add multibyte equivalent of strcasestr(), plus some miscellaneous

cleanups

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@2293 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

diff --git a/ChangeLog b/ChangeLog
index 7c64b4a37db5084a82c234483c95aa45b7d8bfec..eaaf9b1f7acd9b212d95b73e83caf7406267ed96 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -97,8 +97,8 @@ CVS code -
          is_alnum_mbchar(), is_alnum_wchar(), is_blank_mbchar(),
          is_blank_wchar(), is_cntrl_mbchar(), is_cntrl_wchar(),
          control_mbrep(), control_wrep(), mbwidth(), mb_cur_max(),
-         make_mbchar(), mbstrnlen(), mbstrcasecmp(), and
-         mbstrncasecmp(); changes to is_byte() (moved to chars.c),
+         make_mbchar(), mbstrnlen(), mbstrcasecmp(), mbstrncasecmp(),
+         and mbstrcasestr(); changes to is_byte() (moved to chars.c),
          is_blank_char() (moved to chars.c), is_cntrl_char() (moved to
          chars.c), nstricmp() (renamed nstrcasecmp() and moved to
          chars.c), nstrnicmp() (renamed nstrncasecmp() and moved to
@@ -110,8 +110,8 @@ CVS code -
          moved to chars.c), move_right() (renamed move_mbright() and
          moved to chars.c), do_home(), do_verbatim_input(),
          do_delete(), do_tab(), do_next_word(), do_prev_word(),
-         do_input(), do_output(), get_buffer(), unget_input(),
-         unget_kbinput(), get_input(), parse_kbinput(),
+         do_input(), do_output(), strstrwrapper(), get_buffer(),
+         unget_input(), unget_kbinput(), get_input(), parse_kbinput(),
          unparse_kbinput(), parse_verbatim_kbinput(),
          do_statusbar_input(), do_statusbar_home(),
          do_statusbar_verbatim_kbinput(), do_statusbar_output(), and
@@ -137,7 +137,9 @@ CVS code -
        - Use void instead of RETSIGTYPE, as signal handlers are
          supposed to return void anyway.  Also, the value of RETSIGTYPE
          is sometimes misdetected as int, leading to compilation
-         warnings or errors. (David Benbennick)
+         warnings or errors.  Changes to cancel_fork(),
+         handle_hipterm(), do_suspend(), and do_cont(). (David
+         Benbennick)
 - cut.c:
   do_cut_text()
        - If keep_cutbuffer is FALSE, only blow away the text in the

diff --git a/src/chars.c b/src/chars.c
index f1dc503472956bee14ae3a2122e6fb26aa8fe943..a68f8fada6b79f58d903746a1acc3d97cb7bebca 100644 (file)
--- a/src/chars.c
+++ b/src/chars.c
@@ -30,13 +30,14 @@
 #include "proto.h"
 #include "nano.h"
 
-#if defined(HAVE_WCHAR_H) && defined(NANO_WIDE)
+#ifdef NANO_WIDE
+#ifdef HAVE_WCHAR_H
 #include <wchar.h>
 #endif
-
-#if defined(HAVE_WCTYPE_H) && defined(NANO_WIDE)
+#ifdef HAVE_WCTYPE_H
 #include <wctype.h>
 #endif
+#endif
 
 /* Return TRUE if the value of c is in byte range, and FALSE
  * otherwise. */
@@ -262,12 +263,11 @@ int mbwidth(const char *c)
 /* Return the maximum width in bytes of a multibyte character. */
 int mb_cur_max(void)
 {
+    return
 #ifdef NANO_WIDE
-    if (!ISSET(NO_UTF8))
-       return MB_CUR_MAX;
-    else
+       !ISSET(NO_UTF8) ? MB_CUR_MAX :
 #endif
-       return 1;
+       1;
 }
 
 /* Convert the value in chr to a multibyte character with the same
@@ -348,7 +348,11 @@ int parse_mbchar(const char *buf, char *chr
             * and the width in columns of its visible equivalent as
             * returned by control_rep(). */
            else if (is_cntrl_mbchar(buf)) {
-               char *ctrl_buf_mb = charalloc(mb_cur_max());
+               char *ctrl_buf_mb =
+#ifdef NANO_WIDE
+                       !ISSET(NO_UTF8) ? charalloc(MB_CUR_MAX) :
+#endif
+                       charalloc(1);
                int ctrl_buf_mb_len;
 
