2147 lines
46 KiB
C
2147 lines
46 KiB
C
|
/*
|
|||
|
|
|||
|
MIT License
|
|||
|
|
|||
|
Copyright (c) 2017 - 2019 CK Tan
|
|||
|
https://github.com/cktan/tomlc99
|
|||
|
|
|||
|
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|||
|
of this software and associated documentation files (the "Software"), to deal
|
|||
|
in the Software without restriction, including without limitation the rights
|
|||
|
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|||
|
copies of the Software, and to permit persons to whom the Software is
|
|||
|
furnished to do so, subject to the following conditions:
|
|||
|
|
|||
|
The above copyright notice and this permission notice shall be included in all
|
|||
|
copies or substantial portions of the Software.
|
|||
|
|
|||
|
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|||
|
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|||
|
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|||
|
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|||
|
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|||
|
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|||
|
SOFTWARE.
|
|||
|
|
|||
|
*/
|
|||
|
#define _POSIX_C_SOURCE 200809L
|
|||
|
#include <stdio.h>
|
|||
|
#include <setjmp.h>
|
|||
|
#include <stdlib.h>
|
|||
|
#include <assert.h>
|
|||
|
#include <errno.h>
|
|||
|
#include <stdint.h>
|
|||
|
#include <ctype.h>
|
|||
|
#include <string.h>
|
|||
|
#include "toml.h"
|
|||
|
|
|||
|
|
|||
|
static void* (*ppmalloc)(size_t) = malloc;
|
|||
|
static void (*ppfree)(void*) = free;
|
|||
|
static void* (*ppcalloc)(size_t, size_t) = calloc;
|
|||
|
static void* (*pprealloc)(void*, size_t) = realloc;
|
|||
|
|
|||
|
void toml_set_memutil(void* (*xxmalloc)(size_t),
|
|||
|
void (*xxfree)(void*),
|
|||
|
void* (*xxcalloc)(size_t, size_t),
|
|||
|
void* (*xxrealloc)(void*, size_t))
|
|||
|
{
|
|||
|
ppmalloc = xxmalloc;
|
|||
|
ppfree = xxfree;
|
|||
|
ppcalloc = xxcalloc;
|
|||
|
pprealloc = xxrealloc;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
#define MALLOC(a) ppmalloc(a)
|
|||
|
#define FREE(a) ppfree(a)
|
|||
|
#define CALLOC(a,b) ppcalloc(a,b)
|
|||
|
#define REALLOC(a,b) pprealloc(a,b)
|
|||
|
|
|||
|
static char* STRDUP(const char* s)
|
|||
|
{
|
|||
|
int len = strlen(s);
|
|||
|
char* p = MALLOC(len+1);
|
|||
|
if (p) {
|
|||
|
memcpy(p, s, len);
|
|||
|
p[len] = 0;
|
|||
|
}
|
|||
|
return p;
|
|||
|
}
|
|||
|
|
|||
|
static char* STRNDUP(const char* s, size_t n)
|
|||
|
{
|
|||
|
size_t len = strnlen(s, n);
|
|||
|
char* p = MALLOC(len+1);
|
|||
|
if (p) {
|
|||
|
memcpy(p, s, len);
|
|||
|
p[len] = 0;
|
|||
|
}
|
|||
|
return p;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
|
|||
|
/**
|
|||
|
* Convert a char in utf8 into UCS, and store it in *ret.
|
|||
|
* Return #bytes consumed or -1 on failure.
|
|||
|
*/
|
|||
|
int toml_utf8_to_ucs(const char* orig, int len, int64_t* ret)
|
|||
|
{
|
|||
|
const unsigned char* buf = (const unsigned char*) orig;
|
|||
|
unsigned i = *buf++;
|
|||
|
int64_t v;
|
|||
|
|
|||
|
/* 0x00000000 - 0x0000007F:
|
|||
|
0xxxxxxx
|
|||
|
*/
|
|||
|
if (0 == (i >> 7)) {
|
|||
|
if (len < 1) return -1;
|
|||
|
v = i;
|
|||
|
return *ret = v, 1;
|
|||
|
}
|
|||
|
/* 0x00000080 - 0x000007FF:
|
|||
|
110xxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (0x6 == (i >> 5)) {
|
|||
|
if (len < 2) return -1;
|
|||
|
v = i & 0x1f;
|
|||
|
for (int j = 0; j < 1; j++) {
|
|||
|
i = *buf++;
|
|||
|
if (0x2 != (i >> 6)) return -1;
|
|||
|
v = (v << 6) | (i & 0x3f);
|
|||
|
}
|
|||
|
return *ret = v, (const char*) buf - orig;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00000800 - 0x0000FFFF:
|
|||
|
1110xxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (0xE == (i >> 4)) {
|
|||
|
if (len < 3) return -1;
|
|||
|
v = i & 0x0F;
|
|||
|
for (int j = 0; j < 2; j++) {
|
|||
|
i = *buf++;
|
|||
|
if (0x2 != (i >> 6)) return -1;
|
|||
|
v = (v << 6) | (i & 0x3f);
|
|||
|
}
|
|||
|
