v0.3.0

The parser is incomplete until v1.0.0. If you see
an odd error from valid Elixir code, it is most likely due to the incomplete parser. Please subscribe to notifications
or comment on Issue #6 if you want to be notified when the
parser is complete.

Changelog

v0.3.0

• Enhancements
  • #108 - \x is marked as an error in CharLists, CharList Heredocs, Strings, and String Heredocs, but not in any sigils. - @KronicDeth
  • #111 - New Elixir File will automatically underscore the camel case module name when creating the file name and will convert qualifying aliases before the last . to directories - @KronicDeth
• Incompatible Changes
  • #111 - New Elixir File validates that the name is a valid Alias, so each . separated part must start with a capital letter. Previous New Elixir File validated that the name was a valid path, and so forced the name to be lowercase. - @KronicDeth

Upgrading

v0.3.0

The prior version of New > Elixir File validated that the input name was a valid path and would only allow lowercase
names, so you'd be forced to set the name to foo, which would produce a foo.ex file, but foo would also be used
in the file contents:

defmodule foo do

end

In v0.3.0, the validator was corrected so that it only allows Alias (with and without .), so instead of entering the
name as foo, enter it as Foo. The file will still be named foo.ex, but the module name will correctly be Foo
in the file contents:

defmodule Foo do
end

v0.2.1

v0.2.1

• Enhancements
  • #105 - No parentheses function calls can occur as the right operand in binary infix operations or the sole operand of unary prefix operation. - @KronicDeth
  • #74 - @KronicDeth
    • Function calls with neither parentheses nor do blocks that have at least 2 arguments: a positional argument and keyword arguments or 2 or more positional argument(s) followed by optional keyword arguments.
    • Inspection that marks errors for ambiguous commas
    • Inspection that marks errors for ambiguous parentheses
    • Quick Fix for the ambiguous parentheses to remove the space between the function name and the opening parentheses.
  • #75 - @KronicDeth
    • Inspection that marks errors for missing end-of-expressions (; and newlines) between expressions.
    • Quick Fix to insert ; for missing end-of-expression.
    • Quick Fix to insert newline for missing end-of-expression.
• Bug Fixes
  • #74 - Right hand-side of dot_alias and dot_identifier
    was translated incorrectly. Only Aliases and Identifiers are allowed now. @KronicDeth

v0.2.0

This is the first release with a community contributor, @abaire, who I'd like to thank for adding Module creation support. This is the release marks the transition from primarily tokenization to parsing, including error recovery for base integers

v0.2.0

• Enhancements
  • #73 - @KronicDeth
    • New attributes for parts of numbers on Color Settings Page
      • Binary, Decimal, Hexadecimal, and Octal Digits
    • Decimal Exponent, Mark, and Separator
    • Invalid Binary, Decimal, Hexadecimal, and Octal Digits
      • 2-9, A-Z, and a-z will be parsed as invalid binary digits
      • 8-9, A-Z, and a-z will be parsed as invalid octal digits
      • G-Z and g-z will be parsed as invalid hexadecimal digits
    • Non-Decimal Base Prefix
      • Any letter other than b, o, or x, in either case, will be recognized as an invalid whole number base
    • Obsolete Non-Decimal Base Prefix (B for binary and X for hexadecimal)
  • Any digit, 0-9, A-Z, or a-z will be parsed as invalid for invalid whole number based numbers
  • Recovery for non-decimal whole numbers if the prefix is given, but no digits are given
• Incompatible Changes
  • #73: Number attribute has been removed from Color Settings page - @KronicDeth

