source: gs3-extensions/maori-lang-detection/src/org/greenstone/atea/MaoriTextDetector.java@ 33577

/**
 * Class that uses OpenNLP with the Language Detection Model to determine, with a default
 * or configurable level of confidence, whether text (from a file or stdin) is in Māori or not.
 * Internal functions can be used for detecting any of the 103 languages currently supported by
 * the OpenNLP Language Detection Model.
 * 
 * http://opennlp.apache.org/news/model-langdetect-183.html
 * language detector model: http://opennlp.apache.org/models.html
 *        Pre-trained models for OpenNLP 1.5: http://opennlp.sourceforge.net/models-1.5/
 * Use of Apache OpenNLP in general:
 *   http://opennlp.apache.org/docs/1.9.1/manual/opennlp.html#intro.api
 * Use of OpenNLP for language detection:
 * http://opennlp.apache.org/docs/1.9.1/manual/opennlp.html#tools.langdetect
 * 
 * This code was based on the information and sample code at the above links and the links dispersed throughout this file.
 * See also the accompanying README file.
 *
 * July 2019
 */

package org.greenstone.atea;

import java.io.*;
import opennlp.tools.langdetect.*;
import opennlp.tools.util.*;

/**
 * EXPORT OPENNLP_HOME environment variable to be your apache OpenNLP installation.
 * Create a folder called "models" within the $OPENNLP_HOME folder, and put the file "langdetect-183.bin" in there
 *    (which is the language detection model zipped up and renamed to .bin extension).
 *
 * Then, to compile this program, do the following from the "src" folder (the folder containing this java file):
 *    maori-lang-detection/src$ javac -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" org/greenstone/atea/MaoriTextDetector.java
 *
 * To run this program, issue one of the following commands from the "src" folder (the folder containing this java file):
 *
 *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" org.greenstone.atea.MaoriTextDetector --help
 *
 *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" org.greenstone.atea.MaoriTextDetector --file <full/path/to/textfile>
 *
 *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" org.greenstone.atea.MaoriTextDetector -
 *       Press enter. This variant of the program expects text to stream in from standard input.
 *       If entering text manually, then remember to press Ctrl-D to indicate the usual end of StdIn.
 *
 * https://stackoverflow.com/questions/219585/including-all-the-jars-in-a-directory-within-the-java-classpath
 * Also has information on how to run this class if it's in a Java package.
 */
public class MaoriTextDetector {
    /** The 3 letter language code for Maori in ISO 639-2 or ISO 639-3 */ 
    public static final String MAORI_3LETTER_CODE = "mri";
    public static final double DEFAULT_MINIMUM_CONFIDENCE = 0.50;

    /** Configurable: cut off minimum confidence value,
    greater or equal to which determines that the best predicted language is acceptable to user of MaoriTextDetector. */
    public final double MINIMUM_CONFIDENCE;
    
    /** silentMode set to false means MaoriTextDetector won't print helpful messages while running. Set to true to run silently. */
    public final boolean silentMode;

    /** Language Detection Model file for OpenNLP is expected to be at $OPENNLP_HOME/models/langdetect-183.bin */
    private final String LANG_DETECT_MODEL_RELATIVE_PATH = "models" + File.separator + "langdetect-183.bin";

    /**
     * The LanguageDetectorModel object that will do the actual language detection/prediction for us.
     * Created once in the constructor, can be used as often as needed thereafter.
    */
    private LanguageDetector myCategorizer = null;
    
    /** String taken from our university website, https://www.waikato.ac.nz/maori/ */
    public static final String TEST_MRI_INPUT_TEXT = "Ko tēnei te Whare Wānanga o Waikato e whakatau nei i ngā iwi o te ao, ki roto i te riu o te awa e rere nei, ki runga i te whenua e hora nei, ki raro i te taumaru o ngā maunga whakaruru e tau awhi nei.";

    /** test input string for a negative result */
    public static final String TEST_ENG_INPUT_TEXT = "The main program exits with -1 if an Exception occurred when attempting to detect the text's language";
    
    
    public MaoriTextDetector(boolean silentMode) throws Exception {
    this(silentMode, DEFAULT_MINIMUM_CONFIDENCE);
    }
    
    public MaoriTextDetector(boolean silentMode, double min_confidence) throws Exception {
    this.silentMode = silentMode;
    this.MINIMUM_CONFIDENCE = min_confidence;