return *ret = v, (const char*) buf - orig;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00010000 - 0x001FFFFF:
|
|||
|
11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (0x1E == (i >> 3)) {
|
|||
|
if (len < 4) return -1;
|
|||
|
v = i & 0x07;
|
|||
|
for (int j = 0; j < 3; j++) {
|
|||
|
i = *buf++;
|
|||
|
if (0x2 != (i >> 6)) return -1;
|
|||
|
v = (v << 6) | (i & 0x3f);
|
|||
|
}
|
|||
|
return *ret = v, (const char*) buf - orig;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00200000 - 0x03FFFFFF:
|
|||
|
111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (0x3E == (i >> 2)) {
|
|||
|
if (len < 5) return -1;
|
|||
|
v = i & 0x03;
|
|||
|
for (int j = 0; j < 4; j++) {
|
|||
|
i = *buf++;
|
|||
|
if (0x2 != (i >> 6)) return -1;
|
|||
|
v = (v << 6) | (i & 0x3f);
|
|||
|
}
|
|||
|
return *ret = v, (const char*) buf - orig;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x04000000 - 0x7FFFFFFF:
|
|||
|
1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (0x7e == (i >> 1)) {
|
|||
|
if (len < 6) return -1;
|
|||
|
v = i & 0x01;
|
|||
|
for (int j = 0; j < 5; j++) {
|
|||
|
i = *buf++;
|
|||
|
if (0x2 != (i >> 6)) return -1;
|
|||
|
v = (v << 6) | (i & 0x3f);
|
|||
|
}
|
|||
|
return *ret = v, (const char*) buf - orig;
|
|||
|
}
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/**
|
|||
|
* Convert a UCS char to utf8 code, and return it in buf.
|
|||
|
* Return #bytes used in buf to encode the char, or
|
|||
|
* -1 on error.
|
|||
|
*/
|
|||
|
int toml_ucs_to_utf8(int64_t code, char buf[6])
|
|||
|
{
|
|||
|
/* http://stackoverflow.com/questions/6240055/manually-converting-unicode-codepoints-into-utf-8-and-utf-16 */
|
|||
|
/* The UCS code values 0xd800–0xdfff (UTF-16 surrogates) as well
|
|||
|
* as 0xfffe and 0xffff (UCS noncharacters) should not appear in
|
|||
|
* conforming UTF-8 streams.
|
|||
|
*/
|
|||
|
if (0xd800 <= code && code <= 0xdfff) return -1;
|
|||
|
if (0xfffe <= code && code <= 0xffff) return -1;
|
|||
|
|
|||
|
/* 0x00000000 - 0x0000007F:
|
|||
|
0xxxxxxx
|
|||
|
*/
|
|||
|
if (code < 0) return -1;
|
|||
|
if (code <= 0x7F) {
|
|||
|
buf[0] = (unsigned char) code;
|
|||
|
return 1;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00000080 - 0x000007FF:
|
|||
|
110xxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (code <= 0x000007FF) {
|
|||
|
buf[0] = 0xc0 | (code >> 6);
|
|||
|
buf[1] = 0x80 | (code & 0x3f);
|
|||
|
return 2;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00000800 - 0x0000FFFF:
|
|||
|
1110xxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (code <= 0x0000FFFF) {
|
|||
|
buf[0] = 0xe0 | (code >> 12);
|
|||
|
buf[1] = 0x80 | ((code >> 6) & 0x3f);
|
|||
|
buf[2] = 0x80 | (code & 0x3f);
|
|||
|
return 3;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00010000 - 0x001FFFFF:
|
|||
|
11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (code <= 0x001FFFFF) {
|
|||
|
buf[0] = 0xf0 | (code >> 18);
|
|||
|
buf[1] = 0x80 | ((code >> 12) & 0x3f);
|
|||
|
buf[2] = 0x80 | ((code >> 6) & 0x3f);
|
|||
|
buf[3] = 0x80 | (code & 0x3f);
|
|||
|
return 4;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x00200000 - 0x03FFFFFF:
|
|||
|
111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (code <= 0x03FFFFFF) {
|
|||
|
buf[0] = 0xf8 | (code >> 24);
|
|||
|
buf[1] = 0x80 | ((code >> 18) & 0x3f);
|
|||
|
buf[2] = 0x80 | ((code >> 12) & 0x3f);
|
|||
|
buf[3] = 0x80 | ((code >> 6) & 0x3f);
|
|||
|
buf[4] = 0x80 | (code & 0x3f);
|
|||
|
return 5;
|
|||
|
}
|
|||
|
|
|||
|
/* 0x04000000 - 0x7FFFFFFF:
|
|||
|
1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
|
|||
|
*/
|
|||
|
if (code <= 0x7FFFFFFF) {
|
|||
|
buf[0] = 0xfc | (code >> 30);
|
|||
|
buf[1] = 0x80 | ((code >> 24) & 0x3f);
|
|||
|
buf[2] = 0x80 | ((code >> 18) & 0x3f);
|
|||
|
buf[3] = 0x80 | ((code >> 12) & 0x3f);
|
|||
|
buf[4] = 0x80 | ((code >> 6) & 0x3f);
|
|||
|
buf[5] = 0x80 | (code & 0x3f);
|
|||
|
return 6;
|
|||
|
}
|
|||
|
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
/*
|
|||
|
* TOML has 3 data structures: value, array, table.
