,11 +25,11 @@ import java.util.LinkedList; import java.util.List; import java.util.concurrent.ArrayBlockingQueue; +import org.apache.commons.io.IOUtils; import org.apache.tika.batch.FileResource; import org.apache.tika.batch.OutputStreamFactory; import org.apache.tika.batch.ParserFactory; import org.apache.tika.config.TikaConfig; -import org.apache.tika.io.IOUtils; import org.apache.tika.metadata.Metadata; import org.apache.tika.metadata.TikaCoreProperties; import org.apache.tika.metadata.serialization.JsonMetadataList; Modified: tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/strawman/StrawManTikaAppDriver.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/strawman/StrawManTikaAppDriver.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/strawman/StrawManTikaAppDriver.java (original) +++ tika/trunk/tika-batch/src/main/java/org/apache/tika/batch/fs/strawman/StrawManTikaAppDriver.java Thu Aug 20 10:02:19 2015 @@ -34,7 +34,7 @@ import java.util.concurrent.Executors; import java.util.concurrent.Future; import java.util.concurrent.atomic.AtomicInteger; -import org.apache.tika.io.IOUtils; +import org.apache.commons.io.IOUtils; import org.slf4j.Logger; import org.slf4j.LoggerFactory; import org.slf4j.MarkerFactory; Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/CommandLineParserBuilderTest.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/CommandLineParserBuilderTest.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/CommandLineParserBuilderTest.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/CommandLineParserBuilderTest.java Thu Aug 20 10:02:19 2015 @@ -22,9 +22,9 @@ import java.io.FileInputStream; import java.io.InputStream; import org.apache.commons.cli.Options; +import org.apache.commons.io.IOUtils; import org.apache.tika.batch.builders.CommandLineParserBuilder; import org.apache.tika.batch.fs.FSBatchTestBase; -import org.apache.tika.io.IOUtils; import org.junit.Test; Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java Thu Aug 20 10:02:19 2015 @@ -16,6 +16,7 @@ */ package org.apache.tika.batch; +import static java.nio.charset.StandardCharsets.UTF_8; import static org.junit.Assert.assertEquals; import java.io.ByteArrayInputStream; @@ -31,7 +32,6 @@ import java.util.concurrent.ArrayBlockin import org.apache.tika.TikaTest; import org.apache.tika.batch.fs.RecursiveParserWrapperFSConsumer; import org.apache.tika.config.TikaConfig; -import org.apache.tika.io.IOUtils; import org.apache.tika.metadata.Metadata; import org.apache.tika.metadata.TikaCoreProperties; import org.apache.tika.metadata.serialization.JsonMetadataList; @@ -76,7 +76,7 @@ public class RecursiveParserWrapperFSCon IFileProcessorFutureResult result = consumer.call(); mockOSFactory.getStreams().get(0).flush(); byte[] bytes = mockOSFactory.getStreams().get(0).toByteArray(); - List<Metadata> results = JsonMetadataList.fromJson(new InputStreamReader(new ByteArrayInputStream(bytes), IOUtils.UTF_8)); + List<Metadata> results = JsonMetadataList.fromJson(new InputStreamReader(new ByteArrayInputStream(bytes), UTF_8)); assertEquals(4, results.size()); assertContains("another null pointer", @@ -123,7 +123,7 @@ public class RecursiveParserWrapperFSCon IFileProcessorFutureResult result = consumer.call(); mockOSFactory.getStreams().get(0).flush(); byte[] bytes = mockOSFactory.getStreams().get(0).toByteArray(); - List<Metadata> results = JsonMetadataList.fromJson(new InputStreamReader(new ByteArrayInputStream(bytes), IOUtils.UTF_8)); + List<Metadata> results = JsonMetadataList.fromJson(new InputStreamReader(new ByteArrayInputStream(bytes), UTF_8)); assertEquals(2, results.size()); assertContains("another null pointer", results.get(0).get(TikaCoreProperties.TIKA_META_EXCEPTION_PREFIX + "runtime")); Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchDriverTest.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchDriverTest.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchDriverTest.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchDriverTest.java Thu Aug 20 10:02:19 2015 @@ -17,6 +17,7 @@ package org.apache.tika.batch.fs; * limitations under the License. */ +import static java.nio.charset.StandardCharsets.UTF_8; import static org.junit.Assert.assertEquals; import static org.junit.Assert.assertFalse; import static org.junit.Assert.assertNotNull; @@ -29,7 +30,6 @@ import java.util.Map; import org.apache.commons.io.FileUtils; import org.apache.tika.batch.BatchProcessDriverCLI; -import org.apache.tika.io.IOUtils; import org.junit.Test; @@ -53,10 +53,7 @@ public class BatchDriverTest extends FSB assertFalse(driver.getUserInterrupted()); assertEquals(5, outputDir.listFiles().length); assertContains("first test file", - FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), - IOUtils.UTF_8.toString())); - - + FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), UTF_8)); } @Test(timeout = 30000) @@ -74,8 +71,7 @@ public class BatchDriverTest extends FSB assertTrue(driver.getNumRestarts() > 0); assertFalse(driver.getUserInterrupted()); assertContains("first test file", - FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), UTF_8)); } @Test(timeout = 15000) @@ -117,8 +113,7 @@ public class BatchDriverTest extends FSB assertEquals(1, driver.getNumRestarts()); assertFalse(driver.getUserInterrupted()); assertContains("first test file", - FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), UTF_8)); } @Test(timeout = 30000) @@ -136,8 +131,7 @@ public class BatchDriverTest extends FSB assertEquals(3, driver.getNumRestarts()); assertFalse(driver.getUserInterrupted()); assertContains("first test file", - FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), UTF_8)); } @Test(timeout = 30000) Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchProcessTest.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchProcessTest.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchProcessTest.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/BatchProcessTest.java Thu Aug 20 10:02:19 2015 @@ -17,6 +17,7 @@ package org.apache.tika.batch.fs; */ +import static java.nio.charset.StandardCharsets.UTF_8; import static org.junit.Assert.assertEquals; import static org.junit.Assert.assertFalse; import static org.junit.Assert.assertTrue; @@ -29,7 +30,6 @@ import java.util.Map; import org.apache.commons.io.FileUtils; import org.apache.tika.batch.BatchProcess; import org.apache.tika.batch.BatchProcessDriverCLI; -import org.apache.tika.io.IOUtils; import org.junit.Test; public class BatchProcessTest extends FSBatchTestBase { @@ -85,8 +85,7 @@ public class BatchProcessTest extends FS assertEquals(0, hvyHang.length()); } assertContains("This is tika-batch's first test file", - FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test6_ok.xml.xml"), UTF_8)); //key that the process realize that there were no more processable files //in the queue and does not ask for a restart! @@ -137,8 +136,7 @@ public class BatchProcessTest extends FS assertEquals(4, outputDir.listFiles().length); assertContains("This is tika-batch's first test file", - FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test2_ok.xml.xml"), UTF_8)); assertContains(BatchProcess.BATCH_CONSTANTS.BATCH_PROCESS_FATAL_MUST_RESTART.toString(), streamStrings.getErrString()); @@ -191,8 +189,7 @@ public class BatchProcessTest extends FS File[] files = outputDir.listFiles(); assertEquals(1, files.length); assertContains("<p>some content</p>", - FileUtils.readFileToString(new File(outputDir, "test0_sleep.xml.xml"), - IOUtils.UTF_8.toString())); + FileUtils.readFileToString(new File(outputDir, "test0_sleep.xml.xml"), UTF_8)); assertContains("exitStatus="+BatchProcessDriverCLI.PROCESS_RESTART_EXIT_CODE, streamStrings.getOutString()); assertContains("causeForTermination='BATCH_PROCESS_ALIVE_TOO_LONG'", Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/FSBatchTestBase.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/FSBatchTestBase.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/FSBatchTestBase.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/FSBatchTestBase.java Thu Aug 20 10:02:19 2015 @@ -30,12 +30,12 @@ import java.util.concurrent.Future; import java.util.concurrent.TimeUnit; import org.apache.commons.io.FileUtils; +import org.apache.commons.io.IOUtils; import org.apache.tika.TikaTest; import org.apache.tika.batch.BatchProcess; import org.apache.tika.batch.BatchProcessDriverCLI; import org.apache.tika.batch.ParallelFileProcessingResult; import org.apache.tika.batch.builders.BatchProcessBuilder; -import org.apache.tika.io.IOUtils; import org.junit.AfterClass; import org.junit.BeforeClass; Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/HandlerBuilderTest.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/HandlerBuilderTest.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/HandlerBuilderTest.