    // 1. Check we can find the Language Detect Model file in the correct location (check that $OPENNLP_HOME/models/langdetect-183.bin exists);
    String langDetectModelPath = System.getenv("OPENNLP_HOME");
    if(System.getenv("OPENNLP_HOME") == null) {
        throw new Exception("\n\t*** Environment variable OPENNLP_HOME must be set to your Apache OpenNLP installation folder.");
    }   
    langDetectModelPath = langDetectModelPath + File.separator + LANG_DETECT_MODEL_RELATIVE_PATH;
    File langDetectModelBinFile = new File(langDetectModelPath);
    if(!langDetectModelBinFile.exists()) {
        throw new Exception("\n\t*** " + langDetectModelBinFile.getPath() + " doesn't exist."
                + "\n\t*** Ensure the $OPENNLP_HOME folder contains a 'models' folder with the model file 'langdetect-183.bin' in it.");
    }


    // 2. Set up our language detector Model and the Categorizer for language predictions based on the Model.
    // http://opennlp.apache.org/docs/1.9.1/manual/opennlp.html#intro.api
    // https://docs.oracle.com/javase/tutorial/essential/exceptions/tryResourceClose.html
    try (InputStream modelIn = new FileInputStream(langDetectModelPath)) {

        LanguageDetectorModel model = new LanguageDetectorModel(modelIn);

        // http://opennlp.apache.org/docs/1.9.1/manual/opennlp.html#tools.langdetect
        this.myCategorizer = new LanguageDetectorME(model);
    }/*catch(Exception e) {
        e.printStackTrace();
        }*/
    
    // instantiating function should handle critical exceptions. Constructors shouldn't.
    }    

    /**
     * @return true if the input text is Maori (mri) with MINIMUM_CONFIDENCE levels of confidence (if set,
     * else DEFAULT_MINIMUM_CONFIDENCE levels of confidence).
     */
    public boolean isTextInMaori(String text) { 
    return isTextInLanguage(MAORI_3LETTER_CODE, text);
    }

    /** @param langCode is 3 letter language code, ISO 639-2/3 
     * https://www.loc.gov/standards/iso639-2/php/code_list.php
     * https://en.wikipedia.org/wiki/ISO_639-3
     * @return true if the input text is Maori (mri) with MINIMUM_CONFIDENCE levels of confidence (if set,
     * else DEFAULT_MINIMUM_CONFIDENCE levels of confidence).
     */
    public boolean isTextInLanguage(String langCode, String text) {
    // Get the most probable language
    Language bestLanguage = myCategorizer.predictLanguage(text);
    doPrint("Best language: " + bestLanguage.getLang());
    doPrint("Best language confidence: " + bestLanguage.getConfidence());

    return (bestLanguage.getLang().equals(langCode) && bestLanguage.getConfidence() >= this.MINIMUM_CONFIDENCE);
    }
    
    
    /**
     * Handle "smaller" textfiles/streams of text read in.
     * Return value is the same as for isTextInMaori(String text);
     */
    public boolean isTextInMaori(BufferedReader reader) throws Exception {
    return isTextInLanguage(MAORI_3LETTER_CODE, reader);
    }
    /**
     * Handle "smaller" textfiles/streams of text read in.
     * Return value is the same as for isTextInLanguage(String langCode, String text);
     */
    public boolean isTextInLanguage(String langCode, BufferedReader reader) throws Exception {
    // https://stackoverflow.com/questions/326390/how-do-i-create-a-java-string-from-the-contents-of-a-file
    
    StringBuilder text = new StringBuilder();
    String line = null;

    
    while((line = reader.readLine()) != null) { // readLine removes newline separator
        text.append(line + "\n"); // add back (unix style) line ending
    }
    return isTextInLanguage(langCode, text.toString());
    }
    
    /*
     * Need better handling of "larger" textfiles/streams of text read in:
     * what if multiple languages with high confidence every NUM_LINES read in?
     * Does this mean the file is multi-lingual with each section dominated by a different language?
     * How best to convey such information to the user?
     */
    /**
     * Rudimentary attempt to deal with very large files.
     * Return value is the same as for isTextInMaori(String text);
     */
    public boolean isLargeTextInMaori(BufferedReader reader) throws Exception {
    return isLargeTextInLanguage(MAORI_3LETTER_CODE, reader);
    }

    /**
     * Rudimentary attempt to deal with very large files.
     * Return value is the same as for isTextInLanguage(String langCode, String text);
     */    
    public boolean isLargeTextInLanguage(String langCode, BufferedReader reader) throws Exception {
    // https://stackoverflow.com/questions/326390/how-do-i-create-a-java-string-from-the-contents-of-a-file
    
    final int NUM_LINES = 100; // arbitrary 100 lines read, predict language, calculate confidence

