v0.2.69

Improvements and bug fixes

• Adds a very basic error check if whittling fails (not enough inputted words to make the list as long as the user asks for), given input parameters and inputted word list(s)
• Fixes bug where whittling function would get stuck in infinite loop, trying the same take_first over and over again (change would evaluate to 0).
• Improves accuracy of entropy calculations and, more importantly, the tests for the brute force and Shannon lines
• Sorts list alphabetically after print-rand and print-first functions, rather than before.
• Prints recommendation for when to use whittle-to option.

v0.2.66

• Adds ability to skip rows from start and end of inputted files, using --skip-rows-start and/or --skip-rows-end.
• Adds ability to "whittle" a list to an exact length.
• Improves entropy calculation accuracy by avoiding floating-point arithmetic when possible and respecting number of significant digits in Shanon's calculations. Improvements also made to calculations of whether a list is above or below the brute force line and the Shannon line.
• Renames --cut-to to --print-rand and adds a --print-first

v0.2.61

• Fixes a bug where the take-first and take-rand were truncating the inputted list not at the very beginning, but after a number of word modifications/deletions.

v0.2.60

• Attributes can check if a produced list is uniquely decodable. See #13 and this blog post.
• Adds "Schlinkert pruning" functionality (-K flag) to create uniquely decodable lists while preserving as many words as possible. See #13 and this blog post.
• Fixes some dice numbering bugs (though other discrepancies likely remain). See #14.
• Updates readme with, among other things, a helpful table for clearing the brute force line and the Shannon line (see readme).

v0.2.56

• Fixes bug where the convenience delimiter shortcuts of 's' for space and 't' for tab were not being parsed when calculating attributes of the outputted list or when generating sample passphrases from the outputted list.

v0.2.54

• Adds -G and g options to ignore characters on either side of a given one-character delimiter. Handy when dealing with metadata like word frequencies counts that you want to preserve.
• Re-assigns -d and -D to be generic "delete" options. They accept any single character delimiter.
• Re-assigns some other short flags like -C and -S.
• Changes some order of operations to be hopefully more intuitive with fewer surprises to users.
• Bug fixes.

Full Changelog: v0.2.30...v0.2.54

v0.2.30

A test/initial release to see how this works on GitHub. This is very alpha. Recommend you install Tidy via cargo as described in the README.

See issue #10 for more on the Windows.