* string.c (rb_str_coderange_scan_restartable): coderange scaning

  for partial read.

* io.c (read_all): set coderange when not convert encoding.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@15617 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 55 insertions, 0 deletions

diff --git a/string.c b/string.c
index 596459d919..26707ed0c5 100644
--- a/string.c
+++ b/string.c
@@ -201,6 +201,61 @@ coderange_scan(const char *p, long len, rb_encoding *enc)
     return ENC_CODERANGE_VALID;
 }
 
+long
+rb_str_coderange_scan_restartable(const char *s, const char *e, rb_encoding *enc, int *cr)
+{
+    long c;
+    const char *p;
+    int ret;
+    p = s;
+
+    if (*cr == ENC_CODERANGE_BROKEN)
+	return e - s;
+
+    if (rb_enc_to_index(enc) == 0) {
+	/* enc is ASCII-8BIT.  ASCII-8BIT string never be broken. */
+	p = search_nonascii(p, e);
+	*cr = (!p && *cr != ENC_CODERANGE_VALID) ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
+	return e - s;
+    }
+    else if (rb_enc_asciicompat(enc)) {
+	p = search_nonascii(p, e);
+	if (!p) {
+	    if (*cr != ENC_CODERANGE_VALID) *cr = ENC_CODERANGE_7BIT;
+	    return e - s;
+	}
+	while (p < e) {
+	    int ret = rb_enc_precise_mbclen(p, e, enc);
+	    if (!MBCLEN_CHARFOUND_P(ret)) {
+		*cr = MBCLEN_INVALID_P(ret) ? ENC_CODERANGE_BROKEN: ret;
+		return e - s;
+	    }
+	    p += MBCLEN_CHARFOUND_LEN(ret);
+	    if (p < e) {
+		p = search_nonascii(p, e);
+		if (!p) {
+		    *cr = ENC_CODERANGE_VALID;
+		    return e - s;
+		}
+	    }
+	}
+	*cr = e < p ? ENC_CODERANGE_BROKEN: ENC_CODERANGE_VALID;
+	return p - s;
+    }
+    else {
+	while (p < e) {
+	    int ret = rb_enc_precise_mbclen(p, e, enc);
+	    if (!MBCLEN_CHARFOUND_P(ret)) {
+		*cr = MBCLEN_INVALID_P(ret) ? ENC_CODERANGE_BROKEN: ret;
+		return p - s;
+	    }
+	    p += MBCLEN_CHARFOUND_LEN(ret);
+	}
+	*cr = e < p ? ENC_CODERANGE_BROKEN: ENC_CODERANGE_VALID;
+	return p - s;
+    }
+}
+
 static void
 rb_enc_str_copy(VALUE dest, VALUE src)
 {