From f290245b41ff4c5c8844933634ac752f9d6aa1d5 Mon Sep 17 00:00:00 2001 From: Arindam Guha Date: Tue, 14 Mar 2023 15:11:51 +0100 Subject: [PATCH] erstattes submodul med kopi --- .gitmodules | 3 - tomlc99 | 1 - tomlc99/LICENSE | 22 + tomlc99/Makefile | 42 + tomlc99/README.md | 124 ++ tomlc99/test1/.gitignore | 3 + tomlc99/test1/README.md | 9 + tomlc99/test1/build.sh | 9 + tomlc99/test1/extra/array_of_tables.toml | 1 + tomlc99/test1/extra/inline_array.toml | 1 + tomlc99/test1/extra/inline_table.toml | 1 + tomlc99/test1/run.sh | 4 + tomlc99/test2/.gitignore | 1 + tomlc99/test2/Note.txt | 6 + tomlc99/test2/build.sh | 6 + tomlc99/test2/run.sh | 42 + tomlc99/toml.c | 2146 ++++++++++++++++++++++ tomlc99/toml.h | 150 ++ tomlc99/toml_cat.c | 193 ++ tomlc99/toml_json.c | 214 +++ tomlc99/unittest/Makefile | 11 + tomlc99/unittest/t1.c | 57 + 22 files changed, 3042 insertions(+), 4 deletions(-) delete mode 100644 .gitmodules delete mode 160000 tomlc99 create mode 100644 tomlc99/LICENSE create mode 100644 tomlc99/Makefile create mode 100644 tomlc99/README.md create mode 100644 tomlc99/test1/.gitignore create mode 100644 tomlc99/test1/README.md create mode 100644 tomlc99/test1/build.sh create mode 100644 tomlc99/test1/extra/array_of_tables.toml create mode 100644 tomlc99/test1/extra/inline_array.toml create mode 100644 tomlc99/test1/extra/inline_table.toml create mode 100644 tomlc99/test1/run.sh create mode 100644 tomlc99/test2/.gitignore create mode 100644 tomlc99/test2/Note.txt create mode 100644 tomlc99/test2/build.sh create mode 100644 tomlc99/test2/run.sh create mode 100644 tomlc99/toml.c create mode 100644 tomlc99/toml.h create mode 100644 tomlc99/toml_cat.c create mode 100644 tomlc99/toml_json.c create mode 100644 tomlc99/unittest/Makefile create mode 100644 tomlc99/unittest/t1.c diff --git a/.gitmodules b/.gitmodules deleted file mode 100644 index d07db8b..0000000 --- a/.gitmodules +++ /dev/null @@ -1,3 +0,0 @@ -[submodule "tomlc99"] - path = tomlc99 - url = https://github.com/cktan/tomlc99.git diff --git a/tomlc99 b/tomlc99 deleted file mode 160000 index 8949028..0000000 --- a/tomlc99 +++ /dev/null @@ -1 +0,0 @@ -Subproject commit 894902820a3ea2f1ec470cd7fe338bde54045cf5 diff --git a/tomlc99/LICENSE b/tomlc99/LICENSE new file mode 100644 index 0000000..a3292b1 --- /dev/null +++ b/tomlc99/LICENSE @@ -0,0 +1,22 @@ +MIT License + +Copyright (c) 2017 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. diff --git a/tomlc99/Makefile b/tomlc99/Makefile new file mode 100644 index 0000000..fe6a54c --- /dev/null +++ b/tomlc99/Makefile @@ -0,0 +1,42 @@ +HFILES = toml.h +CFILES = toml.c +OBJ = $(CFILES:.c=.o) +EXEC = toml_json toml_cat + +CFLAGS = -std=c99 -Wall -Wextra -fpic +LIB = libtoml.a +LIB_SHARED = libtoml.so + +# to compile for debug: make DEBUG=1 +# to compile for no debug: make +ifdef DEBUG + CFLAGS += -O0 -g +else + CFLAGS += -O2 -DNDEBUG +endif + + +all: $(LIB) $(LIB_SHARED) $(EXEC) + +*.o: $(HFILES) + +libtoml.a: toml.o + ar -rcs $@ $^ + +libtoml.so: toml.o + $(CC) -shared -o $@ $^ + +toml_json: toml_json.c $(LIB) + +toml_cat: toml_cat.c $(LIB) + +prefix ?= /usr/local + +install: all + install -d ${prefix}/include ${prefix}/lib + install toml.h ${prefix}/include + install $(LIB) ${prefix}/lib + install $(LIB_SHARED) ${prefix}/lib + +clean: + rm -f *.o $(EXEC) $(LIB) $(LIB_SHARED) diff --git a/tomlc99/README.md b/tomlc99/README.md new file mode 100644 index 0000000..c3ed79c --- /dev/null +++ b/tomlc99/README.md @@ -0,0 +1,124 @@ +# tomlc99 + +TOML in c99; v1.0 compliant. + +## Usage + +Please see the `toml.h` file for details. What follows is a simple example that +parses this config file: + +```toml +[server] + host = "www.example.com" + port = 80 +``` + +The steps for getting values from our file is usually : + +1. Parse the whole TOML file. +2. Get a single table from the file. +3. Find a value from the table. +4. Convert that value to the appropriate type, i.e., string, int, etc. +5. Then, free up that memory if needed. + +Below is an example of parsing the values from the example table. + +1. Parse the whole TOML file. + +```c +FILE* fp; +toml_table_t* conf; +char errbuf[200]; + +/* Open the file. */ +if (0 == (fp = fopen("path/to/file.toml", "r"))) { + return handle_error(); +} + +/* Run the file through the parser. */ +conf = toml_parse_file(fp, errbuf, sizeof(errbuf)); +if (0 == conf) { + return handle_error(); +} + +fclose(fp); + +/* Alternatively, use `toml_parse` which takes a string rather than a file. */ +conf = toml_parse("A null terminated string that is TOML\0", errbuf, sizeof(errbuf)); +``` + +2. Get a single table from the file. + +```c +toml_table_t* server; + +/* Locate the [server] table. */ +if (0 == (server = toml_table_in(conf, "server"))) { + return handle_error(); +} +``` + +3. Find a value from the table. +4. Convert that value to the appropriate type (I.E. string, int). + +```c +toml_raw_t raw; +char* host; +int64_t port; + +/* Extract 'host' config value. */ +if (0 == (raw = toml_raw_in(server, "host"))) { + return handle_error(); +} + +/* Convert the raw value into a string. */ +if (toml_rtos(raw, &host)) { + return handle_error(); +} + +/* Extract 'port' config value. */ +if (0 == (raw = toml_raw_in(server, "port"))) { + return handle_error(); +} + +/* Convert the raw value into an int. */ +if (toml_rtoi(raw, &port)) { + return handle_error(); +} +``` + +5. Then, free up that memory if needed. + +```c +/* Use `toml_free` on the table returned from `toml_parse[_file]`. */ +toml_free(conf); + +/* Free any values returned from `toml_rto*`. */ +free(host); +``` + +## Building + +A normal *make* suffices. Alternately, you can also simply include the +`toml.c` and `toml.h` files in your project. + +## Testing + +To test against the standard