Release 0.4.0 · BurntSushi/ripgrep

This is a new minor version release of ripgrep that includes a couple very
minor breaking changes, a few new features and lots of bug fixes.

This version of ripgrep upgrades its regex dependency from 0.1 to 0.2,
which includes a few minor syntax changes:

POSIX character classes now require double bracketing. Previously, the regex
[:upper:] would parse as the upper POSIX character class. Now it parses
as the character class containing the characters :upper:. The fix to this
change is to use [[:upper:]] instead. Note that variants like
[[:upper:][:blank:]] continue to work.
The character [ must always be escaped inside a character class.
The characters &, - and ~ must be escaped if any one of them are
repeated consecutively. For example, [&], [\&], [\&\&], [&-&] are all
equivalent while [&&] is illegal. (The motivation for this and the prior
change is to provide a backwards compatible path for adding character class
set notation.)

Feature enhancements:

Added or improved file type filtering for Crystal, Kotlin, Perl, PowerShell,
Ruby, Swig
FEATURE #83:
Type definitions can now include other type definitions.
FEATURE #243:
BREAKING CHANGE: The --column flag now implies --line-number.
FEATURE #263:
Add a new --sort-files flag.
FEATURE #275:
Add a new --path-separator flag. Useful in cygwin.

Bug fixes:

Provide feedback