diff options
Diffstat (limited to 'userspace/extern')
| -rw-r--r-- | userspace/extern/inc/cargs.h | 185 | ||||
| -rw-r--r-- | userspace/extern/inc/toml.h | 175 | ||||
| -rw-r--r-- | userspace/extern/src/cargs.c | 460 | ||||
| -rw-r--r-- | userspace/extern/src/toml.c | 2379 |
4 files changed, 3199 insertions, 0 deletions
diff --git a/userspace/extern/inc/cargs.h b/userspace/extern/inc/cargs.h new file mode 100644 index 0000000..7d9dacd --- /dev/null +++ b/userspace/extern/inc/cargs.h @@ -0,0 +1,185 @@ +/* +MIT License + +Copyright (c) 2022 Leonard Iklé + +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. + */ +#pragma once + +/** + * This is a simple alternative cross-platform implementation of getopt, which + * is used to parse argument strings submitted to the executable (argc and argv + * which are received in the main function). + */ + +#ifndef CAG_LIBRARY_H +#define CAG_LIBRARY_H + +#include <stdbool.h> +#include <stddef.h> +#include <stdio.h> + +#if defined(_WIN32) || defined(__CYGWIN__) +#define CAG_EXPORT __declspec(dllexport) +#define CAG_IMPORT __declspec(dllimport) +#elif __GNUC__ >= 4 +#define CAG_EXPORT __attribute__((visibility("default"))) +#define CAG_IMPORT __attribute__((visibility("default"))) +#else +#define CAG_EXPORT +#define CAG_IMPORT +#endif + +#if defined(CAG_SHARED) +#if defined(CAG_EXPORTS) +#define CAG_PUBLIC CAG_EXPORT +#else +#define CAG_PUBLIC CAG_IMPORT +#endif +#else +#define CAG_PUBLIC +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/** + * An option is used to describe a flag/argument option submitted when the + * program is run. + */ +typedef struct cag_option +{ + const char identifier; + const char *access_letters; + const char *access_name; + const char *value_name; + const char *description; +} cag_option; + +/** + * A context is used to iterate over all options provided. It stores the parsing + * state. + */ +typedef struct cag_option_context +{ + const struct cag_option *options; + size_t option_count; + int argc; + char **argv; + int index; + int inner_index; + bool forced_end; + char identifier; + char *value; +} cag_option_context; + +/** + * This is just a small macro which calculates the size of an array. + */ +#define CAG_ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0])) + +/** + * @brief Prints all options to the terminal. + * + * This function prints all options to the terminal. This can be used to + * generate the output for a "--help" option. + * + * @param options The options which will be printed. + * @param option_count The option count which will be printed. + * @param destination The destination where the output will be printed. + */ +CAG_PUBLIC void cag_option_print(const cag_option *options, size_t option_count, + FILE *destination); + +/** + * @brief Prepare argument options context for parsing. + * + * This function prepares the context for iteration and initializes the context + * with the supplied options and arguments. After the context has been prepared, + * it can be used to fetch arguments from it. + * + * @param context The context which will be initialized. + * @param options The registered options which are available for the program. + * @param option_count The amount of options which are available for the + * program. + * @param argc The amount of arguments the user supplied in the main function. + * @param argv A pointer to the arguments of the main function. + */ +CAG_PUBLIC void cag_option_prepare(cag_option_context *context, + const cag_option *options, size_t option_count, int argc, char **argv); + +/** + * @brief Fetches an option from the argument list. + * + * This function fetches a single option from the argument list. The context + * will be moved to that item. Information can be extracted from the context + * after the item has been fetched. + * The arguments will be re-ordered, which means that non-option arguments will + * be moved to the end of the argument list. After all options have been + * fetched, all non-option arguments will be positioned after the index of + * the context. + * + * @param context The context from which we will fetch the option. + * @return Returns true if there was another option or false if the end is + * reached. + */ +CAG_PUBLIC bool cag_option_fetch(cag_option_context *context); + +/** + * @brief Gets the