ALPHA-VISION® AvET: Regular Expressions

Regular Expressions

SUMMARY

A regular Expression (RE) serves as a template or mask a text to be searched is compared with.

CONTENTS

1. Introduction
2. Literal, Meta Character, Anchor
3. Meta characters that are supported
4. Anchor that are supported

Introduction

A Regular Expression (RE) serves as a template or mask a text to be searched is compared with. Thereby the RE is interpreted und tested whether it matches on any position of the searched text.
Some characters do take a special role, they are called Meta Characters.

Literal, Meta Character, Anchor

Literal are the letters A-Z, a-z and the numerics 0-9 as well as most of the special characters; literal can be taken one to one into a RE.

Meta characters are characters that are not searched for, but describe in connection with others the string that is searched for. By combining meta characters you get a powerful tool to spot several search terms by a single RE.

Literal and meta characters relate to one or more characters while an anchor represents a position, it does not correspond with a character of the text to be searched.

Meta characters that are supported

• *
  The asterisk after a RE-Part allows an arbitrary occurrence of the definition on the actual position.
  e.g. the RE "ab*" recognizes the strings a, ab, abbbb etc.
  
  
• +
  The "+" sign equates the asterisk but expects the RE-Part to appear at least once.
  e.g. the RE "ab+" recognizes the strings ab, abb, abbbb etc., but not the a.
  
  
• ?
  The "?" after a RE-Part allows the singular appearance in the text to be searched.
  e.g. the RE "ab?" recognizes the strings a and ab, but nothing else.
  
  
• [
  Characters embedded by square brackets are defined as “Character Class”. A character is identified if it appears in the defined set. It is essential that exactly one character of the alternatives occurs at the actual position.
  The alternatives are merely lined up in the RE: [abcd01234], standing for the letters “a” to “d “ and the digits “0” to “4”. Intervals are permitted i.e. [a-d0-4] is valid too. By the open square bracket “[“ a new language level that got its own meta characters is opened. The only meta characters used then are “^]” and “-“. Using the meta character “^” (only if on the first position) you get the complement of the quantity: [^0-9] denotes “the next character is not a digit”.
  WARNING

  The interpretation “subsequently no digit is allowed” is wrong, since it is asked for a character out of the complement quantity.
• .
  The decimal point “.” is a special class. It may stand for any ONE character. The RE “s.art” will find ‘smart’, ‘swart’, ‘start’, but not ‘sart’ or 'sitart'
  
  
• (and)
  Round brackets define the evaluation sequence of the RE-Part. It is not searched for the brackets itself;
  e.g. the RE a(b*(c?))d recognizes the strings abbcd, abcd, acd, abbd, abd, ad usw.
  
  
• |
  An alternation is described by (reg1|reg2|...|regn). The alternates are tested from the left to the right, it is considered successfully if one regular sub expression matches
  Note:

  Alternations are few efficient. If the alternatives are single characters it is preferable to use a character class: Write "[Ai]st" instead of "(A|i)st".
  e.g. the RA "(bc|d?e|a*f)" recognizes the strings bc, de, e, f, af, aaf etc.
  
  

Anchor that are supported

• ^
  The “^” make sense at the first position of a RE only, it denotes that the first character must be recognized too; i.e. the string dabbbbcd is recognized by the RE “ab*c” but not by the RE “^ab*c”.
• $
  Analogue to the “^” the “$” demands that the last character must match; i.e. the string abbbbcd is recognized by the RE "ab*c" but not by the RE "ab*c$".