|
|||
|
* Each of them can have identification key.
|
|||
|
*/
|
|||
|
typedef struct toml_keyval_t toml_keyval_t;
|
|||
|
struct toml_keyval_t {
|
|||
|
const char* key; /* key to this value */
|
|||
|
const char* val; /* the raw value */
|
|||
|
};
|
|||
|
|
|||
|
|
|||
|
struct toml_array_t {
|
|||
|
const char* key; /* key to this array */
|
|||
|
int kind; /* element kind: 'v'alue, 'a'rray, or 't'able */
|
|||
|
int type; /* for value kind: 'i'nt, 'd'ouble, 'b'ool, 's'tring, 't'ime, 'D'ate, 'T'imestamp */
|
|||
|
|
|||
|
int nelem; /* number of elements */
|
|||
|
union {
|
|||
|
char** val;
|
|||
|
toml_array_t** arr;
|
|||
|
toml_table_t** tab;
|
|||
|
} u;
|
|||
|
};
|
|||
|
|
|||
|
|
|||
|
struct toml_table_t {
|
|||
|
const char* key; /* key to this table */
|
|||
|
int implicit; /* table was created implicitly */
|
|||
|
|
|||
|
/* key-values in the table */
|
|||
|
int nkval;
|
|||
|
toml_keyval_t** kval;
|
|||
|
|
|||
|
/* arrays in the table */
|
|||
|
int narr;
|
|||
|
toml_array_t** arr;
|
|||
|
|
|||
|
/* tables in the table */
|
|||
|
int ntab;
|
|||
|
toml_table_t** tab;
|
|||
|
};
|
|||
|
|
|||
|
|
|||
|
static inline void xfree(const void* x) { if (x) FREE((void*)(intptr_t)x); }
|
|||
|
|
|||
|
|
|||
|
enum tokentype_t {
|
|||
|
INVALID,
|
|||
|
DOT,
|
|||
|
COMMA,
|
|||
|
EQUAL,
|
|||
|
LBRACE,
|
|||
|
RBRACE,
|
|||
|
NEWLINE,
|
|||
|
LBRACKET,
|
|||
|
RBRACKET,
|
|||
|
STRING,
|
|||
|
};
|
|||
|
typedef enum tokentype_t tokentype_t;
|
|||
|
|
|||
|
typedef struct token_t token_t;
|
|||
|
struct token_t {
|
|||
|
tokentype_t tok;
|
|||
|
int lineno;
|
|||
|
char* ptr; /* points into context->start */
|
|||
|
int len;
|
|||
|
int eof;
|
|||
|
};
|
|||
|
|
|||
|
|
|||
|
typedef struct context_t context_t;
|
|||
|
struct context_t {
|
|||
|
char* start;
|
|||
|
char* stop;
|
|||
|
char* errbuf;
|
|||
|
int errbufsz;
|
|||
|
jmp_buf jmp;
|
|||
|
|
|||
|
token_t tok;
|
|||
|
toml_table_t* root;
|
|||
|
toml_table_t* curtab;
|
|||
|
|
|||
|
struct {
|
|||
|
int top;
|
|||
|
char* key[10];
|
|||
|
token_t tok[10];
|
|||
|
} tpath;
|
|||
|
|
|||
|
};
|
|||
|
|
|||
|
#define STRINGIFY(x) #x
|
|||
|
#define TOSTRING(x) STRINGIFY(x)
|
|||
|
#define FLINE __FILE__ ":" TOSTRING(__LINE__)
|
|||
|
|
|||
|
static tokentype_t next_token(context_t* ctx, int dotisspecial);
|
|||
|
|
|||
|
/* error routines. All these functions longjmp to ctx->jmp */
|
|||
|
static int e_outofmemory(context_t* ctx, const char* fline)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "ERROR: out of memory (%s)", fline);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
static int e_internal_error(context_t* ctx, const char* fline)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "internal error (%s)", fline);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
static int e_syntax_error(context_t* ctx, int lineno, const char* msg)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "line %d: %s", lineno, msg);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
static int e_bad_key_error(context_t* ctx, int lineno)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "line %d: bad key", lineno);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
/*
|
|||
|
static int e_noimpl(context_t* ctx, const char* feature)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "not implemented: %s", feature);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