    StringBuilder text = new StringBuilder();
    String line = null;
    
    double cumulativeConfidence = 0;
    int numLoops = 0;
    
    int i = 0;
    String language = null;
    
    while((line = reader.readLine()) != null) { // readLine removes newline separator
        text.append(line + "\n"); // add back (unix style) line ending
        
        i++; // read nth line of numLoop
        
        
        if(i == NUM_LINES) { // arbitrary 100 lines read, predict language, calculate confidence
        
        
        Language bestLanguage = myCategorizer.predictLanguage(text.toString());
        if(language != null && !bestLanguage.getLang().equals(language)) { // predicted lang of current n lines not the same as predicted lang for prev n lines
            doPrintErr("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
        }
        language = bestLanguage.getLang();
        cumulativeConfidence += bestLanguage.getConfidence();
        
        doPrintErr("Best predicted language for last " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")"); 
        
        // finished analysing language of NUM_LINES of text
        text = new StringBuilder();
        i = 0;
        numLoops++;
        }       
    }
    
    // process any (remaining) text that was less than n NUM_LINES
    if(!text.toString().equals("")) {
        text.append(line + "\n"); // add back (unix style) line ending      
        i++;
        
        Language bestLanguage = myCategorizer.predictLanguage(text.toString());
        
        if(language != null && !bestLanguage.getLang().equals(language)) { // predicted lang of current n lines not the same as predicted lang for prev n lines
        doPrintErr("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
        }
        language = bestLanguage.getLang();
        cumulativeConfidence += bestLanguage.getConfidence();
        doPrintErr("Best predicted language for final " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")");
    }
    
    
    int totalLinesRead = numLoops * NUM_LINES + i; // not used
    double avgConfidence = cumulativeConfidence/(numLoops + 1); // not quite the average as the text processed outside the loop may have fewer lines than NUM_LINES
    
    
    return (language.equals(langCode) && avgConfidence >= this.MINIMUM_CONFIDENCE);
    }
    

    /**
     * Prints to STDOUT the predicted languages of the input text in order of descending confidence.
     * UNUSED.
     */
    public void predictedLanguages(String text) {
    // Get an array with the most probable languages
    
    Language[] languages = myCategorizer.predictLanguages(text);
    
    if(languages == null || languages.length <= 0) {
        doPrintErr("No languages predicted for the input text");
    } else {
        for(int i = 0; i < languages.length; i++) {
        doPrint("Language prediction " + i + ": " + languages[i]);
        }
    }
    
    }

    public void doPrint(String msg) {
    doPrint(this.silentMode, msg);
    }
    public void doPrintErr(String msg) {
    doPrintErr(this.silentMode, msg);
    }

    /********** STATIC METHODS *************/

    public static void doPrint(boolean runSilent, String msg) {
    if(!runSilent) System.out.println(msg); 
    }
    public static void doPrintErr(boolean runSilent, String msg) {
    if(!runSilent) System.err.println(msg); 
    }
    
    public static void printUsage() {
    System.err.println("Run this program with:");
    System.err.println("\t--help (-h)\tfor seeing this usage message again");
    System.err.println("\t-\tto have input text read from STDIN (as always, hit Ctrl-D to mark end of text stream)");
    System.err.println("\t--file (-f)\tto provide an input file path");
    System.err.println("\t--silent (-s): optional, to run silently and just exit with exit value.");
    System.err.println("\t--min-confidence (-c): optional, to override the default minimum confidence value (" + DEFAULT_MINIMUM_CONFIDENCE + ")");
    System.err.println("\t\tof the predicted language that will be considered acceptable.");
    System.err.println();
    System.err.println("This program terminates with exit value:");
    System.err.println("\t0 if the input text is in Maori");
    System.err.println("\t1 if input text is not in Maori");
    System.err.println();
    System.err.println("\t-1 if the input arguments were wrong");
    System.err.println("\t255(!) if an Exception occurred in instantiating the MaoriTextDetector when attempting to detect the text's language");
    System.err.println("\t2 if the user asked to run this program with --help/-h.");
    System.err.println();
    }

    /**
     * The main program exits with:
     *    0 if text is in Maori;
     *    1 if text is not in Maori;
     *
     *    -1 if the input arguments were wrong
     *    255(!) if an Exception occurred in instantiating the MaoriTextDetector when attempting to detect the text's language
     * QTODO: why does the program exit value end up as 255 and not -1 when returnVal remains at -1 on Exception?
     *    2 if the user asked to run this program with --help/-h.
     */
    public static void main(String args[]) {
    int returnVal = -1;

