mirror of
https://codeberg.org/grunfink/snac2.git
synced 2024-11-22 21:35:04 +00:00
530 lines
11 KiB
C
530 lines
11 KiB
C
/* copyright (c) 2022 - 2023 grunfink / MIT license */
|
|
|
|
#ifndef _XS_JSON_H
|
|
|
|
#define _XS_JSON_H
|
|
|
|
xs_str *xs_json_dumps_pp(const xs_val *data, int indent);
|
|
#define xs_json_dumps(data) xs_json_dumps_pp(data, 0)
|
|
xs_val *xs_json_loads(const xs_str *json);
|
|
|
|
|
|
#ifdef XS_IMPLEMENTATION
|
|
|
|
/** IMPLEMENTATION **/
|
|
|
|
/** JSON dumps **/
|
|
|
|
static xs_str *_xs_json_dumps_str(xs_str *s, const char *data)
|
|
/* dumps a string in JSON format */
|
|
{
|
|
unsigned char c;
|
|
s = xs_str_cat(s, "\"");
|
|
|
|
while ((c = *data)) {
|
|
if (c == '\n')
|
|
s = xs_str_cat(s, "\\n");
|
|
else
|
|
if (c == '\r')
|
|
s = xs_str_cat(s, "\\r");
|
|
else
|
|
if (c == '\t')
|
|
s = xs_str_cat(s, "\\t");
|
|
else
|
|
if (c == '\\')
|
|
s = xs_str_cat(s, "\\\\");
|
|
else
|
|
if (c == '"')
|
|
s = xs_str_cat(s, "\\\"");
|
|
else
|
|
if (c < 32) {
|
|
char tmp[10];
|
|
|
|
snprintf(tmp, sizeof(tmp), "\\u%04x", (unsigned int) c);
|
|
s = xs_str_cat(s, tmp);
|
|
}
|
|
else
|
|
s = xs_append_m(s, data, 1);
|
|
|
|
data++;
|
|
}
|
|
|
|
s = xs_str_cat(s, "\"");
|
|
|
|
return s;
|
|
}
|
|
|
|
|
|
static xs_str *_xs_json_indent(xs_str *s, int level, int indent)
|
|
/* adds indentation */
|
|
{
|
|
if (indent) {
|
|
int n;
|
|
|
|
s = xs_str_cat(s, "\n");
|
|
|
|
for (n = 0; n < level * indent; n++)
|
|
s = xs_str_cat(s, " ");
|
|
}
|
|
|
|
return s;
|
|
}
|
|
|
|
|
|
static xs_str *_xs_json_dumps(xs_str *s, const xs_val *s_data, int level, int indent)
|
|
/* dumps partial data as JSON */
|
|
{
|
|
int c = 0;
|
|
xs_val *v;
|
|
xs_val *data = (xs_val *)s_data;
|
|
|
|
switch (xs_type(data)) {
|
|
case XSTYPE_NULL:
|
|
s = xs_str_cat(s, "null");
|
|
break;
|
|
|
|
case XSTYPE_TRUE:
|
|
s = xs_str_cat(s, "true");
|
|
break;
|
|
|
|
case XSTYPE_FALSE:
|
|
s = xs_str_cat(s, "false");
|
|
break;
|
|
|
|
case XSTYPE_NUMBER:
|
|
s = xs_str_cat(s, xs_number_str(data));
|
|
break;
|
|
|
|
case XSTYPE_LIST:
|
|
s = xs_str_cat(s, "[");
|
|
|
|
while (xs_list_iter(&data, &v)) {
|
|
if (c != 0)
|
|
s = xs_str_cat(s, ",");
|
|
|
|
s = _xs_json_indent(s, level + 1, indent);
|
|
s = _xs_json_dumps(s, v, level + 1, indent);
|
|
|
|
c++;
|
|
}
|
|
|
|
s = _xs_json_indent(s, level, indent);
|
|
s = xs_str_cat(s, "]");
|
|
|
|
break;
|
|
|
|
case XSTYPE_DICT:
|
|
s = xs_str_cat(s, "{");