*/
|
|||
|
|
|||
|
static int e_key_exists_error(context_t* ctx, int lineno)
|
|||
|
{
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz,
|
|||
|
"line %d: key exists", lineno);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
return -1;
|
|||
|
}
|
|||
|
|
|||
|
static char* norm_lit_str(const char* src, int srclen,
|
|||
|
int multiline,
|
|||
|
char* errbuf, int errbufsz)
|
|||
|
{
|
|||
|
char* dst = 0; /* will write to dst[] and return it */
|
|||
|
int max = 0; /* max size of dst[] */
|
|||
|
int off = 0; /* cur offset in dst[] */
|
|||
|
const char* sp = src;
|
|||
|
const char* sq = src + srclen;
|
|||
|
int ch;
|
|||
|
|
|||
|
/* scan forward on src */
|
|||
|
for (;;) {
|
|||
|
if (off >= max - 10) { /* have some slack for misc stuff */
|
|||
|
char* x = REALLOC(dst, max += 50);
|
|||
|
if (!x) {
|
|||
|
xfree(dst);
|
|||
|
snprintf(errbuf, errbufsz, "out of memory");
|
|||
|
return 0;
|
|||
|
}
|
|||
|
dst = x;
|
|||
|
}
|
|||
|
|
|||
|
/* finished? */
|
|||
|
if (sp >= sq) break;
|
|||
|
|
|||
|
ch = *sp++;
|
|||
|
/* control characters other than tab is not allowed */
|
|||
|
if ((0 <= ch && ch <= 0x08)
|
|||
|
|| (0x0a <= ch && ch <= 0x1f)
|
|||
|
|| (ch == 0x7f)) {
|
|||
|
if (! (multiline && (ch == '\r' || ch == '\n'))) {
|
|||
|
xfree(dst);
|
|||
|
snprintf(errbuf, errbufsz, "invalid char U+%04x", ch);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
// a plain copy suffice
|
|||
|
dst[off++] = ch;
|
|||
|
}
|
|||
|
|
|||
|
dst[off++] = 0;
|
|||
|
return dst;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
|
|||
|
|
|||
|
/*
|
|||
|
* Convert src to raw unescaped utf-8 string.
|
|||
|
* Returns NULL if error with errmsg in errbuf.
|
|||
|
*/
|
|||
|
static char* norm_basic_str(const char* src, int srclen,
|
|||
|
int multiline,
|
|||
|
char* errbuf, int errbufsz)
|
|||
|
{
|
|||
|
char* dst = 0; /* will write to dst[] and return it */
|
|||
|
int max = 0; /* max size of dst[] */
|
|||
|
int off = 0; /* cur offset in dst[] */
|
|||
|
const char* sp = src;
|
|||
|
const char* sq = src + srclen;
|
|||
|
int ch;
|
|||
|
|
|||
|
/* scan forward on src */
|
|||
|
for (;;) {
|
|||
|
if (off >= max - 10) { /* have some slack for misc stuff */
|
|||
|
char* x = REALLOC(dst, max += 50);
|
|||
|
if (!x) {
|
|||
|
xfree(dst);
|
|||
|
snprintf(errbuf, errbufsz, "out of memory");
|
|||
|
return 0;
|
|||
|
}
|
|||
|
dst = x;
|
|||
|
}
|
|||
|
|
|||
|
/* finished? */
|
|||
|
if (sp >= sq) break;
|
|||
|
|
|||
|
ch = *sp++;
|
|||
|
if (ch != '\\') {
|
|||
|
/* these chars must be escaped: U+0000 to U+0008, U+000A to U+001F, U+007F */
|
|||
|
if ((0 <= ch && ch <= 0x08)
|
|||
|
|| (0x0a <= ch && ch <= 0x1f)
|
|||
|
|| (ch == 0x7f)) {
|
|||
|
if (! (multiline && (ch == '\r' || ch == '\n'))) {
|
|||
|
xfree(dst);
|
|||
|
snprintf(errbuf, errbufsz, "invalid char U+%04x", ch);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
// a plain copy suffice
|
|||
|
dst[off++] = ch;
|
|||
|
continue;
|
|||
|
}
|
|||
|
|
|||
|
/* ch was backslash. we expect the escape char. */
|
|||
|
if (sp >= sq) {
|
|||
|
snprintf(errbuf, errbufsz, "last backslash is invalid");
|
|||
|
xfree(dst);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
|
|||
|
/* for multi-line, we want to kill line-ending-backslash ... */
|
|||
|
if (multiline) {
|
|||
|
|
|||
|
// if there is only whitespace after the backslash ...