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/HandlerBuilderTest.java Thu Aug 20 10:02:19 2015 @@ -17,6 +17,7 @@ package org.apache.tika.batch.fs; * limitations under the License. */ +import static java.nio.charset.StandardCharsets.UTF_8; import static org.junit.Assert.assertFalse; import static org.junit.Assert.assertTrue; @@ -26,7 +27,6 @@ import java.util.Map; import org.apache.commons.io.FileUtils; import org.apache.tika.batch.BatchProcess; import org.apache.tika.batch.ParallelFileProcessingResult; -import org.apache.tika.io.IOUtils; import org.junit.Test; public class HandlerBuilderTest extends FSBatchTestBase { @@ -42,7 +42,7 @@ public class HandlerBuilderTest extends BatchProcess runner = getNewBatchRunner("/tika-batch-config-test.xml", args); ParallelFileProcessingResult result = run(runner); File outputFile = new File(outputDir, "test0.xml.xml"); - String resultString = FileUtils.readFileToString(outputFile, IOUtils.UTF_8.toString()); + String resultString = FileUtils.readFileToString(outputFile, UTF_8); assertTrue(resultString.contains("<html xmlns=\"http://www.w3.org/1999/xhtml\">")); assertTrue(resultString.contains("<?xml version=\"1.0\" encoding=\"UTF-8\"?>")); assertTrue(resultString.contains("This is tika-batch's first test file")); @@ -59,7 +59,7 @@ public class HandlerBuilderTest extends BatchProcess runner = getNewBatchRunner("/tika-batch-config-test.xml", args); ParallelFileProcessingResult result = run(runner); File outputFile = new File(outputDir, "test0.xml.html"); - String resultString = FileUtils.readFileToString(outputFile, IOUtils.UTF_8.toString()); + String resultString = FileUtils.readFileToString(outputFile, UTF_8); assertTrue(resultString.contains("<html xmlns=\"http://www.w3.org/1999/xhtml\">")); assertFalse(resultString.contains("<?xml version=\"1.0\" encoding=\"UTF-8\"?>")); assertTrue(resultString.contains("This is tika-batch's first test file")); @@ -76,7 +76,7 @@ public class HandlerBuilderTest extends BatchProcess runner = getNewBatchRunner("/tika-batch-config-test.xml", args); ParallelFileProcessingResult result = run(runner); File outputFile = new File(outputDir, "test0.xml.txt"); - String resultString = FileUtils.readFileToString(outputFile, IOUtils.UTF_8.toString()); + String resultString = FileUtils.readFileToString(outputFile, UTF_8); assertFalse(resultString.contains("<html xmlns=\"http://www.w3.org/1999/xhtml\">")); assertFalse(resultString.contains("<?xml version=\"1.0\" encoding=\"UTF-8\"?>")); assertTrue(resultString.contains("This is tika-batch's first test file")); @@ -94,7 +94,7 @@ public class HandlerBuilderTest extends ParallelFileProcessingResult result = run(runner); File outputFile = new File(outputDir, "test0.xml.xml"); - String resultString = FileUtils.readFileToString(outputFile, IOUtils.UTF_8.toString()); + String resultString = FileUtils.readFileToString(outputFile, UTF_8); //this is not ideal. How can we change handlers to writeout whatever //they've gotten so far, up to the writeLimit? assertTrue(resultString.equals("")); @@ -112,7 +112,7 @@ public class HandlerBuilderTest extends BatchProcess runner = getNewBatchRunner("/tika-batch-config-test.xml", args); ParallelFileProcessingResult result = run(runner); File outputFile = new File(outputDir, "test0.xml.json"); - String resultString = FileUtils.readFileToString(outputFile, IOUtils.UTF_8.toString()); + String resultString = FileUtils.readFileToString(outputFile, UTF_8); assertTrue(resultString.contains("\"author\":\"Nikolai Lobachevsky\"")); assertTrue(resultString.contains("tika-batch\\u0027s first test file")); } Modified: tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/StringStreamGobbler.java URL: http://svn.apache.org/viewvc/tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/StringStreamGobbler.java?rev=1696749&r1=1696748&r2=1696749&view=diff ============================================================================== --- tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/StringStreamGobbler.java (original) +++ tika/trunk/tika-batch/src/test/java/org/apache/tika/batch/fs/StringStreamGobbler.java Thu Aug 20 10:02:19 2015 @@ -17,13 +17,15 @@ package org.apache.tika.batch.fs; * limitations under the License. */ +import org.apache.commons.io.IOUtils; + import java.io.BufferedInputStream; import java.io.BufferedReader; import java.io.IOException; import java.io.InputStream; import java.io.InputStreamReader; -import org.apache.tika.io.IOUtils; +import static java.nio.charset.StandardCharsets.UTF_8; public class StringStreamGobbler implements Runnable { @@ -33,8 +35,7 @@ public class StringStreamGobbler impleme private final StringBuilder sb = new StringBuilder(); public StringStreamGobbler(InputStream is) { - this.reader = new BufferedReader(new InputStreamReader(new BufferedInputStream(is), - IOUtils.UTF_8)); + this.reader = new BufferedReader(new InputStreamReader(new BufferedInputStream(is), UTF_8)); } @Override