/// <summary>
/// Attempts to identify the locations of (substrings of) the search segment in the result segment.
/// Typically, the search segment is a single search string (such as a concordance search string),
/// and the result segment is a transation unit segment.
/// </summary>
/// <param name="search">The search text.</param>
/// <param name="text">The text segment, usually a sentence or larger block of text.</param>
/// <param name="culture">The culture the search and text belong to.</param>
/// <returns>The locations and other information about the coverage of the search segment
/// in the result segment.</returns>
public static TermFinderResult FindTerms(string search,
string text,
System.Globalization.CultureInfo culture,
bool expectContinuousMatch)
{
if (String.IsNullOrEmpty(search))
{
throw new ArgumentNullException("search");
}
if (String.IsNullOrEmpty(text))
{
throw new ArgumentNullException("text");
}
if (culture == null)
{
throw new ArgumentNullException("culture");
}
Core.Segment searchSegment = new Sdl.LanguagePlatform.Core.Segment(culture);
Core.Segment textSegment = new Sdl.LanguagePlatform.Core.Segment(culture);
searchSegment.Add(search);
textSegment.Add(text);
Lingua.Tokenization.Tokenizer t =
new Sdl.LanguagePlatform.Lingua.Tokenization.Tokenizer(Lingua.Tokenization.TokenizerSetupFactory.Create(culture));
searchSegment.Tokens = t.Tokenize(searchSegment);
textSegment.Tokens = t.Tokenize(textSegment);
return(FindTerms(searchSegment, textSegment, expectContinuousMatch));
}
// process the segment; individuating the content text, tags & comments
public void ProcessSegment(ISegment segment, bool includeTagText)
{
TagCounter = 0;
Segment = new Sdl.LanguagePlatform.Core.Segment();
PlainText = new StringBuilder(string.Empty);
SegmentSections = new List <SegmentSection>();
Comments = new List <Comment>();
TagPairs = new List <ITagPair>();
LockedContentTags = new List <ILockedContent>();
PlaceholderTags = new List <IPlaceholderTag>();
TagUnits = new List <TagUnit>();
IncludeTagText = includeTagText;
VisitChildren(segment);
}
public static string ToHtml(this Sdl.LanguagePlatform.Core.Segment segment)
{
var stringBuilder = new StringBuilder();
foreach (var element in segment.Elements)
{
var text = element as Sdl.LanguagePlatform.Core.Text;
if (text != null)
{
stringBuilder.Append(text.Value.XmlEscape());
}
else
{
var tag = element as Sdl.LanguagePlatform.Core.Tag;
if (tag != null)
{
switch (tag.Type)
{
case Sdl.LanguagePlatform.Core.TagType.Start:
stringBuilder.AppendFormat("<span class='{0}' id='{1}'>",
tag.Type, tag.Anchor);
break;
case Sdl.LanguagePlatform.Core.TagType.End:
stringBuilder.AppendFormat("</span>");
break;
case Sdl.LanguagePlatform.Core.TagType.Standalone:
case Sdl.LanguagePlatform.Core.TagType.TextPlaceholder:
case Sdl.LanguagePlatform.Core.TagType.LockedContent:
//stringBuilder.AppendFormat("<span class='{0}' id='{1}'></span>",=tag.Type, tag.Anchor);
stringBuilder.AppendFormat("<span class='{0}' id='{1}'></span> ",
tag.Type, tag.Anchor);
break;
default:
break;
}
}
}
}
return(stringBuilder.ToString());
}
public static Sdl.LanguagePlatform.Core.Segment ToSegment(this string translatedText, Sdl.LanguagePlatform.Core.Segment sourceSegment)
{
var htmlTagName = "span"; // the only we feed for translation is span, so we expect the translation only has span tags too.
var xmlFragment = "<segment>" + translatedText + "</segment>";
var xmlReader = new System.Xml.XmlTextReader(xmlFragment, System.Xml.XmlNodeType.Element, null);
var tagStack = new Stack <Sdl.LanguagePlatform.Core.Tag>();
var translatedSegment = new Sdl.LanguagePlatform.Core.Segment();
try
{
while (xmlReader.Read())
{
switch (xmlReader.NodeType)
{
case System.Xml.XmlNodeType.Element:
if (xmlReader.Name == htmlTagName)
{
var tagClass = xmlReader.GetAttribute("class");
var tagType = (Sdl.LanguagePlatform.Core.TagType)
Enum.Parse(typeof(Sdl.LanguagePlatform.Core.TagType), tagClass);
int id = Convert.ToInt32(xmlReader.GetAttribute("id"));
Sdl.LanguagePlatform.Core.Tag sourceTag = sourceSegment.FindTag(tagType, id);
if (tagType != Sdl.LanguagePlatform.Core.TagType.Standalone && !xmlReader.IsEmptyElement)
{
tagStack.Push(sourceTag);
}
translatedSegment.Add(sourceTag.Duplicate());
if (tagType != Sdl.LanguagePlatform.Core.TagType.Standalone && xmlReader.IsEmptyElement)
// the API translated <span></span> to <span/> (it does that if the tag is empty).
// must fetch the end tag as there is no EndElement to triger the next case block.
{
var endTag = sourceSegment.FindTag(Sdl.LanguagePlatform.Core.TagType.End, id);
translatedSegment.Add(endTag.Duplicate());
}
}
break;
case System.Xml.XmlNodeType.EndElement:
{
if (xmlReader.Name == htmlTagName)
{
var startTag = tagStack.Pop();
if (startTag.Type != Sdl.LanguagePlatform.Core.TagType.Standalone)
{
var endTag = sourceSegment.FindTag(
Sdl.LanguagePlatform.Core.TagType.End, startTag.Anchor);
if (endTag != null)
{
translatedSegment.Add(endTag.Duplicate());
}
}
}
}
break;
case System.Xml.XmlNodeType.Text:
translatedSegment.Add(xmlReader.Value);
break;
case System.Xml.XmlNodeType.Whitespace:
translatedSegment.Add(xmlReader.Value);
break;
default:
break;
}
}
}
catch (Exception)
{
var paintextSegment = new Sdl.LanguagePlatform.Core.Segment();
string plaitext = Regex.Replace(translatedText, "<[^>]+>", "");
paintextSegment.Add(plaitext);
return(paintextSegment);
}
return(translatedSegment);
}
