Changeset 5677 for trunk

Timestamp:

2003-10-17T09:48:14+13:00 (21 years ago)

Author:

jmt12

Message:

Extended the parsers ability to recognize collectionextras, such as the one containing proper new lines found in the cdl collection. This process isn't fail safe, as a missing closing speechmark might cause the remainder of the collect.cfg to appear as description - but its the best you can do

File:

• trunk/gli/src/org/greenstone/gatherer/collection/CollectionConfiguration.java (modified) (3 diffs)

trunk/gli/src/org/greenstone/gatherer/collection/CollectionConfiguration.java

@@ -6 +6 @@
 import org.greenstone.gatherer.msm.MSMUtils;
 import org.greenstone.gatherer.util.Codec;
+import org.greenstone.gatherer.util.StaticStrings;
 import org.greenstone.gatherer.util.Utility;
 import org.w3c.dom.*;
@@ -72 +73 @@
                 }
                 else if(temp.equalsIgnoreCase("collectionextra")) {
-                    temp = tokenizer.nextToken();
-                    ///ystem.err.println("Read: " + temp);
-                    if(temp.startsWith("[") && temp.endsWith("]")) {
-                    String language = temp.substring(temp.indexOf("=") + 1, temp.length() - 1);
-                    ///ystem.err.println("Try to match " + language + " to " + Gatherer.dictionary.getLanguage());
-                    if(description == null || language.equalsIgnoreCase(Gatherer.dictionary.getLanguage())) {
-                        description = Utility.decodeGreenstone(tokenizer.nextToken());
-                        ///ystem.err.println("Found language match, or first match: " + description);
-                    }
-                    }
-                    else if(description == null) {
-                    description = Codec.transform(temp, Codec.GREENSTONE_TO_TEXT);
-                    ///ystem.err.println("Found first match: " + description);
-                    }
-                    if(description != null && description.startsWith("\"") && description.endsWith("\"")) {
-                    description = description.substring(1, description.length() - 1);
-                    }
+                   // Normal case
+                   temp = tokenizer.nextToken();
+                   ///ystem.err.println("Read: " + temp);
+                      if(temp.startsWith("[") && temp.endsWith("]")) {
+                     String language = temp.substring(temp.indexOf("=") + 1, temp.length() - 1);
+                     ///ystem.err.println("Try to match " + language + " to " + Gatherer.dictionary.getLanguage());
+                        if(description == null || language.equalsIgnoreCase(Gatherer.dictionary.getLanguage())) {
+                           description = Utility.decodeGreenstone(tokenizer.nextToken());
+                           ///ystem.err.println("Found language match, or first match: " + description);
+                          }
+                      }
+                      else if(description == null){
+                     // Some of the collections floating around have collection extras which contain normal newlines, and so their initial description starts with " (the Command Tokenizer hasn't found a closing speech mark). This will cause an error message to be displayed for the collection. If we have read a collectionextra tag, but we have run out of tokens then we do a sneaky and start reading lines until we encounter a closing "
+                     if(temp.startsWith(StaticStrings.SPEECH_CHARACTER) && (!temp.endsWith(StaticStrings.SPEECH_CHARACTER) || temp.length() ==1 )) {
+                        ///ystem.err.println("Encountered a strange Collection Description. Doing my best to parse a sensible text block!");
+                        StringBuffer description_raw = new StringBuffer("");
+                        while(description_raw.indexOf(StaticStrings.SPEECH_CHARACTER) == -1) {
+                           String next_line = br.readLine();
+                           if(next_line != null) {
+                          description_raw.append(next_line);
+                          description_raw.append(StaticStrings.NEW_LINE_CHAR);
+                           }
+                           next_line = null;
+                        }
+                        description = description_raw.substring(0, description_raw.lastIndexOf(StaticStrings.SPEECH_CHARACTER));
+                        description_raw = null;
+                     }
+                     else {
+                        description = Codec.transform(temp, Codec.GREENSTONE_TO_TEXT);
+                     }
+                      }
+                      if(description != null && description.startsWith(StaticStrings.SPEECH_CHARACTER) && description.endsWith(StaticStrings.SPEECH_CHARACTER)) {
+                     description = description.substring(1, description.length() - 1);
+                      }  
                 }
                 }
@@ -163 +181 @@
 
     public String getName() {
-    String result = "Error";
+    String result = StaticStrings.ERROR_STR;
     if(name_element != null) {
         result = MSMUtils.getValue(name_element);