Generalized utf8_icmp_n() and added cstr_icmp(), csview_icmp().

4 files changed, 20 insertions, 9 deletions

diff --git a/include/stc/cstr.h b/include/stc/cstr.h
index 273ad70b..7bbd9f3b 100644
--- a/include/stc/cstr.h
+++ b/include/stc/cstr.h
@@ -240,6 +240,11 @@ STC_INLINE char* cstr_expand_uninit(cstr *self, size_t n) {
 STC_INLINE int cstr_cmp(const cstr* s1, const cstr* s2) 

     { return strcmp(cstr_str(s1), cstr_str(s2)); }

 

+STC_INLINE int cstr_icmp(const cstr* s1, const cstr* s2) {

+    csview x = cstr_sv(s1), y = cstr_sv(s2);

+    return utf8_icmp_n(~(size_t)0, x.str, x.size, y.str, y.size);

+}

+

 STC_INLINE bool cstr_eq(const cstr* s1, const cstr* s2) {

     csview x = cstr_sv(s1), y = cstr_sv(s2);

     return x.size == y.size && !memcmp(x.str, y.str, x.size);

diff --git a/include/stc/csview.h b/include/stc/csview.h
index eebcb9b7..8508db4f 100644
--- a/include/stc/csview.h
+++ b/include/stc/csview.h
@@ -165,6 +165,9 @@ STC_INLINE bool cstr_ends_with_sv(cstr s, csview sub) {
 STC_INLINE int csview_cmp(const csview* x, const csview* y) 

     { return strcmp(x->str, y->str); }

 

+STC_INLINE int csview_icmp(const csview* x, const csview* y)

+    { return utf8_icmp_n(~(size_t)0, x->str, x->size, y->str, y->size); }

+

 STC_INLINE bool csview_eq(const csview* x, const csview* y)

     { return x->size == y->size && !memcmp(x->str, y->str, x->size); }

 

diff --git a/include/stc/utf8.h b/include/stc/utf8.h
index 333a0f92..a3647e3c 100644
--- a/include/stc/utf8.h
+++ b/include/stc/utf8.h
@@ -38,7 +38,8 @@ uint32_t    utf8_toupper(uint32_t c);
 

 bool        utf8_valid(const char* s);

 bool        utf8_valid_n(const char* s, size_t n);

-int         utf8_icmp_n(const char* s1, const char* s2, size_t u8max);

+int         utf8_icmp_n(size_t u8max, const char* s1, size_t n1,

+                                      const char* s2, size_t n2);

 

 /* encode/decode next utf8 codepoint. */

 enum { UTF8_OK = 0, UTF8_ERROR = 4 };

@@ -50,7 +51,7 @@ void        utf8_decode(utf8_decode_t *d, const uint8_t b);
 

 /* case-insensitive utf8 string comparison */

 STC_INLINE int utf8_icmp(const char* s1, const char* s2)

-    { return utf8_icmp_n(s1, s2, (size_t)(-1)); }

+    { return utf8_icmp_n(~(size_t)0, s1, ~(size_t)0, s2, ~(size_t)0); }

 

 /* number of characters in the utf8 codepoint from s */

 STC_INLINE unsigned utf8_codep_size(const char *s) {

diff --git a/src/utf8code.c b/src/utf8code.c
index cef3efca..2f429541 100644
--- a/src/utf8code.c
+++ b/src/utf8code.c
@@ -100,17 +100,19 @@ uint32_t utf8_toupper(uint32_t c) {
     return c;
 }
 
-int utf8_icmp_n(const char* s1, const char* s2, size_t u8max) {
+int utf8_icmp_n(size_t u8max, const char* s1, const size_t n1,
+                              const char* s2, const size_t n2) {
     int ret = 0;
     utf8_decode_t d1 = {UTF8_OK}, d2 = {UTF8_OK};
-    for (; u8max--; s1 += d1.size, s2 += d2.size) {
-        utf8_peek(&d1, s1);
-        utf8_peek(&d2, s2);
+    size_t j1 = 0, j2 = 0;
+    for (; u8max-- && ((j1 < n1) & (j2 < n2)); j1 += d1.size, j2 += d2.size) {
+        utf8_peek(&d1, s1+j1);
+        utf8_peek(&d2, s2+j2);
         ret = utf8_tolower(d1.codep) - utf8_tolower(d2.codep);
-        if (ret || !*s2)
-            break;
+        if (ret || !s2[j2])
+            return ret;
     }
-    return ret;
+    return (j2 < n2) - (j1 < n1);
 }
 
 bool utf8_isupper(uint32_t c) {