identifier of the option. + * + * This function gets the identifier of the option, which should be unique to + * this option and can be used to determine what kind of option this is. + * + * @param context The context from which the option was fetched. + * @return Returns the identifier of the option. + */ +CAG_PUBLIC char cag_option_get(const cag_option_context *context); + +/** + * @brief Gets the value from the option. + * + * This function gets the value from the option, if any. If the option does not + * contain a value, this function will return NULL. + * + * @param context The context from which the option was fetched. + * @return Returns a pointer to the value or NULL if there is no value. + */ +CAG_PUBLIC const char *cag_option_get_value(const cag_option_context *context); + +/** + * @brief Gets the current index of the context. + * + * This function gets the index within the argv arguments of the context. The + * context always points to the next item which it will inspect. This is + * particularly useful to inspect the original argument array, or to get + * non-option arguments after option fetching has finished. + * + * @param context The context from which the option was fetched. + * @return Returns the current index of the context. + */ +CAG_PUBLIC int cag_option_get_index(const cag_option_context *context); + +#ifdef __cplusplus +} // extern "C" +#endif + +#endif diff --git a/userspace/extern/inc/toml.h b/userspace/extern/inc/toml.h new file mode 100644 index 0000000..19dc3d2 --- /dev/null +++ b/userspace/extern/inc/toml.h @@ -0,0 +1,175 @@ +/* + MIT License + + Copyright (c) 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 + +#ifdef _MSC_VER +#pragma warning(disable: 4996) +#endif + +#include <stdint.h> +#include <stdio.h> + +#ifdef __cplusplus +#define TOML_EXTERN extern "C" +#else +#define TOML_EXTERN extern +#endif + +typedef struct toml_timestamp_t toml_timestamp_t; +typedef struct toml_table_t toml_table_t; +typedef struct toml_array_t toml_array_t; +typedef struct toml_datum_t toml_datum_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(). Once + * this function is called, any handles accessed through this tab + * directly or indirectly are no longer valid. + */ +TOML_EXTERN void toml_free(toml_table_t *tab); + +/* 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. + */ +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; +}; + +/*----------------------------------------------------------------- + * Enhanced access methods + */ +struct toml_datum_t { + int ok; + union { + toml_timestamp_t *ts; /* ts must be freed after use */ + char *s; /* string value. s must be freed after use */ + int b; /* bool value */ + int64_t i; /* int value */ + double d; /* double value */ + } u; +}; + +/* on arrays: */ +/* ... retrieve size of array. */ +TOML_EXTERN int toml_array_nelem(const toml_array_t *arr); +/* ... retrieve values using index. */ +TOML_EXTERN toml_datum_t toml_string_at(const toml_array_t *arr, int idx); +TOML_EXTERN toml_datum_t toml_bool_at(const toml_array_t *arr, int idx); +TOML_EXTERN toml_datum_t toml_int_at(const toml_array_t *arr, int idx); +TOML_EXTERN toml_datum_t toml_double_at(const toml_array_t *arr, int idx); +TOML_EXTERN toml_datum_t toml_timestamp_at(const toml_array_t *arr, int idx); +/* ... retrieve array or table using index. */ +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); + +/* on tables: */ +/* ... 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); +/* ... returns 1 if key exists in tab, 0 otherwise */ +TOML_EXTERN int toml_key_exists(const toml_table_t *tab, const char *key); +/* ... retrieve values using key. */ +TOML_EXTERN toml_datum_t toml_string_in(const toml_table_t *arr, + const char *key); +TOML_EXTERN toml_datum_t toml_bool_in(const toml_table_t *arr, const char *key); +TOML_EXTERN toml_datum_t toml_int_in(const toml_table_t *arr, const char *key); +TOML_EXTERN toml_datum_t toml_double_in(const toml_table_t *arr, + const char *key); +TOML_EXTERN toml_datum_t toml_timestamp_in(const toml_table_t *arr, + const char *key); +/* .. retrieve array or table using 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); + +/*----------------------------------------------------------------- + * lesser used + */ +/* Return the array kind: 't'able, 'a'rray, 