summaryrefslogtreecommitdiff
path: root/libguile/srfi-13.c
diff options
context:
space:
mode:
authorMarius Vollmer <mvo@zagadka.de>2004-08-24 22:12:08 +0000
committerMarius Vollmer <mvo@zagadka.de>2004-08-24 22:12:08 +0000
commit63181a97f5d1104e43219548ff71fbf13096c598 (patch)
tree80473344f4a206b8b58c33c0d5c2d49e12a52761 /libguile/srfi-13.c
parent4b0d27f2660eadbddd64871def00edcf8502f567 (diff)
downloadguile-63181a97f5d1104e43219548ff71fbf13096c598.tar.gz
* srfi-13.c, srfi-13.h, srfi-14.c, srfi-14.h: New files.
* strop.h, strop.c: Removed, they are now empty. * Makefile.am: Updated for new and removed files.
Diffstat (limited to 'libguile/srfi-13.c')
-rw-r--r--libguile/srfi-13.c3323
1 files changed, 3323 insertions, 0 deletions
diff --git a/libguile/srfi-13.c b/libguile/srfi-13.c
new file mode 100644
index 000000000..8f1a89ff5
--- /dev/null
+++ b/libguile/srfi-13.c
@@ -0,0 +1,3323 @@
+/* srfi-13.c --- SRFI-13 procedures for Guile
+ *
+ * Copyright (C) 2001, 2004 Free Software Foundation, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+
+#include <string.h>
+#include <ctype.h>
+
+#include "libguile.h"
+
+#include "libguile/srfi-13.h"
+#include "libguile/srfi-14.h"
+
+/* SCM_VALIDATE_SUBSTRING_SPEC_COPY is deprecated since it encourages
+ messing with the internal representation of strings. We define our
+ own version since we use it so much and are messing with Guile
+ internals anyway.
+*/
+
+#define MY_VALIDATE_SUBSTRING_SPEC_COPY(pos_str, str, c_str, \
+ pos_start, start, c_start, \
+ pos_end, end, c_end) \
+ do { \
+ SCM_VALIDATE_STRING (pos_str, str); \
+ c_str = scm_i_string_chars (str); \
+ scm_i_get_substring_spec (scm_i_string_length (str), \
+ start, &c_start, end, &c_end); \
+ } while (0)
+
+#define MY_VALIDATE_SUBSTRING_SPEC(pos_str, str, \
+ pos_start, start, c_start, \
+ pos_end, end, c_end) \
+ do { \
+ SCM_VALIDATE_STRING (pos_str, str); \
+ scm_i_get_substring_spec (scm_i_string_length (str), \
+ start, &c_start, end, &c_end); \
+ } while (0)
+
+/* Likewise for SCM_VALIDATE_STRING_COPY. */
+
+#define MY_VALIDATE_STRING_COPY(pos, str, cvar) \
+ do { \
+ scm_validate_string (pos, str); \
+ cvar = scm_i_string_chars (str); \
+ } while (0)
+
+
+SCM_DEFINE (scm_string_null_p, "string-null?", 1, 0, 0,
+ (SCM str),
+ "Return @code{#t} if @var{str}'s length is zero, and\n"
+ "@code{#f} otherwise.\n"
+ "@lisp\n"
+ "(string-null? \"\") @result{} #t\n"
+ "y @result{} \"foo\"\n"
+ "(string-null? y) @result{} #f\n"
+ "@end lisp")
+#define FUNC_NAME s_scm_string_null_p
+{
+ SCM_VALIDATE_STRING (1, str);
+ return scm_from_bool (scm_i_string_length (str) == 0);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_any, "string-any", 2, 2, 0,
+ (SCM char_pred, SCM s, SCM start, SCM end),
+ "Check if the predicate @var{pred} is true for any character in\n"
+ "the string @var{s}.\n"
+ "\n"
+ "Calls to @var{pred} are made from left to right across @var{s}.\n"
+ "When it returns true (ie.@: non-@code{#f}), that return value\n"
+ "is the return from @code{string-any}.\n"
+ "\n"
+ "The SRFI-13 specification requires that the call to @var{pred}\n"
+ "on the last character of @var{s} (assuming that point is\n"
+ "reached) be a tail call, but currently in Guile this is not the\n"
+ "case.")
+#define FUNC_NAME s_scm_string_any
+{
+ const char *cstr;
+ int cstart, cend;
+ SCM res;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+
+ if (SCM_CHARP (char_pred))
+ {
+ return (memchr (cstr+cstart, (int) SCM_CHAR (char_pred),
+ cend-cstart) == NULL
+ ? SCM_BOOL_F : SCM_BOOL_T);
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ int i;
+ for (i = cstart; i < cend; i++)
+ if (SCM_CHARSET_GET (char_pred, cstr[i]))
+ return SCM_BOOL_T;
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (1, char_pred);
+
+ cstr += cstart;
+ while (cstart < cend)
+ {
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (*cstr));
+ if (scm_is_true (res))
+ return res;
+ cstr++;
+ cstart++;
+ }
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_every, "string-every", 2, 2, 0,
+ (SCM char_pred, SCM s, SCM start, SCM end),
+ "Check if the predicate @var{pred} is true for every character\n"
+ "in the string @var{s}.\n"
+ "\n"
+ "Calls to @var{pred} are made from left to right across @var{s}.\n"
+ "If the predicate is true for every character then the return\n"
+ "value from the last @var{pred} call is the return from\n"
+ "@code{string-every}.\n"
+ "\n"
+ "If there are no characters in @var{s} (ie.@: @var{start} equals\n"
+ "@var{end}) then the return is @code{#t}.\n"
+ "\n"
+ "The SRFI-13 specification requires that the call to @var{pred}\n"
+ "on the last character of @var{s} (assuming that point is\n"
+ "reached) be a tail call, but currently in Guile this is not the\n"
+ "case.")
+#define FUNC_NAME s_scm_string_every
+{
+ const char *cstr;
+ int cstart, cend;
+ SCM res;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ int i;
+ for (i = cstart; i < cend; i++)
+ if (cstr[i] != cchr)
+ return SCM_BOOL_F;
+ return SCM_BOOL_T;
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ int i;
+ for (i = cstart; i < cend; i++)
+ if (! SCM_CHARSET_GET (char_pred, cstr[i]))
+ return SCM_BOOL_F;
+ return SCM_BOOL_T;
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (1, char_pred);
+
+ res = SCM_BOOL_T;
+ cstr += cstart;
+ while (cstart < cend)
+ {
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (*cstr));
+ if (scm_is_false (res))
+ return res;
+ cstr++;
+ cstart++;
+ }
+ return res;
+ }
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_tabulate, "string-tabulate", 2, 0, 0,
+ (SCM proc, SCM len),
+ "@var{proc} is an integer->char procedure. Construct a string\n"
+ "of size @var{len} by applying @var{proc} to each index to\n"
+ "produce the corresponding string element. The order in which\n"
+ "@var{proc} is applied to the indices is not specified.")
+#define FUNC_NAME s_scm_string_tabulate
+{
+ size_t clen, i;
+ SCM res;
+ SCM ch;
+ char *p;
+
+ SCM_VALIDATE_PROC (1, proc);
+ clen = scm_to_size_t (len);
+ SCM_ASSERT_RANGE (2, len, clen >= 0);
+
+ res = scm_i_make_string (clen, &p);
+ i = 0;
+ while (i < clen)
+ {
+ /* The RES string remains untouched since nobody knows about it
+ yet. No need to refetch P.
+ */
+ ch = scm_call_1 (proc, scm_from_int (i));
+ if (!SCM_CHARP (ch))
+ SCM_MISC_ERROR ("procedure ~S returned non-char", scm_list_1 (proc));
+ *p++ = SCM_CHAR (ch);
+ i++;
+ }
+ return res;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_substring_to_list, "string->list", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Convert the string @var{str} into a list of characters.")
+#define FUNC_NAME s_scm_substring_to_list
+{
+ const char *cstr;
+ int cstart, cend;
+ SCM result = SCM_EOL;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ while (cstart < cend)
+ {
+ cend--;
+ result = scm_cons (SCM_MAKE_CHAR (cstr[cend]), result);
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+/* We export scm_substring_to_list as "string->list" since it is
+ compatible and more general. This function remains for the benefit
+ of C code that used it.
