Fixed mishandling of UTF-8 codepoints in the FDF0..FEDF range (these were unduly rejected when extracting names from certificates, thereby preventing use of the extra presentation forms of Arabic).

Author: pornin

src/x509/asn1.t0

@@ -480,7 +480,7 @@ OID: id-at-commonName            2.5.4.3
 \ 66 noncharacters, and also the surrogate range; this function does NOT
 \ check that the value is in the 0..10FFFF range.
 : valid-unicode? ( val -- bool )
-	dup 0xFDD0 0xFEDF between? if drop 0 ret then
+	dup 0xFDD0 0xFDEF between? if drop 0 ret then
 	dup 0xD800 0xDFFF between? if drop 0 ret then
 	0xFFFF and 0xFFFE < ;
 

src/x509/x509_minimal.c

@@ -703,7 +703,7 @@ static const unsigned char t0_codeblock[] = {
 	0x76, 0x00, 0x00, 0x01, 0x00, 0x30, 0x31, 0x0B, 0x42, 0x00, 0x00, 0x01,
 	0x81, 0x70, 0x00, 0x00, 0x01, 0x82, 0x0D, 0x00, 0x00, 0x01, 0x82, 0x22,
 	0x00, 0x00, 0x01, 0x82, 0x05, 0x00, 0x00, 0x01, 0x03, 0x33, 0x01, 0x03,
-	0x33, 0x00, 0x00, 0x25, 0x01, 0x83, 0xFB, 0x50, 0x01, 0x83, 0xFD, 0x5F,
+	0x33, 0x00, 0x00, 0x25, 0x01, 0x83, 0xFB, 0x50, 0x01, 0x83, 0xFB, 0x6F,
 	0x72, 0x06, 0x04, 0x24, 0x01, 0x00, 0x00, 0x25, 0x01, 0x83, 0xB0, 0x00,
 	0x01, 0x83, 0xBF, 0x7F, 0x72, 0x06, 0x04, 0x24, 0x01, 0x00, 0x00, 0x01,
 	0x83, 0xFF, 0x7F, 0x15, 0x01, 0x83, 0xFF, 0x7E, 0x0D, 0x00