'v'alue, 'm'ixed */ +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, 'm'ixed + 0 if unknown +*/ +TOML_EXTERN char toml_array_type(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); + +/*-------------------------------------------------------------- + * 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 *)); + +/*-------------------------------------------------------------- + * deprecated + */ +/* A raw value, must be processed by toml_rto* before using. */ +typedef const char *toml_raw_t; +TOML_EXTERN toml_raw_t toml_raw_in(const toml_table_t *tab, const char *key); +TOML_EXTERN toml_raw_t toml_raw_at(const toml_array_t *arr, int idx); +TOML_EXTERN int toml_rtos(toml_raw_t s, char **ret); +TOML_EXTERN int toml_rtob(toml_raw_t s, int *ret); +TOML_EXTERN int toml_rtoi(toml_raw_t s, int64_t *ret); +TOML_EXTERN int toml_rtod(toml_raw_t s, double *ret); +TOML_EXTERN int toml_rtod_ex(toml_raw_t s, double *ret, char *buf, int buflen); +TOML_EXTERN int toml_rtots(toml_raw_t s, toml_timestamp_t *ret); + +#endif /* TOML_H */ diff --git a/userspace/extern/src/cargs.c b/userspace/extern/src/cargs.c new file mode 100644 index 0000000..4f95e18 --- /dev/null +++ b/userspace/extern/src/cargs.c @@ -0,0 +1,460 @@ +/* +MIT License + +Copyright (c) 2022 Leonard Iklé + +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. + */ +#include <assert.h> +#include <cargs.h> +#include <memory.h> +#include <stdio.h> +#include <string.h> + +#define CAG_OPTION_PRINT_DISTANCE 4 +#define CAG_OPTION_PRINT_MIN_INDENTION 20 + +static void cag_option_print_value(const cag_option *option, + size_t *accessor_length, FILE *destination) +{ + if (option->value_name != NULL) { + *accessor_length += fprintf(destination, "=%s", option->value_name); + } +} + +static void cag_option_print_letters(const cag_option *option, bool *first, + size_t *accessor_length, FILE *destination) +{ + const char *access_letter; + access_letter = option->access_letters; + if (access_letter != NULL) { + while (*access_letter) { + if (*first) { + *accessor_length += fprintf(destination, "-%c", *access_letter); + *first = false; + } else { + *accessor_length += fprintf(destination, ", -%c", *access_letter); + } + ++access_letter; + } + } +} + +static void cag_option_print_name(const cag_option *option, bool *first, + size_t *accessor_length, FILE *destination) +{ + if (option->access_name != NULL) { + if (*first) { + *accessor_length += fprintf(destination, "--%s", option->access_name); + } else { + *accessor_length += fprintf(destination, ", --%s", option->access_name); + } + } +} + +static size_t cag_option_get_print_indention(const cag_option *options, + size_t option_count) +{ + size_t option_index, indention, result; + const cag_option *option; + + result = CAG_OPTION_PRINT_MIN_INDENTION; + + for (option_index = 0; option_index < option_count; ++option_index) { + indention = CAG_OPTION_PRINT_DISTANCE; + option = &options[option_index]; + if (option->access_letters != NULL && *option->access_letters) { + indention += strlen(option->access_letters) * 4 - 2; + if (option->access_name != NULL) { + indention += strlen(option->access_name) + 4; + } + } else if (option->access_name != NULL) { + indention += strlen(option->access_name) + 2; + } + + if (option->value_name != NULL) { + indention += strlen(option->value_name) + 1; + } + + if (indention > result) { + result = indention; + } + } + + return result; +} + +void cag_option_print(const cag_option *options, size_t option_count, + FILE *destination) +{ + size_t option_index, indention, i, accessor_length; + const cag_option *option; + bool first; + + indention = cag_option_get_print_indention(options, option_count); + + for (option_index = 0; option_index < option_count; ++option_index) { + option = &options[option_index]; + accessor_length = 0; + first = true; + + fputs(" ", destination); + + cag_option_print_letters(option, &first, &accessor_length, destination); + cag_option_print_name(option, &first, &accessor_length, destination); + cag_option_print_value(option, &accessor_length, destination); + + for (i = accessor_length; i < indention; ++i) { + fputs(" ", destination); + } + + fputs(" ", destination); + fputs(option->description, destination); + + fprintf(destination, "\n"); + } +} + +void cag_option_prepare(cag_option_context *context, const cag_option *options, + size_t option_count, int argc, char **argv) +{ + // This just initialized the values to the beginning of all the arguments. + context->options = options; + context->option_count = option_count; + context->argc = argc; + context->argv = argv; + context->index = 1; + context->inner_index = 0; + context->forced_end = false; +} + +static const cag_option *cag_option_find_by_name(cag_option_context *context, + char *name, size_t name_size) +{ + const cag_option *option; + size_t i; + + // We loop over all the available options and stop as soon as we have found + // one. We don't use any hash map table, since there won't be that many + // arguments anyway. + for (i = 0; i < context->option_count; ++i) { + option = &context->options[i]; + + // The option might not have an item name, we can just skip those. + if (option->access_name == NULL) { + continue; + } + + // Try to compare the name of the access name. We can use the name_size or + // this comparison, since we are guaranteed to have null-terminated access + // names. + if (strncmp(option->access_name, name, name_size) == 0) { + return option; + } + } + + return NULL; +} + +static const cag_option *cag_option_find_by_letter(cag_option_context *context, + char letter) +{ + const cag_option *option; + size_t i; + + // We loop over all the available options and stop as soon as we have found + // one. We don't use any look up table, since there won't be that many + // arguments anyway. + for (i = 0; i < context->option_count; ++i) { + option = &context->options[i]; + + // If this option doesn't have any access letters we will skip them. + if (option->access_letters == NULL) { + continue; + } + + // Verify whether this option has the access letter in it's access letter + // string. If it does, then this is our option. + if (strchr(option->access_letters, letter) != NULL) { + return option; + } + } + + return NULL; +} + +static void cag_option_parse_value(cag_option_context *context, + const cag_option *option, char **c) +{ + // And now let's check whether this option is supposed to have a value, which + // is the case if there is a value name set. The value can be either submitted + // with a '=' sign or a space, which means we would have to jump over to the + // next argv index. This is somewhat ugly, but we do it to behave the same as + // the other option parsers. + if (option->value_name != NULL) { + if (**c == '=') { + context->value = ++(*c); + } else { + // If the next index is larger or equal to the argument count, then the + // parameter for this option is missing. The user will know about this, + // since the value pointer of the context will be NULL because we don't + // set it here in that case. + if (context->argc > context->index + 1) { + // We consider this argv to be the value, no matter what the contents + // are. + ++context->index; + *c = context->argv[context->index]; + context->value = *c; + } + } + + // Move c to the end of the value, to not confuse the caller about our + // position. + while (**c) { + ++(*c); + } + } +} + +static void cag_option_parse_access_name(cag_option_context *context, char **c) +{ + const cag_option *option; + char *n; + + // Now we need to extract the access name, which is any symbol up to a '=' or + // a '\0'. + n = *c; + while (**c && **c != '=') { + ++*c; + } + + // Now this will obviously always be true, but we are paranoid. Sometimes. It + // doesn't hurt to check. + assert(*c >= n); + + // Figure out which option this name belongs to. This might return NULL if the + // name is not registered, which means the user supplied an unknown option. In + // that case we return true to indicate that we finished with this option. We + // have to skip the value parsing since we don't know whether the user thinks + // this option has one or not. Since we don't set any identifier specifically, + // it will remain '?' within the context. + option = cag_option_find_by_name(context, n, (size_t)(*c - n)); + if (option == NULL) { + // Since this option is invalid, we will move on to the next index. There is + // nothing we can do about this. + ++context->index; + return; + } + + // We found an option and now we can specify the identifier within the + // context. + context->identifier = option->identifier; + + // And now