Author: yamakenz
Date: Tue Jul 10 22:58:08 2007
New Revision: 4693
Modified:
trunk/doc/COMPATIBILITY
trunk/scm/prime.scm
trunk/scm/util.scm
trunk/test/test-uim-util.scm
trunk/uim/uim-util.c
Log:
* uim/uim-util.c
- (uim_strsplit, uim_split_string): Removed
- (string_contains): New static function
- (uim_init_util_subrs):
* Remove definition for string-split
* Add definition for string-contains
* scm/util.scm
- (string-split): New procedure. The specification is changed from C
version. See doc/COMPATIBILITY. All the invocation of string-split
is checked as safe by eyes
* scm/prime.scm
- (prime-engine-send-command): Follow the specification change of
string-split
- (prime-util-string-split): Add a comment about string-split
* test/test-uim-util.scm
- Update the "passed revision" comment
- Add tests for new string-split
- Add tests for string-contains
* doc/COMPATIBILITY
- Add new section "Specification changes of utility procedures"
Modified: trunk/doc/COMPATIBILITY
==============================================================================
--- trunk/doc/COMPATIBILITY (original)
+++ trunk/doc/COMPATIBILITY Tue Jul 10 22:58:08 2007
@@ -57,6 +57,23 @@
The changes are described below in most recently updated order.
------------------------------------------------------------------------------
+Summary: Specification changes of utility procedures
+Affects: uim developers, IM developers
+Updates: Scheme API
+Version: 1.5.0
+Revision: ac4693
+Date: 2007-07-11
+Modifier: YamaKen
+Related:
+URL:
+Changes:
+ (changed) string-split
+Description:
+ Now string-split produces empty strings as follows. See
+ test-uim-utils.scm for further information.
+ uim 1.4: (string-split "hhh" "h") ==> ()
+ uim 1.5: (string-split "hhh" "h") ==> ("" "" "" "")
+------------------------------------------------------------------------------
Summary: SRFI-1 procedures replacement
Affects: uim developers, IM developers
Updates: Scheme API
Modified: trunk/scm/prime.scm
==============================================================================
--- trunk/scm/prime.scm (original)
+++ trunk/scm/prime.scm Tue Jul 10 22:58:08 2007
@@ -743,6 +743,11 @@
;; (string-split "\t\t" "\t") => ().
;; (prime-util-string-split "\t\t" "\t") => ("" "" "").
;; The second argument separator must be a single character string.
+;;
+;; uim 1.5.0 revised the specification of string-split as
+;; follows. Replace prime-util-string-split with the new string-split
+;; if no other problems are remaining. -- YamaKen 2007-07-11
+;; (string-split "\t\t" "\t") => ("" "" "")
(define prime-util-string-split
(lambda (string separator)
(let ((result (list))
@@ -809,11 +814,11 @@
;; problem with unix domain socket.
(define prime-engine-send-command
(lambda (arg-list)
- (cdr
- (string-split
- (prime-send-command
- (prime-util-string-concat arg-list "\t"))
- "\n"))))
+ (let* ((result (prime-send-command
+ (prime-util-string-concat arg-list "\t")))
+ (result-lines (string-split result "\n")))
+ (take! result-lines (- (length result-lines) 1)) ;; drop last "\n"
+ (cdr result-lines)))) ;; drop status line
(define prime-engine-conv-predict
(lambda (prime-session)
Modified: trunk/scm/util.scm
==============================================================================
--- trunk/scm/util.scm (original)
+++ trunk/scm/util.scm Tue Jul 10 22:58:08 2007
@@ -153,6 +153,18 @@
(lambda (sep str-list)
(apply string-append (join sep str-list))))
+(define string-split
+ (lambda (str sep)
+ (let ((slen (string-length str))
+ (seplen (string-length sep)))
+ (let rec ((start 0))
+ (let ((next (and (<= start slen)
+ (string-contains str sep start))))
+ (if next
+ (cons (substring str start next)
+ (rec (+ next seplen)))
+ (list (substring str start slen))))))))
+
(define string-append-map
(lambda args
(apply string-append (apply map args))))
Modified: trunk/test/test-uim-util.scm
==============================================================================
--- trunk/test/test-uim-util.scm (original)
+++ trunk/test/test-uim-util.scm Tue Jul 10 22:58:08 2007
@@ -29,7 +29,7 @@
;;; SUCH DAMAGE.