|
|||
|
if (sp[strspn(sp, " \t\r")] == '\n') {
|
|||
|
/* skip all the following whitespaces */
|
|||
|
sp += strspn(sp, " \t\r\n");
|
|||
|
continue;
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
/* get the escaped char */
|
|||
|
ch = *sp++;
|
|||
|
switch (ch) {
|
|||
|
case 'u': case 'U':
|
|||
|
{
|
|||
|
int64_t ucs = 0;
|
|||
|
int nhex = (ch == 'u' ? 4 : 8);
|
|||
|
for (int i = 0; i < nhex; i++) {
|
|||
|
if (sp >= sq) {
|
|||
|
snprintf(errbuf, errbufsz, "\\%c expects %d hex chars", ch, nhex);
|
|||
|
xfree(dst);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
ch = *sp++;
|
|||
|
int v = ('0' <= ch && ch <= '9')
|
|||
|
? ch - '0'
|
|||
|
: (('A' <= ch && ch <= 'F') ? ch - 'A' + 10 : -1);
|
|||
|
if (-1 == v) {
|
|||
|
snprintf(errbuf, errbufsz, "invalid hex chars for \\u or \\U");
|
|||
|
xfree(dst);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
ucs = ucs * 16 + v;
|
|||
|
}
|
|||
|
int n = toml_ucs_to_utf8(ucs, &dst[off]);
|
|||
|
if (-1 == n) {
|
|||
|
snprintf(errbuf, errbufsz, "illegal ucs code in \\u or \\U");
|
|||
|
xfree(dst);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
off += n;
|
|||
|
}
|
|||
|
continue;
|
|||
|
|
|||
|
case 'b': ch = '\b'; break;
|
|||
|
case 't': ch = '\t'; break;
|
|||
|
case 'n': ch = '\n'; break;
|
|||
|
case 'f': ch = '\f'; break;
|
|||
|
case 'r': ch = '\r'; break;
|
|||
|
case '"': ch = '"'; break;
|
|||
|
case '\\': ch = '\\'; break;
|
|||
|
default:
|
|||
|
snprintf(errbuf, errbufsz, "illegal escape char \\%c", ch);
|
|||
|
xfree(dst);
|
|||
|
return 0;
|
|||
|
}
|
|||
|
|
|||
|
dst[off++] = ch;
|
|||
|
}
|
|||
|
|
|||
|
// Cap with NUL and return it.
|
|||
|
dst[off++] = 0;
|
|||
|
return dst;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/* Normalize a key. Convert all special chars to raw unescaped utf-8 chars. */
|
|||
|
static char* normalize_key(context_t* ctx, token_t strtok)
|
|||
|
{
|
|||
|
const char* sp = strtok.ptr;
|
|||
|
const char* sq = strtok.ptr + strtok.len;
|
|||
|
int lineno = strtok.lineno;
|
|||
|
char* ret;
|
|||
|
int ch = *sp;
|
|||
|
char ebuf[80];
|
|||
|
|
|||
|
/* handle quoted string */
|
|||
|
if (ch == '\'' || ch == '\"') {
|
|||
|
/* if ''' or """, take 3 chars off front and back. Else, take 1 char off. */
|
|||
|
int multiline = 0;
|
|||
|
if (sp[1] == ch && sp[2] == ch) {
|
|||
|
sp += 3, sq -= 3;
|
|||
|
multiline = 1;
|
|||
|
}
|
|||
|
else
|
|||
|
sp++, sq--;
|
|||
|
|
|||
|
if (ch == '\'') {
|
|||
|
/* for single quote, take it verbatim. */
|
|||
|
if (! (ret = STRNDUP(sp, sq - sp))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
} else {
|
|||
|
/* for double quote, we need to normalize */
|
|||
|
ret = norm_basic_str(sp, sq - sp, multiline, ebuf, sizeof(ebuf));
|
|||
|
if (!ret) {
|
|||
|
snprintf(ctx->errbuf, ctx->errbufsz, "line %d: %s", lineno, ebuf);
|
|||
|
longjmp(ctx->jmp, 1);
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
/* newlines are not allowed in keys */
|
|||
|
if (strchr(ret, '\n')) {
|
|||
|
xfree(ret);
|
|||
|
e_bad_key_error(ctx, lineno);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
return ret;
|
|||
|
}
|
|||
|
|
|||
|
/* for bare-key allow only this regex: [A-Za-z0-9_-]+ */
|
|||
|
const char* xp;
|
|||
|
for (xp = sp; xp != sq; xp++) {
|
|||
|
int k = *xp;
|
|||
|
if (isalnum(k)) continue;
|
|||
|
if (k == '_' || k == '-') continue;
|
|||
|
e_bad_key_error(ctx, lineno);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* dup and return it */
|
|||
|
if (! (ret = STRNDUP(sp, sq - sp))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
return ret;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/*
|
|||
|
* Look up key in tab. Return 0 if not found, or
|
|||
|
* 'v'alue, 'a'rray or 't'able depending on the element.
|
|||
|
*/
|
|||
|
static int check_key(toml_table_t* tab, const char* key,
|
|||
|
toml_keyval_t** ret_val,
|
|||
|
toml_array_t** ret_arr,
|
|||
|
toml_table_t** ret_tab)
|
|||
|
{
|
|||
|
int i;
|
|||
|
void* dummy;
|
|||
|
|
|||
|
if (!ret_tab) ret_tab = (toml_table_t**) &dummy;
|
|||
|
if (!ret_arr) ret_arr = (toml_array_t**) &dummy;
|
|||
|
if (!ret_val) ret_val = (toml_keyval_t**) &dummy;
|
|||
|
|
|||
|
*ret_tab = 0; *ret_arr = 0; *ret_val = 0;
|
|||
|
|
|||
|
for (i = 0; i < tab->nkval; i++) {
|
|||
|
if (0 == strcmp(key, tab->kval[i]->key)) {
|
|||
|
*ret_val = tab->kval[i];
|
|||
|
return 'v';
|
|||
|
}
|
|||
|
}
|
|||
|
for (i = 0; i < tab->narr; i++) {
|
|||
|
if (0 == strcmp(key, tab->arr[i]->key)) {
|
|||
|
*ret_arr = tab->arr[i];
|
|||
|
return 'a';
|
|||
|
}
|
|||
|
}
|
|||
|
for (i = 0; i < tab->ntab; i++) {
|
|||
|
if (0 == strcmp(key, tab->tab[i]->key)) {
|
|||
|
*ret_tab = tab->tab[i];
|
|||
|
return 't';
|
|||
|
}
|
|||
|
}
|
|||
|
return 0;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
static int key_kind(toml_table_t* tab, const char* key)
|
|||
|
{
|
|||
|
return check_key(tab, key, 0, 0, 0);
|
|||
|
}
|
|||
|
|
|||
|
/* Create a keyval in the table.
