Abnormal expressions

Abnormal expressions (abnex) is an alternative to regular expressions (regex). This is a Python library but the abnex syntax could be ported to other languages.

Examples

Matching an email address

• Regex

  • ([\w\._-]+)@([\w\.]+)

• Abnex

  • {[w"._-"]1++}"@"{[w"."]1++}

• Abnex (spaced)

  • {[w "._-"]1++} "@" {[w "."]1++}

• Abnex (expanded)

  {
    [w "._-"]1++
  }
  "@"
  {
    [w "."]1++
  }
  

  A more advanced pattern:

  {{{[a-z '_']1++} {[a-z 0-9 '_-.']0++}} '@' {{[a-z 0-9]1++} '.' {[a-z 0-9]1++} {[a-z 0-9 '-_.']0++}} {[a-z 0-9]1++}}.

Why is Abnex Better?

• It's easier to read, write and understand.
• You can use spaces inside of the expression, you can also "expand" it, i.e. write it over multiple lines and use indention.
• You don't have to use a backslashes all the time
• More logical/common symbols like ! for not, {} for groups, 1++, 0++, 0+ for: one or more, zero or more, zero or one.
• It's easier to see if a symbol is an actual symbol you are searching for or if it's a regex character, ex:
  • Regex: [\w-]+@[\w-_]+
  • Abnex: [w "-"]1++ "@" [w "-"]1++

Documentation

Regex on right after -> is the abnex equivalent

Anchors

• Start of string, or start of line in multi-line pattern
  • ^ -> ->
• End of string, or end of line in multi-line pattern
  • $ -> <-
• Start of string
  • \A -> s>
• End of string
  • \Z -> <s
• Word boundary
  • \b -> :
• Not word boundary
  • \B -> !:
• Start of word
  • \< -> w>
• End of word
  • \> -> <w

Character Classes

• Control character
  • \c -> c
• White space
  • \s -> _
• Not white space
  • \S -> !_
• Digit
  • \d -> d
• Not digit
  • \D -> !d
• Word
  • \w -> w
• Not word
  • \W -> !w
• Hexade­cimal digit
  • \x -> x
• Octal digit
  • \o -> o

Quantifiers

• 0 or more
  • * -> 0++
• 1 or more
  • + -> 1++
• 0 or 1
  • ? -> 0+

Groups and Ranges

• Any character except new line (\n)
  • . -> *
• a or b
  • a|b -> "a"|"b"
• Group
  • (...) -> {...}
• Passive (non-c­apt­uring) group
  • (?:...) -> {#...}
• Range (a or b or c)
  • [abc] -> ['abc'] or ["a" "b" "c"]
• Not in set
  • [^...] -> [!...]
• Lower case letter from a to Z
  • [a-q] -> [a-z]
• Upper case letter from A to Q
  • [A-Q] -> [A-Q]
• Digit from 0 to 7
  • [0-7] -> [0-7]

Standards

What is the recommended way to write abnexes

• Use spaces between characters in character sets:
  • Correct: [w "_-"]
  • Incorrect: [w"_-"]
• Put multiple exact characters between the same quotes in character sets:
  • Correct: ["abc"]
  • Incorrect: ["a" "b" "c"], especially incorrect: ["a""b""c"]
• Put spaces between groups:
  • Correct: {w} "." {w}
  • Incorrect: {w}"."{w}

Examples:

Match for an email address:

• Regex:
  • [\w-\._]+@[\w-\.]+
• Abnex (following standards):
  • {[w "-._"]1++} "@" {[w "-."]1++}
• Abnex (not following standards):
  • {[w"-._"]1++}"@"{[w"-."]1++}

Functions (In Python)

Abnex has most functions from the re library, but it also has som extra functionality like: last() & contains().

Common functions between re and abnex

Regex on right after -> is the abnex equivalent

• match() -> match()
• findall() -> all()
• split() -> split()
• sub() -> replace()
• subn() -> replace_count()
• search() -> first()

Special to abnex

• holds(): whether or not a string matches an expression (bool).
• contains(): wheter or not a string contains a match (bool).
• last(): the last match in a string.