Merge pull request #4478 from dannysmyda/4656-allow-user-to-pick-language

4656 - Add OCR language pack configuration

Core/src/org/sleuthkit/autopsy/coreutils/PlatformUtil.java

@@ -36,6 +36,7 @@ import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.List;
 import javax.swing.filechooser.FileSystemView;
+import org.apache.commons.io.FilenameUtils;
 import org.hyperic.sigar.Sigar;
 import org.hyperic.sigar.ptql.ProcessFinder;
 import org.openide.modules.InstalledFileLocator;
@@ -53,6 +54,7 @@ public class PlatformUtil {
     private static final String PYTHON_MODULES_SUBDIRECTORY = "python_modules"; //NON-NLS
     private static final String CLASSIFIERS_SUBDIRECTORY = "object_detection_classifiers"; //NON-NLS
     private static final String OCR_LANGUAGE_SUBDIRECTORY = "ocr_language_packs"; //NON-NLS
+    private static final String OCR_LANGUAGE_PACK_EXT = "traineddata";
     private static String javaPath = null;
     public static final String OS_NAME_UNKNOWN = NbBundle.getMessage(PlatformUtil.class, "PlatformUtil.nameUnknown");
     public static final String OS_VERSION_UNKNOWN = NbBundle.getMessage(PlatformUtil.class, "PlatformUtil.verUnknown");
@@ -127,6 +129,26 @@ public class PlatformUtil {
         return getUserDirectory().getAbsolutePath() + File.separator + OCR_LANGUAGE_SUBDIRECTORY;
     }
     
+    /**
+     * Get the names of the language packs installed at the user directory.
+     * 
+     * @return List of language packs base names
+     */
+    public static List<String> getOcrLanguagePacks() {
+        File languagePackRootDir = new File(getOcrLanguagePacksPath());
+
+        List<String> languagePacks = new ArrayList<>();
+        for (File languagePack : languagePackRootDir.listFiles()) {
+            String fileExt = FilenameUtils.getExtension(languagePack.getName()); 
+            if (!languagePack.isDirectory() && OCR_LANGUAGE_PACK_EXT.equals(fileExt)) {
+                String packageName = FilenameUtils.getBaseName(languagePack.getName());
+                languagePacks.add(packageName);
+            }
+        }
+        
+        return languagePacks;
+    }
+
     /**
      * Get root path where the user's object detection classifiers are stored.
      * 

Core/src/org/sleuthkit/autopsy/textextractors/TikaTextExtractor.java

@@ -29,7 +29,6 @@ import java.io.InputStream;
 import java.io.PushbackReader;
 import java.io.Reader;
 import java.nio.file.Paths;
-import java.util.ArrayList;
 import java.util.List;
 import java.util.Objects;
 import java.util.concurrent.Callable;
@@ -42,7 +41,6 @@ import java.util.concurrent.TimeoutException;
 import java.util.logging.Level;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
-import org.apache.commons.io.FilenameUtils;
 import org.apache.tika.Tika;
 import org.apache.tika.metadata.Metadata;
 import org.apache.tika.parser.AutoDetectParser;
@@ -134,8 +132,7 @@ final class TikaTextExtractor implements TextExtractor {
     private static final String TESSERACT_DIR_NAME = "Tesseract-OCR"; //NON-NLS
     private static final String TESSERACT_EXECUTABLE = "tesseract.exe"; //NON-NLS
     private static final File TESSERACT_PATH = locateTesseractExecutable();
-    private static final String LANGUAGE_PACKS = getLanguagePacks();
+    private String languagePacks = formatLanguagePacks(PlatformUtil.getOcrLanguagePacks());
-    private static final String TESSERACT_LANGUAGE_PACK_EXT = "traineddata"; //NON-NLS
     private static final String TESSERACT_OUTPUT_FILE_NAME = "tess_output"; //NON-NLS
     
     private ProcessTerminator processTerminator;
@@ -203,7 +200,7 @@ final class TikaTextExtractor implements TextExtractor {
                 String tesseractFolder = TESSERACT_PATH.getParent();
                 ocrConfig.setTesseractPath(tesseractFolder);
                 
