Merge pull request #24 from zoho/hawking_dev

Master Merge Hawking Enhancement  v0.1.5

pom.xml

@@ -4,7 +4,7 @@
 
     <groupId>com.zoho</groupId>
     <artifactId>hawking</artifactId>
-    <version>0.1.4</version>
+    <version>0.1.5</version>
     <packaging>jar</packaging>
     <name>Hawking</name>
     <description>Hawking is a natural language date time parser that extracts date and time from text with context and parse to the required format.</description>

src/main/java/com/zoho/hawking/language/english/DateTimeGateWay.java

@@ -105,9 +105,9 @@ private static int findDateType(DateTimeEssentials dateTimeEssentials) {
       for (Triple<String, Integer, Integer> triple : triples) {
         if (triple.first.equals("R")) {
           String relationWord = dateTimeEssentials.getSentence().substring(triple.second, triple.third).replace("[^a-zA-z]", ""); //No I18N
-          if (PrepositionConstants.RELATIONSHIP_RANGE.contains(relationWord)) {
+          if (PrepositionConstants.RELATIONSHIP_RANGE.contains(relationWord.toLowerCase())) {
             dateType = 0;
-          } else if (PrepositionConstants.RELATIONSHIP_SET.contains(relationWord)) {
+          } else if (PrepositionConstants.RELATIONSHIP_SET.contains(relationWord.toLowerCase())) {
             dateType = 1;
           } else {
             dateType = 2;

src/main/java/com/zoho/hawking/language/english/DateTimeWordProperties.java

@@ -195,7 +195,7 @@ public class DateTimeWordProperties {
 
     public static final WordProperty AUGUST = new WordProperty("august", MonthsOfYear.MONTH_EIGHT, new String[]{"aug"});
 
-    public static final WordProperty SEPTEMBER = new WordProperty("september", MonthsOfYear.MONTH_NINE, new String[]{"sep"});
+    public static final WordProperty SEPTEMBER = new WordProperty("september", MonthsOfYear.MONTH_NINE, new String[]{"sep","sept"});
 
     public static final WordProperty OCTOBER = new WordProperty("october", MonthsOfYear.MONTH_TEN, new String[]{"oct"});
 

src/main/java/com/zoho/hawking/language/english/Parser.java

@@ -41,6 +41,8 @@ private static CRFClassifier<CoreLabel> getCRFInstance() {
     }
 
     public static List<Triple<String, Integer, Integer>> parse(String input) {
+        input = input.replaceAll("http","----"); //No I18N
+        input = input.replaceAll("www","---"); //No I18N
         return crf.classifyToCharacterOffsets(input);
     }
 

src/main/java/com/zoho/hawking/utils/RecognizerTagger.java

@@ -37,7 +37,7 @@ public class RecognizerTagger {
   private final static Pattern exact_year = Pattern.compile("^\\d{4}$");
   private final static Pattern exact_number_1 = Pattern.compile(
       "^(first|second|third|fourth|fifth|sixth|seventh|eighth|ninth|tenth|eleventh|twelfth|thirteenth|fourteenth|fifteenth|sixteenth|seventeenth|eightheenth|ninteenth|twentieth|twenty-first|twenty-second|twenty-third|twenty-fourth|twenty-fifth|twenty-sixth|twenty-seventh|twenty-eighth|twenty-ninth|thirtieth|thirty-first|one|two|three|four|five|six|seven|eight|nine|ten|eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen|eighteen|nineteen|twenty|thirty|forty|fifty|st|nd|rd|th)$");
-  private final static Pattern exact_number_2 = Pattern.compile("(\\d+)(st|nd|rd|th|ᵗʰ|ˢᵗ|ⁿᵈ|ʳᵈ)?");
+  private final static Pattern exact_number_2 = Pattern.compile("^((\\d+)(st|nd|rd|th|ᵗʰ|ˢᵗ|ⁿᵈ|ʳᵈ)?)$");
 
   public static String getTagger(String word) {
     String tag = "";