Changeset 33338 for gs3-extensions

Timestamp:

2019-07-20T23:24:46+12:00 (5 years ago)

Author:

ak19

Message:

1.After renaming the java class, changed all occurrences of the old name MaoriDetector to MaoriTextDetector; 2. Now can run in silent mode done, except when the usage is explicitly invoked with the double-minus help or minus-h flag.

Location:

gs3-extensions/maori-lang-detection/src

Files:

• MaoriTextDetector.class (modified) ( previous)
• MaoriTextDetector.java (modified) (22 diffs)

gs3-extensions/maori-lang-detection/src/MaoriTextDetector.java

@@ -17 +17 @@
  * EXPORT OPENNLP_HOME environment variable to be your apache OpenNLP installation.
  * Then, to compile this program:
- *    maori-lang-detection/src$ javac -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" MaoriDetector.java
+ *    maori-lang-detection/src$ javac -cp ".:$OPENNLP_HOME/lib/opennlp-tools-1.9.1.jar" MaoriTextDetector.java
  * To run this program, one of:
  *
- *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriDetector --help
+ *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriTextDetector --help
  *
- *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriDetector --file <full/path/to/textfile>
+ *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriTextDetector --file <full/path/to/textfile>
  *
- *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriDetector -
+ *    maori-lang-detection/src$ java -cp ".:$OPENNLP_HOME/lib/*" MaoriTextDetector -
  *       which 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.
@@ -31 +31 @@
  * Also has information on how to run this class if it's in a Java package.
  */
-public class MaoriDetector {
+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";
@@ -37 +37 @@
 
     /** Configurable: cut off minimum confidence value,
-    greater or equal to which determines that the best predicted language is acceptable to user of MaoriDetector. */
+    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 MaoriDetector won't print helpful messages while running. Set to true to run silently. */
+    /** silentMode set to false means MaoriTextDetector won't print helpful messages while running. Set to true to run silently. */
     public final boolean silentMode;
 
@@ -56 +56 @@
     
     
-    public MaoriDetector(boolean silentMode) throws Exception {
+    public MaoriTextDetector(boolean silentMode) throws Exception {
     this(silentMode, DEFAULT_MINIMUM_CONFIDENCE);
     }
     
-    public MaoriDetector(boolean silentMode, double min_confidence) throws Exception {
+    public MaoriTextDetector(boolean silentMode, double min_confidence) throws Exception {
     this.silentMode = silentMode;
     this.MINIMUM_CONFIDENCE = min_confidence;
@@ -110 +110 @@
     // Get the most probable language
     Language bestLanguage = myCategorizer.predictLanguage(text);
-    System.out.println("Best language: " + bestLanguage.getLang());
-    System.out.println("Best language confidence: " + bestLanguage.getConfidence());
+    doPrint("Best language: " + bestLanguage.getLang());
+    doPrint("Best language confidence: " + bestLanguage.getConfidence());
 
     return (bestLanguage.getLang().equals(langCode) && bestLanguage.getConfidence() >= this.MINIMUM_CONFIDENCE);
@@ -184 +184 @@
         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
-            System.err.println("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
+            doPrintErr("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
         }
         language = bestLanguage.getLang();
         cumulativeConfidence += bestLanguage.getConfidence();
         
-        System.err.println("Best predicted language for last " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")"); 
+        doPrintErr("Best predicted language for last " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")"); 
         
         // finished analysing language of NUM_LINES of text
@@ -206 +206 @@
         
         if(language != null && !bestLanguage.getLang().equals(language)) { // predicted lang of current n lines not the same as predicted lang for prev n lines
-        System.err.println("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
+        doPrintErr("**** WARNING: text seems to contain content in multiple languages or unable to consistently predict the same language.");           
         }
         language = bestLanguage.getLang();
         cumulativeConfidence += bestLanguage.getConfidence();
-        System.err.println("Best predicted language for final " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")");
+        doPrintErr("Best predicted language for final " + NUM_LINES + " lines: " + language + "(confidence: " + bestLanguage.getConfidence() + ")");
     }
     
@@ -220 +220 @@
     return (language.equals(langCode) && avgConfidence >= this.MINIMUM_CONFIDENCE);
     }
-
     