+*/
+
+SCM
+scm_string_to_list (SCM str)
+{
+ return scm_substring_to_list (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+SCM_DEFINE (scm_reverse_list_to_string, "reverse-list->string", 1, 0, 0,
+ (SCM chrs),
+ "An efficient implementation of @code{(compose string->list\n"
+ "reverse)}:\n"
+ "\n"
+ "@smalllisp\n"
+ "(reverse-list->string '(#\\a #\\B #\\c)) @result{} \"cBa\"\n"
+ "@end smalllisp")
+#define FUNC_NAME s_scm_reverse_list_to_string
+{
+ SCM result;
+ long i = scm_ilength (chrs);
+ char *data;
+
+ if (i < 0)
+ SCM_WRONG_TYPE_ARG (1, chrs);
+ result = scm_i_make_string (i, &data);
+
+ {
+
+ data += i;
+ while (!SCM_NULLP (chrs))
+ {
+ SCM elt = SCM_CAR (chrs);
+
+ SCM_VALIDATE_CHAR (SCM_ARGn, elt);
+ data--;
+ *data = SCM_CHAR (elt);
+ chrs = SCM_CDR (chrs);
+ }
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_SYMBOL (scm_sym_infix, "infix");
+SCM_SYMBOL (scm_sym_strict_infix, "strict-infix");
+SCM_SYMBOL (scm_sym_suffix, "suffix");
+SCM_SYMBOL (scm_sym_prefix, "prefix");
+
+SCM_DEFINE (scm_string_join, "string-join", 1, 2, 0,
+ (SCM ls, SCM delimiter, SCM grammar),
+ "Append the string in the string list @var{ls}, using the string\n"
+ "@var{delim} as a delimiter between the elements of @var{ls}.\n"
+ "@var{grammar} is a symbol which specifies how the delimiter is\n"
+ "placed between the strings, and defaults to the symbol\n"
+ "@code{infix}.\n"
+ "\n"
+ "@table @code\n"
+ "@item infix\n"
+ "Insert the separator between list elements. An empty string\n"
+ "will produce an empty list.\n"
+ "@item string-infix\n"
+ "Like @code{infix}, but will raise an error if given the empty\n"
+ "list.\n"
+ "@item suffix\n"
+ "Insert the separator after every list element.\n"
+ "@item prefix\n"
+ "Insert the separator before each list element.\n"
+ "@end table")
+#define FUNC_NAME s_scm_string_join
+{
+#define GRAM_INFIX 0
+#define GRAM_STRICT_INFIX 1
+#define GRAM_SUFFIX 2
+#define GRAM_PREFIX 3
+ SCM tmp;
+ SCM result;
+ int gram = GRAM_INFIX;
+ int del_len = 0, extra_len = 0;
+ int len = 0;
+ char * p;
+ long strings = scm_ilength (ls);
+
+ /* Validate the string list. */
+ if (strings < 0)
+ SCM_WRONG_TYPE_ARG (1, ls);
+
+ /* Validate the delimiter and record its length. */
+ if (SCM_UNBNDP (delimiter))
+ {
+ delimiter = scm_from_locale_string (" ");
+ del_len = 1;
+ }
+ else
+ {
+ SCM_VALIDATE_STRING (2, delimiter);
+ del_len = scm_i_string_length (delimiter);
+ }
+
+ /* Validate the grammar symbol and remember the grammar. */
+ if (SCM_UNBNDP (grammar))
+ gram = GRAM_INFIX;
+ else if (scm_is_eq (grammar, scm_sym_infix))
+ gram = GRAM_INFIX;
+ else if (scm_is_eq (grammar, scm_sym_strict_infix))
+ gram = GRAM_STRICT_INFIX;
+ else if (scm_is_eq (grammar, scm_sym_suffix))
+ gram = GRAM_SUFFIX;
+ else if (scm_is_eq (grammar, scm_sym_prefix))
+ gram = GRAM_PREFIX;
+ else
+ SCM_WRONG_TYPE_ARG (3, grammar);
+
+ /* Check grammar constraints and calculate the space required for
+ the delimiter(s). */
+ switch (gram)
+ {
+ case GRAM_INFIX:
+ if (!SCM_NULLP (ls))
+ extra_len = (strings > 0) ? ((strings - 1) * del_len) : 0;
+ break;
+ case GRAM_STRICT_INFIX:
+ if (strings == 0)
+ SCM_MISC_ERROR ("strict-infix grammar requires non-empty list",
+ SCM_EOL);
+ extra_len = (strings - 1) * del_len;
+ break;
+ default:
+ extra_len = strings * del_len;
+ break;
+ }
+
+ tmp = ls;
+ while (SCM_CONSP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ SCM_VALIDATE_STRING (1, elt);
+ len += scm_i_string_length (elt);
+ tmp = SCM_CDR (tmp);
+ }
+
+ result = scm_i_make_string (len + extra_len, &p);
+
+ tmp = ls;
+ switch (gram)
+ {
+ case GRAM_INFIX:
+ case GRAM_STRICT_INFIX:
+ while (!SCM_NULLP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ memmove (p, scm_i_string_chars (elt),
+ scm_i_string_length (elt));
+ p += scm_i_string_length (elt);
+ if (!SCM_NULLP (SCM_CDR (tmp)) && del_len > 0)
+ {
+ memmove (p, scm_i_string_chars (delimiter), del_len);
+ p += del_len;
+ }
+ tmp = SCM_CDR (tmp);
+ }
+ break;
+ case GRAM_SUFFIX:
+ while (!SCM_NULLP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ memmove (p, scm_i_string_chars (elt),
+ scm_i_string_length (elt));
+ p += scm_i_string_length (elt);
+ if (del_len > 0)
+ {
+ memmove (p, scm_i_string_chars (delimiter), del_len);
+ p += del_len;
+ }
+ tmp = SCM_CDR (tmp);
+ }
+ break;
+ case GRAM_PREFIX:
+ while (!SCM_NULLP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ if (del_len > 0)
+ {
+ memmove (p, scm_i_string_chars (delimiter), del_len);
+ p += del_len;
+ }
+ memmove (p, scm_i_string_chars (elt),
+ scm_i_string_length (elt));
+ p += scm_i_string_length (elt);
+ tmp = SCM_CDR (tmp);
+ }
+ break;
+ }
+ return result;
+#undef GRAM_INFIX
+#undef GRAM_STRICT_INFIX
+#undef GRAM_SUFFIX
+#undef GRAM_PREFIX
+}
+#undef FUNC_NAME
+
+
+/* There are a number of functions to consider here for Scheme and C:
+
+ string-copy STR [start [end]] ;; SRFI-13 variant of R5RS string-copy
+ substring/copy STR start [end] ;; Guile variant of R5RS substring
+
+ scm_string_copy (str) ;; Old function from Guile
+ scm_substring_copy (str, [start, [end]])
+ ;; C version of SRFI-13 string-copy
+ ;; and C version of substring/copy
+
+ The C function underlying string-copy is not exported to C
+ programs. scm_substring_copy is defined in strings.c as the
+ underlying function of substring/copy and allows an optional START
+ argument.
+*/
+
+SCM scm_srfi13_substring_copy (SCM str, SCM start, SCM end);
+
+SCM_DEFINE (scm_srfi13_substring_copy, "string-copy", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Return a freshly allocated copy of the string @var{str}. If\n"
+ "given, @var{start} and @var{end} delimit the portion of\n"
+ "@var{str} which is copied.")
+#define FUNC_NAME s_scm_srfi13_substring_copy
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return scm_c_substring_copy (str, cstart, cend);
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_copy (SCM str)
+{
+ return scm_c_substring (str, 0, scm_c_string_length (str));
+}
+
+SCM_DEFINE (scm_string_copy_x, "string-copy!", 3, 2, 0,
+ (SCM target, SCM tstart, SCM s, SCM start, SCM end),
+ "Copy the sequence of characters from index range [@var{start},\n"
+ "@var{end}) in string @var{s} to string @var{target}, beginning\n"
+ "at index @var{tstart}. The characters are copied left-to-right\n"
+ "or right-to-left as needed -- the copy is guaranteed to work,\n"
+ "even if @var{target} and @var{s} are the same string. It is an\n"
+ "error if the copy operation runs off the end of the target\n"
+ "string.")
+#define FUNC_NAME s_scm_string_copy_x
+{
+ const char *cstr;
+ char *ctarget;
+ size_t cstart, cend, ctstart, dummy, len;
+ SCM sdummy = SCM_UNDEFINED;
+
+ MY_VALIDATE_SUBSTRING_SPEC (1, target,
+ 2, tstart, ctstart,
+ 2, sdummy, dummy);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (3, s, cstr,
+ 4, start, cstart,
+ 5, end, cend);
+ len = cend - cstart;
+ SCM_ASSERT_RANGE (3, s, len <= scm_i_string_length (target) - ctstart);
+
+ ctarget = scm_i_string_writable_chars (target);
+ memmove (ctarget + ctstart, cstr + cstart, len);
+ scm_i_string_stop_writing ();
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_substring_move_x, "substring-move!", 5, 0, 0,
+ (SCM str1, SCM start1, SCM end1, SCM str2, SCM start2),
+ "Copy the substring of @var{str1} bounded by @var{start1} and @var{end1}\n"
+ "into @var{str2} beginning at position @var{start2}.\n"
+ "@var{str1} and @var{str2} can be the same string.")
+#define FUNC_NAME s_scm_substring_move_x
+{
+ return scm_string_copy_x (str2, start2, str1, start1, end1);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_take, "string-take", 2, 0, 0,
+ (SCM s, SCM n),
+ "Return the @var{n} first characters of @var{s}.")
+#define FUNC_NAME s_scm_string_take
+{
+ return scm_substring (s, SCM_INUM0, n);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_drop, "string-drop", 2, 0, 0,
+ (SCM s, SCM n),
+ "Return all but the first @var{n} characters of @var{s}.")
+#define FUNC_NAME s_scm_string_drop
+{
+ return scm_substring (s, n, SCM_UNDEFINED);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_take_right, "string-take-right", 2, 0, 0,
+ (SCM s, SCM n),
+ "Return the @var{n} last characters of @var{s}.")
+#define FUNC_NAME s_scm_string_take_right
+{
+ return scm_substring (s,
+ scm_difference (scm_string_length (s), n),
+ SCM_UNDEFINED);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_drop_right, "string-drop-right", 2, 0, 0,
+ (SCM s, SCM n),
+ "Return all but the last @var{n} characters of @var{s}.")