|
|||
|
*/
|
|||
|
static toml_keyval_t* create_keyval_in_table(context_t* ctx, toml_table_t* tab, token_t keytok)
|
|||
|
{
|
|||
|
/* first, normalize the key to be used for lookup.
|
|||
|
* remember to free it if we error out.
|
|||
|
*/
|
|||
|
char* newkey = normalize_key(ctx, keytok);
|
|||
|
|
|||
|
/* if key exists: error out. */
|
|||
|
toml_keyval_t* dest = 0;
|
|||
|
if (key_kind(tab, newkey)) {
|
|||
|
xfree(newkey);
|
|||
|
e_key_exists_error(ctx, keytok.lineno);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* make a new entry */
|
|||
|
int n = tab->nkval;
|
|||
|
toml_keyval_t** base;
|
|||
|
if (0 == (base = (toml_keyval_t**) REALLOC(tab->kval, (n+1) * sizeof(*base)))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
tab->kval = base;
|
|||
|
|
|||
|
if (0 == (base[n] = (toml_keyval_t*) CALLOC(1, sizeof(*base[n])))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
dest = tab->kval[tab->nkval++];
|
|||
|
|
|||
|
/* save the key in the new value struct */
|
|||
|
dest->key = newkey;
|
|||
|
return dest;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/* Create a table in the table.
|
|||
|
*/
|
|||
|
static toml_table_t* create_keytable_in_table(context_t* ctx, toml_table_t* tab, token_t keytok)
|
|||
|
{
|
|||
|
/* first, normalize the key to be used for lookup.
|
|||
|
* remember to free it if we error out.
|
|||
|
*/
|
|||
|
char* newkey = normalize_key(ctx, keytok);
|
|||
|
|
|||
|
/* if key exists: error out */
|
|||
|
toml_table_t* dest = 0;
|
|||
|
if (check_key(tab, newkey, 0, 0, &dest)) {
|
|||
|
xfree(newkey); /* don't need this anymore */
|
|||
|
|
|||
|
/* special case: if table exists, but was created implicitly ... */
|
|||
|
if (dest && dest->implicit) {
|
|||
|
/* we make it explicit now, and simply return it. */
|
|||
|
dest->implicit = 0;
|
|||
|
return dest;
|
|||
|
}
|
|||
|
e_key_exists_error(ctx, keytok.lineno);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* create a new table entry */
|
|||
|
int n = tab->ntab;
|
|||
|
toml_table_t** base;
|
|||
|
if (0 == (base = (toml_table_t**) REALLOC(tab->tab, (n+1) * sizeof(*base)))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
tab->tab = base;
|
|||
|
|
|||
|
if (0 == (base[n] = (toml_table_t*) CALLOC(1, sizeof(*base[n])))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
dest = tab->tab[tab->ntab++];
|
|||
|
|
|||
|
/* save the key in the new table struct */
|
|||
|
dest->key = newkey;
|
|||
|
return dest;
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/* Create an array in the table.
|
|||
|
*/
|
|||
|
static toml_array_t* create_keyarray_in_table(context_t* ctx,
|
|||
|
toml_table_t* tab,
|
|||
|
token_t keytok,
|
|||
|
char kind)
|
|||
|
{
|
|||
|
/* first, normalize the key to be used for lookup.
|
|||
|
* remember to free it if we error out.
