Starting with version 2.0.0, mapping functionality and type converters are in the separate package FolkerKinzel.CsvTools.Mappings. This ensures that only what is really needed needs to be installed.
You can use this small library independently
- for analyzing CSV files and CSV strings (delimiter, header row, column names, text encoding, and required options for reading non-standard CSV)
- retrieving the appropriate method arguments for exchanging CSV data with Excel
- serializing collections of any data and DataTables as CSV with simple formatting options (Use FolkerKinzel.CsvTools.Mappings for advanced CSV serializing and deserializing.)
- parsing the string content of CSV files and CSV strings, e.g., with Linq. (Use FolkerKinzel.CsvTools.Mappings for deserializing data tables and data types other than strings.)
FolkerKinzel.CsvTools allows you to do things easily and with just a few lines of code. In addition, the library also provides the means to write high-performance code:
- It allows you to parse ultra large CSV files because only one row of the file has to be in memory at a time.
- It makes the fields of the CSV file available as
ReadOnlyMemory<Char>instances. This avoids the allocation of numerous temporary strings. - The
CsvOpts.DisableCachingoption allows reusing the sameCsvRecordinstance for each parsed row of the CSV file. This can avoid further allocations.
(See benchmark results and the associated benchmark code.)