aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--ChangeLog7
-rw-r--r--string.c52
2 files changed, 27 insertions, 32 deletions
diff --git a/ChangeLog b/ChangeLog
index af8ca96453..3623b8a8eb 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,10 @@
+Mon Dec 17 16:39:25 2007 Tanaka Akira <akr@fsij.org>
+
+ * string.c (rb_enc_str_coderange): set ENC_CODERANGE_BROKEN using
+ rb_enc_precise_mbclen.
+ (rb_str_valid_encoding_p): just check coderange is
+ ENC_CODERANGE_BROKEN or not.
+
Mon Dec 17 16:04:16 2007 GOTOU Yuuzou <gotoyuzo@notwork.org>
* ext/openssl/lib/openssl/buffering.rb (Buffering#gets): added second
diff --git a/string.c b/string.c
index 5eea404054..85ea099ea5 100644
--- a/string.c
+++ b/string.c
@@ -105,27 +105,26 @@ rb_enc_str_coderange(VALUE str)
if (cr == ENC_CODERANGE_UNKNOWN) {
rb_encoding *enc = rb_enc_get(str);
- if (!rb_enc_asciicompat(enc)) {
- cr = ENC_CODERANGE_VALID;
- ENC_CODERANGE_SET(str, cr);
- return cr;
- }
- else {
- const char *p = RSTRING_PTR(str);
- const char *e = p + RSTRING_LEN(str);
+ const char *p = RSTRING_PTR(str);
+ const char *e = p + RSTRING_LEN(str);
- cr = ENC_CODERANGE_7BIT;
- while (p < e) {
- int c = (unsigned char)*p;
+ cr = rb_enc_asciicompat(enc) ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
+ while (p < e) {
+ int ret = rb_enc_precise_mbclen(p, e, enc);
+ int len = MBCLEN_CHARFOUND(ret);
- if (!rb_enc_isascii(c, enc)) {
- cr = ENC_CODERANGE_VALID;
- break;
- }
- p++;
- }
- ENC_CODERANGE_SET(str, cr);
- }
+ if (len) {
+ if (len != 1 || !rb_enc_isascii((unsigned char)*p, enc)) {
+ cr = ENC_CODERANGE_VALID;
+ }
+ p += len;
+ }
+ else {
+ cr = ENC_CODERANGE_BROKEN;
+ break;
+ }
+ }
+ ENC_CODERANGE_SET(str, cr);
}
return cr;
}
@@ -5316,20 +5315,9 @@ rb_str_force_encoding(VALUE str, VALUE enc)
static VALUE
rb_str_valid_encoding_p(VALUE str)
{
- char *p = RSTRING_PTR(str);
- char *pend = RSTRING_END(str);
- rb_encoding *enc = rb_enc_get(str);
+ int cr = rb_enc_str_coderange(str);
- while (p < pend) {
- int n;
-
- n = rb_enc_precise_mbclen(p, pend, enc);
- if (!MBCLEN_CHARFOUND(n)) {
- return Qfalse;
- }
- p += n;
- }
- return Qtrue;
+ return cr == ENC_CODERANGE_BROKEN ? Qfalse : Qtrue;
}
/**********************************************************************