                (*col)++;
@@ -452,19 +456,16 @@ int mbstrcasecmp(const char *s1, const char *s2)
 {
 #ifdef NANO_WIDE
     if (!ISSET(NO_UTF8)) {
-       char *s1_mb = charalloc(mb_cur_max());
-       char *s2_mb = charalloc(mb_cur_max());
-       int s1_mb_len, s2_mb_len;
+       char *s1_mb = charalloc(MB_CUR_MAX);
+       char *s2_mb = charalloc(MB_CUR_MAX);
        wchar_t ws1, ws2;
 
        assert(s1 != NULL && s2 != NULL);
 
        while (*s1 != '\0' && *s2 != '\0') {
-           s1_mb_len = parse_mbchar(s1, s1_mb
-#ifdef NANO_WIDE
-               , NULL
-#endif
-               , NULL);
+           int s1_mb_len, s2_mb_len;
+
+           s1_mb_len = parse_mbchar(s1, s1_mb, NULL, NULL);
 
            if (mbtowc(&ws1, s1_mb, s1_mb_len) <= 0) {
                mbtowc(NULL, NULL, 0);
@@ -472,11 +473,7 @@ int mbstrcasecmp(const char *s1, const char *s2)
            }
 
 
-           s2_mb_len = parse_mbchar(s2, s2_mb
-#ifdef NANO_WIDE
-               , NULL
-#endif
-               , NULL);
+           s2_mb_len = parse_mbchar(s2, s2_mb, NULL, NULL);
 
            if (mbtowc(&ws2, s2_mb, s2_mb_len) <= 0) {
                mbtowc(NULL, NULL, 0);
@@ -529,30 +526,23 @@ int mbstrncasecmp(const char *s1, const char *s2, size_t n)
 {
 #ifdef NANO_WIDE
     if (!ISSET(NO_UTF8)) {
-       char *s1_mb = charalloc(mb_cur_max());
-       char *s2_mb = charalloc(mb_cur_max());
-       int s1_mb_len, s2_mb_len;
+       char *s1_mb = charalloc(MB_CUR_MAX);
+       char *s2_mb = charalloc(MB_CUR_MAX);
        wchar_t ws1, ws2;
 
        assert(s1 != NULL && s2 != NULL);
 
        while (n > 0 && *s1 != '\0' && *s2 != '\0') {
-           s1_mb_len = parse_mbchar(s1, s1_mb
-#ifdef NANO_WIDE
-               , NULL
-#endif
-               , NULL);
+           int s1_mb_len, s2_mb_len;
+
+           s1_mb_len = parse_mbchar(s1, s1_mb, NULL, NULL);
 
            if (mbtowc(&ws1, s1_mb, s1_mb_len) <= 0) {
                mbtowc(NULL, NULL, 0);
                ws1 = (unsigned char)*s1_mb;
            }
 
-           s2_mb_len = parse_mbchar(s2, s2_mb
-#ifdef NANO_WIDE
-               , NULL
-#endif
-               , NULL);
+           s2_mb_len = parse_mbchar(s2, s2_mb, NULL, NULL);
 
            if (mbtowc(&ws2, s2_mb, s2_mb_len) <= 0) {
                mbtowc(NULL, NULL, 0);
@@ -602,6 +592,66 @@ const char *nstrcasestr(const char *haystack, const char *needle)
 }
 #endif
 