we try to parse the value. This function will also check whether + // this option is actually supposed to have a value. + cag_option_parse_value(context, option, c); + + // And finally we move on to the next index. + ++context->index; +} + +static void cag_option_parse_access_letter(cag_option_context *context, + char **c) +{ + const cag_option *option; + char *n = *c; + char *v; + + // Figure out which option this letter belongs to. This might return NULL if + // the letter is not registered, which means the user supplied an unknown + // option. In that case we return true to indicate that we finished with this + // option. We have to skip the value parsing since we don't know whether the + // user thinks this option has one or not. Since we don't set any identifier + // specifically, it will remain '?' within the context. + option = cag_option_find_by_letter(context, n[context->inner_index]); + if (option == NULL) { + ++context->index; + context->inner_index = 0; + return; + } + + // We found an option and now we can specify the identifier within the + // context. + context->identifier = option->identifier; + + // And now we try to parse the value. This function will also check whether + // this option is actually supposed to have a value. + v = &n[++context->inner_index]; + cag_option_parse_value(context, option, &v); + + // Check whether we reached the end of this option argument. + if (*v == '\0') { + ++context->index; + context->inner_index = 0; + } +} + +static void cag_option_shift(cag_option_context *context, int start, int option, + int end) +{ + char *tmp; + int a_index, shift_index, shift_count, left_index, right_index; + + shift_count = option - start; + + // There is no shift is required if the start and the option have the same + // index. + if (shift_count == 0) { + return; + } + + // Lets loop through the option strings first, which we will move towards the + // beginning. + for (a_index = option; a_index < end; ++a_index) { + // First remember the current option value, because we will have to save + // that later at the beginning. + tmp = context->argv[a_index]; + + // Let's loop over all option values and shift them one towards the end. + // This will override the option value we just stored temporarily. + for (shift_index = 0; shift_index < shift_count; ++shift_index) { + left_index = a_index - shift_index; + right_index = a_index - shift_index - 1; + context->argv[left_index] = context->argv[right_index]; + } + + // Now restore the saved option value at the beginning. + context->argv[a_index - shift_count] = tmp; + } + + // The new index will be before all non-option values, in such a way that they + // all will be moved again in the next fetch call. + context->index = end - shift_count; +} + +static bool cag_option_is_argument_string(const char *c) +{ + return *c == '-' && *(c + 1) != '\0'; +} + +static int cag_option_find_next(cag_option_context *context) +{ + int next_index, next_option_index; + char *c; + + // Prepare to search the next option at the next index. + next_index = context->index; + next_option_index = next_index; + + // Grab a pointer to the string and verify that it is not the end. If it is + // the end, we have to return false to indicate that we finished. + c = context->argv[next_option_index]; + if (context->forced_end || c == NULL) { + return -1; + } + + // Check whether it is a '-'. We need to find the next option - and an option + // always starts with a '-'. If there is a string "-\0", we don't consider it + // as an option neither. + while (!cag_option_is_argument_string(c)) { + c = context->argv[++next_option_index]; + if (c == NULL) { + // We reached the end and did not find any argument anymore. Let's tell + // our caller that we reached the end. + return -1; + } + } + + // Indicate that we found an option which can be processed. The index of the + // next option will be returned. + return next_option_index; +} + +bool cag_option_fetch(cag_option_context *context) +{ + char *c; + int old_index, new_index; + + // Reset our identifier to a question mark, which indicates an "unknown" + // option. The value is set to NULL, to make sure we are not carrying the + // parameter from the previous option to this one. + context->identifier = '?'; + context->value = NULL; + + // Check whether there