Changelog

Version 0.8.3

  • Allow users to specify output encoding for some CLI commands (thanks to @jbdesbas)

  • Optimize the normal-form detection (thanks to @no23reason)

  • Internal: fix names of C modules

Version 0.8.2

  • Add more type hints to CleverCSV

  • Move the import of the optional tabview dependency to where it’s needed (for #101)

  • Allow inspecting more rows for header detection (fixes #98)

Version 0.8.1

  • Add type hints to CleverCSV

  • Disable 32-bit builds on Windows and Linux

  • Bump minimal Python version to 3.8

  • Minor documentation improvements

Version 0.8.0

  • Improve median runtime by ~68% (~52% on average) by: 1) more caching, 2) implementing a heavy function in C.

  • Redesign computation of consistency measure to a class: ConsistencyDetector.

  • Fix potential memory leak in C code for base abstraction

  • Fixes to escape sequences in regexes (thanks to @JakobGM!)

  • Various improvements to code quality

  • Switch documentation style to furo.

Version 0.7.7

  • Use r-prefix for regex patterns (thanks to @JakobGM!)

  • Fix documentation typo (thanks to @Aritra8438!)

Version 0.7.6

  • Simplify faust-cchardet import for Windows builds

Version 0.7.5

  • Add support for Python 3.11 by fixing a bug regarding empty strings in dialects (thanks to @stefanor!)

  • Fix installation error due to change in internals at setuptools (thanks to @mweinelt!)

  • Migrate to faust-cchardet as cChardet fails to install on Python 3.11 (on Windows, currently only chardet will work for Python 3.11)

  • Migrate to packaging for version comparison

Version 0.7.4

  • Add wrapper for writing a list of dictionaries (write_dicts)

  • Fix bug when writing CSVs using the csv module dialects

  • Add the builtin dialects to CleverCSV (e.g., clevercsv.excel)

Version 0.7.3

  • Release to build wheels for Python 3.10

Version 0.7.2

  • Re-implement command line interface using Wilderness

  • Add man-pages to package

Version 0.7.1

  • Remove deprecated wrapper functions

  • Expand URL regex to support localhost:<port> urls

  • Minor changes to the TypeDetector API

  • Add cChardet as optional dependency (fixes #48)

Version 0.7.0

  • Add a JSON object data type to address a specific failure case (#37).

  • Add support for timezones for time data type

  • Add support for building wheels on non-native architectures (#39).

  • Add a flag to disable skipping type detection using the command line interface.

Version 0.6.8

  • Add a “bytearray” type to address a specific failure case (#35).

  • Minor clarifications to licensing.

Version 0.6.7

  • Updates to release process. This version introduces pre-compiled wheels for Python 3.9.

Version 0.6.6

  • Add an encoding argument to write_table to allow specifying the output encoding. Thanks to @mitchgrogg for reporting issue #27.

Version 0.6.5

  • Add support for standardizing in-place and standardizing multiple files.

  • Add warning on duplicate field names in DictReader

  • Add return value to writers to match the standard library.

Version 0.6.4

  • Various speed ups to constructing the list of potential dialects. This removes a costly step of the detection process that will likely add a few more potential dialects, but has the end result of making overall dialect detection faster.

Version 0.6.3

  • Rename wrapper functions to a more coherent naming scheme. Old names will be available until 0.7.0, but now produce a FutureWarning.

  • Add stream_dicts wrapper function.

  • Improve handling of file encoding for the read_dataframe wrapper: detected encoding is now passed on to Pandas.

  • Fix handling of optional dependency error for TabView on non-Windows platforms.

Version 0.6.2

  • Update URL regex to avoid catastrophic backtracking and increase performance. See issue #13 and issue #15. Thanks to @kaskawu for the fix and @jlumbroso for re-raising the issue.

  • Add num_chars keyword argument to read_as_dicts and csv2df wrappers.

  • Improve documentation w.r.t. handling large files. Thanks to @jlumbroso for raising this issue.

Version 0.6.1

  • Add an explore command to the command line application for CleverCSV. This command makes it easy to start exploring a CSV file using the Python interactive shell.

Version 0.6.0

  • Split the package into a “core” and “full” version. This allows users who only need the improved dialect detection functionality to download a version with a smaller footprint. Fixes issue #10]. Thanks to @seperman.

Version 0.5.6

  • Fix speed of unix_path regex used in type detection. (issue #13). Thanks to @kaskawu.

Version 0.5.5

  • Add stream_csv wrapper that returns a generator over rows

  • Minor update to the URL type detection

  • Documentation updates

Version 0.5.4

  • Fix bugs discovered from fuzz testing (issue #7)

  • Minor changes to readme and code quality

Version 0.5.3

  • Fix using nan as default value when skipping a dialect (issue #5)

Version 0.5.2

  • Bump version to fix wheel building

Version 0.5.1

  • Bump version to fix wheel building

Version 0.5.0

  • Improve type detection for quoted alphanumeric cells (#4)

  • Pass strict dialect property to parser.

Version 0.4.7

  • Bugfix for write_table wrapper on Windows.

  • Move building Windows platform wheels to Travis.

  • Use cibuildwheel version 1.0.0 for building wheels.

Version 0.4.6

  • Add a wrapper function that writes a table to a CSV file.

Version 0.4.5

  • Update CleverCSV to match updated clikit dependency

  • Fix dependency versions for clikit and cleo

Version 0.4.4

  • Update standardize command to use CRLF line endings on all platforms.

  • Add work around for Tabview being unavailable on Windows.

  • Remove packaging and dependency management with Poetry.

  • Add support for building platform wheels on Travis and AppVeyor.

Version 0.4.3

  • Add optional method parameter to dialect detector.

  • Bugfix for clevercsv code command when the delimiter is tab.

Version 0.4.2

  • Fix a failing build due to dependency version mismatch

Version 0.4.1

  • Allow underscore in alphanumeric strings

  • Update unix path regular expression

  • Add more integration tests and log detection method

Version 0.4.0

  • Update URL regular expression and add unit tests

  • Add IPv4 type detection

  • Add tie-breaker for combined quotechar and escapechar ties

Version 0.3.7

  • Bugfix for console script code command

  • Update readme

Version 0.3.6

  • Cleanly handle failure to detect dialect in console application

  • Remove any (partial) support for Python 2

Version 0.3.5

  • Remove Python parser - this speeds up file reading and tie breaking

Version 0.3.4

  • Ensure the C parser is used in the reader.

  • Update integration tests to improve error handling

  • Readme updates

Version 0.3.3

  • Ensure detected encoding is in the generated Python code for the clevercsv code command.

  • Ensure encoding is detected in wrappers.detect_dialect.

  • Bugfix in integration test

  • Expand readme

Version 0.3.2

  • Add documentation on Read the Docs

  • Use requirements.txt file for dependencies when packaging

Version 0.3.1

  • Add help description to each CLI command

  • Update README

  • Add transpose flag for standardize and view commands

Version 0.3.0

  • Rewrite console application using Cleo

  • Add unit tests for console application

  • Add detect_dialect wrapper function

  • Add support for “unix_path” data type in type detection

  • Add encoding and num_chars options to read_csv wrapper

  • Add -p/--pandas flag to code command to generate Pandas output.

Version 0.2.5

  • Rename read_as_lol to read_csv.

Version 0.2.4

  • Allow setting the number of characters to read

  • Simplify printing of skipped potential dialects

Version 0.2.3

  • Add read_as_lol wrapper function.

Version 0.2.2

  • Add code command to clevercsv command line program.

Version 0.2.1

  • Bugfix to update executable to new name

Version 0.2.0

  • Rename package to clevercsv