|
|||
|
*/
|
|||
|
char* newkey = normalize_key(ctx, keytok);
|
|||
|
|
|||
|
/* if key exists: error out */
|
|||
|
if (key_kind(tab, newkey)) {
|
|||
|
xfree(newkey); /* don't need this anymore */
|
|||
|
e_key_exists_error(ctx, keytok.lineno);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* make a new array entry */
|
|||
|
int n = tab->narr;
|
|||
|
toml_array_t** base;
|
|||
|
if (0 == (base = (toml_array_t**) REALLOC(tab->arr, (n+1) * sizeof(*base)))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
tab->arr = base;
|
|||
|
|
|||
|
if (0 == (base[n] = (toml_array_t*) CALLOC(1, sizeof(*base[n])))) {
|
|||
|
xfree(newkey);
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
toml_array_t* dest = tab->arr[tab->narr++];
|
|||
|
|
|||
|
/* save the key in the new array struct */
|
|||
|
dest->key = newkey;
|
|||
|
dest->kind = kind;
|
|||
|
return dest;
|
|||
|
}
|
|||
|
|
|||
|
/* Create an array in an array
|
|||
|
*/
|
|||
|
static toml_array_t* create_array_in_array(context_t* ctx,
|
|||
|
toml_array_t* parent)
|
|||
|
{
|
|||
|
int n = parent->nelem;
|
|||
|
toml_array_t** base;
|
|||
|
if (0 == (base = (toml_array_t**) REALLOC(parent->u.arr, (n+1) * sizeof(*base)))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
parent->u.arr = base;
|
|||
|
|
|||
|
if (0 == (base[n] = (toml_array_t*) CALLOC(1, sizeof(*base[n])))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
return parent->u.arr[parent->nelem++];
|
|||
|
}
|
|||
|
|
|||
|
/* Create a table in an array
|
|||
|
*/
|
|||
|
static toml_table_t* create_table_in_array(context_t* ctx,
|
|||
|
toml_array_t* parent)
|
|||
|
{
|
|||
|
int n = parent->nelem;
|
|||
|
toml_table_t** base;
|
|||
|
if (0 == (base = (toml_table_t**) REALLOC(parent->u.tab, (n+1) * sizeof(*base)))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
parent->u.tab = base;
|
|||
|
|
|||
|
if (0 == (base[n] = (toml_table_t*) CALLOC(1, sizeof(*base[n])))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return 0; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
return parent->u.tab[parent->nelem++];
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
static void skip_newlines(context_t* ctx, int isdotspecial)
|
|||
|
{
|
|||
|
while (ctx->tok.tok == NEWLINE) {
|
|||
|
next_token(ctx, isdotspecial);
|
|||
|
if (ctx->tok.eof) break;
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
#define EAT_TOKEN(ctx, typ, isdotspecial) \
|
|||
|
if ((ctx)->tok.tok != typ) e_internal_error(ctx, FLINE); else next_token(ctx, isdotspecial)
|
|||
|
|
|||
|
|
|||
|
static void parse_keyval(context_t* ctx, toml_table_t* tab);
|
|||
|
|
|||
|
|
|||
|
/* We are at '{ ... }'.
|
|||
|
* Parse the table.
|
|||
|
*/
|
|||
|
static void parse_table(context_t* ctx, toml_table_t* tab)
|
|||
|
{
|
|||
|
EAT_TOKEN(ctx, LBRACE, 1);
|
|||
|
|
|||
|
for (;;) {
|
|||
|
if (ctx->tok.tok == NEWLINE) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "newline not allowed in inline table");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* until } */
|
|||
|
if (ctx->tok.tok == RBRACE) break;
|
|||
|
|
|||
|
if (ctx->tok.tok != STRING) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "syntax error");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
parse_keyval(ctx, tab);
|
|||
|
|
|||
|
if (ctx->tok.tok == NEWLINE) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "newline not allowed in inline table");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* on comma, continue to scan for next keyval */
|
|||
|
if (ctx->tok.tok == COMMA) {
|
|||
|
EAT_TOKEN(ctx, COMMA, 1);
|
|||
|
continue;
|
|||
|
}
|
|||
|
break;
|
|||
|
}
|
|||
|
|
|||
|
if (ctx->tok.tok != RBRACE) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "syntax error");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
EAT_TOKEN(ctx, RBRACE, 1);
|
|||
|
}
|
|||
|
|
|||
|
static int valtype(const char* val)
|
|||
|
{
|
|||
|
toml_timestamp_t ts;
|
|||
|
if (*val == '\'' || *val == '"') return 's';
|
|||
|
if (0 == toml_rtob(val, 0)) return 'b';
|
|||
|
if (0 == toml_rtoi(val, 0)) return 'i';
|
|||
|
if (0 == toml_rtod(val, 0)) return 'd';
|
|||
|
if (0 == toml_rtots(val, &ts)) {
|
|||
|
if (ts.year && ts.hour) return 'T'; /* timestamp */
|
|||
|
if (ts.year) return 'D'; /* date */
|
|||
|
return 't'; /* time */
|
|||
|
}
|
|||
|
return 'u'; /* unknown */
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/* We are at '[...]' */
|
|||
|
static void parse_array(context_t* ctx, toml_array_t* arr)
|
|||
|
{
|
|||
|
EAT_TOKEN(ctx, LBRACKET, 0);
|
|||
|
|
|||
|
for (;;) {
|
|||
|
skip_newlines(ctx, 0);
