public IEnumerable <string> Merge(JsonConversionData jsonFileData, DocxConversionData docxFileData, Boolean ignoreHidden = false)
{
if (docxFileData == null)
{
throw new ArgumentNullException("docxFileData", "DocxConversionData cannot be null.");
}
if (jsonFileData == null)
{
throw new ArgumentNullException("jsonFileData", "JsonFileData cannot be null.");
}
string docxConversionDataFullPath = docxFileData.FullPath;
using (WordprocessingDocument doc = WordprocessingDocument.Open(docxConversionDataFullPath, true))
{
var strings2Translate = textExtractionService.ExtractText(doc, ignoreHidden);
for (int j = 0; j < docxFileData.TotalStrings2Translate; j++)
{
int indexInDocument = j + 1;
var newValue = (string)jsonFileData.JsonData["lines"][j];
strings2Translate.Take(indexInDocument).Last().Text = newValue;
}
IEnumerable <string> outputStrings2Translate = strings2Translate.Select(text => text.Text);
return(outputStrings2Translate);
}
}
public DocxConversionData CreateCleanedCopy(string fullPath, bool ignoreHidden)
{
if (string.IsNullOrEmpty(fullPath))
{
throw new ArgumentNullException("path variable cannot be null or empty");
}
// ensure we have a valid path to a docx file
if (!validator.IsValid(fullPath))
{
return(null);
}
var origFileData = new DocxConversionData(fullPath);
var cleanedFileName = origFileData.FileName + "_cleaned" + origFileData.FileExtension;
var cleanedFileData = new DocxConversionData(Path.Join(origFileData.FilePath, cleanedFileName));
try
{
// ensure there is not an orphaned copy already on disk
File.Delete(cleanedFileData.FullPath);
// Make a copy of the original using the "cleaned" file name
File.Copy(origFileData.FullPath, cleanedFileData.FullPath);
//Clean Document of extra tags
using (WordprocessingDocument doc = WordprocessingDocument.Open(cleanedFileData.FullPath, true))
{
OpenXmlPowerTools.SimplifyMarkupSettings settings = new OpenXmlPowerTools.SimplifyMarkupSettings
{
AcceptRevisions = true,
NormalizeXml = false, //setting this to false reduces translation quality, but if true some documents have XML format errors when opening
RemoveBookmarks = true,
RemoveComments = true,
RemoveContentControls = true,
RemoveEndAndFootNotes = true,
RemoveFieldCodes = false, //FieldCode remove pagination
RemoveGoBackBookmark = true,
//RemoveHyperlinks = false,
RemoveLastRenderedPageBreak = true,
RemoveMarkupForDocumentComparison = true,
RemovePermissions = false,
RemoveProof = true,
RemoveRsidInfo = true,
RemoveSmartTags = true,
RemoveSoftHyphens = true,
RemoveWebHidden = true,
ReplaceTabsWithSpaces = false
};
OpenXmlPowerTools.MarkupSimplifier.SimplifyMarkup(doc, settings);
//Extract Strings for translation
var textBlocks = textExtractor.ExtractText(doc, ignoreHidden);
//var textBlocks = textExtractor.ExtractParagraph(doc, ignoreHidden);
cleanedFileData.Strings2Translate = textBlocks.Select(para => para.InnerText);
}
//Total the number of strings for translation
cleanedFileData.TotalStrings2Translate = cleanedFileData.Strings2Translate.Count();
if (File.Exists(cleanedFileData.FullPath))
{
cleanedFileData.Messages.Add("Success");
}
else
{
cleanedFileData.Messages.Add("Docx data failed to build properly.");
}
}
catch (Exception e)
{
cleanedFileData.Messages.Add(e.Message);
}
return(cleanedFileData);
}