are any options left to parse and remember the old + // index as well as the new index. In the end we will move the option junk to + // the beginning, so that non option arguments can be read. + old_index = context->index; + new_index = cag_option_find_next(context); + if (new_index >= 0) { + context->index = new_index; + } else { + return false; + } + + // Grab a pointer to the beginning of the option. At this point, the next + // character must be a '-', since if it was not the prepare function would + // have returned false. We will skip that symbol and proceed. + c = context->argv[context->index]; + assert(*c == '-'); + ++c; + + // Check whether this is a long option, starting with a double "--". + if (*c == '-') { + ++c; + + // This might be a double "--" which indicates the end of options. If this + // is the case, we will not move to the next index. That ensures that + // another call to the fetch function will not skip the "--". + if (*c == '\0') { + context->forced_end = true; + } else { + // We parse now the access name. All information about it will be written + // to the context. + cag_option_parse_access_name(context, &c); + } + } else { + // This is no long option, so we can just parse an access letter. + cag_option_parse_access_letter(context, &c); + } + + // Move the items so that the options come first followed by non-option + // arguments. + cag_option_shift(context, old_index, new_index, context->index); + + return context->forced_end == false; +} + +char cag_option_get(const cag_option_context *context) +{ + // We just return the identifier here. + return context->identifier; +} + +const char *cag_option_get_value(const cag_option_context *context) +{ + // We just return the internal value pointer of the context. + return context->value; +} + +int cag_option_get_index(const cag_option_context *context) +{ + // Either we point to a value item, + return context->index; +} diff --git a/userspace/extern/src/toml.c b/userspace/extern/src/toml.c new file mode 100644 index 0000000..2f4dc03 --- /dev/null +++ b/userspace/extern/src/toml.c @@ -0,0 +1,2379 @@ +/* + + MIT License + + Copyright (c) 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 "toml.h" +#include <assert.h> +#include <ctype.h> +#include <errno.h> +#include <stdbool.h> +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +static void *(*ppmalloc)(size_t) = malloc; +static void (*ppfree)(void *) = free; + +void toml_set_memutil(void *(*xxmalloc)(size_t), void (*xxfree)(void *)) { + if (xxmalloc) + ppmalloc = xxmalloc; + if (xxfree) + ppfree = xxfree; +} + +#define MALLOC(a) ppmalloc(a) +#define FREE(a) ppfree(a) + +#define malloc(x) error - forbidden - use MALLOC instead +#define free(x) error - forbidden - use FREE instead +#define calloc(x, y) error - forbidden - use CALLOC instead + +static void *CALLOC(size_t nmemb, size_t sz) { + int nb = sz * nmemb; + void *p = MALLOC(nb); + if (p) { + memset(p, 0, nb); + } + return p; +} + +// some old platforms define strdup macro -- drop it. +#undef strdup +#define strdup(x) error - forbidden - use STRDUP instead + +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; +} + +// some old platforms define strndup macro -- drop it. +#undef strndup +#define strndup(x) error - forbiden - use STRNDUP instead + +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] = (unsigned char) (0xc0 | (code >> 6)); + buf[1] = (unsigned char) (0x80 | (code & 0x3f)); + return 2; + } + + /* 0x00000800 - 0x0000FFFF: + 1110xxxx 10xxxxxx 10xxxxxx + */ + if (code <= 0x0000FFFF) { + buf[0] = (unsigned char) (0xe0 | (code >> 12)); + buf[1] = (unsigned char) (0x80 | ((code >> 6) & 0x3f)); + buf[2] = (unsigned char) (0x80 | (code & 0x3f)); + return 3; + } + + /* 0x00010000 - 0x001FFFFF: + 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx + */ + if (code <= 0x001FFFFF) { + buf[0] = (unsigned char) (0xf0 | (code >> 18)); + buf[1] = (unsigned char) (0x80 | ((code >> 12) & 0x3f)); + buf[2] = (unsigned char) (0x80 | ((code >> 6) & 0x3f)); + buf[3] = (unsigned char) (0x80 | (code & 0x3f)); + return 4; + } + + /* 0x00200000 - 0x03FFFFFF: + 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx + */ + if (code <= 0x03FFFFFF) { + buf[0] = (unsigned char) (0xf8 | (code >> 24)); + buf[1] = (unsigned char) (0x80 | ((code >> 18) & 0x3f)); + buf[2] = (unsigned