public ParallelQuery <CsvMappingResult <TEntity> > Parse(IEnumerable <string> csvData)
{
if (csvData == null)
{
throw new ArgumentNullException("csvData");
}
var query = csvData
.Skip(options.SkipHeader ? 1 : 0)
.AsParallel();
// If you want to get the same order as in the CSV file, this option needs to be set:
if (options.KeepOrder)
{
query = query.AsOrdered();
}
query = query
.WithDegreeOfParallelism(options.DegreeOfParallelism)
.Where(line => !string.IsNullOrWhiteSpace(line));
// Ignore Lines, that start with a comment character:
if (!string.IsNullOrWhiteSpace(options.CommentCharacter))
{
query = query.Where(line => !line.StartsWith(options.CommentCharacter));
}
return(query
.Select(line => options.Tokenizer.Tokenize(line))
.Select(fields => mapping.Map(fields)));
}
public ParallelQuery <ParsedRow <TEntity> > Parse(IEnumerable <string> csvData)
{
if (csvData == null)
{
throw new ArgumentNullException("csvData");
}
var query = csvData
.Select((x, pos) => new Row(pos, x))
.Skip(options.SkipHeader ? 1 : 0)
.AsParallel();
// If you want to get the same order as in the CSV file, this option needs to be set:
if (options.KeepOrder)
{
query = query.AsOrdered();
}
query = query
.WithDegreeOfParallelism(options.DegreeOfParallelism)
.Where(row => !string.IsNullOrWhiteSpace(row.Data));
// Ignore Lines, that start with a comment character:
if (!string.IsNullOrWhiteSpace(options.CommentCharacter))
{
query = query.Where(row => !row.Data.StartsWith(options.CommentCharacter));
}
return(query
.Select(row => new TokenizedRow(row.Position, options.Tokenizer.Tokenize(row.Data)))
.Select(tokenizedRow => new ParsedRow <TEntity>(tokenizedRow.Position, mapping.Map(tokenizedRow.Data))));
}