+#define FUNC_NAME s_scm_string_drop_right
+{
+ return scm_substring (s,
+ SCM_INUM0,
+ scm_difference (scm_string_length (s), n));
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_pad, "string-pad", 2, 3, 0,
+ (SCM s, SCM len, SCM chr, SCM start, SCM end),
+ "Take that characters from @var{start} to @var{end} from the\n"
+ "string @var{s} and return a new string, right-padded by the\n"
+ "character @var{chr} to length @var{len}. If the resulting\n"
+ "string is longer than @var{len}, it is truncated on the right.")
+#define FUNC_NAME s_scm_string_pad
+{
+ char cchr;
+ const char *cstr;
+ size_t cstart, cend, clen;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 4, start, cstart,
+ 5, end, cend);
+ clen = scm_to_size_t (len);
+
+ if (SCM_UNBNDP (chr))
+ cchr = ' ';
+ else
+ {
+ SCM_VALIDATE_CHAR (3, chr);
+ cchr = SCM_CHAR (chr);
+ }
+ if (clen < (cend - cstart))
+ return scm_c_substring (s, cend - clen, cend);
+ else
+ {
+ SCM result;
+ char *dst;
+
+ result = scm_i_make_string (clen, &dst);
+ memset (dst, cchr, (clen - (cend - cstart)));
+ memmove (dst + clen - (cend - cstart), cstr + cstart, cend - cstart);
+ return result;
+ }
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_pad_right, "string-pad-right", 2, 3, 0,
+ (SCM s, SCM len, SCM chr, SCM start, SCM end),
+ "Take that characters from @var{start} to @var{end} from the\n"
+ "string @var{s} and return a new string, left-padded by the\n"
+ "character @var{chr} to length @var{len}. If the resulting\n"
+ "string is longer than @var{len}, it is truncated on the left.")
+#define FUNC_NAME s_scm_string_pad_right
+{
+ char cchr;
+ const char *cstr;
+ size_t cstart, cend, clen;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 4, start, cstart,
+ 5, end, cend);
+ clen = scm_to_size_t (len);
+
+ if (SCM_UNBNDP (chr))
+ cchr = ' ';
+ else
+ {
+ SCM_VALIDATE_CHAR (3, chr);
+ cchr = SCM_CHAR (chr);
+ }
+ if (clen < (cend - cstart))
+ return scm_c_substring (s, cstart, cstart + clen);
+ else
+ {
+ SCM result;
+ char *dst;
+
+ result = scm_i_make_string (clen, &dst);
+ memset (dst + (cend - cstart), cchr, clen - (cend - cstart));
+ memmove (dst, cstr + cstart, cend - cstart);
+ return result;
+ }
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_trim, "string-trim", 1, 3, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Trim @var{s} by skipping over all characters on the left\n"
+ "that satisfy the parameter @var{char_pred}:\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "if it is the character @var{ch}, characters equal to\n"
+ "@var{ch} are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a procedure @var{pred} characters that\n"
+ "satisfy @var{pred} are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a character set, characters in that set are trimmed.\n"
+ "@end itemize\n"
+ "\n"
+ "If called without a @var{char_pred} argument, all whitespace is\n"
+ "trimmed.")
+#define FUNC_NAME s_scm_string_trim
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_UNBNDP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!isspace((int) (unsigned char) cstr[cstart]))
+ break;
+ cstart++;
+ }
+ }
+ else if (SCM_CHARP (char_pred))
+ {
+ char chr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (chr != cstr[cstart])
+ break;
+ cstart++;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[cstart]))
+ break;
+ cstart++;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cstart]));
+ if (scm_is_false (res))
+ break;
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ }
+ return scm_c_substring (s, cstart, cend);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_trim_right, "string-trim-right", 1, 3, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Trim @var{s} by skipping over all characters on the rightt\n"
+ "that satisfy the parameter @var{char_pred}:\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "if it is the character @var{ch}, characters equal to @var{ch}\n"
+ "are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a procedure @var{pred} characters that satisfy\n"
+ "@var{pred} are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a character sets, all characters in that set are\n"
+ "trimmed.\n"
+ "@end itemize\n"
+ "\n"
+ "If called without a @var{char_pred} argument, all whitespace is\n"
+ "trimmed.")
+#define FUNC_NAME s_scm_string_trim_right
+{
+ const char *cstr;
+ int cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_UNBNDP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!isspace((int) (unsigned char) cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ }
+ else if (SCM_CHARP (char_pred))
+ {
+ char chr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (chr != cstr[cend - 1])
+ break;
+ cend--;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cend - 1]));
+ if (scm_is_false (res))
+ break;
+ cstr = scm_i_string_chars (s);
+ cend--;
+ }
+ }
+ return scm_c_substring (s, cstart, cend);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_trim_both, "string-trim-both", 1, 3, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Trim @var{s} by skipping over all characters on both sides of\n"
+ "the string that satisfy the parameter @var{char_pred}:\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "if it is the character @var{ch}, characters equal to @var{ch}\n"
+ "are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a procedure @var{pred} characters that satisfy\n"
+ "@var{pred} are trimmed,\n"
+ "\n"
+ "@item\n"
+ "if it is a character set, the characters in the set are\n"
+ "trimmed.\n"
+ "@end itemize\n"
+ "\n"
+ "If called without a @var{char_pred} argument, all whitespace is\n"
+ "trimmed.")
+#define FUNC_NAME s_scm_string_trim_both
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_UNBNDP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!isspace((int) (unsigned char) cstr[cstart]))
+ break;
+ cstart++;
+ }
+ while (cstart < cend)
+ {
+ if (!isspace((int) (unsigned char) cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ }
+ else if (SCM_CHARP (char_pred))
+ {
+ char chr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (chr != cstr[cstart])
+ break;
+ cstart++;
+ }
+ while (cstart < cend)
+ {
+ if (chr != cstr[cend - 1])
+ break;
+ cend--;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[cstart]))
+ break;
+ cstart++;
+ }
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cstart]));
+ if (scm_is_false (res))
+ break;
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ while (cstart < cend)
+ {
+ SCM res;
+
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cend - 1]));
+ if (scm_is_false (res))
+ break;
+ cstr = scm_i_string_chars (s);
+ cend--;
+ }
+ }
+ return scm_c_substring (s, cstart, cend);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_substring_fill_x, "string-fill!", 2, 2, 0,
+ (SCM str, SCM chr, SCM start, SCM end),
+ "Stores @var{chr} in every element of the given @var{str} and\n"
+ "returns an unspecified value.")
+#define FUNC_NAME s_scm_substring_fill_x
+{
+ char *cstr;
+ size_t cstart, cend;
+ int c;
+ size_t k;
+
+ /* Older versions of Guile provided the function
+ scm_substring_fill_x with the following order of arguments:
+
+ str, start, end, chr
+
+ We accomodate this here by detecting such a usage and reordering
+ the arguments.
+ */
+ if (SCM_CHARP (end))
+ {
+ SCM tmp = end;
+ end = start;
+ start = chr;
+ chr = tmp;
+ }
+
+ MY_VALIDATE_SUBSTRING_SPEC (1, str,
+ 3, start, cstart,
+ 4, end, cend);
+ SCM_VALIDATE_CHAR_COPY (2, chr, c);
+
+ cstr = scm_i_string_writable_chars (str);
+ for (k = cstart; k < cend; k++)
+ cstr[k] = c;
+ scm_i_string_stop_writing ();
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_fill_x (SCM str, SCM chr)
+{
+ return scm_substring_fill_x (str, chr, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+SCM_DEFINE (scm_string_compare, "string-compare", 5, 4, 0,
+ (SCM s1, SCM s2, SCM proc_lt, SCM proc_eq, SCM proc_gt, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Apply @var{proc_lt}, @var{proc_eq}, @var{proc_gt} to the\n"
+ "mismatch index, depending upon whether @var{s1} is less than,\n"
+ "equal to, or greater than @var{s2}. The mismatch index is the\n"
+ "largest index @var{i} such that for every 0 <= @var{j} <\n"
+ "@var{i}, @var{s1}[@var{j}] = @var{s2}[@var{j}] -- that is,\n"
+ "@var{i} is the first position that does not match.")
+#define FUNC_NAME s_scm_string_compare
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 6, start1, cstart1,
+ 7, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 8, start2, cstart2,
+ 9, end2, cend2);
+ SCM_VALIDATE_PROC (3, proc_lt);
+ SCM_VALIDATE_PROC (4, proc_eq);
+ SCM_VALIDATE_PROC (5, proc_gt);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ return scm_call_1 (proc_lt, scm_from_size_t (cstart1));
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ return scm_call_1 (proc_gt, scm_from_size_t (cstart1));
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ return scm_call_1 (proc_gt, scm_from_size_t (cstart1));
+ else if (cstart2 < cend2)
+ return scm_call_1 (proc_lt, scm_from_size_t (cstart1));
+ else
+ return scm_call_1 (proc_eq, scm_from_size_t (cstart1));
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_compare_ci, "string-compare-ci", 5, 4, 0,
+ (SCM s1, SCM s2, SCM proc_lt, SCM proc_eq, SCM proc_gt, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Apply @var{proc_lt}, @var{proc_eq}, @var{proc_gt} to the\n"
+ "mismatch index, depending upon whether @var{s1} is less than,\n"
+ "equal to, or greater than @var{s2}. The mismatch index is the\n"
+ "largest index @var{i} such that for every 0 <= @var{j} <\n"
+ "@var{i}, @var{s1}[@var{j}] = @var{s2}[@var{j}] -- that is,\n"
+ "@var{i} is the first position that does not match. The\n"
+ "character comparison is done case-insensitively.")