|
|
|
|
xs_str *k;
|
|
while (xs_dict_iter(&data, &k, &v)) {
|
|
if (c != 0)
|
|
s = xs_str_cat(s, ",");
|
|
|
|
s = _xs_json_indent(s, level + 1, indent);
|
|
|
|
s = _xs_json_dumps_str(s, k);
|
|
s = xs_str_cat(s, ":");
|
|
|
|
if (indent)
|
|
s = xs_str_cat(s, " ");
|
|
|
|
s = _xs_json_dumps(s, v, level + 1, indent);
|
|
|
|
c++;
|
|
}
|
|
|
|
s = _xs_json_indent(s, level, indent);
|
|
s = xs_str_cat(s, "}");
|
|
break;
|
|
|
|
case XSTYPE_STRING:
|
|
s = _xs_json_dumps_str(s, data);
|
|
break;
|
|
|
|
default:
|
|
break;
|
|
}
|
|
|
|
return s;
|
|
}
|
|
|
|
|
|
xs_str *xs_json_dumps_pp(const xs_val *data, int indent)
|
|
/* dumps a piece of data as JSON */
|
|
{
|
|
xstype t = xs_type(data);
|
|
xs_str *s = NULL;
|
|
|
|
if (t == XSTYPE_LIST || t == XSTYPE_DICT) {
|
|
s = xs_str_new(NULL);
|
|
s = _xs_json_dumps(s, data, 0, indent);
|
|
}
|
|
|
|
return s;
|
|
}
|
|
|
|
|
|
/** JSON loads **/
|
|
|
|
/* this code comes mostly from the Minimum Profit Text Editor (MPDM) */
|
|
|
|
typedef enum {
|
|
JS_ERROR = -1,
|
|
JS_INCOMPLETE,
|
|
JS_OCURLY,
|
|
JS_OBRACK,
|
|
JS_CCURLY,
|
|
JS_CBRACK,
|
|
JS_COMMA,
|
|
JS_COLON,
|
|
JS_VALUE,
|
|
JS_STRING,
|
|
JS_INTEGER,
|
|
JS_REAL,
|
|
JS_TRUE,
|
|
JS_FALSE,
|
|
JS_NULL,
|
|
JS_ARRAY,
|
|
JS_OBJECT
|
|
} js_type;
|
|
|
|
|
|
static xs_val *_xs_json_loads_lexer(const char **json, js_type *t)
|
|
{
|
|
char c;
|
|
const char *s = *json;
|
|
xs_val *v = NULL;
|
|
|
|
/* skip blanks */
|
|
while (*s == L' ' || *s == L'\t' || *s == L'\n' || *s == L'\r')
|
|
s++;
|
|
|
|
c = *s++;
|
|
|
|
if (c == '{')
|
|
*t = JS_OCURLY;
|
|
else
|
|
if (c == '}')
|
|
*t = JS_CCURLY;
|
|
else
|
|
if (c == '[')
|
|
*t = JS_OBRACK;
|
|
else
|
|
if (c == ']')
|
|
*t = JS_CBRACK;
|
|
else
|
|
if (c == ',')
|
|
*t = JS_COMMA;
|
|
else
|
|
if (c == ':')
|
|
*t = JS_COLON;
|
|
else
|
|
if (c == '"') {
|
|
*t = JS_STRING;
|
|
|
|
v = xs_str_new(NULL);
|
|
|
|
while ((c = *s) != '"' && c != '\0') {
|
|
char tmp[5];
|
|
int cp, i;
|
|
|
|
if (c == '\\') {
|
|
s++;
|
|
c = *s;
|
|
switch (c) {
|
|
case 'n': c = '\n'; break;
|
|
case 'r': c = '\r'; break;
|
|
case 't': c = '\t'; break;
|
|
case 'u': /* Unicode codepoint as an hex char */
|
|
s++;
|
|
strncpy(tmp, s, 4);
|
|
tmp[4] = '\0';
|
|
|
|
if (strlen(tmp) != 4) {
|
|
*t = JS_ERROR;
|
|
break;
|
|
}
|
|
|
|
s += 3; /* skip as it was one byte */
|
|
|
|
sscanf(tmp, "%04x", &i);
|
|
|
|
if (i >= 0xd800 && i <= 0xdfff) {