|
|||
|
|
|||
|
/* until ] */
|
|||
|
if (ctx->tok.tok == RBRACKET) break;
|
|||
|
|
|||
|
switch (ctx->tok.tok) {
|
|||
|
case STRING:
|
|||
|
{
|
|||
|
char* val = ctx->tok.ptr;
|
|||
|
int vlen = ctx->tok.len;
|
|||
|
|
|||
|
/* set array kind if this will be the first entry */
|
|||
|
if (arr->kind == 0) arr->kind = 'v';
|
|||
|
/* check array kind */
|
|||
|
if (arr->kind != 'v') {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno,
|
|||
|
"a string array can only contain strings");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
/* make a new value in array */
|
|||
|
char** tmp = (char**) REALLOC(arr->u.val, (arr->nelem+1) * sizeof(*tmp));
|
|||
|
if (!tmp) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
arr->u.val = tmp;
|
|||
|
if (! (val = STRNDUP(val, vlen))) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
arr->u.val[arr->nelem++] = val;
|
|||
|
|
|||
|
/* set array type if this is the first entry, or check that the types matched. */
|
|||
|
if (arr->nelem == 1)
|
|||
|
arr->type = valtype(arr->u.val[0]);
|
|||
|
else if (arr->type != valtype(val)) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno,
|
|||
|
"array type mismatch while processing array of values");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
EAT_TOKEN(ctx, STRING, 0);
|
|||
|
break;
|
|||
|
}
|
|||
|
|
|||
|
case LBRACKET:
|
|||
|
{ /* [ [array], [array] ... ] */
|
|||
|
/* set the array kind if this will be the first entry */
|
|||
|
if (arr->kind == 0) arr->kind = 'a';
|
|||
|
/* check array kind */
|
|||
|
if (arr->kind != 'a') {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno,
|
|||
|
"array type mismatch while processing array of arrays");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
parse_array(ctx, create_array_in_array(ctx, arr));
|
|||
|
break;
|
|||
|
}
|
|||
|
|
|||
|
case LBRACE:
|
|||
|
{ /* [ {table}, {table} ... ] */
|
|||
|
/* set the array kind if this will be the first entry */
|
|||
|
if (arr->kind == 0) arr->kind = 't';
|
|||
|
/* check array kind */
|
|||
|
if (arr->kind != 't') {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno,
|
|||
|
"array type mismatch while processing array of tables");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
parse_table(ctx, create_table_in_array(ctx, arr));
|
|||
|
break;
|
|||
|
}
|
|||
|
|
|||
|
default:
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "syntax error");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
skip_newlines(ctx, 0);
|
|||
|
|
|||
|
/* on comma, continue to scan for next element */
|
|||
|
if (ctx->tok.tok == COMMA) {
|
|||
|
EAT_TOKEN(ctx, COMMA, 0);
|
|||
|
continue;
|
|||
|
}
|
|||
|
break;
|
|||
|
}
|
|||
|
|
|||
|
if (ctx->tok.tok != RBRACKET) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "expect a right bracket");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
EAT_TOKEN(ctx, RBRACKET, 1);
|
|||
|
}
|
|||
|
|
|||
|
|
|||
|
/* handle lines like these:
|
|||
|
key = "value"
|
|||
|
key = [ array ]
|
|||
|
key = { table }
|
|||
|
*/
|
|||
|
static void parse_keyval(context_t* ctx, toml_table_t* tab)
|
|||
|
{
|
|||
|
token_t key = ctx->tok;
|
|||
|
EAT_TOKEN(ctx, STRING, 1);
|
|||
|
|
|||
|
if (ctx->tok.tok == DOT) {
|
|||
|
/* handle inline dotted key.
|
|||
|
e.g.
|
|||
|
physical.color = "orange"
|
|||
|
physical.shape = "round"
|
|||
|
*/
|
|||
|
toml_table_t* subtab = 0;
|
|||
|
{
|
|||
|
char* subtabstr = normalize_key(ctx, key);
|
|||
|
subtab = toml_table_in(tab, subtabstr);
|
|||
|
xfree(subtabstr);
|
|||
|
}
|
|||
|
if (!subtab) {
|
|||
|
subtab = create_keytable_in_table(ctx, tab, key);
|
|||
|
}
|
|||
|
next_token(ctx, 1);
|
|||
|
parse_keyval(ctx, subtab);
|
|||
|
return;
|
|||
|
}
|
|||
|
|
|||
|
if (ctx->tok.tok != EQUAL) {
|
|||
|
e_syntax_error(ctx, ctx->tok.lineno, "missing =");
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
next_token(ctx, 0);
|
|||
|
|
|||
|
switch (ctx->tok.tok) {
|
|||
|
case STRING:
|
|||
|
{ /* key = "value" */
|
|||
|
toml_keyval_t* keyval = create_keyval_in_table(ctx, tab, key);
|
|||
|
token_t val = ctx->tok;
|
|||
|
assert(keyval->val == 0);
|
|||
|
keyval->val = STRNDUP(val.ptr, val.len);
|
|||
|
if (! keyval->val) {
|
|||
|
e_outofmemory(ctx, FLINE);
|
|||
|
return; /* not reached */
|
|||
|
}
|
|||
|
|
|||
|
next_token(ctx, 1);
|
|||
|
|
|||
|
return;
|
|||
|
}
|
|||
|
|
|||
|
case LBRACKET:
|
|||
|
{ /* key = [ array ] */
|
|||
|
toml_array_t* arr = create_keyarray_in_table(ctx, tab, key, 0);
|
|||
|
parse_array(ctx, arr);
|
|||
|
return;
|
|||
|
}
|
|||
|
|
|||
|
case LBRACE:
|
|||
|