Commit | Line | Data |
---|---|---|
1a179b03 | 1 | ;;;; Copyright (C) 1997, 1999, 2001 Free Software Foundation, Inc. |
87fefc1c | 2 | ;;;; |
73be1d9e MV |
3 | ;;;; This library is free software; you can redistribute it and/or |
4 | ;;;; modify it under the terms of the GNU Lesser General Public | |
5 | ;;;; License as published by the Free Software Foundation; either | |
6 | ;;;; version 2.1 of the License, or (at your option) any later version. | |
7 | ;;;; | |
8 | ;;;; This library is distributed in the hope that it will be useful, | |
400d7382 | 9 | ;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of |
73be1d9e MV |
10 | ;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
11 | ;;;; Lesser General Public License for more details. | |
12 | ;;;; | |
13 | ;;;; You should have received a copy of the GNU Lesser General Public | |
14 | ;;;; License along with this library; if not, write to the Free Software | |
15 | ;;;; Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | |
87fefc1c TTN |
16 | ;;;; |
17 | ||
18 | ;;; Commentary: | |
19 | ||
20 | ;; These procedures are exported: | |
21 | ;; (match:count match) | |
22 | ;; (match:string match) | |
23 | ;; (match:prefix match) | |
24 | ;; (match:suffix match) | |
25 | ;; (regexp-match? match) | |
26 | ;; (regexp-quote string) | |
27 | ;; (match:start match . submatch-num) | |
28 | ;; (match:end match . submatch-num) | |
29 | ;; (match:substring match . submatch-num) | |
30 | ;; (string-match pattern str . start) | |
31 | ;; (regexp-substitute port match . items) | |
32 | ;; (fold-matches regexp string init proc . flags) | |
33 | ;; (list-matches regexp string . flags) | |
34 | ;; (regexp-substitute/global port regexp string . items) | |
35 | ||
36 | ;;; Code: | |
400d7382 JB |
37 | \f |
38 | ;;;; POSIX regex support functions. | |
39 | ||
1a179b03 MD |
40 | (define-module (ice-9 regex) |
41 | :export (match:count match:string match:prefix match:suffix | |
42 | regexp-match? regexp-quote match:start match:end match:substring | |
43 | string-match regexp-substitute fold-matches list-matches | |
44 | regexp-substitute/global)) | |
05817d9e | 45 | |
400d7382 JB |
46 | ;;; FIXME: |
47 | ;;; It is not clear what should happen if a `match' function | |
48 | ;;; is passed a `match number' which is out of bounds for the | |
49 | ;;; regexp match: return #f, or throw an error? These routines | |
50 | ;;; throw an out-of-range error. | |
51 | ||
52 | ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; | |
53 | ;;;; These procedures are not defined in SCSH, but I found them useful. | |
54 | ||
1a179b03 | 55 | (define (match:count match) |
400d7382 JB |
56 | (- (vector-length match) 1)) |
57 | ||
1a179b03 | 58 | (define (match:string match) |
400d7382 JB |
59 | (vector-ref match 0)) |
60 | ||
1a179b03 | 61 | (define (match:prefix match) |
4e15fee8 | 62 | (substring (match:string match) 0 (match:start match 0))) |
400d7382 | 63 | |
1a179b03 | 64 | (define (match:suffix match) |
4e15fee8 | 65 | (substring (match:string match) (match:end match 0))) |
400d7382 JB |
66 | |
67 | ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; | |
68 | ;;;; SCSH compatibility routines. | |
69 | ||
1a179b03 | 70 | (define (regexp-match? match) |
400d7382 JB |
71 | (and (vector? match) |
72 | (string? (vector-ref match 0)) | |
73 | (let loop ((i 1)) | |
74 | (cond ((>= i (vector-length match)) #t) | |
75 | ((and (pair? (vector-ref match i)) | |
76 | (integer? (car (vector-ref match i))) | |
77 | (integer? (cdr (vector-ref match i)))) | |
78 | (loop (+ 1 i))) | |
79 | (else #f))))) | |
80 | ||
87fefc1c | 81 | (define (regexp-quote string) |
400d7382 JB |
82 | (call-with-output-string |
83 | (lambda (p) | |
84 | (let loop ((i 0)) | |
87fefc1c | 85 | (and (< i (string-length string)) |
400d7382 | 86 | (begin |
87fefc1c | 87 | (case (string-ref string i) |
400d7382 JB |
88 | ((#\* #\. #\( #\) #\+ #\? #\\ #\^ #\$ #\{ #\}) |
89 | (write-char #\\ p))) | |
87fefc1c | 90 | (write-char (string-ref string i) p) |
400d7382 JB |
91 | (loop (1+ i)))))))) |
92 | ||
1a179b03 | 93 | (define (match:start match . args) |
400d7382 JB |
94 | (let* ((matchnum (if (pair? args) |
95 | (+ 1 (car args)) | |
96 | 1)) | |
97 | (start (car (vector-ref match matchnum)))) | |
98 | (if (= start -1) #f start))) | |
99 | ||
1a179b03 | 100 | (define (match:end match . args) |
400d7382 JB |
101 | (let* ((matchnum (if (pair? args) |
102 | (+ 1 (car args)) | |
103 | 1)) | |
104 | (end (cdr (vector-ref match matchnum)))) | |
105 | (if (= end -1) #f end))) | |
106 | ||
1a179b03 | 107 | (define (match:substring match . args) |
400d7382 JB |
108 | (let* ((matchnum (if (pair? args) |
109 | (car args) | |
110 | 0)) | |
111 | (start (match:start match matchnum)) | |
112 | (end (match:end match matchnum))) | |