test set provided by BurntSushi/toml-test: + +```sh +% make +% cd test1 +% bash build.sh # do this once +% bash run.sh # this will run the test suite +``` + + +To test against the standard test set provided by iarna/toml: + +```sh +% make +% cd test2 +% bash build.sh # do this once +% bash run.sh # this will run the test suite +``` diff --git a/tomlc99/test1/.gitignore b/tomlc99/test1/.gitignore new file mode 100644 index 0000000..f39c742 --- /dev/null +++ b/tomlc99/test1/.gitignore @@ -0,0 +1,3 @@ +/goworkspace +/toml-test +/toml-test-decoder diff --git a/tomlc99/test1/README.md b/tomlc99/test1/README.md new file mode 100644 index 0000000..bdf6a30 --- /dev/null +++ b/tomlc99/test1/README.md @@ -0,0 +1,9 @@ +How to run the tests +=== + +``` +% bash build.sh +% bash run.sh +77 passed, 0 failed +``` + diff --git a/tomlc99/test1/build.sh b/tomlc99/test1/build.sh new file mode 100644 index 0000000..9007489 --- /dev/null +++ b/tomlc99/test1/build.sh @@ -0,0 +1,9 @@ + +DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" + +mkdir -p $DIR/goworkspace +export GOPATH=$DIR/goworkspace +go get github.com/BurntSushi/toml-test # install test suite +go get github.com/BurntSushi/toml/cmd/toml-test-decoder # e.g., install my parser +cp $GOPATH/bin/* . + diff --git a/tomlc99/test1/extra/array_of_tables.toml b/tomlc99/test1/extra/array_of_tables.toml new file mode 100644 index 0000000..86ed1f0 --- /dev/null +++ b/tomlc99/test1/extra/array_of_tables.toml @@ -0,0 +1 @@ +x = [ {'a'= 1}, {'a'= 2} ] diff --git a/tomlc99/test1/extra/inline_array.toml b/tomlc99/test1/extra/inline_array.toml new file mode 100644 index 0000000..a0a971d --- /dev/null +++ b/tomlc99/test1/extra/inline_array.toml @@ -0,0 +1 @@ +x = [1,2,3] diff --git a/tomlc99/test1/extra/inline_table.toml b/tomlc99/test1/extra/inline_table.toml new file mode 100644 index 0000000..7673f2e --- /dev/null +++ b/tomlc99/test1/extra/inline_table.toml @@ -0,0 +1 @@ +x = {'a'= 1, 'b'= 2 } diff --git a/tomlc99/test1/run.sh b/tomlc99/test1/run.sh new file mode 100644 index 0000000..2bb548b --- /dev/null +++ b/tomlc99/test1/run.sh @@ -0,0 +1,4 @@ +DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" +export GOPATH=$DIR/goworkspace # if it isn't already set +# $GOPATH/bin/toml-test $GOPATH/bin/toml-test-decoder # e.g., run tests on my parser +$GOPATH/bin/toml-test ../toml_json diff --git a/tomlc99/test2/.gitignore b/tomlc99/test2/.gitignore new file mode 100644 index 0000000..21335a0 --- /dev/null +++ b/tomlc99/test2/.gitignore @@ -0,0 +1 @@ +/toml-spec-tests \ No newline at end of file diff --git a/tomlc99/test2/Note.txt b/tomlc99/test2/Note.txt new file mode 100644 index 0000000..6ca7347 --- /dev/null +++ b/tomlc99/test2/Note.txt @@ -0,0 +1,6 @@ +Note that we utilize the 'jq' command to normalize json files for comparison. Depending the +verson of jq on your system, some tests may generate error that looks like this: + + parse error: Exceeds depth limit for parsing at line 1 + +This is an error in 'jq', and it does not indicate that the tests themselves fail. diff --git a/tomlc99/test2/build.sh b/tomlc99/test2/build.sh new file mode 100644 index 0000000..74da4d0 --- /dev/null +++ b/tomlc99/test2/build.sh @@ -0,0 +1,6 @@ +set -e + +DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" + +[ -d toml-spec-tests ] || git clone https://github.com/cktan/toml-spec-tests.git + diff --git a/tomlc99/test2/run.sh b/tomlc99/test2/run.sh new file mode 100644 index 0000000..e1d93df --- /dev/null +++ b/tomlc99/test2/run.sh @@ -0,0 +1,42 @@ +if ! (which jq >& /dev/null); then + echo "ERROR: please install the 'jq' utility" + exit 1 +fi + +# +# POSITIVE tests +# +for i in toml-spec-tests/values/*.toml; do + fname="$i" + ext="${fname##*.}" + fname="${fname%.*}" + echo -n $fname ' ' + res='[OK]' + if (../toml_json $fname.toml >& $fname.json.out); then + jq -S . $fname.json.out > t.json + mv t.json $fname.json.out + if [ -f $fname.json ]; then + if ! (diff $fname.json $fname.json.out >& /dev/null); then + res='[FAILED]' + else + rm -f $fname.json.out + fi + else + res='[??]' + fi + fi + echo ... $res +done + + +# +# NEGATIVE tests +# +for i in toml-spec-tests/errors/*.toml; do + echo -n $i ' ' + res='[OK]' + if (../toml_json $i >& $i.json.out); then + res='[FAILED]' + fi + echo ... $res +done diff --git a/tomlc99/toml.c b/tomlc99/toml.c new file mode 100644 index 0000000..a5c8052 --- /dev/null +++ b/tomlc99/toml.c @@ -0,0 +1,2146 @@ +/* + + 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 +#include +#include +#include +#include +#include +#include +#include +#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: + { /* key = { table } */ + toml_table_t* nxttab = create_keytable_in_table(ctx, tab, key); + parse_table(ctx, nxttab); + return; + } + + default: + e_syntax_error(ctx, ctx->tok.lineno, "syntax error"); + return; /* not reached */ + } +} + + +typedef struct tabpath_t tabpath_t; +struct tabpath_t { + int cnt; + token_t key[10]; +}; + +/* at [x.y.z] or [[x.y.z]] + * Scan forward and fill tabpath until it enters ] or ]] + * There will be at least one entry on return. + */ +static void fill_tabpath(context_t* ctx) +{ + int lineno = ctx->tok.lineno; + int i; + + /* clear tpath */ + for (i = 0; i < ctx->tpath.top; i++) { + char** p = &ctx->tpath.key[i]; + xfree(*p); + *p = 0; + } + ctx->tpath.top = 0; + + for (;;) { + if (ctx->tpath.top >= 10) { + e_syntax_error(ctx, lineno, "table path is too deep; max allowed is 10."); + return; /* not reached */ + } + + if (ctx->tok.tok != STRING) { + e_syntax_error(ctx, lineno, "invalid or missing key"); + return; /* not reached */ + } + + ctx->tpath.tok[ctx->tpath.top] = ctx->tok; + ctx->tpath.key[ctx->tpath.top] = normalize_key(ctx, ctx->tok); + ctx->tpath.top++; + + next_token(ctx, 1); + + if (ctx->tok.tok == RBRACKET) break; + + if (ctx->tok.tok != DOT) { + e_syntax_error(ctx, lineno, "invalid key"); + return; /* not reached */ + } + + next_token(ctx, 1); + } + + if (ctx->tpath.top <= 0) { + e_syntax_error(ctx, lineno, "empty table selector"); + return; /* not reached */ + } +} + + +/* Walk tabpath from the root, and create new tables on the way. + * Sets ctx->curtab to the final table. + */ +static void walk_tabpath(context_t* ctx) +{ + /* start from root */ + toml_table_t* curtab = ctx->root; + + for (int i = 0; i < ctx->tpath.top; i++) { + const char* key = ctx->tpath.key[i]; + + toml_keyval_t* nextval = 0; + toml_array_t* nextarr = 0; + toml_table_t* nexttab = 0; + switch (check_key(curtab, key, &nextval, &nextarr, &nexttab)) { + case 't': + /* found a table. nexttab is where we will go next. */ + break; + + case 'a': + /* found an array. nexttab is the last table in the array. */ + if (nextarr->kind != 't') { + e_internal_error(ctx, FLINE); + return; /* not reached */ + } + if (nextarr->nelem == 0) { + e_internal_error(ctx, FLINE); + return; /* not reached */ + } + nexttab = nextarr->u.tab[nextarr->nelem-1]; + break; + + case 'v': + e_key_exists_error(ctx, ctx->tpath.tok[i].lineno); + return; /* not reached */ + + default: + { /* Not found. Let's create an implicit table. */ + int n = curtab->ntab; + toml_table_t** base = (toml_table_t**) REALLOC(curtab->tab, (n+1) * sizeof(*base)); + if (0 == base) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + curtab->tab = base; + + if (0 == (base[n] = (toml_table_t*) CALLOC(1, sizeof(*base[n])))) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + + if (0 == (base[n]->key = STRDUP(key))) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + + nexttab = curtab->tab[curtab->ntab++]; + + /* tabs created by walk_tabpath are considered implicit */ + nexttab->implicit = 1; + } + break; + } + + /* switch to next tab */ + curtab = nexttab; + } + + /* save it */ + ctx->curtab = curtab; +} + + +/* handle lines like [x.y.z] or [[x.y.z]] */ +static void parse_select(context_t* ctx) +{ + assert(ctx->tok.tok == LBRACKET); + + /* true if [[ */ + int llb = (ctx->tok.ptr + 1 < ctx->stop && ctx->tok.ptr[1] == '['); + /* need to detect '[[' on our own because next_token() will skip whitespace, + and '[ [' would be taken as '[[', which is wrong. */ + + /* eat [ or [[ */ + EAT_TOKEN(ctx, LBRACKET, 1); + if (llb) { + assert(ctx->tok.tok == LBRACKET); + EAT_TOKEN(ctx, LBRACKET, 1); + } + + fill_tabpath(ctx); + + /* For [x.y.z] or [[x.y.z]], remove z from tpath. + */ + token_t z = ctx->tpath.tok[ctx->tpath.top-1]; + xfree(ctx->tpath.key[ctx->tpath.top-1]); + ctx->tpath.top--; + + /* set up ctx->curtab */ + walk_tabpath(ctx); + + if (! llb) { + /* [x.y.z] -> create z = {} in x.y */ + ctx->curtab = create_keytable_in_table(ctx, ctx->curtab, z); + } else { + /* [[x.y.z]] -> create z = [] in x.y */ + toml_array_t* arr = 0; + { + char* zstr = normalize_key(ctx, z); + arr = toml_array_in(ctx->curtab, zstr); + xfree(zstr); + } + if (!arr) { + arr = create_keyarray_in_table(ctx, ctx->curtab, z, 't'); + if (!arr) { + e_internal_error(ctx, FLINE); + return; + } + } + if (arr->kind != 't') { + e_syntax_error(ctx, z.lineno, "array mismatch"); + return; /* not reached */ + } + + /* add to z[] */ + toml_table_t* dest; + { + int n = arr->nelem; + toml_table_t** base = REALLOC(arr->u.tab, (n+1) * sizeof(*base)); + if (0 == base) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + arr->u.tab = base; + + if (0 == (base[n] = CALLOC(1, sizeof(*base[n])))) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + + if (0 == (base[n]->key = STRDUP("__anon__"))) { + e_outofmemory(ctx, FLINE); + return; /* not reached */ + } + + dest = arr->u.tab[arr->nelem++]; + } + + ctx->curtab = dest; + } + + if (ctx->tok.tok != RBRACKET) { + e_syntax_error(ctx, ctx->tok.lineno, "expects ]"); + return; /* not reached */ + } + if (llb) { + if (! (ctx->tok.ptr + 1 < ctx->stop && ctx->tok.ptr[1] == ']')) { + e_syntax_error(ctx, ctx->tok.lineno, "expects ]]"); + return; /* not reached */ + } + EAT_TOKEN(ctx, RBRACKET, 1); + } + EAT_TOKEN(ctx, RBRACKET, 1); + + if (ctx->tok.tok != NEWLINE) { + e_syntax_error(ctx, ctx->tok.lineno, "extra chars after ] or ]]"); + return; /* not reached */ + } +} + + + + +toml_table_t* toml_parse(char* conf, + char* errbuf, + int errbufsz) +{ + context_t ctx; + + // clear errbuf + if (errbufsz <= 0) errbufsz = 0; + if (errbufsz > 0) errbuf[0] = 0; + + // init context + memset(&ctx, 0, sizeof(ctx)); + ctx.start = conf; + ctx.stop = ctx.start + strlen(conf); + ctx.errbuf = errbuf; + ctx.errbufsz = errbufsz; + + // start with an artificial newline of length 0 + ctx.tok.tok = NEWLINE; + ctx.tok.lineno = 1; + ctx.tok.ptr = conf; + ctx.tok.len = 0; + + // make a root table + if (0 == (ctx.root = CALLOC(1, sizeof(*ctx.root)))) { + /* do not call outofmemory() here... setjmp not done yet */ + snprintf(ctx.errbuf, ctx.errbufsz, "ERROR: out of memory (%s)", FLINE); + return 0; + } + + // set root as default table + ctx.curtab = ctx.root; + + if (0 != setjmp(ctx.jmp)) { + // Got here from a long_jmp. Something bad has happened. + // Free resources and return error. + for (int i = 0; i < ctx.tpath.top; i++) xfree(ctx.tpath.key[i]); + toml_free(ctx.root); + return 0; + } + + /* Scan forward until EOF */ + for (token_t tok = ctx.tok; ! tok.eof ; tok = ctx.tok) { + switch (tok.tok) { + + case NEWLINE: + next_token(&ctx, 1); + break; + + case STRING: + parse_keyval(&ctx, ctx.curtab); + if (ctx.tok.tok != NEWLINE) { + e_syntax_error(&ctx, ctx.tok.lineno, "extra chars after value"); + return 0; /* not reached */ + } + + EAT_TOKEN(&ctx, NEWLINE, 1); + break; + + case LBRACKET: /* [ x.y.z ] or [[ x.y.z ]] */ + parse_select(&ctx); + break; + + default: + snprintf(ctx.errbuf, ctx.errbufsz, "line %d: syntax error", tok.lineno); + longjmp(ctx.jmp, 1); + } + } + + /* success */ + for (int i = 0; i < ctx.tpath.top; i++) xfree(ctx.tpath.key[i]); + return ctx.root; +} + + +toml_table_t* toml_parse_file(FILE* fp, + char* errbuf, + int errbufsz) +{ + int bufsz = 0; + char* buf = 0; + int off = 0; + + /* prime the buf[] */ + bufsz = 1000; + if (! (buf = MALLOC(bufsz + 1))) { + snprintf(errbuf, errbufsz, "out of memory"); + return 0; + } + + /* read from fp into buf */ + while (! feof(fp)) { + bufsz += 1000; + + /* Allocate 1 extra byte because we will tag on a NUL */ + char* x = REALLOC(buf, bufsz + 1); + if (!x) { + snprintf(errbuf, errbufsz, "out of memory"); + xfree(buf); + return 0; + } + buf = x; + + errno = 0; + int n = fread(buf + off, 1, bufsz - off, fp); + if (ferror(fp)) { + snprintf(errbuf, errbufsz, "%s", + errno ? strerror(errno) : "Error reading file"); + xfree(buf); + return 0; + } + off += n; + } + + /* tag on a NUL to cap the string */ + buf[off] = 0; /* we accounted for this byte in the REALLOC() above. */ + + /* parse it, cleanup and finish */ + toml_table_t* ret = toml_parse(buf, errbuf, errbufsz); + xfree(buf); + return ret; +} + + +static void xfree_kval(toml_keyval_t* p) +{ + if (!p) return; + xfree(p->key); + xfree(p->val); + xfree(p); +} + +static void xfree_tab(toml_table_t* p); + +static void xfree_arr(toml_array_t* p) +{ + if (!p) return; + + xfree(p->key); + switch (p->kind) { + case 'v': + for (int i = 0; i < p->nelem; i++) xfree(p->u.val[i]); + xfree(p->u.val); + break; + + case 'a': + for (int i = 0; i < p->nelem; i++) xfree_arr(p->u.arr[i]); + xfree(p->u.arr); + break; + + case 't': + for (int i = 0; i < p->nelem; i++) xfree_tab(p->u.tab[i]); + xfree(p->u.tab); + break; + } + + xfree(p); +} + + +static void xfree_tab(toml_table_t* p) +{ + int i; + + if (!p) return; + + xfree(p->key); + + for (i = 0; i < p->nkval; i++) xfree_kval(p->kval[i]); + xfree(p->kval); + + for (i = 0; i < p->narr; i++) xfree_arr(p->arr[i]); + xfree(p->arr); + + for (i = 0; i < p->ntab; i++) xfree_tab(p->tab[i]); + xfree(p->tab); + + xfree(p); +} + + +void toml_free(toml_table_t* tab) +{ + xfree_tab(tab); +} + + +static tokentype_t ret_token(context_t* ctx, tokentype_t tok, int lineno, char* ptr, int len) +{ + token_t t; + t.tok = tok; + t.lineno = lineno; + t.ptr = ptr; + t.len = len; + t.eof = 0; + ctx->tok = t; + return tok; +} + +static tokentype_t ret_eof(context_t* ctx, int lineno) +{ + ret_token(ctx, NEWLINE, lineno, ctx->stop, 0); + ctx->tok.eof = 1; + return ctx->tok.tok; +} + + +/* Scan p for n digits compositing entirely of [0-9] */ +static int scan_digits(const char* p, int n) +{ + int ret = 0; + for ( ; n > 0 && isdigit(*p); n--, p++) { + ret = 10 * ret + (*p - '0'); + } + return n ? -1 : ret; +} + +static int scan_date(const char* p, int* YY, int* MM, int* DD) +{ + int year, month, day; + year = scan_digits(p, 4); + month = (year >= 0 && p[4] == '-') ? scan_digits(p+5, 2) : -1; + day = (month >= 0 && p[7] == '-') ? scan_digits(p+8, 2) : -1; + if (YY) *YY = year; + if (MM) *MM = month; + if (DD) *DD = day; + return (year >= 0 && month >= 0 && day >= 0) ? 0 : -1; +} + +static int scan_time(const char* p, int* hh, int* mm, int* ss) +{ + int hour, minute, second; + hour = scan_digits(p, 2); + minute = (hour >= 0 && p[2] == ':') ? scan_digits(p+3, 2) : -1; + second = (minute >= 0 && p[5] == ':') ? scan_digits(p+6, 2) : -1; + if (hh) *hh = hour; + if (mm) *mm = minute; + if (ss) *ss = second; + return (hour >= 0 && minute >= 0 && second >= 0) ? 0 : -1; +} + + +static tokentype_t scan_string(context_t* ctx, char* p, int lineno, int dotisspecial) +{ + char* orig = p; + if (0 == strncmp(p, "'''", 3)) { + p = strstr(p + 3, "'''"); + if (0 == p) { + e_syntax_error(ctx, lineno, "unterminated triple-s-quote"); + return 0; /* not reached */ + } + + return ret_token(ctx, STRING, lineno, orig, p + 3 - orig); + } + + if (0 == strncmp(p, "\"\"\"", 3)) { + int hexreq = 0; /* #hex required */ + int escape = 0; + int qcnt = 0; /* count quote */ + for (p += 3; *p && qcnt < 3; p++) { + if (escape) { + escape = 0; + if (strchr("btnfr\"\\", *p)) continue; + if (*p == 'u') { hexreq = 4; continue; } + if (*p == 'U') { hexreq = 8; continue; } + if (p[strspn(p, " \t\r")] == '\n') continue; /* allow for line ending backslash */ + e_syntax_error(ctx, lineno, "bad escape char"); + return 0; /* not reached */ + } + if (hexreq) { + hexreq--; + if (strchr("0123456789ABCDEF", *p)) continue; + e_syntax_error(ctx, lineno, "expect hex char"); + return 0; /* not reached */ + } + if (*p == '\\') { escape = 1; continue; } + qcnt = (*p == '"') ? qcnt + 1 : 0; + } + if (qcnt != 3) { + e_syntax_error(ctx, lineno, "unterminated triple-quote"); + return 0; /* not reached */ + } + + return ret_token(ctx, STRING, lineno, orig, p - orig); + } + + if ('\'' == *p) { + for (p++; *p && *p != '\n' && *p != '\''; p++); + if (*p != '\'') { + e_syntax_error(ctx, lineno, "unterminated s-quote"); + return 0; /* not reached */ + } + + return ret_token(ctx, STRING, lineno, orig, p + 1 - orig); + } + + if ('\"' == *p) { + int hexreq = 0; /* #hex required */ + int escape = 0; + for (p++; *p; p++) { + if (escape) { + escape = 0; + if (strchr("btnfr\"\\", *p)) continue; + if (*p == 'u') { hexreq = 4; continue; } + if (*p == 'U') { hexreq = 8; continue; } + e_syntax_error(ctx, lineno, "bad escape char"); + return 0; /* not reached */ + } + if (hexreq) { + hexreq--; + if (strchr("0123456789ABCDEF", *p)) continue; + e_syntax_error(ctx, lineno, "expect hex char"); + return 0; /* not reached */ + } + if (*p == '\\') { escape = 1; continue; } + if (*p == '\n') break; + if (*p == '"') break; + } + if (*p != '"') { + e_syntax_error(ctx, lineno, "unterminated quote"); + return 