[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


> I've found string hacking to be a fundamental component of many different
> kinds of hacking I've done -- web servers and scripting in particular. So
> I'm very interested in developing a rich, carefully-thought-out library
> that will assist this kind of programming.

In my web server scripting I often have to escape strings into HTML or
LaTeX.  I understand some people have to escape strings to safely pass
to a Unix shell.  All of these are cases of escaping single characters
into multiple characters for the output string.  I think the string
library would be more useful with some efficient mechanism for
implementing such escapes.

Here's a sample implementation to help fuel discussion:

;; examples of use

(define html-escape (string-escaper '((#\< . "&lt;")
				      (#\> . "&gt;")
				      (#\& . "&amp;"))))

(define scheme-escape (string-escaper '((#\\ . "\\\\")
					(#\" . "\\\""))))

(define latex-escape (string-escaper '((#\\ . "\\\\")
				       (#\~ . "\\~")
				       (#\# . "\\#")
				       (#\$ . "\\$")
				       (#\% . "\\%")
				       (#\^ . "\\^")
				       (#\& . "\\&")
				       (#\{ . "\\{")
				       (#\} . "\\}")
				       (#\_ . "\\_"))))

;; example implementation

(define (string-escaper esc)
  (let ((spec (char-escape-spec esc)))
    (lambda (str) (string-escape str spec))))

(define (string-needs-escape? str esc)
  (let ((len (string-length str)))
    (let loop ((i 0))
      (if (= i len)
	  (let ((c (string-ref str i)))
	    (if (and (char>=? c (car esc))
		     (char<=? c (cadr esc)))
		(loop (+ 1 i))))))))

(define (string-escape str esc)
  (if (string-needs-escape? str esc)
	(let ((len (string-length str)))
	  (let loop ((i 0)
		     (li '()))
	    (if (= i len)
		(loop (+ 1 i)
		      (let ((c (string-ref str i)))
			(if (and (char>=? c (car esc))
				 (char<=? c (cadr esc)))
			    (let ((li2 (vector-ref
					(caddr esc)
					(- (char->integer c)
					   (char->integer (car esc))))))
			      (if li2
				  (append li2 li)
				  (cons c li)))
			    (cons c li)))))))))

(define (char-escape-spec speclist)
  (let ((minchar (caar speclist))
	(maxchar (caar speclist)))
    (let loop ((li (cdr speclist)))
      (if (not (null? li))
	    (let ((testchar (caar li)))
	      (if (char<? testchar minchar)
		  (set! minchar testchar))
	      (if (char>? testchar maxchar)
		  (set! maxchar testchar)))
	    (loop (cdr li)))))
     (let ((specv (make-vector (+ 1 (- (char->integer maxchar)
				       (char->integer minchar))) #f)))
      (map (lambda (specpair)
	     (vector-set! specv
			  (- (char->integer (car specpair))
			     (char->integer minchar))
			  (reverse (string->list (cdr specpair)))))