char) (0x80 | ((code >> 12) & 0x3f)); + buf[3] = (unsigned char) (0x80 | ((code >> 6) & 0x3f)); + buf[4] = (unsigned char) (0x80 | (code & 0x3f)); + return 5; + } + + /* 0x04000000 - 0x7FFFFFFF: + 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx + */ + if (code <= 0x7FFFFFFF) { + buf[0] = (unsigned char) (0xfc | (code >> 30)); + buf[1] = (unsigned char) (0x80 | ((code >> 24) & 0x3f)); + buf[2] = (unsigned char) (0x80 | ((code >> 18) & 0x3f)); + buf[3] = (unsigned char) (0x80 | ((code >> 12) & 0x3f)); + buf[4] = (unsigned char) (0x80 | ((code >> 6) & 0x3f)); + buf[5] = (unsigned char) (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 */ +}; + +typedef struct toml_arritem_t toml_arritem_t; +struct toml_arritem_t { + int valtype; /* for value kind: 'i'nt, 'd'ouble, 'b'ool, 's'tring, 't'ime, + 'D'ate, 'T'imestamp */ + char *val; + toml_array_t *arr; + toml_table_t *tab; +}; + +struct toml_array_t { + const char *key; /* key to this array */ + int kind; /* element kind: 'v'alue, 'a'rray, or 't'able, 'm'ixed */ + int type; /* for value kind: 'i'nt, 'd'ouble, 'b'ool, 's'tring, 't'ime, + 'D'ate, 'T'imestamp, 'm'ixed */ + + int nitem; /* number of elements */ + toml_arritem_t *item; +}; + +struct toml_table_t { + const char *key; /* key to this table */ + bool implicit; /* table was created implicitly */ + bool readonly; /* no more modification allowed */ + + /* 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; + + 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 int next_token(context_t *ctx, int dotisspecial); + +/* + Error reporting. Call when an error is detected. Always return -1. +*/ +static int e_outofmemory(context_t *ctx, const char *fline) { + snprintf(ctx->errbuf, ctx->errbufsz, "ERROR: out of memory (%s)", fline); + return -1; +} + +static int e_internal(context_t *ctx, const char *fline) { + snprintf(ctx->errbuf, ctx->errbufsz, "internal error (%s)", fline); + return -1; +} + +static int e_syntax(context_t *ctx, int lineno, const char *msg) { + snprintf(ctx->errbuf, ctx->errbufsz, "line %d: %s", lineno, msg); + return -1; +} + +static int e_badkey(context_t *ctx, int lineno) { + snprintf(ctx->errbuf, ctx->errbufsz, "line %d: bad key", lineno); + return -1; +} + +static int e_keyexists(context_t *ctx, int lineno) { + snprintf(ctx->errbuf, ctx->errbufsz, "line %d: key exists", lineno); + return -1; +} + +static int e_forbid(context_t *ctx, int lineno, const char *msg) { + snprintf(ctx->errbuf, ctx->errbufsz, "line %d: %s", lineno, msg); + return -1; +} + +static void *expand(void *p, int sz, int newsz) { + void *s = MALLOC(newsz); + if (!s) + return 0; + + memcpy(s, p, sz); + FREE(p); + return s; +} + +static void **expand_ptrarr(void **p, int n) { + void **s = MALLOC((n + 1) * sizeof(void *)); + if (!s) + return 0; + + s[n] = 0; + memcpy(s, p, n * sizeof(void *)); + FREE(p); + return s; +} + +static toml_arritem_t *expand_arritem(toml_arritem_t *p, int n) { + toml_arritem_t *pp = expand(p, n * sizeof(*p), (n + 1) * sizeof(*p)); + if (!pp) + return 0; + + memset(&pp[n], 0, sizeof(pp[n])); + return pp; +} + +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 */ + int newmax = max + 50; + char *x = expand(dst, max, newmax); + if (!x) { + xfree(dst); + snprintf(errbuf, errbufsz, "out of memory"); + return 0; + } + dst = x; + max = newmax; + } + + /* 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 */ + int newmax = max + 50; + char *x = expand(dst, max, newmax); + if (!x) { + xfree(dst); + snprintf(errbuf, errbufsz, "out of memory"); + return 0; + } + dst = x; + max = newmax; + } + + /* 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; + } + } else { + /* for double quote, we need to normalize */ + ret = norm_basic_str(sp, sq - sp, multiline, ebuf, sizeof(ebuf)); + if (!ret) { + e_syntax(ctx, lineno, ebuf); + return 0; + } + } + + /* newlines are not allowed in keys */ + if (strchr(ret, '\n')) { + xfree(ret); + e_badkey(ctx, lineno); + return 0; + } + 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_badkey(ctx, lineno); + return 0; + } + + /* dup and return it */ + if (!