0; /* not reached */ + } + + return ret_token(ctx, STRING, lineno, orig, p + 1 - orig); + } + + /* check for timestamp without quotes */ + if (0 == scan_date(p, 0, 0, 0) || 0 == scan_time(p, 0, 0, 0)) { + // forward thru the timestamp + for ( ; strchr("0123456789.:+-T Z", toupper(*p)); p++); + // squeeze out any spaces at end of string + for ( ; p[-1] == ' '; p--); + // tokenize + return ret_token(ctx, STRING, lineno, orig, p - orig); + } + + /* literals */ + for ( ; *p && *p != '\n'; p++) { + int ch = *p; + if (ch == '.' && dotisspecial) break; + if ('A' <= ch && ch <= 'Z') continue; + if ('a' <= ch && ch <= 'z') continue; + if (strchr("0123456789+-_.", ch)) continue; + break; + } + + return ret_token(ctx, STRING, lineno, orig, p - orig); +} + + +static tokentype_t next_token(context_t* ctx, int dotisspecial) +{ + int lineno = ctx->tok.lineno; + char* p = ctx->tok.ptr; + int i; + + /* eat this tok */ + for (i = 0; i < ctx->tok.len; i++) { + if (*p++ == '\n') + lineno++; + } + + /* make next tok */ + while (p < ctx->stop) { + /* skip comment. stop just before the \n. */ + if (*p == '#') { + for (p++; p < ctx->stop && *p != '\n'; p++); + continue; + } + + if (dotisspecial && *p == '.') + return ret_token(ctx, DOT, lineno, p, 1); + + switch (*p) { + case ',': return ret_token(ctx, COMMA, lineno, p, 1); + case '=': return ret_token(ctx, EQUAL, lineno, p, 1); + case '{': return ret_token(ctx, LBRACE, lineno, p, 1); + case '}': return ret_token(ctx, RBRACE, lineno, p, 1); + case '[': return ret_token(ctx, LBRACKET, lineno, p, 1); + case ']': return ret_token(ctx, RBRACKET, lineno, p, 1); + case '\n': return ret_token(ctx, NEWLINE, lineno, p, 1); + case '\r': case ' ': case '\t': + /* ignore white spaces */ + p++; + continue; + } + + return scan_string(ctx, p, lineno, dotisspecial); + } + + return ret_eof(ctx, lineno); +} + + +const char* toml_key_in(const toml_table_t* tab, int keyidx) +{ + if (keyidx < tab->nkval) return tab->kval[keyidx]->key; + + keyidx -= tab->nkval; + if (keyidx < tab->narr) return tab->arr[keyidx]->key; + + keyidx -= tab->narr; + if (keyidx < tab->ntab) return tab->tab[keyidx]->key; + + return 0; +} + +toml_raw_t toml_raw_in(const toml_table_t* tab, const char* key) +{ + int i; + for (i = 0; i < tab->nkval; i++) { + if (0 == strcmp(key, tab->kval[i]->key)) + return tab->kval[i]->val; + } + return 0; +} + +toml_array_t* toml_array_in(const toml_table_t* tab, const char* key) +{ + int i; + for (i = 0; i < tab->narr; i++) { + if (0 == strcmp(key, tab->arr[i]->key)) + return tab->arr[i]; + } + return 0; +} + + +toml_table_t* toml_table_in(const toml_table_t* tab, const char* key) +{ + int i; + for (i = 0; i < tab->ntab; i++) { + if (0 == strcmp(key, tab->tab[i]->key)) + return tab->tab[i]; + } + return 0; +} + +toml_raw_t toml_raw_at(const toml_array_t* arr, int idx) +{ + if (arr->kind != 'v') + return 0; + if (! (0 <= idx && idx < arr->nelem)) + return 0; + return arr->u.val[idx]; +} + +char toml_array_kind(const toml_array_t* arr) +{ + return arr->kind; +} + +char toml_array_type(const toml_array_t* arr) +{ + if (arr->kind != 'v') + return 0; + + if (arr->nelem == 0) + return 0; + + return arr->type; +} + + +int toml_array_nelem(const toml_array_t* arr) +{ + return arr->nelem; +} + +const char* toml_array_key(const toml_array_t* arr) +{ + return arr ? arr->key : (const char*) NULL; +} + +int toml_table_nkval(const toml_table_t* tab) +{ + return tab->nkval; +} + +int toml_table_narr(const toml_table_t* tab) +{ + return tab->narr; +} + +int toml_table_ntab(const toml_table_t* tab) +{ + return tab->ntab; +} + +const char* toml_table_key(const toml_table_t* tab) +{ + return tab ? tab->key : (const char*) NULL; +} + +toml_array_t* toml_array_at(const toml_array_t* arr, int idx) +{ + if (arr->kind != 'a') + return 0; + if (! (0 <= idx && idx < arr->nelem)) + return 0; + return arr->u.arr[idx]; +} + +toml_table_t* toml_table_at(const toml_array_t* arr, int idx) +{ + if (arr->kind != 't') + return 0; + if (! (0 <= idx && idx < arr->nelem)) + return 0; + return arr->u.tab[idx]; +} + + +int toml_rtots(toml_raw_t src_, toml_timestamp_t* ret) +{ + if (! src_) return -1; + + const char* p = src_; + int must_parse_time = 0; + + memset(ret, 0, sizeof(*ret)); + + int* year = &ret->__buffer.year; + int* month = &ret->__buffer.month; + int* day = &ret->__buffer.day; + int* hour = &ret->__buffer.hour; + int* minute = &ret->__buffer.minute; + int* second = &ret->__buffer.second; + int* millisec = &ret->__buffer.millisec; + + /* parse date YYYY-MM-DD */ + if (0 == scan_date(p, year, month, day)) { + ret->year = year; + ret->month = month; + ret->day = day; + + p += 10; + if (*p) { + // parse the T or space separator + if (*p != 'T' && *p != ' ') return -1; + must_parse_time = 1; + p++; + } + } + + /* parse time HH:MM:SS */ + if (0 == scan_time(p, hour, minute, second)) { + ret->hour = hour; + ret->minute = minute; + ret->second = second; + + /* optionally, parse millisec */ + p += 8; + if (*p == '.') { + char* qq; + p++; + errno = 0; + *millisec = strtol(p, &qq, 0); + if (errno) { + return -1; + } + while (*millisec > 999) { + *millisec /= 10; + } + + ret->millisec = millisec; + p = qq; + } + + if (*p) { + /* parse and copy Z */ + char* z = ret->__buffer.z; + ret->z = z; + if (*p == 'Z' || *p == 'z') { + *z++ = 'Z'; p++; + *z = 0; + + } else if (*p == '+' || *p == '-') { + *z++ = *p++; + + if (! (isdigit(p[0]) && isdigit(p[1]))) return -1; + *z++ = *p++; + *z++ = *p++; + + if (*p == ':') { + *z++ = *p++; + + if (! (isdigit(p[0]) && isdigit(p[1]))) return -1; + *z++ = *p++; + *z++ = *p++; + } + + *z = 0; + } + } + } + if (*p != 0) + return -1; + + if (must_parse_time && !ret->hour) + return -1; + + return 0; +} + + +/* Raw to boolean */ +int toml_rtob(toml_raw_t