+#define FUNC_NAME s_scm_string_compare_ci
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 6, start1, cstart1,
+ 7, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 8, start2, cstart2,
+ 9, end2, cend2);
+ SCM_VALIDATE_PROC (3, proc_lt);
+ SCM_VALIDATE_PROC (4, proc_eq);
+ SCM_VALIDATE_PROC (5, proc_gt);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ return scm_call_1 (proc_lt, scm_from_size_t (cstart1));
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ return scm_call_1 (proc_gt, scm_from_size_t (cstart1));
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ return scm_call_1 (proc_gt, scm_from_size_t (cstart1));
+ else if (cstart2 < cend2)
+ return scm_call_1 (proc_lt, scm_from_size_t (cstart1));
+ else
+ return scm_call_1 (proc_eq, scm_from_size_t (cstart1));
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_eq, "string=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} and @var{s2} are not equal, a true\n"
+ "value otherwise.")
+#define FUNC_NAME s_scm_string_eq
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ if ((cend1 - cstart1) != (cend2 - cstart2))
+ goto false;
+
+ while (cstart1 < cend1)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto false;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_neq, "string<>", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} and @var{s2} are equal, a true\n"
+ "value otherwise.")
+#define FUNC_NAME s_scm_string_neq
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto true;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_lt, "string<", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is greater or equal to @var{s2}, a\n"
+ "true value otherwise.")
+#define FUNC_NAME s_scm_string_lt
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto true;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto false;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_gt, "string>", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is less or equal to @var{s2}, a\n"
+ "true value otherwise.")
+#define FUNC_NAME s_scm_string_gt
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto false;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto false;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_le, "string<=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is greater to @var{s2}, a true\n"
+ "value otherwise.")
+#define FUNC_NAME s_scm_string_le
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto true;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto false;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto true;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ge, "string>=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is less to @var{s2}, a true value\n"
+ "otherwise.")
+#define FUNC_NAME s_scm_string_ge
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] < cstr2[cstart2])
+ goto false;
+ else if (cstr1[cstart1] > cstr2[cstart2])
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto false;
+ else
+ goto true;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_eq, "string-ci=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} and @var{s2} are not equal, a true\n"
+ "value otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_eq
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto false;
+ else if (cstart2 < cend2)
+ goto false;
+ else
+ goto true;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_neq, "string-ci<>", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} and @var{s2} are equal, a true\n"
+ "value otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_neq
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_lt, "string-ci<", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is greater or equal to @var{s2}, a\n"
+ "true value otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_lt
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto false;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_gt, "string-ci>", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is less or equal to @var{s2}, a\n"
+ "true value otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_gt
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto false;
+ else
+ goto false;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_le, "string-ci<=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is greater to @var{s2}, a true\n"
+ "value otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_le
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto false;
+ else if (cstart2 < cend2)
+ goto true;
+ else
+ goto true;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_ci_ge, "string-ci>=", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return @code{#f} if @var{s1} is less to @var{s2}, a true value\n"
+ "otherwise. The character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_ci_ge
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) < scm_c_downcase (cstr2[cstart2]))
+ goto false;
+ else if (scm_c_downcase (cstr1[cstart1]) > scm_c_downcase (cstr2[cstart2]))
+ goto true;
+ cstart1++;
+ cstart2++;
+ }
+ if (cstart1 < cend1)
+ goto true;
+ else if (cstart2 < cend2)
+ goto false;
+ else
+ goto true;
+
+ true:
+ scm_remember_upto_here_2 (s1, s2);
+ return scm_from_size_t (cstart1);
+
+ false:
+ scm_remember_upto_here_2 (s1, s2);
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_substring_hash, "string-hash", 1, 3, 0,
+ (SCM s, SCM bound, SCM start, SCM end),
+ "Compute a hash value for @var{S}. the optional argument "
+ "@var{bound} is a non-negative exact "
+ "integer specifying the range of the hash function. "
+ "A positive value restricts the return value to the "
+ "range [0,bound).")
+#define FUNC_NAME s_scm_substring_hash
+{
+ if (SCM_UNBNDP (bound))
+ bound = scm_from_intmax (SCM_MOST_POSITIVE_FIXNUM);
+ if (SCM_UNBNDP (start))
+ start = SCM_INUM0;
+ return scm_hash (scm_substring_shared (s, start, end), bound);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_substring_hash_ci, "string-hash-ci", 1, 3, 0,
+ (SCM s, SCM bound, SCM start, SCM end),
+ "Compute a hash value for @var{S}. the optional argument "
+ "@var{bound} is a non-negative exact "
+ "integer specifying the range of the hash function. "
+ "A positive value restricts the return value to the "
+ "range [0,bound).")
+#define FUNC_NAME s_scm_substring_hash_ci
+{
+ return scm_substring_hash (scm_substring_downcase (s, start, end),
+ bound,
+ SCM_UNDEFINED, SCM_UNDEFINED);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_prefix_length, "string-prefix-length", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return the length of the longest common prefix of the two\n"
+ "strings.")
+#define FUNC_NAME s_scm_string_prefix_length
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] != cstr2[cstart2])
+ return scm_from_size_t (len);
+ len++;
+ cstart1++;
+ cstart2++;
+ }
+ return scm_from_size_t (len);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_prefix_length_ci, "string-prefix-length-ci", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return the length of the longest common prefix of the two\n"
+ "strings, ignoring character case.")
+#define FUNC_NAME s_scm_string_prefix_length_ci
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) != scm_c_downcase (cstr2[cstart2]))
+ return scm_from_size_t (len);
+ len++;
+ cstart1++;
+ cstart2++;
+ }
+ return scm_from_size_t (len);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_suffix_length, "string-suffix-length", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return the length of the longest common suffix of the two\n"
+ "strings.")
+#define FUNC_NAME s_scm_string_suffix_length
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ cend1--;
+ cend2--;
+ if (cstr1[cend1] != cstr2[cend2])
+ return scm_from_size_t (len);
+ len++;
+ }
+ return scm_from_size_t (len);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_suffix_length_ci, "string-suffix-length-ci", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return the length of the longest common suffix of the two\n"
+ "strings, ignoring character case.")
+#define FUNC_NAME s_scm_string_suffix_length_ci
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ cend1--;
+ cend2--;
+ if (scm_c_downcase (cstr1[cend1]) != scm_c_downcase (cstr2[cend2]))
+ return scm_from_size_t (len);
+ len++;
+ }
+ return scm_from_size_t (len);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_prefix_p, "string-prefix?", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Is @var{s1} a prefix of @var{s2}?")
+#define FUNC_NAME s_scm_string_prefix_p
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0, len1;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len1 = cend1 - cstart1;
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (cstr1[cstart1] != cstr2[cstart2])
+ return scm_from_bool (len == len1);
+ len++;
+ cstart1++;
+ cstart2++;
+ }
+ return scm_from_bool (len == len1);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_prefix_ci_p, "string-prefix-ci?", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Is @var{s1} a prefix of @var{s2}, ignoring character case?")
+#define FUNC_NAME s_scm_string_prefix_ci_p
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0, len1;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len1 = cend1 - cstart1;
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ if (scm_c_downcase (cstr1[cstart1]) != scm_c_downcase (cstr2[cstart2]))
+ return scm_from_bool (len == len1);
+ len++;
+ cstart1++;
+ cstart2++;
+ }
+ return scm_from_bool (len == len1);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_suffix_p, "string-suffix?", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Is @var{s1} a suffix of @var{s2}?")
+#define FUNC_NAME s_scm_string_suffix_p
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0, len1;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len1 = cend1 - cstart1;
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ cend1--;
+ cend2--;
+ if (cstr1[cend1] != cstr2[cend2])
+ return scm_from_bool (len == len1);
+ len++;
+ }
+ return scm_from_bool (len == len1);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_suffix_ci_p, "string-suffix-ci?", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Is @var{s1} a suffix of @var{s2}, ignoring character case?")
