merge_2lang_tmx

/* :name = Merge 2 TMX with different source  :description=
 * Purpose: To merge two TMX files in the /tm folder, taking the source text of the first one and the target text of the second one, and combining them in one third TMX file.
 * #Files:  Expectes first file to be called Translator1.tmx and the second one to be called Translator2.tmx
            Renames Translator2.tmx as Translator2.tmx.fra-bak and outputs new Translator2.tmx
 * #Format: TMX v.1.4
 *
 * @author  Manuel Souto Pico, Kos Ivantsov
 * @date    2019-04-12
 * @version 0.2
 */

import groovy.swing.SwingBuilder
import groovy.util.XmlSlurper
import groovy.util.GroovyCollections
// import groovy.xml.XmlUtil
import java.awt.FlowLayout
import java.nio.file.Files
import java.nio.file.Paths
import java.util.zip.ZipFile
import javax.swing.JOptionPane
import javax.swing.WindowConstants as WC
import org.apache.commons.io.FileUtils
import org.omegat.util.Preferences
import org.omegat.util.StaticUtils
import org.omegat.util.StringUtil
import static javax.swing.JOptionPane.*
import static org.omegat.util.Platform.*

utils = (StringUtil.getMethods().toString().findAll("makeValidXML")) ? StringUtil : StaticUtils

def prop = project.projectProperties;
def targetLang = prop.getTargetLanguage();
def tmDir = prop.getTMRoot();
// console.println(targetLang);

tmxOne = new File(tmDir.toString() + File.separator + "Translator1.tmx");
tmxTwo = new File(tmDir.toString() + File.separator + "Translator2.tmx");

console.println('Merging source text from\n' + tmxOne + '\nwith target text from\n' + tmxTwo + '...\n')

parser=new XmlSlurper()
parser.setFeature("http://apache.org/xml/features/disallow-doctype-decl", false)
parser.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);

if (tmxOne.exists()) {
    // def t1EngTmx = new XmlSlurper().parse(tmxOne)
    tmxOneEntries = parser.parse(tmxOne)
}
if (tmxTwo.exists()) {
    // def t1EngTmx = new XmlSlurper().parse(tmxOne)
    tmxTwoEntries = parser.parse(tmxTwo)
}
//console.println(tmxOneEntries)    // outputs long string
// console.println tmxOneEntries.getClass()

finalTMX_contents = new StringWriter()
preamble = '''<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE tmx SYSTEM "tmx11.dtd">
<tmx version="1.1">
  <header creationtool="OmegaT" o-tmf="OmegaT TMX" adminlang="EN-US" datatype="plaintext" creationtoolversion="4.1.5_3_10485" segtype="sentence" srclang="en-ZZ"/>
  <body>'''
end = '''  </body>
</tmx>'''

def count = 0
while (count < tmxOneEntries.body.tu.size()) {
    //console.println(tmxOneEntries.body.tu[count].tuv[0].seg)
    //console.println(tmxTwoEntries.body.tu[count].tuv[1].seg)

    // Converting node to string, which also escapes <, > and & (but not ' and ")
    def src_node = tmxOneEntries.body.tu[count].tuv[0].seg
    //console.println src.getClass() // gives class groovy.util.slurpersupport.NodeChildren
    def src_str = new groovy.xml.StreamingMarkupBuilder().bindNode(src_node) as String
    // console.println src_str.getClass() // gives class java.lang.String
    // console.println(src_str)

    def tgt_node = tmxOneEntries.body.tu[count].tuv[0].seg
    def tgt_str = new groovy.xml.StreamingMarkupBuilder().bindNode(tgt_node) as String

    // def tgt = tmxTwoEntries.body.tu[count].tuv[1].seg

    finalTMX_contents << """    <tu>
      <tuv lang="en-ZZ">
        ${src_str}
      </tuv>
      <tuv lang="${targetLang}">
        ${tgt_str}
      </tuv>
    </tu>
"""
    count++
}

//console.println(finalTMX_contents)

String origPath = tmxTwo
tmxTwo.renameTo origPath.concat(".fra-bak")

finalTMX =  new File(tmDir + File.separator + "Translator2.tmx")
finalTMX.write(preamble + finalTMX_contents + end, "UTF-8")

console.println("Translator2.tmx has been re-built now with English as the source language.")

//return console.println(tmxOneEntries.body.tu[0].tuv[0].seg)