X-Git-Url: http://git.kpe.io/?a=blobdiff_plain;f=strings.lisp;h=e9a73a3c9c7f1c3041680fc6cc9b5466bb08fe03;hb=e2450e52a0813b8583395d7ac88f24e896f46e13;hp=6ec3f5ce434465fae964e79855d57da64bd5a88a;hpb=91e3d8f38e4ccdfbfb9c351e05b487ee899c9e47;p=kmrcl.git

diff --git a/strings.lisp b/strings.lisp
index 6ec3f5c..e9a73a3 100644
--- a/strings.lisp
+++ b/strings.lisp
@@ -7,7 +7,7 @@
 ;;;; Programmer:    Kevin M. Rosenberg
 ;;;; Date Started:  Apr 2000
 ;;;;
-;;;; $Id: strings.lisp,v 1.49 2003/07/19 20:32:48 kevin Exp $
+;;;; $Id$
 ;;;;
 ;;;; This file, part of KMRCL, is Copyright (c) 2002 by Kevin M. Rosenberg
 ;;;;
@@ -150,22 +150,32 @@
       (setq hash (+ hash (char-code (char str i)))))
     (logand hash bitmask)))
 
-(defun string-not-null? (str)
-  (and str (not (zerop (length str)))))
-  
-(defun whitespace? (c) 
-  (declare (character c))
-  (locally (declare (optimize (speed 3) (safety 0)))
-    (or (char= c #\Space) (char= c #\Tab) (char= c #\Return)
-	(char= c #\Linefeed))))
+(defun is-string-empty (str)
+  (zerop (length str)))
+
+(defvar *whitespace-chars* '(#\space #\tab #\return #\linefeed
+			     #+allegro #\%space
+			     #+lispworks #\No-Break-Space))
 
-(defun not-whitespace? (c)
-  (not (whitespace? c)))
+(defun is-char-whitespace (c) 
+  (declare (character c) (optimize (speed 3) (safety 0)))
+  (or (char= c #\Space) (char= c #\Tab) (char= c #\Return)
+      (char= c #\Linefeed)
+      #+allegro (char= c #\%space)
+      #+lispworks (char= c #\No-Break-Space)))
 
-(defun string-ws? (str)
+(defun is-string-whitespace (str)
   "Return t if string is all whitespace"
-  (when (stringp str)
-    (null (find-if #'not-whitespace? str))))
+  (every #'is-char-whitespace str))
+
+(defun string-right-trim-whitespace (str)
+  (string-right-trim *whitespace-chars* str))
+
+(defun string-left-trim-whitespace (str)
+  (string-left-trim *whitespace-chars* str))
+
+(defun string-trim-whitespace (str)
+  (string-trim *whitespace-chars* str))
 
 (defun replaced-string-length (str repl-alist)
   (declare (simple-string str)
@@ -417,7 +427,7 @@ for characters in a string"
 	(+ 10 (- code +char-code-upper-a+))
 	(- code +char-code-0+))))
 
-(defun escape-uri-field (query)
+(defun encode-uri-string (query)
   "Escape non-alphanumeric characters for URI fields"
   (declare (simple-string query)
 	   (optimize (speed 3) (safety 0) (space 0)))
@@ -440,7 +450,7 @@ for characters in a string"
 	    (setf (schar str dpos) (hexchar (logand c 15))))
 	(setf (schar str dpos) ch)))))
 
-(defun unescape-uri-field (query)
+(defun decode-uri-string (query)
   "Unescape non-alphanumeric characters for URI fields"
   (declare (simple-string query)
 	   (optimize (speed 3) (safety 0) (space 0)))
@@ -551,13 +561,69 @@ for characters in a string"
 (defun shrink-vector (str size)
   #+allegro
   (excl::.primcall 'sys::shrink-svector str size)
-  #+sbcl
-  (sb-kernel:shrink-vector str size)
   #+cmu
   (lisp::shrink-vector str size)
   #+lispworks
   (system::shrink-vector$vector str size)
-  #+(or allegro cmu sbcl lispworks)
-  str
-  #-(or allegro cmu sbcl lispworks)
-  (subseq str 0 size))
+  #+sbcl
+  (sb-kernel:shrink-vector str size)
+  #+scl
+  (common-lisp::shrink-vector str size)
+  #-(or allegro cmu lispworks sbcl scl)
+  (setq str (subseq str 0 size))
+  str)
+
+(defun lex-string (string &key (whitespace '(#\space #\newline)))
+  "Separates a string at whitespace and returns a list of strings"
+  (flet ((is-sep (char) (member char whitespace :test #'char=)))
+    (let ((tokens nil))
+      (do* ((token-start
+             (position-if-not #'is-sep string) 
+             (when token-end
+               (position-if-not #'is-sep string :start (1+ token-end))))
+            (token-end
+             (when token-start
+               (position-if #'is-sep string :start token-start))
+             (when token-start
+               (position-if #'is-sep string :start token-start))))
+           ((null token-start) (nreverse tokens))
+        (push (subseq string token-start token-end) tokens)))))
+
+(defun split-alphanumeric-string (string)
+  "Separates a string at any non-alphanumeric chararacter"
+  (flet ((is-sep (char) (non-alphanumericp char)))
+    (let ((tokens nil))
+      (do* ((token-start
+             (position-if-not #'is-sep string) 
+             (when token-end
+               (position-if-not #'is-sep string :start (1+ token-end))))
+            (token-end
+             (when token-start
+               (position-if #'is-sep string :start token-start))
+             (when token-start
+               (position-if #'is-sep string :start token-start))))
+           ((null token-start) (nreverse tokens))
+        (push (subseq string token-start token-end) tokens)))))
+
+
+
+(defun collapse-whitespace (s)
+  "Convert multiple whitespace characters to a single space character."
+  (declare (simple-string s)
+	   (optimize (speed 3) (safety 0)))
+  (with-output-to-string (stream)
+    (do ((pos 0 (1+ pos))
+	 (in-white nil)
+	 (len (length s)))
+	((= pos len))
+      (declare (fixnum pos len))
+      (let ((c (schar s pos)))
+	(declare (character c))
+	(cond
+	 ((kl:is-char-whitespace c)
+	  (unless in-white
+	    (write-char #\space stream))
+	  (setq in-white t))
+	 (t
+	  (setq in-white nil)
+	  (write-char c stream)))))))