 
@@ -233 +232 @@
     
     if(languages == null || languages.length <= 0) {
-        System.err.println("No languages predicted for the input text");
+        doPrintErr("No languages predicted for the input text");
     } else {
         for(int i = 0; i < languages.length; i++) {
-        System.out.println("Language prediction " + i + ": " + languages[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:");
@@ -256 +271 @@
     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 MaoriDetector when attempting to detect the text's language");
+    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();
@@ -267 +282 @@
      *
      *    -1 if the input arguments were wrong
-     *    255(!) if an Exception occurred in instantiating the MaoriDetector when attempting to detect the text's language
+     *    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.
@@ -292 +307 @@
         i++;
         if(i >= args.length) {
-            System.err.println("ERROR: No minimum confidence value provided with --min-confidence|-c flag.\n");
+            doPrintErr(runSilent, "ERROR: No minimum confidence value provided with --min-confidence|-c flag.\n");
             printUsage = true;
             returnVal = -1;
@@ -302 +317 @@
             }
             } catch(NumberFormatException nfe) {
-            System.err.println("ERROR: value for min-confidence is the wrong format or out of range. It must be a (decimal point) number between 0-1.\n");
+            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;
@@ -313 +328 @@
         i++;
         if(i >= args.length) {
-            System.err.println("ERROR: No input file provided with --file|-f flag.\n");
+            doPrintErr(runSilent, "ERROR: No input file provided with --file|-f flag.\n");
             printUsage = true;
             returnVal = -1;
@@ -320 +335 @@
             inFile = new File(filePath);
             if(!inFile.isFile()) {
-            System.err.println("ERROR: Can't read text. Input file argument provided does not exist or is not a file.\n");
+            doPrintErr(runSilent, "ERROR: Can't read text. Input file argument provided does not exist or is not a file.\n");
             printUsage = true;
             returnVal = -1;
@@ -326 +341 @@
         }
         } else { // unrecognised input argument
-        System.err.println("ERROR: Unrecognised " + i + "th argument to this program.\n");
+        doPrintErr(runSilent, "ERROR: Unrecognised " + i + "th argument to this program.\n");
         printUsage = true;
         returnVal = -1;
@@ -333 +348 @@
 
     if(!readFromStdIn && inFile == null) { // at least one input source must be provided
-        System.err.println("ERROR: must specify source to read text from, either STDIN (-) or input file (--file <file>).\n");
+        doPrintErr(runSilent, "ERROR: must specify source to read text from, either STDIN (-) or input file (--file <file>).\n");
         printUsage = true;
         returnVal = -1;
@@ -339 +354 @@
     
     if(readFromStdIn && inFile != null) { // this program can't be asked to read from stdin and from an input file
-        System.err.println("ERROR: instructed to read from both STDIN and from an input file. Not possible.\n");
+        doPrintErr(runSilent, "ERROR: instructed to read from both STDIN and from an input file. Not possible.\n");
         printUsage = true;
         returnVal = -1;
@@ -345 +360 @@
     
     if(printUsage) {
-        if(!runSilent || returnVal == 2) { // if expressly asked for help or not running silent
+        // 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();
         }
@@ -352 +369 @@
     
     try {
-        MaoriDetector maoriTextDetector = null;
+        MaoriTextDetector maoriTextDetector = null;
         if(minConfidence == -1) {
-        maoriTextDetector = new MaoriDetector(runSilent);
+        maoriTextDetector = new MaoriTextDetector(runSilent);
         } else {
-        maoriTextDetector = new MaoriDetector(runSilent, minConfidence);
-        }
-        
-        //boolean textIsInMaori = maoriTextDetector.isTextInMaori(TEST_MRI_INPUT_TEXT);
+        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) {
-        System.err.println("Reading text from file " + inFile.getPath());
+        doPrint(runSilent, "Reading text from file " + inFile.getPath());
         try (BufferedReader reader = new BufferedReader(new FileReader(inFile))) {
             textIsInMaori = maoriTextDetector.isTextInMaori(reader);
@@ -370 +387 @@
         }
         else if (readFromStdIn) {
-        System.err.println("Waiting to read text from STDIN... (press Ctrl-D when done entering text)>");
+        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);            
@@ -386 +403 @@
         
     } finally {
-        System.err.println("Exitting program with returnVal " + returnVal + "...\n");
+        doPrint(runSilent, "Exitting program with returnVal " + returnVal + "...\n");
         System.exit(returnVal);
     }
     }
-
-    // test hardcoded string
-    public static void oldMain(String args[]) {
-    int returnVal = -1;
-    boolean silentMode = false;
-    
-    try {
-        MaoriDetector maoriTextDetector = new MaoriDetector(silentMode);
-        
-        boolean textIsInMaori = maoriTextDetector.isTextInMaori(TEST_MRI_INPUT_TEXT);
-        if(textIsInMaori) {
-        returnVal = 0;
-        } else {
-        returnVal = 1;
-        }
-        
-    } catch(Exception e) {
-        e.printStackTrace();
-    } finally {
-        System.err.println("Exitting program with returnVal " + returnVal + "...\n");
-        System.exit(returnVal);
-    }
-    }
-
     
 }