v0.1.4

• Enhancements
  • #17: All valid escape sequences (\<character>, \x<hexadecimal>, \x{<hexadecimal>} are recognized. - @KronicDeth
  • #18: Support for creation of Elixir modules - @abaire
  • #21: Use pygments' elixir_example.ex supplied by @alco for Color Settings Page - @KronicDeth
  • #25: ? before any character or valid escape sequence will be recognized as a character token. - @KronicDeth
  • #35: ; is recognized as EOL. \r\n and \n style EOL can be escaped with \ and will be treated as whitespace. - @KronicDeth
  • #38: Operator arity, associativity, and precedence - @KronicDeth
  • #39: Decimal integers and floats - @KronicDeth
  • #40: Identifiers (variable, function, and macro names) - @KronicDeth
  • #41: ... identifier - @KronicDeth
  • #42: Aliases (module names) - @KronicDeth
  • #45: Keyword identifiers - @KronicDeth
  • #49: Empty Parentheses - @KronicDeth
  • #52: In Operator - @KronicDeth
  • #54: Dot Operator - @KronicDeth
  • #56: Keyword Lists - @KronicDeth
  • #70: Matched Expressions - @KronicDeth
  • #72: Regular Keywords (end, false, fn, nil, and true) - @KronicDeth
• Bug Fixes
  • #17: Sigil terminator escapes are recognized, so that sigils are no longer prematurely terminated. - @KronicDeth
  • #24: Comments do not consume EOL, so trailing comments don't cause error parsing expression on following line. - @KronicDeth
  • #36: Sigil modifiers now work on groups in addition to heredocs. - @KronicDeth
  • #47: ; is separate from EOL and either or both can separate expressions, but only EOL can separate operators and operands for operations - @KronicDeth

v0.1.3

• Bug Fixes
  • Blank lines are properly parsed as whitespace instead of bad characters.
  • EOL is parsed as bad character in sigil name (after ~) instead of causing the lexer to fail to match, which raised exceptions in Event Log.

v0.1.2

• Enhancements
  • Atoms with highlighting
    • Atom with double or single quotes to allow interpolation. Double quotes are highlighted as 'String' while single
      quotes are highlighted as 'Char List'. This may be changed in the future.
    • Literal atoms highlighted as 'Atom'.
    • Operator atoms highlighted as 'Atom'.

v0.1.1

• Bug Fixes
  • Build using JDK 6 instead of 7 so that plugin will work by default on OSX Mavericks.

v0.1.0

v0.1.0

• Enhancements
  • Literal and interpolated sigils with highlighting
    • CharList Sigils (~c and ~C) highlighted as 'Char List' in Settings.
    • Regex Sigils (~r and ~R) highlighted as 'Sigil' in Settings. NOTE: Regex syntax is not internally highlighted yet
    • String Sigils (~s and ~S) highlighted as 'String' in Settings.
    • Word Sigils (~w and ~W) highlighted as 'Sigil' in Settings.
    • Custom Sigils highlighted as 'Sigil' in Settings.
    • Modifiers are highlighted on Regex, Word, and Custom while modifiers aren't allowed on CharList and String Sigils.
• Bug Fixes
  • Single-quoted strings are correctly referred to as 'Character List' now instead of 'String' in Settings.
  • Double-quoted strings are correctly referred to as 'String' now instead of 'Interpolated String' in Settings.
  • Non-Heredoc CharLists and Strings can be multiline.
  • CharLists and Strings support interpolation and escape sequences.

v0.0.3

v0.0.3

• Enhancements
  • Single quoted strings with highlighting. ('String' in Color Settings.)
  • Double quoted strings with highlighting. ('Interpolated String' in Color Settings.)
    • Interpolation (#{ and }) with highlighting. ('Expression Substitution Mark' in Color Settings.)
    • Escape sequences for " and # with highlighting. ('Escape Sequence' in Color Settings.)

v0.0.2

• Enhancements
  • Binary, Hexadecimal, and Octal numbers (including deprecated syntax) are recognized as numbers.
  • Syntax Highlighting for numbers.
  • Color Settings page for changing the color of comments and numbers for Elixir (Preferences > Editor > Colors & Fonts > Elixir).
• Bug Fixes
  • Parser no longer freezes IDE on tokens it doesn't understand.
  • White space at beginning of lines no longer leads to annotation errors.
  • White space and EOLs at beginning of file no longer lead to annotation errors.