4e15fee8 | 113 | (and start end (substring (match:string match) start end)))) |
400d7382 | 114 | |
1a179b03 | 115 | (define (string-match pattern str . args) |
400d7382 JB |
116 | (let ((rx (make-regexp pattern)) |
117 | (start (if (pair? args) (car args) 0))) | |
118 | (regexp-exec rx str start))) | |
119 | ||
1a179b03 | 120 | (define (regexp-substitute port match . items) |
400d7382 JB |
121 | ;; If `port' is #f, send output to a string. |
122 | (if (not port) | |
123 | (call-with-output-string | |
124 | (lambda (p) | |
125 | (apply regexp-substitute p match items))) | |
126 | ||
127 | ;; Otherwise, process each substitution argument in `items'. | |
128 | (for-each (lambda (obj) | |
129 | (cond ((string? obj) (display obj port)) | |
130 | ((integer? obj) (display (match:substring match obj) port)) | |
131 | ((eq? 'pre obj) (display (match:prefix match) port)) | |
132 | ((eq? 'post obj) (display (match:suffix match) port)) | |
133 | (else (error 'wrong-type-arg obj)))) | |
134 | items))) | |
135 | ||
50ff2ecb JB |
136 | ;;; If we call fold-matches, below, with a regexp that can match the |
137 | ;;; empty string, it's not obvious what "all the matches" means. How | |
138 | ;;; many empty strings are there in the string "a"? Our answer: | |
139 | ;;; | |
140 | ;;; This function applies PROC to every non-overlapping, maximal | |
141 | ;;; match of REGEXP in STRING. | |
142 | ;;; | |
143 | ;;; "non-overlapping": There are two non-overlapping matches of "" in | |
144 | ;;; "a" --- one before the `a', and one after. There are three | |
145 | ;;; non-overlapping matches of "q|x*" in "aqb": the empty strings | |
146 | ;;; before `a' and after `b', and `q'. The two empty strings before | |
147 | ;;; and after `q' don't count, because they overlap with the match of | |
148 | ;;; "q". | |
149 | ;;; | |
150 | ;;; "maximal": There are three distinct maximal matches of "x*" in | |
151 | ;;; "axxxb": one before the `a', one covering `xxx', and one after the | |
152 | ;;; `b'. Around or within `xxx', only the match covering all three | |
153 | ;;; x's counts, because the rest are not maximal. | |
154 | ||
1a179b03 | 155 | (define (fold-matches regexp string init proc . flags) |
50ff2ecb JB |
156 | (let ((regexp (if (regexp? regexp) regexp (make-regexp regexp))) |
157 | (flags (if (null? flags) 0 flags))) | |
158 | (let loop ((start 0) | |
159 | (value init) | |
160 | (abuts #f)) ; True if start abuts a previous match. | |
161 | (let ((m (if (> start (string-length string)) #f | |
162 | (regexp-exec regexp string start flags)))) | |
163 | (cond | |
164 | ((not m) value) | |
165 | ((and (= (match:start m) (match:end m)) abuts) | |
166 | ;; We matched an empty string, but that would overlap the | |
167 | ;; match immediately before. Try again at a position | |
168 | ;; further to the right. | |
169 | (loop (+ start 1) value #f)) | |
170 | (else | |
171 | (loop (match:end m) (proc m value) #t))))))) | |
172 | ||
1a179b03 | 173 | (define (list-matches regexp string . flags) |
50ff2ecb JB |
174 | (reverse! (apply fold-matches regexp string '() cons flags))) |
175 | ||
1a179b03 | 176 | (define (regexp-substitute/global port regexp string . items) |
50ff2ecb | 177 | |
400d7382 JB |
178 | ;; If `port' is #f, send output to a string. |
179 | (if (not port) | |
180 | (call-with-output-string | |
181 | (lambda (p) | |
182 | (apply regexp-substitute/global p regexp string items))) | |
183 | ||
50ff2ecb JB |
184 | ;; Walk the set of non-overlapping, maximal matches. |
185 | (let next-match ((matches (list-matches regexp string)) | |
186 | (start 0)) | |
f5641401 | 187 | (if (null? matches) |
4e15fee8 | 188 | (display (substring string start) port) |
50ff2ecb JB |
189 | (let ((m (car matches))) |
190 | ||
191 | ;; Process all of the items for this match. Don't use | |
192 | ;; for-each, because we need to make sure 'post at the | |
193 | ;; end of the item list is a tail call. | |
194 | (let next-item ((items items)) | |
87fefc1c | 195 | |
50ff2ecb JB |
196 | (define (do-item item) |
197 | (cond | |
198 | ((string? item) (display item port)) | |
199 | ((integer? item) (display (match:substring m item) port)) | |
200 | ((procedure? item) (display (item m) port)) | |
87fefc1c | 201 | ((eq? item 'pre) |
50ff2ecb | 202 | (display |
4e15fee8 | 203 | (substring string start (match:start m)) |
50ff2ecb JB |
204 | port)) |
205 | ((eq? item 'post) | |
f5641401 | 206 | (next-match (cdr matches) (match:end m))) |
50ff2ecb JB |
207 | (else (error 'wrong-type-arg item)))) |
208 | ||
209 | (if (pair? items) | |
210 | (if (null? (cdr items)) | |
211 | (do-item (car items)) ; This is a tail call. | |
212 | (begin | |
213 | (do-item (car items)) ; This is not. | |
214 | (next-item (cdr items))))))))))) |