(*******************************************************************) (** Release history: **) (** 17th January 1985 version 1.0.0 First port to new world **) (*******************************************************************) DEFINITION MODULE ScanS; (* Title : ScanS.ParseF? - simple string parsing utility LastEdit : Wed Jan 9 16:58:15 1985 Author : Graham Toal Acorn Computers VLSI Design Aids Group *) FROM Strings IMPORT String; EXPORT QUALIFIED ParseF0, ParseF1, ParseF2, ParseF3, ParseF4; PROCEDURE ParseF0 (Pattern :ARRAY OF CHAR; Data :String) :BOOLEAN; PROCEDURE ParseF1 (Pattern :ARRAY OF CHAR; Data :String; Replace1 :ARRAY OF CHAR; VAR Result1 :String ) :BOOLEAN; PROCEDURE ParseF2 (Pattern :ARRAY OF CHAR; Data :String; Replace1 :ARRAY OF CHAR; VAR Result1 :String; Replace2 :ARRAY OF CHAR; VAR Result2 :String ) :BOOLEAN; PROCEDURE ParseF3 (Pattern :ARRAY OF CHAR; Data :String; Replace1 :ARRAY OF CHAR; VAR Result1 :String; Replace2 :ARRAY OF CHAR; VAR Result2 :String; Replace3 :ARRAY OF CHAR; VAR Result3 :String ) :BOOLEAN; PROCEDURE ParseF4 (Pattern :ARRAY OF CHAR; Data :String; Replace1 :ARRAY OF CHAR; VAR Result1 :String; Replace2 :ARRAY OF CHAR; VAR Result2 :String; Replace3 :ARRAY OF CHAR; VAR Result3 :String; Replace4 :ARRAY OF CHAR; VAR Result4 :String ) :BOOLEAN; (*------------------------------------------------------------------------*) (* ParseF?: *) (* ParseF is similar to the find/replace command used in both the Acorn *) (* text-editors. The interface is procedural, and the data it works *) (* on is a single string. *) (* ParseF1 takes as arguments a pattern string (usually a literal *) (* in your program) and a data string. The procedure tries to match *) (* the pattern to the data, and any parts of the the data which both *) (* matched the pattern and were wild (e.g. *[A-Z]) will be remembered *) (* for insertion into the replace string. *) (* ParseF2, ParseF3 & ParseF4 offer identical facilities, but with *) (* multiple replace strings - to take advantage of the ability to *) (* perform the parse function once only. *) (* The pattern string can contain both literal text, and a subset *) (* of the type of regular expression which editors like 'sed' offer. *) (* The special characters available are: *) (* *) (* Valid in both PATTERN & REPLACE fields: *) (* |a - control-A *) (* |!|a - control-A + 128 *) (* $ - newline *) (* \ - ensure that the next character is taken verbatim. *) (* *) (* Valid in PATTERN field only: *) (* . - any character *) (* @ - A-Z, a-z, 0-9, and '_'. *) (* # - 0-9 *) (* [xyz] - any one of x, y, or z *) (* a-z - any one in the range a to z. *) (* *) (* The following prefixes can be applied to the above to *) (* modify their effect: *) (* * - the shortest string of 0 or more *) (* ^ - the longest string of 1 or more *) (* ~ - anything but the following *) (* And the following special case: *) (* ` - invert the case sensitivity of searches for pattern *) (* text following (default is case is insensitive) *) (* *) (* Valid in REPLACE field only *) (* % - replace with appropriate field (see below) *) (* & - replace with entire string matched (special case of %&) *) (* *) (* Field selectors take the following form: *) (* optional '+' or '-', followed by *) (* one of '.', '@', '#', '[', '-', '*', '^' or '~', followed by *) (* a digit in the range 0 to 9. *) (* *) (* The significance of these fields is as follows: *) (* +/- - force the case of the replacement text to be upper or *) (* lower. *) (* NOTE: +/- Not yet implemented - will be within 2 days. (at 5-Dec-84) *) (* '^' etc. - select the first '^'-ed field, e.g. %* means *) (* insert the data which matched the first field *) (* which had a star in it. *) (* 0,1, etc. - Modify the above to read 'Nth' instead of 'first': *) (* N=0 is the first, N=1 the second etc. *) (* (This is a bad nomenclature; we only use it here *) (* to be compatible with the ROM-ed Acorn editors.) *) (* *) (* An example Pattern/replace string, and their effect on a data *) (* string are given here: *) (* *) (* Command := Strings.CopyCS ("draw fred.23 "); *) (* IF ParseF1 ("* draw* ^@\.^#* $", Command, *) (* "Drawing cell %^0 vsn %^1$", Message) THEN *) (* *) (* The fields are interpreted as: *) (* Verify a sequence of spaces terminated by the first non-space. *) (* match the text 'draw'. *) (* Verify another sequence of spaces. *) (* Match the longest string of AlphaNumerics you can find. *) (* Check that the next character is a dot (note the '\' escape *) (* to avoid the special significance of '.'. *) (* Match the longest string of numbers you can find. *) (* Verify a (possibly null, as above) sequence of spaces. *) (* Check that we are now at the end of the line. *) (* *) (* The replacement string is generated from: *) (* The literal text 'Drawing cell ', *) (* The text which matched the first field which had a '^' in it, *) (* i.e. 'fred' *) (* The literal text ' vsn ', *) (* The text which matched the second field which had a '^' in it, *) (* i.e. '23' *) (* And an end-of-line character. *) (* *) (*------------------------------------------------------------------------*) END ScanS.