;;;;
-;; These tests are passed at revision 4692 (new repository)
+;; These tests are passed at revision 4693 (new repository)
(use test.unit)
@@ -67,7 +67,47 @@
(assert-true (uim-bool '(unsetenv "UIM_NONEXISTING_ENV")))
(assert-false (uim-bool '(getenv "UIM_NONEXISTING_ENV"))))
- ("test string-split"
+ ;; See "Specification changes of utility procedures" of doc/COMPATIBILITY
+;; ("test string-split (uim 1.4)"
+;; ;; ordinary split
+;; (assert-equal '("h" "geh" "ge")
+;; (uim '(string-split "hogehoge" "o")))
+;; ;; case sensitive
+;; (assert-equal '("hogehoge")
+;; (uim '(string-split "hogehoge" "O")))
+;; ;; split by sequence
+;; (assert-equal '("h" "eh" "e")
+;; (uim '(string-split "hogehoge" "og")))
+;; ;; split by first character
+;; (assert-equal '("oge" "oge")
+;; (uim '(string-split "hogehoge" "h")))
+;; ;; split by first sequence
+;; (assert-equal '("ge" "ge")
+;; (uim '(string-split "hogehoge" "ho")))
+;; ;; split by last character
+;; (assert-equal '("hog" "hog")
+;; (uim '(string-split "hogehoge" "e")))
+;; ;; split by last sequence
+;; (assert-equal '("ho" "ho")
+;; (uim '(string-split "hogehoge" "ge")))
+;; ;; split by whole string
+;; (assert-equal ()
+;; (uim '(string-split "hogehoge" "hogehoge")))
+;; ;; repeated splitter
+;; (assert-equal ()
+;; (uim '(string-split "hhh" "h")))
+;; ;; split by space
+;; (assert-equal '("h" "o" "g" "e" "hoge")
+;; (uim '(string-split " h o g e hoge" " ")))
+;; ;; split by symbolic character
+;; (assert-equal '("h" "o" "g" "e" "hoge")
+;; (uim '(string-split "|h|o|g|e|hoge" "|")))
+;; ;; split by non existent character
+;; (assert-equal '("hogehoge")
+;; (uim '(string-split "hogehoge" "|"))))
+
+ ;; See "Specification changes of utility procedures" of doc/COMPATIBILITY
+ ("test string-split (uim 1.5)"
;; ordinary split
(assert-equal '("h" "geh" "ge")
(uim '(string-split "hogehoge" "o")))
@@ -78,25 +118,28 @@
(assert-equal '("h" "eh" "e")
(uim '(string-split "hogehoge" "og")))
;; split by first character
- (assert-equal '("oge" "oge")
+ (assert-equal '("" "oge" "oge")
(uim '(string-split "hogehoge" "h")))
;; split by first sequence
- (assert-equal '("ge" "ge")
+ (assert-equal '("" "ge" "ge")
(uim '(string-split "hogehoge" "ho")))
;; split by last character
- (assert-equal '("hog" "hog")
+ (assert-equal '("hog" "hog" "")
(uim '(string-split "hogehoge" "e")))
;; split by last sequence
- (assert-equal '("ho" "ho")
+ (assert-equal '("ho" "ho" "")
(uim '(string-split "hogehoge" "ge")))
;; split by whole string
- (assert-equal ()
+ (assert-equal '("" "")
(uim '(string-split "hogehoge" "hogehoge")))
+ ;; repeated splitter
+ (assert-equal '("" "" "" "")
+ (uim '(string-split "hhh" "h")))
;; split by space
- (assert-equal '("h" "o" "g" "e" "hoge")
+ (assert-equal '("" "h" "o" "g" "e" "hoge")
(uim '(string-split " h o g e hoge" " ")))
;; split by symbolic character
- (assert-equal '("h" "o" "g" "e" "hoge")
+ (assert-equal '("" "h" "o" "g" "e" "hoge")
(uim '(string-split "|h|o|g|e|hoge" "|")))
;; split by non existent character
(assert-equal '("hogehoge")
@@ -124,6 +167,17 @@
; (assert-equal '("c" " " " " "b" " " "a")
; (uim '(string-to-list "a b c")))
)
+
+ ("test string-contains"
+ (assert-equal 0 (uim '(string-contains "" "" 0)))
+ (assert-false (uim '(string-contains "" "f" 0)))
+ (assert-equal 0 (uim '(string-contains "foo" "" 0)))
+ (assert-equal 0 (uim '(string-contains "foo" "f" 0)))
+ (assert-equal 1 (uim '(string-contains "foo" "o" 0)))
+ (assert-equal 1 (uim '(string-contains "foo" "oo" 0)))
+ (assert-false (uim '(string-contains "foo" "oof" 0)))
+ (assert-equal 1 (uim '(string-contains "foo" "o" 1)))
+ (assert-equal 2 (uim '(string-contains "foo" "o" 2))))
("test string-prefix?"
