# CSVLoader **`class CSVLoader(config)`** Bases: [`Loader`](squirro.lib.nlp.steps.loaders.Loader.md#squirro.lib.nlp.steps.loaders.Loader) The CSV [`Loader`](squirro.lib.nlp.steps.loaders.Loader.md#squirro.lib.nlp.steps.loaders.Loader) loads CSV files and transforms the data into [`Document`](../technical/libnlp/base.md#squirro.lib.nlp.document.Document) items. Parameters - `type` ([`str`](https://docs.python.org/3.11/library/stdtypes.html#str)) – csv - `delimiter` ([`str`](https://docs.python.org/3.11/library/stdtypes.html#str)`, ``'``,``'`) – delimiter between fields - `encoding` ([`str`](https://docs.python.org/3.11/library/stdtypes.html#str)`, ``'utf-8'`) – given file’s encoding - `quotechar` ([`str`](https://docs.python.org/3.11/library/stdtypes.html#str)`, ``'"'`) – character used around values to quote them - `replace_new_lines` ([`bool`](https://docs.python.org/3.11/library/functions.html#bool)`, ``True`) – indicates if replacement of new lines characters is applied **Example** ```json { "step": "loader", "type": "csv", "fields": [ "text", "label" ] } ``` Methods SummaryMethods Documentation **`process(directory)`** Process a directory of csv files and yield documents. Parameters `directory` ([`str`](https://docs.python.org/3.11/library/stdtypes.html#str)) – path to directory with *.csv files Returns Generator of Documents read from CSV Return type generator([Document](../technical/libnlp/base.md#squirro.lib.nlp.document.Document))