+#define FUNC_NAME s_scm_string_suffix_ci_p
+{
+ const char *cstr1, *cstr2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len = 0, len1;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len1 = cend1 - cstart1;
+ while (cstart1 < cend1 && cstart2 < cend2)
+ {
+ cend1--;
+ cend2--;
+ if (scm_c_downcase (cstr1[cend1]) != scm_c_downcase (cstr2[cend2]))
+ return scm_from_bool (len == len1);
+ len++;
+ }
+ return scm_from_bool (len == len1);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_index, "string-index", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Search through the string @var{s} from left to right, returning\n"
+ "the index of the first occurence of a character which\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "equals @var{char_pred}, if it is character,\n"
+ "\n"
+ "@item\n"
+ "satisifies the predicate @var{char_pred}, if it is a procedure,\n"
+ "\n"
+ "@item\n"
+ "is in the set @var{char_pred}, if it is a character set.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_index
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (cchr == cstr[cstart])
+ return scm_from_size_t (cstart);
+ cstart++;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (SCM_CHARSET_GET (char_pred, cstr[cstart]))
+ return scm_from_size_t (cstart);
+ cstart++;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cstart]));
+ if (scm_is_true (res))
+ return scm_from_size_t (cstart);
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_index_right, "string-index-right", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Search through the string @var{s} from right to left, returning\n"
+ "the index of the last occurence of a character which\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "equals @var{char_pred}, if it is character,\n"
+ "\n"
+ "@item\n"
+ "satisifies the predicate @var{char_pred}, if it is a procedure,\n"
+ "\n"
+ "@item\n"
+ "is in the set if @var{char_pred} is a character set.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_index_right
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ cend--;
+ if (cchr == cstr[cend])
+ return scm_from_size_t (cend);
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ cend--;
+ if (SCM_CHARSET_GET (char_pred, cstr[cend]))
+ return scm_from_size_t (cend);
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+ cend--;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cend]));
+ if (scm_is_true (res))
+ return scm_from_size_t (cend);
+ cstr = scm_i_string_chars (s);
+ }
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_rindex (SCM str, SCM chr, SCM frm, SCM to)
+{
+ return scm_string_index_right (str, chr, frm, to);
+}
+
+SCM_DEFINE (scm_string_skip, "string-skip", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Search through the string @var{s} from left to right, returning\n"
+ "the index of the first occurence of a character which\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "does not equal @var{char_pred}, if it is character,\n"
+ "\n"
+ "@item\n"
+ "does not satisify the predicate @var{char_pred}, if it is a\n"
+ "procedure,\n"
+ "\n"
+ "@item\n"
+ "is not in the set if @var{char_pred} is a character set.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_skip
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (cchr != cstr[cstart])
+ return scm_from_size_t (cstart);
+ cstart++;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[cstart]))
+ return scm_from_size_t (cstart);
+ cstart++;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cstart]));
+ if (scm_is_false (res))
+ return scm_from_size_t (cstart);
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_skip_right, "string-skip-right", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Search through the string @var{s} from right to left, returning\n"
+ "the index of the last occurence of a character which\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "does not equal @var{char_pred}, if it is character,\n"
+ "\n"
+ "@item\n"
+ "does not satisfy the predicate @var{char_pred}, if it is a\n"
+ "procedure,\n"
+ "\n"
+ "@item\n"
+ "is not in the set if @var{char_pred} is a character set.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_skip_right
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ cend--;
+ if (cchr != cstr[cend])
+ return scm_from_size_t (cend);
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ cend--;
+ if (!SCM_CHARSET_GET (char_pred, cstr[cend]))
+ return scm_from_size_t (cend);
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+ cend--;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cend]));
+ if (scm_is_false (res))
+ return scm_from_size_t (cend);
+ cstr = scm_i_string_chars (s);
+ }
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_count, "string-count", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Return the count of the number of characters in the string\n"
+ "@var{s} which\n"
+ "\n"
+ "@itemize @bullet\n"
+ "@item\n"
+ "equals @var{char_pred}, if it is character,\n"
+ "\n"
+ "@item\n"
+ "satisifies the predicate @var{char_pred}, if it is a procedure.\n"
+ "\n"
+ "@item\n"
+ "is in the set @var{char_pred}, if it is a character set.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_count
+{
+ const char *cstr;
+ size_t cstart, cend;
+ size_t count = 0;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ char cchr = SCM_CHAR (char_pred);
+ while (cstart < cend)
+ {
+ if (cchr == cstr[cstart])
+ count++;
+ cstart++;
+ }
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ while (cstart < cend)
+ {
+ if (SCM_CHARSET_GET (char_pred, cstr[cstart]))
+ count++;
+ cstart++;
+ }
+ }
+ else
+ {
+ SCM_VALIDATE_PROC (2, char_pred);
+ while (cstart < cend)
+ {
+ SCM res;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[cstart]));
+ if (scm_is_true (res))
+ count++;
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ }
+ return scm_from_size_t (count);
+}
+#undef FUNC_NAME
+
+
+/* FIXME::martin: This should definitely get implemented more
+ efficiently -- maybe with Knuth-Morris-Pratt, like in the reference
+ implementation. */
+SCM_DEFINE (scm_string_contains, "string-contains", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Does string @var{s1} contain string @var{s2}? Return the index\n"
+ "in @var{s1} where @var{s2} occurs as a substring, or false.\n"
+ "The optional start/end indices restrict the operation to the\n"
+ "indicated substrings.")
+#define FUNC_NAME s_scm_string_contains
+{
+ const char *cs1, * cs2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len2, i, j;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cs1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cs2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len2 = cend2 - cstart2;
+ while (cstart1 <= cend1 - len2)
+ {
+ i = cstart1;
+ j = cstart2;
+ while (i < cend1 && j < cend2 && cs1[i] == cs2[j])
+ {
+ i++;
+ j++;
+ }
+ if (j == cend2)
+ return scm_from_size_t (cstart1);
+ cstart1++;
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+/* FIXME::martin: This should definitely get implemented more
+ efficiently -- maybe with Knuth-Morris-Pratt, like in the reference
+ implementation. */
+SCM_DEFINE (scm_string_contains_ci, "string-contains-ci", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Does string @var{s1} contain string @var{s2}? Return the index\n"
+ "in @var{s1} where @var{s2} occurs as a substring, or false.\n"
+ "The optional start/end indices restrict the operation to the\n"
+ "indicated substrings. Character comparison is done\n"
+ "case-insensitively.")
+#define FUNC_NAME s_scm_string_contains_ci
+{
+ const char *cs1, * cs2;
+ size_t cstart1, cend1, cstart2, cend2;
+ size_t len2, i, j;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cs1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cs2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ len2 = cend2 - cstart2;
+ while (cstart1 <= cend1 - len2)
+ {
+ i = cstart1;
+ j = cstart2;
+ while (i < cend1 && j < cend2 &&
+ scm_c_downcase (cs1[i]) == scm_c_downcase (cs2[j]))
+ {
+ i++;
+ j++;
+ }
+ if (j == cend2)
+ return scm_from_size_t (cstart1);
+ cstart1++;
+ }
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+
+/* Helper function for the string uppercase conversion functions.
+ * No argument checking is performed. */
+static SCM
+string_upcase_x (SCM v, int start, int end)
+{
+ size_t k;
+ char *dst;
+
+ dst = scm_i_string_writable_chars (v);
+ for (k = start; k < end; ++k)
+ dst[k] = scm_c_upcase (dst[k]);
+ scm_i_string_stop_writing ();
+
+ return v;
+}
+
+SCM_DEFINE (scm_substring_upcase_x, "string-upcase!", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Destructively upcase every character in @code{str}.\n"
+ "\n"
+ "@lisp\n"
+ "(string-upcase! y)\n"
+ "@result{} \"ARRDEFG\"\n"
+ "y\n"
+ "@result{} \"ARRDEFG\"\n"
+ "@end lisp")
+#define FUNC_NAME s_scm_substring_upcase_x
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_upcase_x (str, cstart, cend);
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_upcase_x (SCM str)
+{
+ return scm_substring_upcase_x (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+SCM_DEFINE (scm_substring_upcase, "string-upcase", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Upcase every character in @code{str}.")
+#define FUNC_NAME s_scm_substring_upcase
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_upcase_x (scm_string_copy (str), cstart, cend);
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_upcase (SCM str)
+{
+ return scm_substring_upcase (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+/* Helper function for the string lowercase conversion functions.
+ * No argument checking is performed. */
+static SCM
+string_downcase_x (SCM v, int start, int end)
+{
+ size_t k;
+ char *dst;
+
+ dst = scm_i_string_writable_chars (v);
+ for (k = start; k < end; ++k)
+ dst[k] = scm_c_downcase (dst[k]);
+ scm_i_string_stop_writing ();
+
+ return v;
+}
+
+SCM_DEFINE (scm_substring_downcase_x, "string-downcase!", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Destructively downcase every character in @var{str}.\n"
+ "\n"
+ "@lisp\n"
+ "y\n"
+ "@result{} \"ARRDEFG\"\n"
+ "(string-downcase! y)\n"
+ "@result{} \"arrdefg\"\n"
+ "y\n"
+ "@result{} \"arrdefg\"\n"
+ "@end lisp")
+#define FUNC_NAME s_scm_substring_downcase_x
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_downcase_x (str, cstart, cend);
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_downcase_x (SCM str)
+{
+ return scm_substring_downcase_x (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+SCM_DEFINE (scm_substring_downcase, "string-downcase", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Downcase every character in @var{str}.")
+#define FUNC_NAME s_scm_substring_downcase
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_downcase_x (scm_string_copy (str), cstart, cend);
+}
+#undef FUNC_NAME
+
+SCM
+scm_string_downcase (SCM str)
+{
+ return scm_substring_downcase (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+
+/* Helper function for the string capitalization functions.
+ * No argument checking is performed. */
+static SCM
+string_titlecase_x (SCM str, int start, int end)
+{
+ unsigned char *sz;
+ size_t i;
+ int in_word = 0;
+
+ sz = scm_i_string_writable_chars (str);
+ for(i = start; i < end; i++)
+ {
+ if (scm_is_true (scm_char_alphabetic_p (SCM_MAKE_CHAR (sz[i]))))
+ {
+ if (!in_word)
+ {
+ sz[i] = scm_c_upcase(sz[i]);
+ in_word = 1;
+ }
+ else
+ {
+ sz[i] = scm_c_downcase(sz[i]);
+ }
+ }
+ else
+ in_word = 0;
+ }
+ scm_i_string_stop_writing ();
+
+ return str;
+}
+
+
+SCM_DEFINE (scm_string_titlecase_x, "string-titlecase!", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Destructively titlecase every first character in a word in\n"
+ "@var{str}.")
+#define FUNC_NAME s_scm_string_titlecase_x
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_titlecase_x (str, cstart, cend);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_titlecase, "string-titlecase", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Titlecase every first character in a word in @var{str}.")
