added to handle a character literal

This commit is contained in:
hyung-hwan 2018-04-23 14:30:00 +00:00
parent b6df00f2d2
commit 527f48fe80
5 changed files with 221 additions and 45 deletions

View File

@ -58,7 +58,8 @@ enum hcl_json_state_t
HCL_JSON_STATE_IN_WORD_VALUE, HCL_JSON_STATE_IN_WORD_VALUE,
HCL_JSON_STATE_IN_NUMERIC_VALUE, HCL_JSON_STATE_IN_NUMERIC_VALUE,
HCL_JSON_STATE_IN_QUOTED_VALUE HCL_JSON_STATE_IN_STRING_VALUE,
HCL_JSON_STATE_IN_CHARACTER_VALUE
}; };
typedef enum hcl_json_state_t hcl_json_state_t; typedef enum hcl_json_state_t hcl_json_state_t;
@ -73,6 +74,7 @@ enum hcl_json_inst_t
HCL_JSON_INST_KEY, HCL_JSON_INST_KEY,
HCL_JSON_INST_CHARACTER, /* there is no such element as character in real JSON */
HCL_JSON_INST_STRING, HCL_JSON_INST_STRING,
HCL_JSON_INST_NUMBER, HCL_JSON_INST_NUMBER,
HCL_JSON_INST_NIL, HCL_JSON_INST_NIL,