src, int* ret_) +{ + if (!src) return -1; + int dummy; + int* ret = ret_ ? ret_ : &dummy; + + if (0 == strcmp(src, "true")) { + *ret = 1; + return 0; + } + if (0 == strcmp(src, "false")) { + *ret = 0; + return 0; + } + return -1; +} + + +/* Raw to integer */ +int toml_rtoi(toml_raw_t src, int64_t* ret_) +{ + if (!src) return -1; + + char buf[100]; + char* p = buf; + char* q = p + sizeof(buf); + const char* s = src; + int base = 0; + int64_t dummy; + int64_t* ret = ret_ ? ret_ : &dummy; + + + /* allow +/- */ + if (s[0] == '+' || s[0] == '-') + *p++ = *s++; + + /* disallow +_100 */ + if (s[0] == '_') + return -1; + + /* if 0 ... */ + if ('0' == s[0]) { + switch (s[1]) { + case 'x': base = 16; s += 2; break; + case 'o': base = 8; s += 2; break; + case 'b': base = 2; s += 2; break; + case '\0': return *ret = 0, 0; + default: + /* ensure no other digits after it */ + if (s[1]) return -1; + } + } + + /* just strip underscores and pass to strtoll */ + while (*s && p < q) { + int ch = *s++; + switch (ch) { + case '_': + // disallow '__' + if (s[0] == '_') return -1; + continue; /* skip _ */ + default: + break; + } + *p++ = ch; + } + if (*s || p == q) return -1; + + /* last char cannot be '_' */ + if (s[-1] == '_') return -1; + + /* cap with NUL */ + *p = 0; + + /* Run strtoll on buf to get the integer */ + char* endp; + errno = 0; + *ret = strtoll(buf, &endp, base); + return (errno || *endp) ? -1 : 0; +} + + +int toml_rtod_ex(toml_raw_t src, double* ret_, char* buf, int buflen) +{ + if (!src) return -1; + + char* p = buf; + char* q = p + buflen; + const char* s = src; + double dummy; + double* ret = ret_ ? ret_ : &dummy; + + + /* allow +/- */ + if (s[0] == '+' || s[0] == '-') + *p++ = *s++; + + /* disallow +_1.00 */ + if (s[0] == '_') + return -1; + + /* disallow +.99 */ + if (s[0] == '.') + return -1; + + /* zero must be followed by . or 'e', or NUL */ + if (s[0] == '0' && s[1] && !strchr("eE.", s[1])) + return -1; + + /* just strip underscores and pass to strtod */ + while (*s && p < q) { + int ch = *s++; + switch (ch) { + case '.': + if (s[-2] == '_') return -1; + if (s[0] == '_') return -1; + break; + case '_': + // disallow '__' + if (s[0] == '_') return -1; + continue; /* skip _ */ + default: + break; + } + *p++ = ch; + } + if (*s || p == q) return -1; /* reached end of string or buffer is full? */ + + /* last char cannot be '_' */ + if (s[-1] == '_') return -1; + + if (p != buf && p[-1] == '.') + return -1; /* no trailing zero */ + + /* cap with NUL */ + *p = 0; + + /* Run strtod on buf to get the value */ + char* endp; + errno = 0; + *ret = strtod(buf, &endp); + return (errno || *endp) ? -1 : 0; +} + +int toml_rtod(toml_raw_t src, double* ret_) +{ + char buf[100]; + return toml_rtod_ex(src, ret_, buf, sizeof(buf)); +} + + + + +int toml_rtos(toml_raw_t src, char** ret) +{ + int multiline = 0; + const char* sp; + const char* sq; + + *ret = 0; + if (!src) return -1; + + int qchar = src[0]; + int srclen = strlen(src); + if (! (qchar == '\'' || qchar == '"')) { + return -1; + } + + // triple quotes? + if (qchar == src[1] && qchar == src[2]) { + multiline = 1; + sp = src + 3; + sq = src + srclen - 3; + /* last 3 chars in src must be qchar */ + if (! (sp <= sq && sq[0] == qchar && sq[1] == qchar && sq[2] == qchar)) + return -1; + + /* skip new line immediate after qchar */ + if (sp[0] == '\n') + sp++; + else if (sp[0] == '\r' && sp[1] == '\n') + sp += 2; + + } else { + sp = src + 1; + sq = src + srclen - 1; + /* last char in src must be qchar */ + if (! (sp <= sq && *sq == qchar)) + return -1; + } + + if (qchar == '\'') { + *ret = norm_lit_str(sp, sq - sp, + multiline, + 0, 0); + } else { + *ret = norm_basic_str(sp, sq - sp, + multiline, + 0, 0); + } + + return *ret ? 0 : -1; +} diff --git a/tomlc99/toml.h b/tomlc99/toml.h new file mode 100644 index 0000000..ff851ba --- /dev/null +++ b/tomlc99/toml.h @@ -0,0 +1,150 @@ +/* + 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. +*/ +#ifndef TOML_H +#define TOML_H + + +#include +#include + + +#ifdef __cplusplus +#define TOML_EXTERN extern "C" +#else +#define TOML_EXTERN extern +#endif + +typedef struct toml_table_t toml_table_t; +typedef struct toml_array_t toml_array_t; + +/* A raw value, must be processed by toml_rto* before using. */ +typedef const char* toml_raw_t; + +/* Parse a file. Return a table on success, or 0 otherwise. + * Caller must toml_free(the-return-value) after use. + */ +TOML_EXTERN toml_table_t* toml_parse_file(FILE* fp, + char* errbuf, + int errbufsz); + +/* Parse a string containing the full config. + * Return a table on success, or 0 otherwise. + * Caller must toml_free(the-return-value) after use. + */ +TOML_EXTERN toml_table_t* toml_parse(char* conf, /* NUL terminated, please. */ + char* errbuf, + int errbufsz); + +/* Free the table returned by toml_parse() or toml_parse_file(). */ +TOML_EXTERN void toml_free(toml_table_t* tab); + +/* Retrieve the key in table at keyidx. Return 0 if out of range. */ +TOML_EXTERN const char* toml_key_in(const toml_table_t* tab, int keyidx); + +/* Lookup table by key. Return the element or 0 if not found. */ +TOML_EXTERN toml_raw_t toml_raw_in(const toml_table_t* tab, const char* key); +TOML_EXTERN toml_array_t* toml_array_in(const toml_table_t* tab, + const char* key); +TOML_EXTERN toml_table_t* toml_table_in(const toml_table_t* tab, + const char* key); + +/* Return the array kind: 't'able, 'a'rray, 'v'alue */ +TOML_EXTERN char toml_array_kind(const toml_array_t* arr); + +/* For array kind 'v'alue, return the type of values + i:int, d:double, b:bool, s:string, t:time, D:date, T:timestamp + 0 if unknown +*/ +TOML_EXTERN char toml_array_type(const toml_array_t* arr); + + +/* Return the number of elements in the array */ +TOML_EXTERN int toml_array_nelem(const