+/* This function is equivalent to strcasestr() for multibyte strings. */
+const char *mbstrcasestr(const char *haystack, const char *needle)
+{
+#ifdef NANO_WIDE
+    if (!ISSET(NO_UTF8)) {
+       char *p_mb = charalloc(MB_CUR_MAX);
+       char *q_mb = charalloc(MB_CUR_MAX);
+       wchar_t wp, wq;
+       bool found_needle = FALSE;
+
+       assert(haystack != NULL && needle != NULL);
+
+       while (*haystack != '\0') {
+           const char *p = haystack, *q = needle;
+           int p_mb_len, q_mb_len;
+
+           while (*q != '\0') {
+               p_mb_len = parse_mbchar(p, p_mb, NULL, NULL);
+
+               if (mbtowc(&wp, p_mb, p_mb_len) <= 0) {
+                   mbtowc(NULL, NULL, 0);
+                   wp = (unsigned char)*p;
+               }
+
+               q_mb_len = parse_mbchar(q, q_mb, NULL, NULL);
+
+               if (mbtowc(&wq, q_mb, q_mb_len) <= 0) {
+                   mbtowc(NULL, NULL, 0);
+                   wq = (unsigned char)*q;
+               }
+
+               if (towlower(wp) != towlower(wq))
+                   break;
+
+               p += p_mb_len;
+               q += q_mb_len;
+           }
+
+           if (*q == '\0') {
+               found_needle = TRUE;
+               break;
+           }
+
+           haystack += parse_mbchar(haystack, NULL, NULL, NULL);
+       }
+
+       free(p_mb);
+       free(q_mb);
+
+       return found_needle ? haystack : NULL;
+    } else
+#endif
+       return
+#ifdef HAVE_STRCASESTR
+               strcasestr(haystack, needle);
+#else
+               nstrcasestr(haystack, needle);
+#endif
+}
+
 #ifndef NANO_SMALL
 /* This function is equivalent to strstr(), except in that it scans the
  * string in reverse. */
@@ -667,15 +717,11 @@ size_t mbstrnlen(const char *s, size_t maxlen)
 #ifdef NANO_WIDE
     if (!ISSET(NO_UTF8)) {
        size_t n = 0;
-       char *s_mb = charalloc(mb_cur_max());
+       char *s_mb = charalloc(MB_CUR_MAX);
        int s_mb_len;
 
        while (*s != '\0') {
-           s_mb_len = parse_mbchar(s + n, s_mb
-#ifdef NANO_WIDE
-               , NULL
-#endif
-               , NULL);
+           s_mb_len = parse_mbchar(s + n, s_mb, NULL, NULL);
 
            if (maxlen == 0)
                break;

diff --git a/src/nano.h b/src/nano.h
index b8814d5ab85f970cbbe36a281d7364736ed7db05..345839a6be406910bf6d39cae90ac8805606555d 100644 (file)
--- a/src/nano.h
+++ b/src/nano.h
@@ -100,16 +100,10 @@
 #endif
 #endif
 
-/* If no strcasestr(), getdelim(), or getline(), use the versions we
- * have. */
-#ifndef HAVE_STRCASESTR
-#define strcasestr nstrcasestr
-#endif
-
+/* If no getdelim() or getline(), use the versions we have. */
 #ifndef HAVE_GETDELIM
 #define getdelim ngetdelim
 #endif
-
 #ifndef HAVE_GETLINE
 #define getline ngetline
 #endif

diff --git a/src/proto.h b/src/proto.h
index 0b1581e90bffdd350b2342e35941c42b66c7eea2..dd07858013c13c62348be6ec866584cba829adb3 100644 (file)
--- a/src/proto.h
+++ b/src/proto.h
@@ -193,6 +193,7 @@ int mbstrncasecmp(const char *s1, const char *s2, size_t n);
 #ifndef HAVE_STRCASESTR
 const char *nstrcasestr(const char *haystack, const char *needle);
 #endif
+const char *mbstrcasestr(const char *haystack, const char *needle);
 #ifndef NANO_SMALL
 const char *revstrstr(const char *haystack, const char *needle, const
        char *rev_start);

diff --git a/src/utils.c b/src/utils.c
index 9f7d2005b6e875c1e419c4807a18218c93f4609c..cbd678207a2f957ae8a7c52a8e0f8bccecda7165 100644 (file)
--- a/src/utils.c
+++ b/src/utils.c
@@ -245,7 +245,7 @@ const char *strstrwrapper(const char *haystack, const char *needle,
     else if (ISSET(REVERSE_SEARCH))
        return revstrcasestr(haystack, needle, start);
 #endif
-    return strcasestr(start, needle);
+    return mbstrcasestr(start, needle);
 }
 
 /* This is a wrapper for the perror() function.  The wrapper takes care