Re: [Qemu-devel] [PATCH v2 28/60] json: Fix \uXXXX for surrogate pairs

[Eric Blake]

On 08/17/2018 10:05 AM, Markus Armbruster wrote:
> The JSON parser treats each half of a surrogate pair as unpaired
> surrogate.  Fix it to recognize surrogate pairs.
>
> Signed-off-by: Markus Armbruster <address@hidden>
> Reviewed-by: Eric Blake <address@hidden>

I might have dropped the R-b, to ensure the changes since v1 get 
re-reviewed.

> ---
>   qobject/json-parser.c | 60 ++++++++++++++++++++++++++++---------------
>   tests/check-qjson.c   |  3 +--
>   2 files changed, 40 insertions(+), 23 deletions(-)

> @@ -157,22 +169,28 @@ static QString *parse_string(JSONParserContext *ctxt, 
> JSONToken *token)
>                   qstring_append_chr(str, '\t');
>                   break;
>               case 'u':
> -                cp = 0;
> -                for (i = 0; i < 4; i++) {
> -                    if (!qemu_isxdigit(*ptr)) {
> -                        parse_error(ctxt, token,
> -                                    "invalid hex escape sequence in string");
> -                        goto out;
> +                cp = cvt4hex(ptr);
> +                ptr += 4;
> +
> +                /* handle surrogate pairs */
> +                if (cp >= 0xD800 && cp <= 0xDBFF
> +                    && ptr[0] == '\\' && ptr[1] == 'u') {
> +                    /* leading surrogate followed by \u */
> +                    cp = 0x10000 + ((cp & 0x3FF) << 10);
> +                    trailing = cvt4hex(ptr + 2);
> +                    if (trailing >= 0xDC00 && trailing <= 0xDFFF) {
> +                        /* followed by trailing surrogate */
> +                        cp |= trailing & 0x3FF;
> +                        ptr += 6;
> +                    } else {
> +                        cp = -1; /* invalid */
>                       }
> -                    cp <<= 4;
> -                    cp |= hex2decimal(*ptr);
> -                    ptr++;
>                   }
>   
>                   if (mod_utf8_encode(utf8_buf, sizeof(utf8_buf), cp) < 0) {
>                       parse_error(ctxt, token,
> -                                "\\u%.4s is not a valid Unicode character",
> -                                ptr - 3);
> +                                "%.*s is not a valid Unicode character",
> +                                (int)(ptr - beg), beg);

The error reporting here has indeed been improved over v1.

Reviewed-by: Eric Blake <address@hidden>

--
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org