private void ComposeNgramString(uint[] ngram, int count, StringBuilder sb, int keyCount, TermGetter termGetter)
{
Host.AssertValue(sb);
Host.AssertValue(ngram);
Host.Assert(keyCount > 0);
sb.Clear();
DvText term = default(DvText);
string sep = "";
for (int iterm = 0; iterm < count; iterm++)
{
sb.Append(sep);
sep = "|";
var unigram = ngram[iterm];
if (unigram <= 0 || unigram > keyCount)
{
sb.Append("*");
}
else
{
termGetter((int)unigram - 1, ref term);
term.AddToStringBuilder(sb);
}
}
}
/// <summary>
/// Gets the mapping from T into a StringBuilder representation, using various heuristics.
/// This StringBuilder representation will be a component of the composed KeyValues for the
/// hash outputs.
/// </summary>
public static ValueMapper <T, StringBuilder> GetSimpleMapper <T>(ISchema schema, int col)
{
Contracts.AssertValue(schema);
Contracts.Assert(0 <= col && col < schema.ColumnCount);
var type = schema.GetColumnType(col).ItemType;
Contracts.Assert(type.RawType == typeof(T));
var conv = Conversion.Conversions.Instance;
// First: if not key, then get the standard string converison.
if (!type.IsKey)
{
return(conv.GetStringConversion <T>(type));
}
bool identity;
// Second choice: if key, utilize the KeyValues metadata for that key, if it has one and is text.
if (schema.HasKeyNames(col, type.KeyCount))
{
// REVIEW: Non-textual KeyValues are certainly possible. Should we handle them?
// Get the key names.
VBuffer <DvText> keyValues = default(VBuffer <DvText>);
schema.GetMetadata(MetadataUtils.Kinds.KeyValues, col, ref keyValues);
DvText value = default(DvText);
// REVIEW: We could optimize for identity, but it's probably not worthwhile.
var keyMapper = conv.GetStandardConversion <T, uint>(type, NumberType.U4, out identity);
return
((ref T src, ref StringBuilder dst) =>
{
ClearDst(ref dst);
uint intermediate = 0;
keyMapper(ref src, ref intermediate);
if (intermediate == 0)
{
return;
}
keyValues.GetItemOrDefault((int)(intermediate - 1), ref value);
value.AddToStringBuilder(dst);
});
}
// Third choice: just use the key value itself, subject to offsetting by the min.
return(conv.GetKeyStringConversion <T>(type.AsKey));
}