|
|
/* it's a surrogate pair */
|
|
cp = (i & 0x3ff) << 10;
|
|
|
|
/* skip to the next value (last char + \ + u) */
|
|
s++;
|
|
if (memcmp(s, "\\u", 2) != 0) {
|
|
*t = JS_ERROR;
|
|
break;
|
|
}
|
|
s += 2;
|
|
|
|
strncpy(tmp, s, 4);
|
|
tmp[4] = '\0';
|
|
|
|
if (strlen(tmp) != 4) {
|
|
*t = JS_ERROR;
|
|
break;
|
|
}
|
|
|
|
s += 3; /* skip as it was one byte */
|
|
|
|
sscanf(tmp, "%04x", &i);
|
|
cp |= (i & 0x3ff);
|
|
cp += 0x10000;
|
|
}
|
|
else
|
|
cp = i;
|
|
|
|
/* replace dangerous control codes with their visual representations */
|
|
if (cp >= '\0' && cp < ' ' && !strchr("\r\n\t", cp))
|
|
cp += 0x2400;
|
|
|
|
v = xs_utf8_enc(v, cp);
|
|
c = '\0';
|
|
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (c)
|
|
v = xs_append_m(v, &c, 1);
|
|
|
|
s++;
|
|
}
|
|
|
|
if (c != '\0')
|
|
s++;
|
|
}
|
|
else
|
|
if (c == '-' || (c >= '0' && c <= '9') || c == '.') {
|
|
xs *vn = NULL;
|
|
|
|
*t = JS_INTEGER;
|
|
|
|
vn = xs_str_new(NULL);
|
|
vn = xs_append_m(vn, &c, 1);
|
|
|
|
while (((c = *s) >= '0' && c <= '9') || c == '.') {
|
|
if (c == '.')
|
|
*t = JS_REAL;
|
|
|
|
vn = xs_append_m(vn, &c, 1);
|
|
s++;
|
|
}
|
|
|
|
/* convert to XSTYPE_NUMBER */
|
|
v = xs_number_new(atof(vn));
|
|
}
|
|
else
|
|
if (c == 't' && strncmp(s, "rue", 3) == 0) {
|
|
s += 3;
|
|
*t = JS_TRUE;
|
|
|
|
v = xs_val_new(XSTYPE_TRUE);
|
|
}
|
|
else
|
|
if (c == 'f' && strncmp(s, "alse", 4) == 0) {
|
|
s += 4;
|
|
*t = JS_FALSE;
|
|
|
|
v = xs_val_new(XSTYPE_FALSE);
|
|
}
|
|
else
|
|
if (c == 'n' && strncmp(s, "ull", 3) == 0) {
|
|
s += 3;
|
|
*t = JS_NULL;
|
|
|
|
v = xs_val_new(XSTYPE_NULL);
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
|
|
*json = s;
|
|
|
|
return v;
|
|
}
|
|
|
|
|
|
static xs_list *_xs_json_loads_array(const char **json, js_type *t);
|
|
static xs_dict *_xs_json_loads_object(const char **json, js_type *t);
|
|
|
|
static xs_val *_xs_json_loads_value(const char **json, js_type *t, xs_val *v)
|
|
/* parses a JSON value */
|
|
{
|
|
if (*t == JS_OBRACK)
|
|
v = _xs_json_loads_array(json, t);
|
|
else
|
|
if (*t == JS_OCURLY)
|
|
v = _xs_json_loads_object(json, t);
|
|
|
|
if (*t >= JS_VALUE)
|
|
*t = JS_VALUE;
|
|
else
|
|
*t = JS_ERROR;
|
|
|
|
return v;
|
|
}
|
|
|
|
|
|
static xs_list *_xs_json_loads_array(const char **json, js_type *t)
|
|
/* parses a JSON array */
|
|
{
|
|
const char *s = *json;
|
|
xs *v;
|
|
xs_list *l;
|
|
js_type tt;
|
|
|
|
l = xs_list_new();
|
