public ParadigmsCollection ReadFromFile(string fileName, IParadigmStrings tags)
{
ParadigmsCollection result;
using (FileStream stream = File.OpenRead(fileName))
result = ReadFromStream(stream, tags);
return(result);
}
// This method is long, but I prefer not to split it because it performs a rather low level file parsing,
// and it will be easier to maintain this way
public ParadigmsCollection ReadFromStream(Stream stream, IParadigmStrings tags)
{
var paradigms = new List <Paradigm>();
using (BinaryReader reader = new BinaryReader(stream))
{
// In the binary file, we first read the number of paradigms present
int paradigmsCount = reader.ReadUInt16();
for (int i = 0; i < paradigmsCount; i++)
{
// We read the length of the current paradigm
int paradigmLength = reader.ReadUInt16();
// A paradigm has n entries and n is a multiple of 3
var declensions = new List <Declension>(paradigmLength / 3);
// This loop is not efficient, but since we do it only once
// and the filesize is relatively small,
// we will stick to this more readable variant
// First we have n/3 suffix indexes
List <int> suffixIndexes = new List <int>(paradigmLength / 3);
for (int j = 0; j < paradigmLength / 3; j++)
{
suffixIndexes.Add(reader.ReadUInt16());
}
// Then we have n/3 tag indexes
List <int> tagIndexes = new List <int>(paradigmLength / 3);
for (int j = 0; j < paradigmLength / 3; j++)
{
tagIndexes.Add(reader.ReadUInt16());
}
// The we have n/3 prefix indexes
List <int> prefixIndexes = new List <int>(paradigmLength / 3);
for (int j = 0; j < paradigmLength / 3; j++)
{
prefixIndexes.Add(reader.ReadUInt16());
}
// Now we asseble our declensions
for (int j = 0; j < paradigmLength / 3; j++)
{
declensions.Add(new Declension("prefix", new Tag(tags.GetString(tagIndexes[j])), "suffix"));
}
paradigms.Add(new Paradigm(declensions));
}
}
return(new ParadigmsCollection(paradigms));
}