+#define FUNC_NAME s_scm_string_titlecase
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ return string_titlecase_x (scm_string_copy (str), cstart, cend);
+}
+#undef FUNC_NAME
+
+/* Old names, the functions.
+ */
+
+SCM_DEFINE (scm_string_capitalize_x, "string-capitalize!", 1, 0, 0,
+ (SCM str),
+ "Upcase the first character of every word in @var{str}\n"
+ "destructively and return @var{str}.\n"
+ "\n"
+ "@lisp\n"
+ "y @result{} \"hello world\"\n"
+ "(string-capitalize! y) @result{} \"Hello World\"\n"
+ "y @result{} \"Hello World\"\n"
+ "@end lisp")
+#define FUNC_NAME s_scm_string_capitalize_x
+{
+ return scm_string_titlecase_x (str, SCM_UNDEFINED, SCM_UNDEFINED);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_capitalize, "string-capitalize", 1, 0, 0,
+ (SCM str),
+ "Return a freshly allocated string with the characters in\n"
+ "@var{str}, where the first character of every word is\n"
+ "capitalized.")
+#define FUNC_NAME s_scm_string_capitalize
+{
+ return scm_string_capitalize_x (scm_string_copy (str));
+}
+#undef FUNC_NAME
+
+
+/* Reverse the portion of @var{str} between str[cstart] (including)
+ and str[cend] excluding. */
+static void
+string_reverse_x (char * str, int cstart, int cend)
+{
+ char tmp;
+
+ cend--;
+ while (cstart < cend)
+ {
+ tmp = str[cstart];
+ str[cstart] = str[cend];
+ str[cend] = tmp;
+ cstart++;
+ cend--;
+ }
+}
+
+
+SCM_DEFINE (scm_string_reverse, "string-reverse", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Reverse the string @var{str}. The optional arguments\n"
+ "@var{start} and @var{end} delimit the region of @var{str} to\n"
+ "operate on.")
+#define FUNC_NAME s_scm_string_reverse
+{
+ const char *cstr;
+ char *ctarget;
+ size_t cstart, cend;
+ SCM result;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, str, cstr,
+ 2, start, cstart,
+ 3, end, cend);
+ result = scm_string_copy (str);
+ ctarget = scm_i_string_writable_chars (result);
+ string_reverse_x (ctarget, cstart, cend);
+ scm_i_string_stop_writing ();
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_reverse_x, "string-reverse!", 1, 2, 0,
+ (SCM str, SCM start, SCM end),
+ "Reverse the string @var{str} in-place. The optional arguments\n"
+ "@var{start} and @var{end} delimit the region of @var{str} to\n"
+ "operate on. The return value is unspecified.")
+#define FUNC_NAME s_scm_string_reverse_x
+{
+ char *cstr;
+ size_t cstart, cend;
+
+ MY_VALIDATE_SUBSTRING_SPEC (1, str,
+ 2, start, cstart,
+ 3, end, cend);
+
+ cstr = scm_i_string_writable_chars (str);
+ string_reverse_x (cstr, cstart, cend);
+ scm_i_string_stop_writing ();
+
+ scm_remember_upto_here_1 (str);
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_append_shared, "string-append/shared", 0, 0, 1,
+ (SCM ls),
+ "Like @code{string-append}, but the result may share memory\n"
+ "with the argument strings.")
+#define FUNC_NAME s_scm_string_append_shared
+{
+ long i;
+
+ SCM_VALIDATE_REST_ARGUMENT (ls);
+
+ /* Optimize the one-argument case. */
+ i = scm_ilength (ls);
+ if (i == 1)
+ return SCM_CAR (ls);
+ else
+ return scm_string_append (ls);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_concatenate, "string-concatenate", 1, 0, 0,
+ (SCM ls),
+ "Append the elements of @var{ls} (which must be strings)\n"
+ "together into a single string. Guaranteed to return a freshly\n"
+ "allocated string.")
+#define FUNC_NAME s_scm_string_concatenate
+{
+ return scm_string_append (ls);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_concatenate_reverse, "string-concatenate-reverse", 1, 2, 0,
+ (SCM ls, SCM final_string, SCM end),
+ "Without optional arguments, this procedure is equivalent to\n"
+ "\n"
+ "@smalllisp\n"
+ "(string-concatenate (reverse ls))\n"
+ "@end smalllisp\n"
+ "\n"
+ "If the optional argument @var{final_string} is specified, it is\n"
+ "consed onto the beginning to @var{ls} before performing the\n"
+ "list-reverse and string-concatenate operations. If @var{end}\n"
+ "is given, only the characters of @var{final_string} up to index\n"
+ "@var{end} are used.\n"
+ "\n"
+ "Guaranteed to return a freshly allocated string.")
+#define FUNC_NAME s_scm_string_concatenate_reverse
+{
+ long strings;
+ SCM tmp, result;
+ size_t len = 0;
+ char * p;
+ size_t cend = 0;
+
+ /* Check the optional arguments and calculate the additional length
+ of the result string. */
+ if (!SCM_UNBNDP (final_string))
+ {
+ SCM_VALIDATE_STRING (2, final_string);
+ if (!SCM_UNBNDP (end))
+ {
+ cend = scm_to_unsigned_integer (end,
+ 0,
+ scm_i_string_length (final_string));
+ }
+ else
+ {
+ cend = scm_i_string_length (final_string);
+ }
+ len += cend;
+ }
+ strings = scm_ilength (ls);
+ /* Validate the string list. */
+ if (strings < 0)
+ SCM_WRONG_TYPE_ARG (1, ls);
+
+ /* Calculate the length of the result string. */
+ tmp = ls;
+ while (!SCM_NULLP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ SCM_VALIDATE_STRING (1, elt);
+ len += scm_i_string_length (elt);
+ tmp = SCM_CDR (tmp);
+ }
+
+ result = scm_i_make_string (len, &p);
+
+ p += len;
+
+ /* Construct the result string, possibly by using the optional final
+ string. */
+ if (!SCM_UNBNDP (final_string))
+ {
+ p -= cend;
+ memmove (p, scm_i_string_chars (final_string), cend);
+ }
+ tmp = ls;
+ while (!SCM_NULLP (tmp))
+ {
+ SCM elt = SCM_CAR (tmp);
+ p -= scm_i_string_length (elt);
+ memmove (p, scm_i_string_chars (elt),
+ scm_i_string_length (elt));
+ tmp = SCM_CDR (tmp);
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_concatenate_shared, "string-concatenate/shared", 1, 0, 0,
+ (SCM ls),
+ "Like @code{string-concatenate}, but the result may share memory\n"
+ "with the strings in the list @var{ls}.")
+#define FUNC_NAME s_scm_string_concatenate_shared
+{
+ return scm_string_append_shared (ls);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_concatenate_reverse_shared, "string-concatenate-reverse/shared", 1, 2, 0,
+ (SCM ls, SCM final_string, SCM end),
+ "Like @code{string-concatenate-reverse}, but the result may\n"
+ "share memory with the the strings in the @var{ls} arguments.")
+#define FUNC_NAME s_scm_string_concatenate_reverse_shared
+{
+ /* Just call the non-sharing version. */
+ return scm_string_concatenate_reverse (ls, final_string, end);
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_map, "string-map", 2, 2, 0,
+ (SCM proc, SCM s, SCM start, SCM end),
+ "@var{proc} is a char->char procedure, it is mapped over\n"
+ "@var{s}. The order in which the procedure is applied to the\n"
+ "string elements is not specified.")
+#define FUNC_NAME s_scm_string_map
+{
+ const char *cstr;
+ char *p;
+ size_t cstart, cend;
+ SCM result;
+
+ SCM_VALIDATE_PROC (1, proc);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ result = scm_i_make_string (cend - cstart, &p);
+ while (cstart < cend)
+ {
+ unsigned int c = (unsigned char) cstr[cstart];
+ SCM ch = scm_call_1 (proc, SCM_MAKE_CHAR (c));
+ if (!SCM_CHARP (ch))
+ SCM_MISC_ERROR ("procedure ~S returned non-char", scm_list_1 (proc));
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ *p++ = SCM_CHAR (ch);
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_map_x, "string-map!", 2, 2, 0,
+ (SCM proc, SCM s, SCM start, SCM end),
+ "@var{proc} is a char->char procedure, it is mapped over\n"
+ "@var{s}. The order in which the procedure is applied to the\n"
+ "string elements is not specified. The string @var{s} is\n"
+ "modified in-place, the return value is not specified.")
+#define FUNC_NAME s_scm_string_map_x
+{
+ size_t cstart, cend;
+
+ SCM_VALIDATE_PROC (1, proc);
+ MY_VALIDATE_SUBSTRING_SPEC (2, s,
+ 3, start, cstart,
+ 4, end, cend);
+ while (cstart < cend)
+ {
+ SCM ch = scm_call_1 (proc, scm_c_string_ref (s, cstart));
+ if (!SCM_CHARP (ch))
+ SCM_MISC_ERROR ("procedure ~S returned non-char", scm_list_1 (proc));
+ scm_c_string_set_x (s, cstart, ch);
+ cstart++;
+ }
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_fold, "string-fold", 3, 2, 0,
+ (SCM kons, SCM knil, SCM s, SCM start, SCM end),
+ "Fold @var{kons} over the characters of @var{s}, with @var{knil}\n"
+ "as the terminating element, from left to right. @var{kons}\n"
+ "must expect two arguments: The actual character and the last\n"
+ "result of @var{kons}' application.")