-                ocrConfig.setLanguage(LANGUAGE_PACKS);
+                ocrConfig.setLanguage(languagePacks);
                 ocrConfig.setTessdataPath(PlatformUtil.getOcrLanguagePacksPath());
                 parseContext.set(TesseractOCRConfig.class, ocrConfig);
 
@@ -289,7 +286,7 @@ final class TikaTextExtractor implements TextExtractor {
                     String.format("\"%s\"", outputFilePath),
                     "--tessdata-dir", PlatformUtil.getOcrLanguagePacksPath(),
                     //language pack command flag
-                    "-l", LANGUAGE_PACKS);
+                    "-l", languagePacks);
 
             //If the ProcessTerminator was supplied during 
             //configuration apply it here.
@@ -443,23 +440,11 @@ final class TikaTextExtractor implements TextExtractor {
     }
 
     /**
-     * Retrieves all of the installed language packs from their designated
+     * Formats language packs to be parseable from the command line.
-     * directory location to be used to configure Tesseract OCR.
      *
      * @return String of all language packs available for Tesseract to use
      */
-    private static String getLanguagePacks() {
+    private static String formatLanguagePacks(List<String> languagePacks) {
-        File languagePackRootDir = new File(PlatformUtil.getOcrLanguagePacksPath());
-
-        List<String> languagePacks = new ArrayList<>();
-        for (File languagePack : languagePackRootDir.listFiles()) {
-            String fileExt = FilenameUtils.getExtension(languagePack.getName()); 
-            if (!languagePack.isDirectory() && TESSERACT_LANGUAGE_PACK_EXT.equals(fileExt)) {
-                String packageName = FilenameUtils.getBaseName(languagePack.getName());
-                languagePacks.add(packageName);
-            }
-        }
-
         return String.join("+", languagePacks);
     }
 
@@ -499,10 +484,16 @@ final class TikaTextExtractor implements TextExtractor {
     public void setExtractionSettings(Lookup context) {
         if (context != null) {
             ImageConfig configInstance = context.lookup(ImageConfig.class);
-            if (configInstance != null && Objects.nonNull(configInstance.getOCREnabled())) {
+            if (configInstance != null) {
+                if(Objects.nonNull(configInstance.getOCREnabled())) {
                     this.tesseractOCREnabled = configInstance.getOCREnabled();
                 }
                 
+                if(Objects.nonNull(configInstance.getOCRLanguages())) {
+                    this.languagePacks = formatLanguagePacks(configInstance.getOCRLanguages());
+                }
+            }
+
             ProcessTerminator terminatorInstance = context.lookup(ProcessTerminator.class);
             if (terminatorInstance != null) {
                 this.processTerminator = terminatorInstance;

Core/src/org/sleuthkit/autopsy/textextractors/configs/ImageConfig.java

@@ -18,19 +18,22 @@
  */
 package org.sleuthkit.autopsy.textextractors.configs;
 
+import java.util.List;
+
 /**
- * Allows for configuration of OCR on image files. Extractors that use ImageConfig
+ * Allows for configuration of OCR on image files. Extractors that use
- * can be obtained through TextExtractoryFactory.getExtractor().
+ * ImageConfig can be obtained through TextExtractoryFactory.getExtractor().
  *
  * @see org.openide.util.Lookup
  */
 public class ImageConfig {
 
     private Boolean OCREnabled;
+    private List<String> ocrLanguages;
 
     /**
-     * Enables OCR to be run on the text reader responsible for handling
+     * Enables OCR to be run on the text reader responsible for handling image
-     * image files.
+     * files.
      *
      * @param enabled Flag indicating if OCR is enabled.
      */
@@ -46,4 +49,24 @@ public class ImageConfig {
     public boolean getOCREnabled() {
         return this.OCREnabled;
     }
+
+    /**
+     * Sets languages for OCR.
+     * 
+     * See PlatformUtil for list of installed language packs.
+     * 
+     * @param languages List of languages to use
+     */
+    public void setOCRLanguages(List<String> languages) {
+        this.ocrLanguages = languages;
+    }
+
+    /**
+     * Gets the list of languages OCR should perform.
+     * 
+     * @return Collection of OCR languages
+     */
+    public List<String> getOCRLanguages() {
+        return this.ocrLanguages;
+    }
 }