Generating the Paragraph Units

In this chapter you will learn to extend the writer class to output the actual paragraph units from the intermediary (SDLXliff) file to the target BIL document.

Modify the Function for Processing the Paragraph Units

First, we will add the required application logic to the method CreateParagraphUnit(), which loops through the paragraph units in the intermediary document. The function is supposed to work like this: the id of the current paragraph unit is retrieved from the context metadata. Remember that we added the unit id as hidden meta information (see Adding Context Information). Using the id we will determine the corresponding unit in the original BIL document. We then call a separate helper function, which replaces the segments from the original BIL document with the paragraph unit content, and thus the segment pairs from the intermediary document.

The method should look as shown below:

public void ProcessParagraphUnit(IParagraphUnit paragraphUnit)
{
    string unitId = paragraphUnit.Properties.Contexts.Contexts[1].GetMetaData("UnitID");
    XmlNode xmlUnit = _targetFile.SelectSingleNode("//unit[@id='" + unitId + "']");

    // call helper function to generate the paragraph unit
    CreateParagraphUnit(paragraphUnit, xmlUnit);
}

Note that for this simple implementation we just assume that every paragraph unit in the intermediary document corresponds to a unit id from the original BIL file. In a 'real-life' implementation you should, of course, make sure to catch cases in which a given id does not exist.

Add the Helper Function to Generate the Target Paragraph Units

Now it is time to implement the helper function that generates the new BIL unit elements from the source target segments in the intermediary file. This function leverages the BilTextExtractor class, which we created previously (see Adding the Text Extractor Class). Therefore, you should declare a new global text extractor member derived from this class:

private BilTextExtractor _textExtractor;

Then you use the the Initialize() method to create a text extractor object:

public void Initialize(IDocumentProperties documentInfo)
{
    _textExtractor = new BilTextExtractor();
}

Now you can add the following helper function, which works as follows: the source and the target elements are taken from the current BIL unit element. Then, the function loops through the segment pairs of the current paragraph unit from the intermediary (SDLXliff) document. The text extractor object is used to retrieve the plain text and the tag pairs for each segment pair. The source and the target elements are then set to the strings that were retrieved through the text extractor object.

private void CreateParagraphUnit(IParagraphUnit paragraphUnit, XmlNode xmlUnit)
{
    XmlNode source = xmlUnit.SelectSingleNode("source");
    XmlNode target = xmlUnit.SelectSingleNode("target");

    //iterate all segment pairs
    foreach (ISegmentPair segmentPair in paragraphUnit.SegmentPairs)
    {
        source.InnerXml = _textExtractor.GetPlainText(segmentPair.Source);
        target.InnerXml = _textExtractor.GetPlainText(segmentPair.Target);
    }
}

Note that we loop through all segment pairs, although in our simplified implementation each unit can be assumed to contain only one segment pair. A paragraph unit in the intermediary (SDLXliff) document would contain several segment pairs if our fictitious BIL format allowed for more than one seg element within a source or target node.

Putting it All Together

The enhanced writer class should now look as shown below:

using System;
using System.Xml;
using Sdl.FileTypeSupport.Framework.BilingualApi;
using Sdl.FileTypeSupport.Framework.NativeApi;

namespace Sdk.Snippets.Bilingual
{
    class BilWriter2 : AbstractBilingualFileTypeComponent, IBilingualWriter, INativeOutputSettingsAware
    {
        #region "global"
        private IPersistentFileConversionProperties _originalFileProperties;
        private INativeOutputFileProperties _nativeFileProperties;
        private XmlDocument _targetFile;
        private BilTextExtractor _textExtractor;
        #endregion

        #region "INativeOutputSettingsAware members"
        public void GetProposedOutputFileInfo(IPersistentFileConversionProperties fileProperties, IOutputFileInfo proposedFileInfo)
        {
            _originalFileProperties = fileProperties;
        }

        public void SetOutputProperties(INativeOutputFileProperties properties)
        {
            _nativeFileProperties = properties;
        }
        #endregion


        #region "IBilingualWriter members"


        #region "load file"
        public void SetFileProperties(IFileProperties fileInfo)
        {
            _targetFile = new XmlDocument();
            _targetFile.Load(_originalFileProperties.OriginalFilePath);
        }

        #region "initialize"
        public void Initialize(IDocumentProperties documentInfo)
        {
            _textExtractor = new BilTextExtractor();
        }
        #endregion

        #endregion

        #region "paragraphs"
        public void ProcessParagraphUnit(IParagraphUnit paragraphUnit)
        {
            string unitId = paragraphUnit.Properties.Contexts.Contexts[1].GetMetaData("UnitID");
            XmlNode xmlUnit = _targetFile.SelectSingleNode("//unit[@id='" + unitId + "']");

            // call helper function to generate the paragraph unit
            CreateParagraphUnit(paragraphUnit, xmlUnit);
        }
        #endregion

        #region "create paragraph"
        private void CreateParagraphUnit(IParagraphUnit paragraphUnit, XmlNode xmlUnit)
        {
            XmlNode source = xmlUnit.SelectSingleNode("source");
            XmlNode target = xmlUnit.SelectSingleNode("target");

            //iterate all segment pairs
            foreach (ISegmentPair segmentPair in paragraphUnit.SegmentPairs)
            {
                source.InnerXml = _textExtractor.GetPlainText(segmentPair.Source);
                target.InnerXml = _textExtractor.GetPlainText(segmentPair.Target);
            }
        }
        #endregion

        #region "save file and complete"
        public void FileComplete()
        {
            _targetFile.Save(_nativeFileProperties.OutputFilePath);
            _targetFile = null;
        }

        public void Complete()
        {

        }
        #endregion

        #endregion

        #region Implementation of IDisposable

        public void Dispose()
        {
            throw new NotImplementedException();
        }

        #endregion
    }
}

Generating the Paragraph Units

Modify the Function for Processing the Paragraph Units

Add the Helper Function to Generate the Target Paragraph Units

Putting it All Together

See Also

Note