+#define FUNC_NAME s_scm_string_fold
+{
+ const char *cstr;
+ size_t cstart, cend;
+ SCM result;
+
+ SCM_VALIDATE_PROC (1, kons);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (3, s, cstr,
+ 4, start, cstart,
+ 5, end, cend);
+ result = knil;
+ while (cstart < cend)
+ {
+ unsigned int c = (unsigned char) cstr[cstart];
+ result = scm_call_2 (kons, SCM_MAKE_CHAR (c), result);
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_fold_right, "string-fold-right", 3, 2, 0,
+ (SCM kons, SCM knil, SCM s, SCM start, SCM end),
+ "Fold @var{kons} over the characters of @var{s}, with @var{knil}\n"
+ "as the terminating element, from right to left. @var{kons}\n"
+ "must expect two arguments: The actual character and the last\n"
+ "result of @var{kons}' application.")
+#define FUNC_NAME s_scm_string_fold_right
+{
+ const char *cstr;
+ size_t cstart, cend;
+ SCM result;
+
+ SCM_VALIDATE_PROC (1, kons);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (3, s, cstr,
+ 4, start, cstart,
+ 5, end, cend);
+ result = knil;
+ while (cstart < cend)
+ {
+ unsigned int c = (unsigned char) cstr[cend - 1];
+ result = scm_call_2 (kons, SCM_MAKE_CHAR (c), result);
+ cstr = scm_i_string_chars (s);
+ cend--;
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_unfold, "string-unfold", 4, 2, 0,
+ (SCM p, SCM f, SCM g, SCM seed, SCM base, SCM make_final),
+ "@itemize @bullet\n"
+ "@item @var{g} is used to generate a series of @emph{seed}\n"
+ "values from the initial @var{seed}: @var{seed}, (@var{g}\n"
+ "@var{seed}), (@var{g}^2 @var{seed}), (@var{g}^3 @var{seed}),\n"
+ "@dots{}\n"
+ "@item @var{p} tells us when to stop -- when it returns true\n"
+ "when applied to one of these seed values.\n"
+ "@item @var{f} maps each seed value to the corresponding\n"
+ "character in the result string. These chars are assembled\n"
+ "into the string in a left-to-right order.\n"
+ "@item @var{base} is the optional initial/leftmost portion\n"
+ "of the constructed string; it default to the empty\n"
+ "string.\n"
+ "@item @var{make_final} is applied to the terminal seed\n"
+ "value (on which @var{p} returns true) to produce\n"
+ "the final/rightmost portion of the constructed string.\n"
+ "It defaults to @code{(lambda (x) "")}.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_unfold
+{
+ SCM res, ans;
+
+ SCM_VALIDATE_PROC (1, p);
+ SCM_VALIDATE_PROC (2, f);
+ SCM_VALIDATE_PROC (3, g);
+ if (!SCM_UNBNDP (base))
+ {
+ SCM_VALIDATE_STRING (5, base);
+ ans = base;
+ }
+ else
+ ans = scm_i_make_string (0, NULL);
+ if (!SCM_UNBNDP (make_final))
+ SCM_VALIDATE_PROC (6, make_final);
+
+ res = scm_call_1 (p, seed);
+ while (scm_is_false (res))
+ {
+ SCM str;
+ char *ptr;
+ SCM ch = scm_call_1 (f, seed);
+ if (!SCM_CHARP (ch))
+ SCM_MISC_ERROR ("procedure ~S returned non-char", scm_list_1 (f));
+ str = scm_i_make_string (1, &ptr);
+ *ptr = SCM_CHAR (ch);
+
+ ans = scm_string_append (scm_list_2 (ans, str));
+ seed = scm_call_1 (g, seed);
+ res = scm_call_1 (p, seed);
+ }
+ if (!SCM_UNBNDP (make_final))
+ {
+ res = scm_call_1 (make_final, seed);
+ return scm_string_append (scm_list_2 (ans, res));
+ }
+ else
+ return ans;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_unfold_right, "string-unfold-right", 4, 2, 0,
+ (SCM p, SCM f, SCM g, SCM seed, SCM base, SCM make_final),
+ "@itemize @bullet\n"
+ "@item @var{g} is used to generate a series of @emph{seed}\n"
+ "values from the initial @var{seed}: @var{seed}, (@var{g}\n"
+ "@var{seed}), (@var{g}^2 @var{seed}), (@var{g}^3 @var{seed}),\n"
+ "@dots{}\n"
+ "@item @var{p} tells us when to stop -- when it returns true\n"
+ "when applied to one of these seed values.\n"
+ "@item @var{f} maps each seed value to the corresponding\n"
+ "character in the result string. These chars are assembled\n"
+ "into the string in a right-to-left order.\n"
+ "@item @var{base} is the optional initial/rightmost portion\n"
+ "of the constructed string; it default to the empty\n"
+ "string.\n"
+ "@item @var{make_final} is applied to the terminal seed\n"
+ "value (on which @var{p} returns true) to produce\n"
+ "the final/leftmost portion of the constructed string.\n"
+ "It defaults to @code{(lambda (x) "")}.\n"
+ "@end itemize")
+#define FUNC_NAME s_scm_string_unfold_right
+{
+ SCM res, ans;
+
+ SCM_VALIDATE_PROC (1, p);
+ SCM_VALIDATE_PROC (2, f);
+ SCM_VALIDATE_PROC (3, g);
+ if (!SCM_UNBNDP (base))
+ {
+ SCM_VALIDATE_STRING (5, base);
+ ans = base;
+ }
+ else
+ ans = scm_i_make_string (0, NULL);
+ if (!SCM_UNBNDP (make_final))
+ SCM_VALIDATE_PROC (6, make_final);
+
+ res = scm_call_1 (p, seed);
+ while (scm_is_false (res))
+ {
+ SCM str;
+ char *ptr;
+ SCM ch = scm_call_1 (f, seed);
+ if (!SCM_CHARP (ch))
+ SCM_MISC_ERROR ("procedure ~S returned non-char", scm_list_1 (f));
+ str = scm_i_make_string (1, &ptr);
+ *ptr = SCM_CHAR (ch);
+
+ ans = scm_string_append (scm_list_2 (str, ans));
+ seed = scm_call_1 (g, seed);
+ res = scm_call_1 (p, seed);
+ }
+ if (!SCM_UNBNDP (make_final))
+ {
+ res = scm_call_1 (make_final, seed);
+ return scm_string_append (scm_list_2 (res, ans));
+ }
+ else
+ return ans;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_for_each, "string-for-each", 2, 2, 0,
+ (SCM proc, SCM s, SCM start, SCM end),
+ "@var{proc} is mapped over @var{s} in left-to-right order. The\n"
+ "return value is not specified.")
+#define FUNC_NAME s_scm_string_for_each
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ SCM_VALIDATE_PROC (1, proc);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ while (cstart < cend)
+ {
+ unsigned int c = (unsigned char) cstr[cstart];
+ scm_call_1 (proc, SCM_MAKE_CHAR (c));
+ cstr = scm_i_string_chars (s);
+ cstart++;
+ }
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_for_each_index, "string-for-each-index", 2, 2, 0,
+ (SCM proc, SCM s, SCM start, SCM end),
+ "@var{proc} is mapped over @var{s} in left-to-right order. The\n"
+ "return value is not specified.")
+#define FUNC_NAME s_scm_string_for_each_index
+{
+ const char *cstr;
+ size_t cstart, cend;
+
+ SCM_VALIDATE_PROC (1, proc);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ while (cstart < cend)
+ {
+ scm_call_1 (proc, scm_from_size_t (cstart));
+ cstart++;
+ }
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_xsubstring, "xsubstring", 2, 3, 0,
+ (SCM s, SCM from, SCM to, SCM start, SCM end),
+ "This is the @emph{extended substring} procedure that implements\n"
+ "replicated copying of a substring of some string.\n"
+ "\n"
+ "@var{s} is a string, @var{start} and @var{end} are optional\n"
+ "arguments that demarcate a substring of @var{s}, defaulting to\n"
+ "0 and the length of @var{s}. Replicate this substring up and\n"
+ "down index space, in both the positive and negative directions.\n"
+ "@code{xsubstring} returns the substring of this string\n"
+ "beginning at index @var{from}, and ending at @var{to}, which\n"
+ "defaults to @var{from} + (@var{end} - @var{start}).")
+#define FUNC_NAME s_scm_xsubstring
+{
+ const char *cs;
+ char *p;
+ size_t cstart, cend, cfrom, cto;
+ SCM result;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cs,
+ 4, start, cstart,
+ 5, end, cend);
+ cfrom = scm_to_size_t (from);
+ if (SCM_UNBNDP (to))
+ cto = cfrom + (cend - cstart);
+ else
+ cto = scm_to_size_t (to);
+ if (cstart == cend && cfrom != cto)
+ SCM_MISC_ERROR ("start and end indices must not be equal", SCM_EOL);
+
+ result = scm_i_make_string (cto - cfrom, &p);
+
+ while (cfrom < cto)
+ {
+ int t = ((cfrom < 0) ? -cfrom : cfrom) % (cend - cstart);
+ if (cfrom < 0)
+ *p = cs[(cend - cstart) - t];
+ else
+ *p = cs[t];
+ cfrom++;
+ p++;
+ }
+ scm_remember_upto_here_1 (s);
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_xcopy_x, "string-xcopy!", 4, 3, 0,
+ (SCM target, SCM tstart, SCM s, SCM sfrom, SCM sto, SCM start, SCM end),
+ "Exactly the same as @code{xsubstring}, but the extracted text\n"
+ "is written into the string @var{target} starting at index\n"
+ "@var{tstart}. The operation is not defined if @code{(eq?\n"
+ "@var{target} @var{s})} or these arguments share storage -- you\n"
+ "cannot copy a string on top of itself.")