View File

@ -61,7 +61,13 @@ struct hcl_json_state_node_t
int escaped; int escaped;
int digit_count; int digit_count;
hcl_ooch_t acc; hcl_ooch_t acc;
} qv; } sv;
struct
{
int escaped;
int digit_count;
hcl_ooch_t acc;
} cv;
struct struct
{ {
int dotted; int dotted;
@ -272,64 +278,88 @@ static int invoke_data_inst (hcl_json_t* json, hcl_json_inst_t inst)
return 0; return 0;
} }
static int handle_quoted_value_char (hcl_json_t* json, hcl_ooci_t c) static int handle_string_value_char (hcl_json_t* json, hcl_ooci_t c)
{ {
if (json->state_stack->u.qv.escaped >= 2) int ret = 1;
if (json->state_stack->u.sv.escaped == 3)
{
if (c >= '0' && c <= '7')
{
json->state_stack->u.sv.acc = json->state_stack->u.sv.acc * 8 + c - '0';
json->state_stack->u.sv.digit_count++;
if (json->state_stack->u.sv.digit_count >= json->state_stack->u.sv.escaped) goto add_sv_acc;
}
else
{
ret = 0;
goto add_sv_acc;
}
}
else if (json->state_stack->u.sv.escaped >= 2)
{ {
if (c >= '0' && c <= '9') if (c >= '0' && c <= '9')
{ {
json->state_stack->u.qv.acc = json->state_stack->u.qv.acc * 16 + c - '0'; json->state_stack->u.sv.acc = json->state_stack->u.sv.acc * 16 + c - '0';
json->state_stack->u.qv.digit_count++; json->state_stack->u.sv.digit_count++;
if (json->state_stack->u.qv.digit_count >= json->state_stack->u.qv.escaped) goto add_qv_acc; if (json->state_stack->u.sv.digit_count >= json->state_stack->u.sv.escaped) goto add_sv_acc;
} }
else if (c >= 'a' && c <= 'f') else if (c >= 'a' && c <= 'f')
{ {
json->state_stack->u.qv.acc = json->state_stack->u.qv.acc * 16 + c - 'a' + 10; json->state_stack->u.sv.acc = json->state_stack->u.sv.acc * 16 + c - 'a' + 10;
json->state_stack->u.qv.digit_count++; json->state_stack->u.sv.digit_count++;
if (json->state_stack->u.qv.digit_count >= json->state_stack->u.qv.escaped) goto add_qv_acc; if (json->state_stack->u.sv.digit_count >= json->state_stack->u.sv.escaped) goto add_sv_acc;
} }
else if (c >= 'A' && c <= 'F') else if (c >= 'A' && c <= 'F')
{ {
json->state_stack->u.qv.acc = json->state_stack->u.qv.acc * 16 + c - 'A' + 10; json->state_stack->u.sv.acc = json->state_stack->u.sv.acc * 16 + c - 'A' + 10;
json->state_stack->u.qv.digit_count++; json->state_stack->u.sv.digit_count++;
if (json->state_stack->u.qv.digit_count >= json->state_stack->u.qv.escaped) goto add_qv_acc; if (json->state_stack->u.sv.digit_count >= json->state_stack->u.sv.escaped) goto add_sv_acc;
} }
else else
{ {
add_qv_acc: ret = 0;
if (add_char_to_token(json, json->state_stack->u.qv.acc) <= -1) return -1; add_sv_acc:
json->state_stack->u.qv.escaped = 0; if (add_char_to_token(json, json->state_stack->u.sv.acc) <= -1) return -1;
json->state_stack->u.sv.escaped = 0;
} }
} }
else if (json->state_stack->u.qv.escaped == 1) else if (json->state_stack->u.sv.escaped == 1)
{ {
if (c == 'x') if (c >= '0' && c <= '8')
{ {
json->state_stack->u.qv.escaped = 2; json->state_stack->u.sv.escaped = 3;
json->state_stack->u.qv.digit_count = 0; json->state_stack->u.sv.digit_count = 0;
json->state_stack->u.qv.acc = 0; json->state_stack->u.sv.acc = c - '0';
}
else if (c == 'x')
{
json->state_stack->u.sv.escaped = 2;
json->state_stack->u.sv.digit_count = 0;
