EZCSV

Some utility functions to do the obvious things you want to do with CSV files.

csv is easy, this is EZ.

Install it: pip install ezcsv

Should be compatible with Python 2.5+ and 3.x

I say "should be" because there are no tests! Please see 'Contributing', below.

Reference

read_dicts(file_path)

Returns the data at file_path as a list of dicts.

• file_path string | Path

read_lists(file_path)

Returns the data at file_path as a list of lists. Ignores blank lines.

• file_path string | Path

Example Usage:

>>> from easycsv import csv_as_list
>>> filepath = 'C:\\path\\to\\file.csv'
>>> print(filepath)
C:\path\to\file.csv
>>> l = csv_as_list(filepath)
>>> print(l[0])
First,Line,Of,File,Prints,Here

write_dicts(data, csv_path, mkdir=False, silent_fail=False)

Writes the input data to a CSV file

Always produces a Unix-style CSV file, regardless of whether on Windows or not.

• data list[dict]

  One dict per line. The first dict is expected to contain all of the keys that will form the header line, unless fieldnames is provided.

• csv_path string | Path

  Should have a .csv extension. mkdir will malfunction if it doesn't.

• mkdir bool

  If true, creates the directory tree leading up to csv_path, ignoring if directories already exist. Default: False

• silent_fail: bool

  If true, causes the method to do nothing if data happens to be empty. Default: False

• fieldnames: list[str]

  If set, provides the list of field names to be passed to csv.DictWriter. These are written as the header, regardless of what keys are present in the first dict in data. Default: None

write_lists(data, csv_path, mkdir=False)

Writes the input data to a CSV file

Always produces a Unix-style CSV file, regardless of whether on Windows or not.

• data list[list]

  • csv_path string | Path

Should have a .csv extension. mkdir will malfunction if it doesn't.

• mkdir bool

  If true, creates the directory tree leading up to csv_path, ignoring if directories already exist.

Contributing

This could really use some tests!

Feel free to fork and open a pull request. Please, one pull request per functional change.