+#define FUNC_NAME s_scm_string_xcopy_x
+{
+ char *p;
+ const char *cs;
+ size_t ctstart, csfrom, csto, cstart, cend;
+ SCM dummy = SCM_UNDEFINED;
+ int cdummy;
+
+ MY_VALIDATE_SUBSTRING_SPEC (1, target,
+ 2, tstart, ctstart,
+ 2, dummy, cdummy);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (3, s, cs,
+ 6, start, cstart,
+ 7, end, cend);
+ csfrom = scm_to_size_t (sfrom);
+ if (SCM_UNBNDP (sto))
+ csto = csfrom + (cend - cstart);
+ else
+ csto = scm_to_size_t (sto);
+ if (cstart == cend && csfrom != csto)
+ SCM_MISC_ERROR ("start and end indices must not be equal", SCM_EOL);
+ SCM_ASSERT_RANGE (1, tstart,
+ ctstart + (csto - csfrom) <= scm_i_string_length (target));
+
+ p = scm_i_string_writable_chars (target) + ctstart;
+ while (csfrom < csto)
+ {
+ int t = ((csfrom < 0) ? -csfrom : csfrom) % (cend - cstart);
+ if (csfrom < 0)
+ *p = cs[(cend - cstart) - t];
+ else
+ *p = cs[t];
+ csfrom++;
+ p++;
+ }
+ scm_i_string_stop_writing ();
+
+ scm_remember_upto_here_2 (target, s);
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_replace, "string-replace", 2, 4, 0,
+ (SCM s1, SCM s2, SCM start1, SCM end1, SCM start2, SCM end2),
+ "Return the string @var{s1}, but with the characters\n"
+ "@var{start1} @dots{} @var{end1} replaced by the characters\n"
+ "@var{start2} @dots{} @var{end2} from @var{s2}.")
+#define FUNC_NAME s_scm_string_replace
+{
+ const char *cstr1, *cstr2;
+ char *p;
+ size_t cstart1, cend1, cstart2, cend2;
+ SCM result;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s1, cstr1,
+ 3, start1, cstart1,
+ 4, end1, cend1);
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (2, s2, cstr2,
+ 5, start2, cstart2,
+ 6, end2, cend2);
+ result = scm_i_make_string (cstart1 + (cend2 - cstart2) +
+ scm_i_string_length (s1) - cend1, &p);
+ memmove (p, cstr1, cstart1 * sizeof (char));
+ memmove (p + cstart1, cstr2 + cstart2, (cend2 - cstart2) * sizeof (char));
+ memmove (p + cstart1 + (cend2 - cstart2),
+ cstr1 + cend1,
+ (scm_i_string_length (s1) - cend1) * sizeof (char));
+ scm_remember_upto_here_2 (s1, s2);
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_tokenize, "string-tokenize", 1, 3, 0,
+ (SCM s, SCM token_set, SCM start, SCM end),
+ "Split the string @var{s} into a list of substrings, where each\n"
+ "substring is a maximal non-empty contiguous sequence of\n"
+ "characters from the character set @var{token_set}, which\n"
+ "defaults to @code{char-set:graphic}.\n"
+ "If @var{start} or @var{end} indices are provided, they restrict\n"
+ "@code{string-tokenize} to operating on the indicated substring\n"
+ "of @var{s}.")
+#define FUNC_NAME s_scm_string_tokenize
+{
+ const char *cstr;
+ size_t cstart, cend;
+ SCM result = SCM_EOL;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+
+ if (SCM_UNBNDP (token_set))
+ token_set = scm_char_set_graphic;
+
+ if (SCM_CHARSETP (token_set))
+ {
+ int idx;
+
+ while (cstart < cend)
+ {
+ while (cstart < cend)
+ {
+ if (SCM_CHARSET_GET (token_set, cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ if (cstart >= cend)
+ break;
+ idx = cend;
+ while (cstart < cend)
+ {
+ if (!SCM_CHARSET_GET (token_set, cstr[cend - 1]))
+ break;
+ cend--;
+ }
+ result = scm_cons (scm_c_substring (s, cend, idx), result);
+ cstr = scm_i_string_chars (s);
+ }
+ }
+ else SCM_WRONG_TYPE_ARG (2, token_set);
+ scm_remember_upto_here_1 (s);
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_split, "string-split", 2, 0, 0,
+ (SCM str, SCM chr),
+ "Split the string @var{str} into the a list of the substrings delimited\n"
+ "by appearances of the character @var{chr}. Note that an empty substring\n"
+ "between separator characters will result in an empty string in the\n"
+ "result list.\n"
+ "\n"
+ "@lisp\n"
+ "(string-split \"root:x:0:0:root:/root:/bin/bash\" #\\:)\n"
+ "@result{}\n"
+ "(\"root\" \"x\" \"0\" \"0\" \"root\" \"/root\" \"/bin/bash\")\n"
+ "\n"
+ "(string-split \"::\" #\\:)\n"
+ "@result{}\n"
+ "(\"\" \"\" \"\")\n"
+ "\n"
+ "(string-split \"\" #\\:)\n"
+ "@result{}\n"
+ "(\"\")\n"
+ "@end lisp")
+#define FUNC_NAME s_scm_string_split
+{
+ long idx, last_idx;
+ const char * p;
+ int ch;
+ SCM res = SCM_EOL;
+
+ SCM_VALIDATE_STRING (1, str);
+ SCM_VALIDATE_CHAR (2, chr);
+
+ idx = scm_i_string_length (str);
+ p = scm_i_string_chars (str);
+ ch = SCM_CHAR (chr);
+ while (idx >= 0)
+ {
+ last_idx = idx;
+ while (idx > 0 && p[idx - 1] != ch)
+ idx--;
+ if (idx >= 0)
+ {
+ res = scm_cons (scm_c_substring (str, idx, last_idx), res);
+ p = scm_i_string_chars (str);
+ idx--;
+ }
+ }
+ scm_remember_upto_here_1 (str);
+ return res;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_filter, "string-filter", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Filter the string @var{s}, retaining only those characters that\n"
+ "satisfy the @var{char_pred} argument. If the argument is a\n"
+ "procedure, it is applied to each character as a predicate, if\n"
+ "it is a character, it is tested for equality and if it is a\n"
+ "character set, it is tested for membership.")
+#define FUNC_NAME s_scm_string_filter
+{
+ const char *cstr;
+ size_t cstart, cend;
+ SCM result;
+ size_t idx;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ SCM ls = SCM_EOL;
+ char chr;
+
+ chr = SCM_CHAR (char_pred);
+ idx = cstart;
+ while (idx < cend)
+ {
+ if (cstr[idx] == chr)
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ SCM ls = SCM_EOL;
+
+ idx = cstart;
+ while (idx < cend)
+ {
+ if (SCM_CHARSET_GET (char_pred, cstr[idx]))
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ else
+ {
+ SCM ls = SCM_EOL;
+
+ SCM_VALIDATE_PROC (2, char_pred);
+ idx = cstart;
+ while (idx < cend)
+ {
+ SCM res;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[idx]));
+ if (scm_is_true (res))
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ scm_remember_upto_here_1 (s);
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_string_delete, "string-delete", 2, 2, 0,
+ (SCM s, SCM char_pred, SCM start, SCM end),
+ "Filter the string @var{s}, retaining only those characters that\n"
+ "do not satisfy the @var{char_pred} argument. If the argument\n"
+ "is a procedure, it is applied to each character as a predicate,\n"
+ "if it is a character, it is tested for equality and if it is a\n"
+ "character set, it is tested for membership.")
+#define FUNC_NAME s_scm_string_delete
+{
+ const char *cstr;
+ size_t cstart, cend;
+ SCM result;
+ size_t idx;
+
+ MY_VALIDATE_SUBSTRING_SPEC_COPY (1, s, cstr,
+ 3, start, cstart,
+ 4, end, cend);
+ if (SCM_CHARP (char_pred))
+ {
+ SCM ls = SCM_EOL;
+ char chr;
+
+ chr = SCM_CHAR (char_pred);
+ idx = cstart;
+ while (idx < cend)
+ {
+ if (cstr[idx] != chr)
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ else if (SCM_CHARSETP (char_pred))
+ {
+ SCM ls = SCM_EOL;
+
+ idx = cstart;
+ while (idx < cend)
+ {
+ if (!SCM_CHARSET_GET (char_pred, cstr[idx]))
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ else
+ {
+ SCM ls = SCM_EOL;
+
+ SCM_VALIDATE_PROC (2, char_pred);
+ idx = cstart;
+ while (idx < cend)
+ {
+ SCM res;
+ res = scm_call_1 (char_pred, SCM_MAKE_CHAR (cstr[idx]));
+ if (scm_is_false (res))
+ ls = scm_cons (SCM_MAKE_CHAR (cstr[idx]), ls);
+ cstr = scm_i_string_chars (s);
+ idx++;
+ }
+ result = scm_reverse_list_to_string (ls);
+ }
+ return result;
+}
+#undef FUNC_NAME
+
+
+/* Initialize the SRFI-13 module. This function will be called by the
+ loading Scheme module. */
+void
+scm_init_srfi_13 (void)
+{
+#include "libguile/srfi-13.x"
+}
+
+/* End of srfi-13.c. */