json->state_stack->u.sv.acc = 0;
} }
else if (c == 'u') else if (c == 'u')
{ {
json->state_stack->u.qv.escaped = 4; json->state_stack->u.sv.escaped = 4;
json->state_stack->u.qv.digit_count = 0; json->state_stack->u.sv.digit_count = 0;
json->state_stack->u.qv.acc = 0; json->state_stack->u.sv.acc = 0;
} }
else if (c == 'U') else if (c == 'U')
{ {
json->state_stack->u.qv.escaped = 8; json->state_stack->u.sv.escaped = 8;
json->state_stack->u.qv.digit_count = 0; json->state_stack->u.sv.digit_count = 0;
json->state_stack->u.qv.acc = 0; json->state_stack->u.sv.acc = 0;
} }
else else
{ {
json->state_stack->u.qv.escaped = 0; json->state_stack->u.sv.escaped = 0;
if (add_char_to_token(json, unescape(c)) <= -1) return -1; if (add_char_to_token(json, unescape(c)) <= -1) return -1;
} }
} }
else if (c == '\\') else if (c == '\\')
{ {
json->state_stack->u.qv.escaped = 1; json->state_stack->u.sv.escaped = 1;
} }
else if (c == '\"') else if (c == '\"')
{ {
@ -341,7 +371,114 @@ static int handle_quoted_value_char (hcl_json_t* json, hcl_ooci_t c)
if (add_char_to_token(json, c) <= -1) return -1; if (add_char_to_token(json, c) <= -1) return -1;
} }
return 1; return ret;
}
static int handle_character_value_char (hcl_json_t* json, hcl_ooci_t c)
{
/* The real JSON dones't support character literal. this is HCL's own extension. */
int ret = 1;
if (json->state_stack->u.cv.escaped == 3)
{
if (c >= '0' && c <= '7')
{
json->state_stack->u.cv.acc = json->state_stack->u.cv.acc * 8 + c - '0';
json->state_stack->u.cv.digit_count++;
if (json->state_stack->u.cv.digit_count >= json->state_stack->u.cv.escaped) goto add_cv_acc;
}
else
{
ret = 0;
goto add_cv_acc;
}
}
if (json->state_stack->u.cv.escaped >= 2)
{
if (c >= '0' && c <= '9')
{
json->state_stack->u.cv.acc = json->state_stack->u.cv.acc * 16 + c - '0';
json->state_stack->u.cv.digit_count++;
if (json->state_stack->u.cv.digit_count >= json->state_stack->u.cv.escaped) goto add_cv_acc;
}
else if (c >= 'a' && c <= 'f')
{
json->state_stack->u.cv.acc = json->state_stack->u.cv.acc * 16 + c - 'a' + 10;
json->state_stack->u.cv.digit_count++;
if (json->state_stack->u.cv.digit_count >= json->state_stack->u.cv.escaped) goto add_cv_acc;
}
else if (c >= 'A' && c <= 'F')
{
json->state_stack->u.cv.acc = json->state_stack->u.cv.acc * 16 + c - 'A' + 10;
json->state_stack->u.cv.digit_count++;
if (json->state_stack->u.cv.digit_count >= json->state_stack->u.cv.escaped) goto add_cv_acc;
}
else
{
ret = 0;
add_cv_acc:
if (add_char_to_token(json, json->state_stack->u.cv.acc) <= -1) return -1;
json->state_stack->u.cv.escaped = 0;
}
}
else if (json->state_stack->u.cv.escaped == 1)
{
if (c >= '0' && c <= '8')
{
json->state_stack->u.cv.escaped = 3;
json->state_stack->u.cv.digit_count = 0;
json->state_stack->u.cv.acc = c - '0';
}
else if (c == 'x')
{
json->state_stack->u.cv.escaped = 2;
json->state_stack->u.cv.digit_count = 0;
json->state_stack->u.cv.acc = 0;
}
else if (c == 'u')
{
json->state_stack->u.cv.escaped = 4;
json->state_stack->u.cv.digit_count = 0;
json->state_stack->u.cv.acc = 0;
}
else if (c == 'U')
{
json->state_stack->u.cv.escaped = 8;
json->state_stack->u.cv.digit_count = 0;
json->state_stack->u.cv.acc = 0;