    // 1. Check input arguments
    boolean printUsage = false;
    boolean readFromStdIn = false;
    File inFile = null;
    boolean runSilent = false;
    double minConfidence = -1;

    for (int i = 0; !printUsage && i < args.length; i++) {
        
        // check for help first and quit after printing usage
        if(args[i].equals("--help") || args[i].equals("-h")) {
        printUsage = true;
        returnVal = 2;      
        } else if(args[i].equals("--silent") || args[i].equals("-s")) {
        runSilent = true;
        } else if(args[i].equals("--min-confidence") || args[i].equals("-c")) {
        i++;
        if(i >= args.length) {
            doPrintErr(runSilent, "ERROR: No minimum confidence value provided with --min-confidence|-c flag.\n");
            printUsage = true;
            returnVal = -1;
        } else {
            try {
            minConfidence = Double.parseDouble(args[i]);
            if(minConfidence < 0 || minConfidence > 1) {
                throw new NumberFormatException("Number out of range, must be between 0-1");
            }
            } catch(NumberFormatException nfe) {
            doPrintErr(runSilent, "ERROR: value for min-confidence is the wrong format or out of range. It must be a (decimal point) number between 0-1.\n");
            printUsage = true;
            returnVal = -1;
            }
        }
        } else if(args[i].equals("-")) {
        readFromStdIn = true;
        //break; // don't bother continuing to check input arguments for any --file flag if we're told to read from stdin
        } else if(args[i].equals("--file") || args[i].equals("-f")) {
        i++;
        if(i >= args.length) {
            doPrintErr(runSilent, "ERROR: No input file provided with --file|-f flag.\n");
            printUsage = true;
            returnVal = -1;
        } else {
            String filePath = args[i];
            inFile = new File(filePath);
            if(!inFile.isFile()) {
            doPrintErr(runSilent, "ERROR: Can't read text. Input file argument provided does not exist or is not a file.\n");
            printUsage = true;
            returnVal = -1;
            }
        }
        } else { // unrecognised input argument
        doPrintErr(runSilent, "ERROR: Unrecognised " + i + "th argument to this program.\n");
        printUsage = true;
        returnVal = -1;
        }
    }

    if(returnVal != 2) { // returnVal == 2 for help. Only if the user did not request --help/-h, do we continue to make sure the arguments provided are sane
        if(!readFromStdIn && inFile == null) { // at least one input source must be provided
        doPrintErr(runSilent, "ERROR: must specify source to read text from, either STDIN (-) or input file (--file <file>).\n");
        printUsage = true;
        returnVal = -1;
        }
        
        if(readFromStdIn && inFile != null) { // this program can't be asked to read from stdin and from an input file
        doPrintErr(runSilent, "ERROR: instructed to read from both STDIN and from an input file. Not possible.\n");
        printUsage = true;
        returnVal = -1;
        }
    }
    
    if(printUsage) {
        // If not running silent print usage.
        // OR if expressly asked for help, then it doesn't matter if we're running silent: still print usage to stderr.
        if(returnVal == 2 || !runSilent) {          
        printUsage();
        }
        System.exit(returnVal);
    }   


    // 2. Finally, we can now do the actual language detection
    try {
        MaoriTextDetector maoriTextDetector = null;
        if(minConfidence == -1) {
        maoriTextDetector = new MaoriTextDetector(runSilent);
        } else {
        maoriTextDetector = new MaoriTextDetector(runSilent, minConfidence);
        }
        
        //boolean textIsInMaori = maoriTextDetector.isTextInMaori(TEST_MRI_INPUT_TEXT); // test hardcoded string
        boolean textIsInMaori = false;
        
        // Using try with resources, https://docs.oracle.com/javase/tutorial/essential/exceptions/tryResourceClose.html
        if(inFile != null) {
        doPrint(runSilent, "Reading text from file " + inFile.getPath());
        try (BufferedReader reader = new BufferedReader(new FileReader(inFile))) {
            textIsInMaori = maoriTextDetector.isTextInMaori(reader);
        } // let outer try deal with any file/reading exceptions
        }
        else if (readFromStdIn) {
        doPrint(runSilent, "Waiting to read text from STDIN... (press Ctrl-D when done entering text)>");
        try (BufferedReader reader = new BufferedReader(new InputStreamReader(System.in))) {
            textIsInMaori = maoriTextDetector.isTextInMaori(reader);            
        } // let outer try deal with any file/reading exceptions
        }
    
        if(textIsInMaori) {
        returnVal = 0;
        } else {
        returnVal = 1;
        }
        
    } catch(Exception e) {
        e.printStackTrace();
        
    } finally {
        doPrint(runSilent, "Exitting program with returnVal " + returnVal + "...\n");
        System.exit(returnVal);
    }
    }
    
}