|
|
|
*t = JS_INCOMPLETE;
|
|
|
|
v = _xs_json_loads_lexer(&s, &tt);
|
|
|
|
if (tt == JS_CBRACK)
|
|
*t = JS_ARRAY;
|
|
else {
|
|
v = _xs_json_loads_value(&s, &tt, v);
|
|
|
|
if (tt == JS_VALUE) {
|
|
l = xs_list_append(l, v);
|
|
|
|
while (*t == JS_INCOMPLETE) {
|
|
xs_free(_xs_json_loads_lexer(&s, &tt));
|
|
|
|
if (tt == JS_CBRACK)
|
|
*t = JS_ARRAY;
|
|
else
|
|
if (tt == JS_COMMA) {
|
|
xs *v2;
|
|
|
|
v2 = _xs_json_loads_lexer(&s, &tt);
|
|
v2 = _xs_json_loads_value(&s, &tt, v2);
|
|
|
|
if (tt == JS_VALUE)
|
|
l = xs_list_append(l, v2);
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
|
|
if (*t == JS_ERROR)
|
|
l = xs_free(l);
|
|
|
|
*json = s;
|
|
|
|
return l;
|
|
}
|
|
|
|
|
|
static xs_dict *_xs_json_loads_object(const char **json, js_type *t)
|
|
/* parses a JSON object */
|
|
{
|
|
const char *s = *json;
|
|
xs *k1;
|
|
xs_dict *d;
|
|
js_type tt;
|
|
|
|
d = xs_dict_new();
|
|
|
|
*t = JS_INCOMPLETE;
|
|
|
|
k1 = _xs_json_loads_lexer(&s, &tt);
|
|
|
|
if (tt == JS_CCURLY)
|
|
*t = JS_OBJECT;
|
|
else
|
|
if (tt == JS_STRING) {
|
|
xs_free(_xs_json_loads_lexer(&s, &tt));
|
|
|
|
if (tt == JS_COLON) {
|
|
xs *v1;
|
|
|
|
v1 = _xs_json_loads_lexer(&s, &tt);
|
|
v1 = _xs_json_loads_value(&s, &tt, v1);
|
|
|
|
if (tt == JS_VALUE) {
|
|
d = xs_dict_append(d, k1, v1);
|
|
|
|
while (*t == JS_INCOMPLETE) {
|
|
xs_free(_xs_json_loads_lexer(&s, &tt));
|
|
|
|
if (tt == JS_CCURLY)
|
|
*t = JS_OBJECT;
|
|
else
|
|
if (tt == JS_COMMA) {
|
|
xs *k = _xs_json_loads_lexer(&s, &tt);
|
|
|
|
if (tt == JS_STRING) {
|
|
xs_free(_xs_json_loads_lexer(&s, &tt));
|
|
|
|
if (tt == JS_COLON) {
|
|
xs *v;
|
|
|
|
v = _xs_json_loads_lexer(&s, &tt);
|
|
v = _xs_json_loads_value(&s, &tt, v);
|
|
|
|
if (tt == JS_VALUE)
|
|
d = xs_dict_append(d, k, v);
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
}
|
|
else
|
|
*t = JS_ERROR;
|
|
|
|
if (*t == JS_ERROR)
|
|
d = xs_free(d);
|
|
|
|
*json = s;
|
|
|
|
return d;
|
|
}
|
|
|
|
|
|
xs_val *xs_json_loads(const xs_str *json)
|
|
/* loads a string in JSON format and converts to a multiple data */
|
|
{
|
|
xs_val *v = NULL;
|
|
js_type t;
|
|
|
|
xs_free(_xs_json_loads_lexer(&json, &t));
|
|
|
|
if (t == JS_OBRACK)
|
|
v = _xs_json_loads_array(&json, &t);
|
|
else
|
|
if (t == JS_OCURLY)
|
|
v = _xs_json_loads_object(&json, &t);
|
|
else
|
|
t = JS_ERROR;
|
|
|
|
return v;
|
|
}
|
|
|
|
#endif /* XS_IMPLEMENTATION */
|
|
|
|
#endif /* _XS_JSON_H */
|