[ < < < Home ] [ < < Reference Start ] [ < Reference Contents ]
[ < Previous=PDGREPPE Magic Patterns ] [ Next=PDGREPPE Search Operators > ]
PDGREPPE Expression Parts

Magic Expressions consist of

	Operators, Terms, Factors and/or Atoms and Controls

Operators:

A reference or comparison or functional feat performed on some
item of data can be considered an operator.

Operators are the "doers".  They DO something provided they have
something to work with.

An expression may have math operators like Boolean AND and OR
operators that test for one thing versus another.  Or there may
be quantity determination operators that test for some number of
repeated data.

Each operator can work with some simple or complex data element
called a term.

Terms:

A Term is one or more items of data.  It is acted upon by ONE
operator such as OR or AND.

Terms are the "things" that "doers" can work with.

Each Term of an Expression can be composed of one or more
Factors or Atoms.

Factor:

Each factor is a data part or position that can determine if
there is a match to a section of data

A factor can be some small "atomic" part like a single byte or
block of bytes or a position.

Or a Factor can itself be an Expression.  This is what allows
complex yet powerful pattern-matching to be done.

So a Factor can be either the smallest through the LARGEST
potential thing in a pattern.

Atom:

An atom, like a factor is part of a pattern, but refers to the
SMALLEST part of something definite to be searched for.

An atom can be the "point" that is found at every position in
a file, some other position, a specific byte, character or
multiple byte section.

What follows is a semi-ordered group of explanations of various
pattern matching Actions and their associated Symbols that
represent Operators working with Terms that contain one or more
Factors.

Controls:

These control something being done while searching or replacing
but do not themselves define any search/replace item.

Controls are like functions rather than data.

Precedence:

Different parts of a magic pattern are taken into consideration
before others.

For example:

Groups like

	"{c|d}" of ab{c|d}ef

or

	"(c|d)" of ab(c|d)ef

that isolate "ab" from "c", and "ef" from "d", have a higher
precedence than parts outside groups.  The innermost group has
the highest precedence.

Quantity operations like

	abc+

have higher precedence than adding or joining atoms so that the
item

	"c+"

is drawn out for the repeat "+" function first before having
the string "ab" added to the expression.  Therefore the repeat
or quantity operation has HIGHER precedence than adding atoms.

Or a pattern like

	"abc|def"

makes two separate sections of

	abc

and

	def

before creating an expression that involves the OR "|" operator.

So the adding of adjoining atoms to make sections takes
place before emplacing a OR or AND operation.  This is then a
HIGHER level of precedence than OR and AND operations.

Also, there is a direction the pattern is processed when parts
are placed by preference, sometimes left to right and sometimes
right to left.

Following is a table that shows various pattern parts, direction
of interpretation:

PART		ELEMENTS			DIRECTION

Expression	() {} #<n> definition		left  to right

Atom		byte [<set>] position file	left  to right
		multibyte

Quantity	;?-!?*+ or [;?-!?*+]{j,k}	right to left

Additive	ie concatenation (eg abc)	left  to right

OR/AND		| ~ &				left  to right

	Table of Pattern Parts Precedence (Highest to Lowest)
[ < < < Home ] [ < < Reference Start ] [ < Reference Contents ]
[ < Previous=PDGREPPE Magic Patterns ] [ Next=PDGREPPE Search Operators > ]