/// <summary>
/// Creates a list of ExtractedResult which contain all the choices with
/// their corresponding score where higher is more similar
/// </summary>
/// <param name="query"></param>
/// <param name="choices"></param>
/// <param name="processor"></param>
/// <param name="scorer"></param>
/// <param name="cutoff"></param>
/// <returns></returns>
public static IEnumerable <ExtractedResult <T> > ExtractAll <T>(
T query,
IEnumerable <T> choices,
Func <T, string> processor,
IRatioScorer scorer = null,
int cutoff = 0)
{
if (scorer == null)
{
scorer = s_defaultScorer;
}
return(ResultExtractor.ExtractWithoutOrder(query, choices, processor, scorer, cutoff));
}
/// <summary>
/// Find the single best match above a score in a list of choices.
/// </summary>
/// <param name="query"></param>
/// <param name="choices"></param>
/// <param name="processor"></param>
/// <param name="scorer"></param>
/// <param name="cutoff"></param>
/// <returns></returns>
public static ExtractedResult <T> ExtractOne <T>(
T query,
IEnumerable <T> choices,
Func <T, string> processor,
IRatioScorer scorer,
int cutoff)
{
if (scorer == null)
{
scorer = s_defaultScorer;
}
return(ResultExtractor.ExtractOne(query, choices, processor, scorer, cutoff));
}
public void TestScoringEmptyString()
{
var scorerType = typeof(IRatioScorer);
var scorerTypes = AppDomain.CurrentDomain.GetAssemblies().SelectMany(s => s.GetTypes()).Where(p => scorerType.IsAssignableFrom(p) && p.IsClass && !p.IsAbstract);
MethodInfo getScorerCacheMethodInfo = typeof(ScorerCache).GetMethod("Get");
string nullString = null; //Null doesnt seem to be handled by any scorer
string emptyString = "";
string whitespaceString = " ";
string[] nullOrWhitespaceStrings = { emptyString, whitespaceString };
foreach (Type t in scorerTypes)
{
System.Diagnostics.Debug.WriteLine($"Testing {t.Name}");
MethodInfo m = getScorerCacheMethodInfo.MakeGenericMethod(t);
IRatioScorer scorer = m.Invoke(this, new object[] { }) as IRatioScorer;
foreach (string s in nullOrWhitespaceStrings)
{
System.Diagnostics.Debug.WriteLine($"Testing string '{s}'");
try
{
scorer.Score(s, "TEST");
}
catch (InvalidOperationException e)
{
Assert.Fail($"{t.Name}.score failed with empty string as first parameter");
}
try
{
scorer.Score("TEST", s);
} catch (InvalidOperationException e)
{
Assert.Fail($"{t.Name}.score failed with empty string as second parameter");
}
try
{
scorer.Score(s, s);
}
catch (InvalidOperationException e)
{
Assert.Fail($"{t.Name}.score failed with empty string as both parameters");
}
}
}
}
/// <summary>
/// Find the single best match above a score in a list of choices.
/// </summary>
/// <param name="query"></param>
/// <param name="choices"></param>
/// <param name="processor"></param>
/// <param name="scorer"></param>
/// <param name="cutoff"></param>
/// <returns></returns>
public static ExtractedResult <string> ExtractOne(
string query,
IEnumerable <string> choices,
Func <string, string> processor = null,
IRatioScorer scorer = null,
int cutoff = 0)
{
if (processor == null)
{
processor = s_defaultStringProcessor;
}
if (scorer == null)
{
scorer = s_defaultScorer;
}
return(ResultExtractor.ExtractOne(query, choices, processor, scorer, cutoff));
}
/// <summary>
/// Creates a sorted list of ExtractedResult which contain the
/// top limit most similar choices
/// </summary>
/// <param name="query"></param>
/// <param name="choices"></param>
/// <param name="processor"></param>
/// <param name="scorer"></param>
/// <param name="limit"></param>
/// <param name="cutoff"></param>
/// <returns></returns>
public static IEnumerable <ExtractedResult <string> > ExtractTop(
string query,
IEnumerable <string> choices,
Func <string, string> processor,
IRatioScorer scorer,
int limit,
int cutoff)
{
if (processor == null)
{
processor = s_defaultStringProcessor;
}
if (scorer == null)
{
scorer = s_defaultScorer;
}
return(ResultExtractor.ExtractTop(query, choices, processor, scorer, limit, cutoff));
}
/// <summary>
/// Creates a list of ExtractedResult which contain all the choices with
/// their corresponding score where higher is more similar
/// </summary>
/// <param name="query"></param>
/// <param name="choices"></param>
/// <param name="processor"></param>
/// <param name="scorer"></param>
/// <param name="cutoff"></param>
/// <returns></returns>
public static IEnumerable <ExtractedResult <string> > ExtractAll(
string query,
IEnumerable <string> choices,
Func <string, string> processor,
IRatioScorer scorer,
int cutoff)
{
if (processor == default(Func <string, string>))
{
processor = s_defaultStringProcessor;
}
if (scorer == default(IRatioScorer))
{
scorer = s_defaultScorer;
}
if (cutoff == default(int))
{
cutoff = 0;
}
return(ResultExtractor.ExtractWithoutOrder(query, choices, processor, scorer, cutoff));
}
public static IEnumerable <ExtractedResult <T> > ExtractTop <T>(T query, IEnumerable <T> choices, Func <T, string> processor, IRatioScorer calculator, int limit, int cutoff = 0)
{
return(ExtractWithoutOrder(query, choices, processor, calculator, cutoff).MaxN(limit).Reverse());
}
public static IEnumerable <ExtractedResult <T> > ExtractSorted <T>(T query, IEnumerable <T> choices, Func <T, string> processor, IRatioScorer calculator, int cutoff = 0)
{
return(ExtractWithoutOrder(query, choices, processor, calculator, cutoff).OrderByDescending(r => r.Score));
}
public static ExtractedResult <T> ExtractOne <T>(T query, IEnumerable <T> choices, Func <T, string> processor, IRatioScorer calculator, int cutoff = 0)
{
return(ExtractWithoutOrder(query, choices, processor, calculator, cutoff).Max());
}
public static IEnumerable <ExtractedResult <T> > ExtractWithoutOrder <T>(T query, IEnumerable <T> choices, Func <T, string> processor, IRatioScorer scorer, int cutoff = 0)
{
int index = 0;
var processedQuery = processor(query);
foreach (var choice in choices)
{
double score = scorer.Score(processedQuery, processor(choice));
if (score >= cutoff)
{
yield return(new ExtractedResult <T>(choice, score, index));
}
index++;
}
}
public void SetUp()
{
_scorer = ScorerCache.Get <TokenSetScorer>();
_partialScorer = ScorerCache.Get <PartialTokenSetScorer>();
}
