Faster-than-CSV

Library	Time (Speed)
Pandas read_csv()	20.09
NumPy fromfile()	3.88
NumPy genfromtxt()	4.00
NumPy loadtxt()	1.26
csv (std lib)	0.40
csv (list)	0.38
csv (map)	0.37
Faster_than_csv	0.08

• This CSV Lib is ~300 Lines of Code.

• Benchmarks run on Docker from Dockerfile on this repo.
• Speed is IRL time to complete 10000 CSV Parsings.
• Lines Of Code counted using CLOC.
• Direct dependencies of the package when ready to run.
• Benchmarks run on Docker from Dockerfile on this repo.
• Stats as of year 2021.
• x86_64 64Bit AMD, SSD, Arch Artix Linux.

Use

import faster_than_csv as csv

csv.csv2list("example.csv")                     # See Docs for more info.
                                                # Custom Separators supported.
csv.csv2json("example.csv", indentation=4)      # CSV to JSON, Pretty-Printed.

csv.csv2htmltable("example.csv")                # CSV to HTML+CSS Table (No JavaScript).

csv.read_clipboard()                            # CSV from the Clipboard.

csv.diff_csvs("example.csv", "anotherfile.csv") # Diff optimized for CSVs.

• Input: CSV, TSV, Clipboard, File, URL, Custom.
• Output: CSV, TSV, HTML, JSON, NDJSON, Diff, File, Custom.

csv2dict()

Description: Takes a path of a CSV file string, process CSV and returns a list of dictionaries. This is very similar to pandas.read_csv(filename).

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: Data from the CSV, dict type.

csv2list()

Description: Takes a path of a CSV file string, process CSV and returns a list.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: Data from the CSV, list type.

read_clipboard()

Description: Reads CSV string from Clipboard, process CSV and returns a list of dictionaries. This is very similar to pandas.read_clipboard(). This works on Linux, Mac, Windows.

Arguments:

• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: Data from the CSV, dict type.

csv2json()

Description: Takes a path of a CSV file string, process CSV and returns JSON.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.
• indentation Pretty-Printed or Minified JSON output, int type, optional, 0 is Minified, 4 is Pretty-Printed, you can use any integer to adjust the indentation.

Returns: Data from the CSV as JSON Minified Single-line string computer-friendly, str type.

csv2ndjson()

Description: Takes a path of a CSV file string, process CSV and returns NDJSON.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• ndjson_file_path path of the NDJSON file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: None. Data from the CSV as NDJSON https://github.com/ndjson/ndjson-spec, str type.

csv2htmltable()

Description: Takes a path of a CSV file string, process CSV and returns the data rendered on HTML Table.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string, defaults to "", if its empty string then No file is written.
• html_file_path path of the CSV file, str type, optional, can be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.
• header_html HTML Header, str type, optional, defaults to Bulma CSS, can be empty string.

Returns: Data from the CSV as HTML Table, str type, raw HTML (no style at all).

csv2karax()

Description: Takes a path of a CSV file string, process CSV and returns the data rendered as a Karax HTML Table.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: Karax DSL, str type.

csv2terminal()

Description: Takes a path of a CSV file string, process CSV and prints to terminal a colored prety-printed table.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string, defaults to "", if its empty string then No file is written.
• column_width column width of the wider column, required, int type, must not be 0, must not be negative.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: None.

csv2xml()

Description: Takes a path of a CSV file string, process CSV and returns a Valid XML string. Output is guaranteed to be always Valid XML.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator Separator character of the CSV data, str type, optional, defaults to ',', must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.
• header_xml XML Header of the XML string, str type, optional, can be empty string, defaults to "<?xml version=\"1.0\" encoding=\"UTF-8\" ?>\n".

Returns: XML, str type.

tsv2csv()

Description: Takes a path of a CSV file string, process CSV and returns a TSV.

Arguments:

• csv_file_path path of the CSV file, str type, required, must not be empty string.
• separator1 Separator character of the CSV data, str type, optional, must not be empty string.
• separator2 Separator character of the CSV data, str type, optional, must not be empty string.
• quote Quote character of the CSV data, str type, optional, defaults to '"', must not be empty string.

Returns: Data from the CSV as TSV, str type.

diff_csvs()

Description: Takes 2 paths of 2 CSV files, process CSV and returns the Diff of the 2 CSV.

Arguments:

• csv_file_path0 path of the CSV file, str type, required, must not be empty string, file must exist.
• csv_file_path1 path of the CSV file, str type, required, must not be empty string, file must exist.

Returns: Diff.

For more Examples check the Examples and Tests.

Instead of having a pair of functions with a lot of arguments that you should provide to make it work, we have tiny functions with very few arguments that do one thing and do it as fast as possible.

Install

• pip install faster_than_csv

Docker

• Make a quick test drive on Docker!.

$ ./build-docker.sh
$ ./run-docker.sh
$ ./run-benchmark.sh  # Inside Docker.

Dependencies

• None

Platforms

• ✅ Linux
• ✅ Windows
• ✅ Mac
• ✅ Android
• ✅ Raspberry Pi
• ✅ BSD

Requisites

• Python 3.6+ 64Bit.

Windows

• If installation fails on Windows, just use the Source Code:

• Git Clone and Compile on Windows 10 on just 2 commands!.
• Alternatively you can try Docker for Windows.
• Alternatively you can try WSL for Windows.
• The file extension must be .pyd, NOT .dll.

Stars

Contributors

• SekouDiaoNlp

FAQ

• Whats the idea, inspiration, reason, etc ?.

Feel free to Fork, Clone, Download, Improve, Reimplement, Play with this Open Source. Make it 10 times faster, 10 times smaller.

• This requires Cython ?.

No.

• This runs on PyPy ?.

No.

• This runs on Python2 ?.

I dunno. (Not supported)

• How can I Install it ?.

https://github.com/juancarlospaco/faster-than-csv/releases

If you dont understand how to install it, you can just download, extract, put the files on the same folder as your *.py file and you are good to go.

• How can be faster than NumPy ?.

I dunno.

• How can be faster than Pandas ?.

I dunno.

• Why needs 64Bit ?.

Maybe it works on 32Bit, but is not supported, integer sizes are too small, and performance can be worse.

• Why needs Python 3 ?.

Maybe it works on Python 2, but is not supported, and performance can be worse, we suggest to migrate to Python3.

• Can I wrap the functions on a try: except: block ?.

Functions do not have internal try: except: blocks, so you can wrap them inside try: except: blocks if you need very resilient code.

• PIP fails to install or fails build the wheel ?.

Add at the end of the PIP install command:

--isolated --disable-pip-version-check --no-cache-dir --no-binary :all:

Not my Bug.

• How to Build the project ?.

build.sh

• How to Package the project ?.

package.sh

• This requires Nimble ?.

No.

• Whats the unit of measurement for speed ?.

Unmmodified raw output of Python timeit module.

Please send Pull Request to Python to improve the output of timeit.