}
else
{
json->state_stack->u.cv.escaped = 0;
if (add_char_to_token(json, unescape(c)) <= -1) return -1;
}
}
else if (c == '\\')
{
json->state_stack->u.cv.escaped = 1;
}
else if (c == '\'')
{
pop_state (json);
if (json->tok.len < 1)
{
hcl_json_seterrbfmt (json, HCL_EINVAL, "no character in a character literal");
return -1;
}
if (invoke_data_inst(json, HCL_JSON_INST_CHARACTER) <= -1) return -1;
}
else
{
if (add_char_to_token(json, c) <= -1) return -1;
}
if (json->tok.len >= 1)
{
hcl_json_seterrbfmt (json, HCL_EINVAL, "too many characters in a character literal - %.*js", json->tok.len, json->tok.ptr);
return -1;
}
return ret;
} }
static int handle_numeric_value_char (hcl_json_t* json, hcl_ooci_t c) static int handle_numeric_value_char (hcl_json_t* json, hcl_ooci_t c)
@ -388,7 +525,7 @@ static int handle_word_value_char (hcl_json_t* json, hcl_ooci_t c)
else if (hcl_comp_oochars_bcstr(json->tok.ptr, json->tok.len, "false") == 0) inst = HCL_JSON_INST_FALSE; else if (hcl_comp_oochars_bcstr(json->tok.ptr, json->tok.len, "false") == 0) inst = HCL_JSON_INST_FALSE;
else else
{ {
hcl_json_seterrbfmt (json, HCL_EINVAL, "invalue word value - %.*js", json->tok.len, json->tok.ptr); hcl_json_seterrbfmt (json, HCL_EINVAL, "invalid word value - %.*js", json->tok.len, json->tok.ptr);
return -1; return -1;
} }
@ -459,12 +596,17 @@ static int handle_char_in_array (hcl_json_t* json, hcl_ooci_t c)
if (c == '\"') if (c == '\"')
{ {
if (push_state(json, HCL_JSON_STATE_IN_QUOTED_VALUE) <= -1) return -1; if (push_state(json, HCL_JSON_STATE_IN_STRING_VALUE) <= -1) return -1;
clear_token (json); clear_token (json);
return 1; return 1;
} }
/* TOOD: else if (c == '\'') HCL character else if (c == '\'')
* else if (c == '#') HCL radixed number {
if (push_state(json, HCL_JSON_STATE_IN_CHARACTER_VALUE) <= -1) return -1;
clear_token (json);
return 1;
}
/* TOOD: else if (c == '#') HCL radixed number
*/ */
else if (is_digitchar(c) || c == '+' || c == '-') else if (is_digitchar(c) || c == '+' || c == '-')
{ {
@ -549,12 +691,17 @@ static int handle_char_in_dic (hcl_json_t* json, hcl_ooci_t c)
if (c == '\"') if (c == '\"')
{ {
if (push_state(json, HCL_JSON_STATE_IN_QUOTED_VALUE) <= -1) return -1; if (push_state(json, HCL_JSON_STATE_IN_STRING_VALUE) <= -1) return -1;
clear_token (json); clear_token (json);
return 1; return 1;
} }
/* TOOD: else if (c == '\'') HCL character else if (c == '\'')
* else if (c == '#') HCL radixed number {
if (push_state(json, HCL_JSON_STATE_IN_CHARACTER_VALUE) <= -1) return -1;
clear_token (json);
return 1;
}
/* TOOD: else if (c == '#') HCL radixed number
*/ */
else if (is_digitchar(c) || c == '+' || c == '-') else if (is_digitchar(c) || c == '+' || c == '-')
{ {
@ -630,8 +777,12 @@ start_over:
x = handle_word_value_char(json, c); x = handle_word_value_char(json, c);
break; break;
case HCL_JSON_STATE_IN_QUOTED_VALUE: case HCL_JSON_STATE_IN_STRING_VALUE:
x = handle_quoted_value_char(json, c); x = handle_string_value_char(json, c);
break;
case HCL_JSON_STATE_IN_CHARACTER_VALUE:
x = handle_character_value_char(json, c);
break; break;
case HCL_JSON_STATE_IN_NUMERIC_VALUE: case HCL_JSON_STATE_IN_NUMERIC_VALUE:

