Could that CSV file be encoded in something other than ASCII? No problem.
This package provides support for reading CSV files that use arbitrary text encodings. It is built on top of Python's standard csv and codecs packages, and it uses Daniel Blanchard's chardet universal encoding detector to guess the encoding for a file, if necessary.
Note that utf-8-sig (UTF-8 with leading Byte Order Mark) is supported. This format is used by recent versions of Microsoft Excel when the user selects "Save As ..." and chooses the "CSV UTF-8."
pip install encoded_csv
There's just one function: get_csv(), as follows:
encoded_csv.get_csv(csv_file, skip_lines=0, encoding='', dialect='', fieldnames=[], sample_lines=100)Code in the tests/ directory provides usage examples. The function returns a tuple, in which the first item is a list of the field names. The second item is a list of ordered dictionaries, each containing the data read from a given line of the CSV file.
The first row (after discarding any header lines) is assumed to contain column names.
Keyword arguments:
csv_file-- path to CSV file to openskip_header_lines-- (optional) number of lines to discard in the assumption that they constitute a file header of some sort (default is to skip no lines)encoding-- (optional) specifies the encoding which is to be used for the file; the standard pythoncodecsmodule is used, so any of the standard encodings may be specified; default behavior is to attempt best guess usingchardet)dialect-- (optional) a set of parameters specific to a particular CSV dialect; the standard pythoncsvmodule is used, so the standard, predefineddialectvalues or formatting parameters must be used; default behavior is to attempt best guess usingcsv.Sniffer.fieldnames-- (optional) is used to force the csv.DictReader to use a particular set of fieldnames.sample_lines-- (optional) integer used to prepare the sample given tocsv.Sniffer()when attempting to detect the CSV dialect in use; default is 100 lines or the entire file, whichever is fewer.
Bug reports and feature requests are welcome, but really I'd prefer pull requests.