(assert-true (uim-bool '(string-prefix? "" "foo_bar")))
Modified: trunk/uim/uim-util.c
==============================================================================
--- trunk/uim/uim-util.c (original)
+++ trunk/uim/uim-util.c Tue Jul 10 22:58:08 2007
@@ -179,95 +179,29 @@
return uim_scm_t();
}
-static char **
-uim_strsplit(const char *splittee, const char *splitter)
-{
- const char *cur, *tmp;
- int nr_token = 0;
- int in_token = 0;
- char **res;
- int len;
- int i;
-
- if (!splittee || !splitter)
- return NULL;
-
-
- /* count the number of token */
- cur = splittee;
- while (*cur) {
- if (strchr(splitter, *cur)) {
- in_token = 0;
- } else {
- if (!in_token) {
- nr_token ++;
- }
- in_token = 1;
- }
- cur ++;
- }
- /* allocate buffer */
- res = (char **)malloc(sizeof(char *) * (nr_token + 1) );
- if (!res) {
- return NULL;
- }
- /**/
- cur = splittee;
- for (i = 0; i < nr_token; i++) {
- /* find current token's start */
- while (strchr(splitter, *cur)) {
- cur ++;
- }
- /* calc length */
- len = 0;
- tmp = cur;
- while (!strchr(splitter, *tmp)) {
- len ++;
- tmp ++;
- }
- /* store */
- res[i] = malloc(sizeof(char) * (len + 1));
- strlcpy(res[i], cur, len + 1);
- cur = tmp;
- }
- /**/
- res[nr_token] = NULL;
-
- return res;
-}
-
+/* Limited version of SRFI-13 string-contains. The number of args are
+ * fixed to 3. */
static uim_lisp
-uim_split_string(uim_lisp _splittee, uim_lisp _splitter)
+string_contains(uim_lisp s1_, uim_lisp s2_, uim_lisp start1_)
{
- const char *splittee = uim_scm_refer_c_str(_splittee);
- const char *splitter = uim_scm_refer_c_str(_splitter);
- char **strs;
- uim_lisp l = uim_scm_null_list();
- int i;
- int n_strs;
-
- if (!uim_scm_stringp(_splittee) || !uim_scm_stringp(_splitter))
- return uim_scm_f();
-
- if (splittee == NULL || splitter == NULL)
- return uim_scm_f();
+ const char *s1, *s2, *found;
+ int start1;
+ size_t s1len;
- strs = uim_strsplit(splittee, splitter);
+ if (!uim_scm_stringp(s1_) || !uim_scm_stringp(s2_))
+ return uim_scm_f(); /* FIXME: uim_scm_error() */
- if (!strs)
- return uim_scm_f();
+ s1 = uim_scm_refer_c_str(s1_);
+ s2 = uim_scm_refer_c_str(s2_);
+ start1 = uim_scm_c_int(start1_);
+ s1len = strlen(s1);
- if (!*strs)
- return uim_scm_null_list();
+ if (start1 < 0 || s1len < (size_t)start1)
+ return uim_scm_f(); /* FIXME: uim_scm_error() */
- for (n_strs = 0; strs[n_strs] != '\0'; n_strs++);
+ found = strstr(&s1[start1], s2);
- l = uim_scm_c_strs_into_list(n_strs, (const char *const *)strs);
- for (i = n_strs - 1; i >= 0; i--) {
- free(strs[i]);
- }
- free(strs);
- return l;
+ return (found) ? uim_scm_make_int(found - s1) : uim_scm_f();
}
static uim_lisp
@@ -350,19 +284,22 @@
uim_scm_init_subr_0("sys-pkglibdir", sys_pkglibdir);
uim_scm_init_subr_0("sys-datadir", sys_datadir);
uim_scm_init_subr_0("sys-pkgdatadir", sys_pkgdatadir);
+
uim_scm_init_subr_1("file-readable?", file_readablep);
uim_scm_init_subr_1("file-writable?", file_writablep);
uim_scm_init_subr_1("file-executable?", file_executablep);
uim_scm_init_subr_1("file-regular?", file_regularp);
uim_scm_init_subr_1("file-directory?", file_directoryp);
uim_scm_init_subr_1("file-mtime", file_mtime);
+
uim_scm_init_subr_0("setugid?", setugidp);
+
uim_scm_init_subr_1("getenv", c_getenv);
uim_scm_init_subr_3("setenv", c_setenv);
uim_scm_init_subr_1("unsetenv", c_unsetenv);
- /* these procedures should be replaced with standard ones of R5RS or SRFIs */
- uim_scm_init_subr_2("string-split", uim_split_string);
+ /* SRFI-13 */
+ uim_scm_init_subr_3("string-contains", string_contains);
uim_scm_init_subr_2("string-prefix?", string_prefixp);
uim_scm_init_subr_2("string-prefix-ci?", string_prefix_cip);
}