View File

@ -290,7 +290,10 @@ static void log_write (hcl_client_t* client, unsigned int mask, const hcl_ooch_t
else if (n <= -1) else if (n <= -1)
{ {
/* conversion error */ /* conversion error */
break; if (bcslen <= 0) break;
if (write_log(client, logfd, buf, bcslen) <= -1) break;
msgidx += ucslen;
len -= ucslen;
} }
} }
#else #else

View File

@ -21,6 +21,8 @@ struct json_xtn_t
hcl_bch_t buf[4096]; hcl_bch_t buf[4096];
hcl_oow_t len; hcl_oow_t len;
} logbuf; } logbuf;
int depth;
}; };
/* ========================================================================= */ /* ========================================================================= */
@ -88,7 +90,6 @@ static int write_log (hcl_json_t* json, int fd, const hcl_bch_t* ptr, hcl_oow_t
{ {
json_xtn_t* xtn; json_xtn_t* xtn;
xtn = hcl_json_getxtn(json); xtn = hcl_json_getxtn(json);
while (len > 0) while (len > 0)
@ -240,7 +241,10 @@ static void log_write (hcl_json_t* json, unsigned int mask, const hcl_ooch_t* ms
else if (n <= -1) else if (n <= -1)
{ {
/* conversion error */ /* conversion error */
break; if (bcslen <= 0) break;
if (write_log(json, logfd, buf, bcslen) <= -1) break;
msgidx += ucslen;
len -= ucslen;
} }
} }
#else #else
@ -257,18 +261,26 @@ static void log_write (hcl_json_t* json, unsigned int mask, const hcl_ooch_t* ms
static int instcb (hcl_json_t* json, hcl_json_inst_t it, const hcl_oocs_t* str) static int instcb (hcl_json_t* json, hcl_json_inst_t it, const hcl_oocs_t* str)
{ {
json_xtn_t* json_xtn;
json_xtn = hcl_json_getxtn(json);
switch (it) switch (it)
{ {
case HCL_JSON_INST_START_ARRAY: case HCL_JSON_INST_START_ARRAY:
json_xtn->depth++;
hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "[\n"); hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "[\n");
break; break;
case HCL_JSON_INST_END_ARRAY: case HCL_JSON_INST_END_ARRAY:
json_xtn->depth--;
hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "]\n"); hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "]\n");
break; break;
case HCL_JSON_INST_START_DIC: case HCL_JSON_INST_START_DIC:
json_xtn->depth++;
hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "{\n"); hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "{\n");
break; break;
case HCL_JSON_INST_END_DIC: case HCL_JSON_INST_END_DIC:
json_xtn->depth--;
hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "}\n"); hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "}\n");
break; break;
@ -276,6 +288,7 @@ static int instcb (hcl_json_t* json, hcl_json_inst_t it, const hcl_oocs_t* str)
hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "%.*js: ", str->len, str->ptr); hcl_json_logbfmt (json, HCL_LOG_INFO | HCL_LOG_APP, "%.*js: ", str->len, str->ptr);
break; break;
case HCL_JSON_INST_CHARACTER:
case HCL_JSON_INST_STRING: case HCL_JSON_INST_STRING:
case HCL_JSON_INST_NUMBER: case HCL_JSON_INST_NUMBER:
case HCL_JSON_INST_TRUE: case HCL_JSON_INST_TRUE:
@ -307,11 +320,15 @@ int main (int argc, char* argv[])
json_xtn = hcl_json_getxtn(json); json_xtn = hcl_json_getxtn(json);
json_xtn->logmask = HCL_LOG_ALL_LEVELS | HCL_LOG_ALL_TYPES; json_xtn->logmask = HCL_LOG_ALL_LEVELS | HCL_LOG_ALL_TYPES;
p = "[ \"ab\\xab\\uC88B\\uC544\\uC6A9c\", \"kaden\", \"iron\", true, { \"null\": \"abc\", \"123\": \"AA20AA\", \"10\": -0.123 } ]"; p = "[ \"ab\\xab\\uC88B\\uC544\\uC6A9c\", \"kaden\", \"iron\", true, { \"null\": \"a\\1bc\", \"123\": \"AA20AA\", \"10\": -0.123, \"way\": '\\uC88B' } ]";
if (hcl_json_feed(json, p, strlen(p), &xlen) <= -1) if (hcl_json_feed(json, p, strlen(p), &xlen) <= -1)
{ {
hcl_json_logbfmt (json, HCL_LOG_FATAL | HCL_LOG_APP, "ERROR: %js\n", hcl_json_geterrmsg(json)); hcl_json_logbfmt (json, HCL_LOG_FATAL | HCL_LOG_APP, "ERROR: unable to process - %js\n", hcl_json_geterrmsg(json));
}
else if (json_xtn->depth != 0)
{
hcl_json_logbfmt (json, HCL_LOG_FATAL | HCL_LOG_APP, "ERROR: incomplete input\n");
} }
hcl_json_close (json); hcl_json_close (json);

View File

@ -291,7 +291,10 @@ static void log_write (hcl_server_t* server, hcl_oow_t wid, unsigned int mask, c
else if (n <= -1) else if (n <= -1)
{ {
/* conversion error */ /* conversion error */
break; if (bcslen <= 0) break;
if (write_log(server, logfd, buf, bcslen) <= -1) break;
msgidx += ucslen;
len -= ucslen;
} }
} }
#else #else