/// <summary>
/// Attempts to parse a string on the <see cref="TagDelimiter"/>
/// followed by a conversion of the string array into <see cref="TagsetBase"/>
/// list.
/// </summary>
/// <param name="taggedString">The string to be parsed into a list of <see cref="TagsetBase"/>.</param>
/// <param name="tagset">The resulting list, will be null when return value is false.</param>
/// <remarks>
/// The argument <see cref="taggedString"/> is split first on the space character (0x20) then each
/// value in the resulting array is split on the <see cref="TagDelimiter"/>. Non-matching results are
/// assigned to a <see cref="NullTagset"/> instance - this is only the case when there was at least one
/// valid match in the results.
/// </remarks>
/// <returns>True if even one match was found, false otherwise.</returns>
public static bool TryParse(string taggedString, out TagsetBase[] tagset)
{
if(string.IsNullOrWhiteSpace(taggedString))
{
tagset = null;
return false;
}
if(!taggedString.Contains(TagDelimiter.ToString()))
{
tagset = null;
return false;
}
try
{
tagset = Parse(taggedString);
}
catch
{
tagset = null;
return false;
}
if (tagset.Length == 0 || tagset.All(t => t.GetType() == typeof(NullTagset)))
{
tagset = null;
return false;
}
return true;
}
/// <summary>
/// Attempts to parse a string on the <see cref="TagDelimiter"/>
/// followed by a conversion of the string array of array's into <see cref="TagsetBase"/>
/// list.
/// </summary>
/// <param name="taggedString"></param>
/// <param name="tagsets"></param>
/// <returns></returns>
public static bool TryParse(string taggedString, out TagsetBase[][] tagsets)
{
if (string.IsNullOrWhiteSpace(taggedString))
{
tagsets = null;
return false;
}
try
{
//get all tags as a single array
TagsetBase[] allTags;
if(!TryParse(taggedString,out allTags))
{
tagsets = null;
return false;
}
//determine how many sentence breaks the array contains
var numberOfSentences = allTags.Count(tag => tag.GetType() == typeof (SentenceBreakPunctuation));
//instantiate the out variable
tagsets = new TagsetBase[numberOfSentences+1][];//the sentence break and end-of-data may not be coterminus
var sentenceIndex = 0;
var tagBuffer = new List<TagsetBase>();
for(var i = 0; i< allTags.Length; i++)
{
var tagEntry = allTags[i];
tagBuffer.Add(tagEntry);
//flush the buffer on each sentence break or at the end of the array
if (tagEntry.GetType() == typeof(SentenceBreakPunctuation) || i == allTags.Length - 1)
{
tagsets[sentenceIndex] = tagBuffer.ToArray();
tagBuffer.Clear();
//increment the sentence counter
sentenceIndex += 1;
}
}
}
catch
{
tagsets = null;
return false;
}
//protect calling assemlby from having to check for nulls when returning a jagged-array
for (var i = 0; i < tagsets.Length; i++ )
{
if (tagsets[i] == null)
tagsets[i] = new TagsetBase[] {new NullTagset()};
}
return true;
}