Changeset 25285

Timestamp:

2012-03-23T16:46:29+13:00 (12 years ago)

Author:

ak19

Message:

Updated to deal with paperspast.dm: added a new module into gti.pl for this. Because getting chunks from the paperspast.dm file results in attribute like values in the source and target strings, there are now functions in ApplyXSLT.java to remove and retrieve these. Finally, the function to remove these attribute-like values in the source and target strings is called in the gti-generate-tmx-xml.xslt file.

Location:

main/trunk/greenstone2

Files:

• bin/script/gti.pl (modified) (1 diff)
• build-src/src/java/org/nzdl/gsdl/ApplyXSLT.java (modified) (1 diff)

main/trunk/greenstone2/bin/script/gti.pl

@@ -54 +54 @@
     'target_file' => "macros/{bn:bengali;fa:farsi;gd:gaelic;id:indo;lv:latvian;pt-br:port-br;pt-pt:port-pt;zh-tr:chinese-trad;iso_639_1_target_language_name}2.dm" },
 
+{ 'key' => "paperspastdm",
+    'file_type' => "macrofile",
+    'source_file' => "macros/paperspast-english.dm",
+    'target_file' => "macros/paperspast-{bn:bengali;fa:farsi;gd:gaelic;id:indo;lv:latvian;pt-br:port-br;pt-pt:port-pt;zh-tr:chinese-trad;iso_639_1_target_language_name}.dm" },
+
 # GLI dictionary
 { 'key' => "glidict",

main/trunk/greenstone2/build-src/src/java/org/nzdl/gsdl/ApplyXSLT.java

@@ -336 +336 @@
   }
 
+    // Necessary for paperspast.dm, but can be used generally. 
+    // The get-chunks cmd of gti.pl perl script when run over paperspast.dm returns XML with source and target lines 
+    // like: [c=paperspast] {All newspapers} for source and [c=paperspast,l=mi] {Niupepa katoa} for target
+    // This function returns just the 'string' portion of the chunk of data: e.g 'All newspapers' and 'Niupepa katoa'
+    static public String getChunkString(String target_file_text)
+    {
+    int startindex = target_file_text.indexOf("[");
+    if(startindex != 0) {
+        return target_file_text;
+    } // to test that the input requires processing
+
+    // else
+    startindex = target_file_text.indexOf("{");
+    int endindex = target_file_text.lastIndexOf("}");
+    if(startindex != -1 && endindex != -1) {
+        return target_file_text.substring(startindex+1, endindex); // skips { and }
+    } else {
+        return target_file_text;
+    }
+    
+    }
+
+    // Necessary for paperspast.dm, but can be used generally. 
+    // The get-chunks cmd of gti.pl perl script when run over paperspast.dm returns XML with source and target lines 
+    // like: [c=paperspast] {All newspapers} for source and [c=paperspast,l=mi] {Niupepa katoa} for target
+    // This function returns just the 'attribute' portion of the chunk of data: e.g 'c=paperspast' and 'c=paperspast,l=mi'
+    static public String getChunkAttr(String target_file_text)
+    {
+    int startindex = target_file_text.indexOf("[");
+    if(startindex != 0) {
+        return target_file_text;
+    } // to test that the input requires processing
+
+    // else
+    startindex = target_file_text.indexOf("{");
+    int endindex = target_file_text.lastIndexOf("}");
+    if(startindex != -1 && endindex != -1) {
+        endindex = target_file_text.lastIndexOf("]", startindex); // look for ] preceding the {
+        if(endindex > 1) { //if(endindex != -1) {
+                       // so there's something to substring between [ and ]                        
+        return target_file_text.substring(1, endindex).trim(); // skips [ and ]
+        }
+    } 
+    return target_file_text;
+    }
 
   public static void main(String[] args)