toml_array_t* arr); + +/* Return the key of an array */ +TOML_EXTERN const char* toml_array_key(const toml_array_t* arr); + +/* Return the number of key-values in a table */ +TOML_EXTERN int toml_table_nkval(const toml_table_t* tab); + +/* Return the number of arrays in a table */ +TOML_EXTERN int toml_table_narr(const toml_table_t* tab); + +/* Return the number of sub-tables in a table */ +TOML_EXTERN int toml_table_ntab(const toml_table_t* tab); + +/* Return the key of a table*/ +TOML_EXTERN const char* toml_table_key(const toml_table_t* tab); + +/* Deref array by index. Return the element at idx or 0 if out of range. */ +TOML_EXTERN toml_raw_t toml_raw_at(const toml_array_t* arr, int idx); +TOML_EXTERN toml_array_t* toml_array_at(const toml_array_t* arr, int idx); +TOML_EXTERN toml_table_t* toml_table_at(const toml_array_t* arr, int idx); + +/* Raw to String. Caller must call free(ret) after use. + * Return 0 on success, -1 otherwise. + */ +TOML_EXTERN int toml_rtos(toml_raw_t s, char** ret); + +/* Raw to Boolean. Return 0 on success, -1 otherwise. */ +TOML_EXTERN int toml_rtob(toml_raw_t s, int* ret); + +/* Raw to Integer. Return 0 on success, -1 otherwise. */ +TOML_EXTERN int toml_rtoi(toml_raw_t s, int64_t* ret); + +/* Raw to Double. Return 0 on success, -1 otherwise. */ +TOML_EXTERN int toml_rtod(toml_raw_t s, double* ret); + +/* Same as toml_rtod, but return the sanitized double in string form as well */ +TOML_EXTERN int toml_rtod_ex(toml_raw_t s, double* ret, char* buf, int buflen); + +/* Timestamp types. The year, month, day, hour, minute, second, z + * fields may be NULL if they are not relevant. e.g. In a DATE + * type, the hour, minute, second and z fields will be NULLs. + */ +typedef struct toml_timestamp_t toml_timestamp_t; +struct toml_timestamp_t { + struct { /* internal. do not use. */ + int year, month, day; + int hour, minute, second, millisec; + char z[10]; + } __buffer; + int *year, *month, *day; + int *hour, *minute, *second, *millisec; + char* z; +}; + +/* Raw to Timestamp. Return 0 on success, -1 otherwise. */ +TOML_EXTERN int toml_rtots(toml_raw_t s, toml_timestamp_t* ret); + +/* misc */ +TOML_EXTERN int toml_utf8_to_ucs(const char* orig, int len, int64_t* ret); +TOML_EXTERN int toml_ucs_to_utf8(int64_t code, char buf[6]); +TOML_EXTERN void toml_set_memutil(void* (*xxmalloc)(size_t), + void (*xxfree)(void*), + void* (*xxcalloc)(size_t, size_t), + void* (*xxrealloc)(void*, size_t)); + +#endif /* TOML_H */ diff --git a/tomlc99/toml_cat.c b/tomlc99/toml_cat.c new file mode 100644 index 0000000..186ebf4 --- /dev/null +++ b/tomlc99/toml_cat.c @@ -0,0 +1,193 @@ +/* +MIT License + +Copyright (c) 2017 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. +*/ + +#ifdef NDEBUG +#undef NDEBUG +#endif + +#include +#include +#include +#include +#include +#include +#include "toml.h" + +typedef struct node_t node_t; +struct node_t { + const char* key; + toml_table_t* tab; +}; + +node_t stack[20]; +int stacktop = 0; + + +static void print_table_title(const char* arrname) +{ + int i; + printf("%s", arrname ? "[[" : "["); + for (i = 1; i < stacktop; i++) { + printf("%s", stack[i].key); + if (i + 1 < stacktop) + printf("."); + } + if (arrname) + printf(".%s]]\n", arrname); + else + printf("]\n"); +} + + +static void print_array_of_tables(toml_array_t* arr, const char* key); +static void print_array(toml_array_t* arr); + + +static void print_table(toml_table_t* curtab) +{ + int i; + const char* key; + const char* raw; + toml_array_t* arr; + toml_table_t* tab; + + + for (i = 0; 0 != (key = toml_key_in(curtab, i)); i++) { + if (0 != (raw = toml_raw_in(curtab, key))) { + printf("%s = %s\n", key, raw); + } else if (0 != (arr = toml_array_in(curtab, key))) { + if (toml_array_kind(arr) == 't') { + print_array_of_tables(arr, key); + } + else { + printf("%s = [\n", key); + print_array(arr); + printf(" ]\n"); + } + } else if (0 != (tab = toml_table_in(curtab, key))) { + stack[stacktop].key = key; + stack[stacktop].tab = tab; + stacktop++; + print_table_title(0); + print_table(tab); + stacktop--; + } else { + abort(); + } + } +} + +static void print_array_of_tables(toml_array_t* arr, const char* key) +{ + int i; + toml_table_t* tab; + printf("\n"); + for (i = 0; 0 != (tab = toml_table_at(arr, i)); i++) { + print_table_title(key); + print_table(tab); + printf("\n"); + } +} + + +static void print_array(toml_array_t* curarr) +{ + toml_array_t* arr; + const char* raw; + toml_table_t* tab; + int i; + + switch (toml_array_kind(curarr)) { + + case 'v': + for (i = 0; 0 != (raw = toml_raw_at(curarr, i)); i++) { + printf(" %d: %s,\n", i, raw); + } + break; + + case 'a': + for (i = 0; 0 != (arr = toml_array_at(curarr, i)); i++) { + printf(" %d: \n", i); + print_array(arr); + } + break; + + case 't': + for (i = 0; 0 != (tab = toml_table_at(curarr, i)); i++) { + print_table(tab); + } + printf("\n"); + break; + + case '\0': + break; + + default: + abort(); + } +} + + + +static void cat(FILE* fp) +{ + char errbuf[200]; + + toml_table_t* tab = toml_parse_file(fp, errbuf, sizeof(errbuf)); + if (!tab) { + fprintf(stderr, "ERROR: %s\n", errbuf); + return; + } + + stack[stacktop].tab = tab; + stack[stacktop].key = ""; + stacktop++; + print_table(tab); + stacktop--; + + toml_free(tab); +} + + +int main(int argc, const char* argv[]) +{ + int i; + if (argc == 1) { + cat(stdin); + } else { + for (i = 1; i < argc; i++) { + + FILE* fp = fopen(argv[i], "r"); + if (!fp) { + fprintf(stderr, "ERROR: cannot open %s: %s\n", + argv[i], strerror(errno)); + exit(1); + } + cat(fp); + fclose(fp); + } + } + return 0; +} diff --git a/tomlc99/toml_json.c b/tomlc99/toml_json.c new file mode 100644 index 0000000..d443cb5 --- /dev/null +++ b/tomlc99/toml_json.c @@ -0,0 +1,214 @@ +/* + MIT License + + Copyright (c) 2017 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. +*/ +#ifdef NDEBUG +#undef NDEBUG +#endif + +#include +#include +#include +#include +#include +#include "toml.h" + + +static void print_escape_string(const char* s) +{ + for ( ; *s; s++) { + int ch = *s; + switch (ch) { + case '\b': printf("\\b"); break; + case '\t': printf("\\t"); break; + case '\n': printf("\\n"); break; + case '\f': printf("\\f"); break; + case '\r': printf("\\r"); break; + case '"': printf("\\\""); break; + case '\\': printf("\\\\"); break; + default: printf("%c", ch); break; + } + } +} + +static void print_raw(const char* s) +{ + char* sval; + int64_t ival; + int bval; + double dval; + toml_timestamp_t ts; + char dbuf[100]; + + if (0 == toml_rtos(s, &sval)) { + printf("{\"type\":\"string\",\"value\":\""); + print_escape_string(sval); + printf("\"}"); + free(sval); + } else if (0 == toml_rtoi(s, &ival)) { + printf("{\"type\":\"integer\",\"value\":\"%" PRId64 "\"}", ival); + } else if (0 == toml_rtob(s, &bval)) { + printf("{\"type\":\"bool\",\"value\":\"%s\"}", bval ? "true" : "false"); + } else if (0 == toml_rtod_ex(s, &dval, dbuf, sizeof(dbuf))) { + printf("{\"type\":\"float\",\"value\":\"%s\"}", dbuf); + } else if (0 == toml_rtots(s, &ts)) { + char millisec[10]; + if (ts.millisec) + sprintf(millisec, ".%d", *ts.millisec); + else + millisec[0] = 0; + if (ts.year && ts.hour) { + printf("{\"type\":\"datetime\",\"value\":\"%04d-%02d-%02dT%02d:%02d:%02d%s%s\"}", + *ts.year, *ts.month, *ts.day, *ts.hour, *ts.minute, *ts.second, + millisec, + (ts.z ? ts.z : "")); + } else if (ts.year) { + printf("{\"type\":\"date\",\"value\":\"%04d-%02d-%02d\"}", + *ts.year, *ts.month, *ts.day); + } else if (ts.hour) { + printf("{\"type\":\"time\",\"value\":\"%02d:%02d:%02d%s\"}", + *ts.hour, *ts.minute, *ts.second, millisec); + } + } else { + fprintf(stderr, "unknown type\n"); + exit(1); + } +} + + +static void print_array(toml_array_t* arr); +static void print_table(toml_table_t* curtab) +{ + int i; + const char* key; + const char* raw; + toml_array_t* arr; + toml_table_t* tab; + + + printf("{"); + for (i = 0; 0 != (key = toml_key_in(curtab, i)); i++) { + + printf("%s\"", i > 0 ? "," : ""); + print_escape_string(key); + printf("\":"); + + if (0 != (raw = toml_raw_in(curtab, key))) { + print_raw(raw); + } else if (0 != (arr = toml_array_in(curtab, key))) { + print_array(arr); + } else if (0 != (tab = toml_table_in(curtab, key))) { + print_table(tab); + } else { + abort(); + } + } + printf("}"); +} + +static void print_table_array(toml_array_t* curarr) +{ + int i; + toml_table_t* tab; + + printf("["); + for (i = 0; 0 != (tab = toml_table_at(curarr, i)); i++) { + printf("%s", i > 0 ? "," : ""); + print_table(tab); + } + printf("]"); +} + +static void print_array(toml_array_t* curarr) +{ + toml_array_t* arr; + const char* raw; + int i; + + if (toml_array_kind(curarr) == 't') { + print_table_array(curarr); + return; + } + + printf("{\"type\":\"array\",\"value\":["); + switch (toml_array_kind(curarr)) { + + case 'v': + for (i = 0; 0 != (raw = toml_raw_at(curarr, i)); i++) { + printf("%s", i > 0 ? "," : ""); + print_raw(raw); + } + break; + + case 'a': + for (i = 0; 0 != (arr = toml_array_at(curarr, i)); i++) { + printf("%s", i > 0 ? "," : ""); + print_array(arr); + } + break; + + default: + break; + } + printf("]}"); +} + + + +static void cat(FILE* fp) +{ + char errbuf[200]; + + toml_table_t* tab = toml_parse_file(fp, errbuf, sizeof(errbuf)); + if (!tab) { + fprintf(stderr, "ERROR: %s\n", errbuf); + exit(1); + } + + print_table(tab); + printf("\n"); + + toml_free(tab); +} + + +int main(int argc, const char* argv[]) +{ + int i; + if (argc == 1) { + cat(stdin); + } else { + for (i = 1; i < argc; i++) { + + FILE* fp = fopen(argv[i], "r"); + if (!fp) { + fprintf(stderr, "ERROR: cannot open %s: %s\n", + argv[i], strerror(errno)); + exit(1); + } + cat(fp); + fclose(fp); + } + } + return 0; +} diff --git a/tomlc99/unittest/Makefile b/tomlc99/unittest/Makefile new file mode 100644 index 0000000..0fb102f --- /dev/null +++ b/tomlc99/unittest/Makefile @@ -0,0 +1,11 @@ +CFLAGS = -g -I.. + +TESTS = t1 + +all: $(TESTS) + +t1: t1.c ../toml.c + +clean: + rm -f $(TESTS) + diff --git a/tomlc99/unittest/t1.c b/tomlc99/unittest/t1.c new file mode 100644 index 0000000..065ba0a --- /dev/null +++ b/tomlc99/unittest/t1.c @@ -0,0 +1,57 @@ +#include +#include +#include +#include +#include "../toml.h" + + + +int main(int argc, const char* argv[]) +{ + char xxbuf[6], buf[6]; + int64_t xxcode, code; + int xxsize; + + + xxsize = 2, xxcode = 0x80; memcpy(xxbuf, "\xc2\x80", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 2, xxcode = 0x7ff; memcpy(xxbuf, "\xdf\xbf", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 3, xxcode = 0x800; memcpy(xxbuf, "\xe0\xa0\x80", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 3, xxcode = 0xfffd; memcpy(xxbuf, "\xef\xbf\xbd", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 4, xxcode = 0x10000; memcpy(xxbuf, "\xf0\x90\x80\x80", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 4, xxcode = 0x1fffff; memcpy(xxbuf, "\xf7\xbf\xbf\xbf", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 5, xxcode = 0x200000; memcpy(xxbuf, "\xf8\x88\x80\x80\x80", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 5, xxcode = 0x3ffffff; memcpy(xxbuf, "\xfb\xbf\xbf\xbf\xbf", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 6, xxcode = 0x4000000; memcpy(xxbuf, "\xfc\x84\x80\x80\x80\x80", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + xxsize = 6, xxcode = 0x7fffffff; memcpy(xxbuf, "\xfd\xbf\xbf\xbf\xbf\xbf", xxsize); + assert(toml_ucs_to_utf8(xxcode, buf) == xxsize && 0 == memcmp(buf, xxbuf, xxsize)); + assert(toml_utf8_to_ucs(buf, xxsize, &code) == xxsize && code == xxcode); + + return 0; +}