(ret = STRNDUP(sp, sq - sp))) { + e_outofmemory(ctx, FLINE); + return 0; + } + 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 (!newkey) + return 0; + + /* if key exists: error out. */ + toml_keyval_t *dest = 0; + if (key_kind(tab, newkey)) { + xfree(newkey); + e_keyexists(ctx, keytok.lineno); + return 0; + } + + /* make a new entry */ + int n = tab->nkval; + toml_keyval_t **base; + if (0 == (base = (toml_keyval_t **)expand_ptrarr((void **)tab->kval, n))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + tab->kval = base; + + if (0 == (base[n] = (toml_keyval_t *)CALLOC(1, sizeof(*base[n])))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + 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 (!newkey) + return 0; + + /* 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 = false; + return dest; + } + e_keyexists(ctx, keytok.lineno); + return 0; + } + + /* create a new table entry */ + int n = tab->ntab; + toml_table_t **base; + if (0 == (base = (toml_table_t **)expand_ptrarr((void **)tab->tab, n))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + tab->tab = base; + + if (0 == (base[n] = (toml_table_t *)CALLOC(1, sizeof(*base[n])))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + 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 (!newkey) + return 0; + + /* if key exists: error out */ + if (key_kind(tab, newkey)) { + xfree(newkey); /* don't need this anymore */ + e_keyexists(ctx, keytok.lineno); + return 0; + } + + /* make a new array entry */ + int n = tab->narr; + toml_array_t **base; + if (0 == (base = (toml_array_t **)expand_ptrarr((void **)tab->arr, n))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + tab->arr = base; + + if (0 == (base[n] = (toml_array_t *)CALLOC(1, sizeof(*base[n])))) { + xfree(newkey); + e_outofmemory(ctx, FLINE); + return 0; + } + toml_array_t *dest = tab->arr[tab->narr++]; + + /* save the key in the new array struct */ + dest->key = newkey; + dest->kind = kind; + return dest; +} + +static toml_arritem_t *create_value_in_array(context_t *ctx, + toml_array_t *parent) { + const int n = parent->nitem; + toml_arritem_t *base = expand_arritem(parent->item, n); + if (!base) { + e_outofmemory(ctx, FLINE); + return 0; + } + parent->item = base; + parent->nitem++; + return &parent->item[n]; +} + +/* Create an array in an array + */ +static toml_array_t *create_array_in_array(context_t *ctx, + toml_array_t *parent) { + const int n = parent->nitem; + toml_arritem_t *base = expand_arritem(parent->item, n); + if (!base) { + e_outofmemory(ctx, FLINE); + return 0; + } + toml_array_t *ret = (toml_array_t *)CALLOC(1, sizeof(toml_array_t)); + if (!ret) { + e_outofmemory(ctx, FLINE); + return 0; + } + base[n].arr = ret; + parent->item = base; + parent->nitem++; + return ret; +} + +/* Create a table in an array + */ +static toml_table_t *create_table_in_array(context_t *ctx, + toml_array_t *parent) { + int n = parent->nitem; + toml_arritem_t *base = expand_arritem(parent->item, n); + if (!base) { + e_outofmemory(ctx, FLINE); + return 0; + } + toml_table_t *ret = (toml_table_t *)CALLOC(1, sizeof(toml_table_t)); + if (!ret) { + e_outofmemory(ctx, FLINE); + return 0; + } + base[n].tab = ret; + parent->item = base; + parent->nitem++; + return ret; +} + +static int skip_newlines(context_t *ctx, int isdotspecial) { + while (ctx->tok.tok == NEWLINE) { + if (next_token(ctx, isdotspecial)) + return -1; + if (ctx->tok.eof) + break; + } + return 0; +} + +static int parse_keyval(context_t *ctx, toml_table_t *tab); + +static inline int eat_token(context_t *ctx, tokentype_t typ, int isdotspecial, + const char *fline) { + if (ctx->tok.tok != typ) + return e_internal(ctx, fline); + + if (next_token(ctx, isdotspecial)) + return -1; + + return 0; +} + +/* We are at '{ ... }'. + * Parse the table. + */ +static int parse_inline_table(context_t *ctx, toml_table_t *tab) { + if (eat_token(ctx, LBRACE, 1, FLINE)) + return -1; + + for (;;) { + if (ctx->tok.tok == NEWLINE) + return e_syntax(ctx, ctx->tok.lineno, + "newline not allowed in inline table"); + + /* until } */ + if (ctx->tok.tok == RBRACE) + break; + + if (ctx->tok.tok != STRING) + return e_syntax(ctx, ctx->tok.lineno, "expect a string"); + + if (parse_keyval(ctx, tab)) + return -1; + + if (ctx->tok.tok == NEWLINE) + return e_syntax(ctx, ctx->tok.lineno, + "newline not allowed in inline table"); + + /* on comma, continue to scan for next keyval */ + if (ctx->tok.tok == COMMA) { + if (eat_token(ctx, COMMA, 1, FLINE)) + return -1; + continue; + } + break; + } + + if (eat_token(ctx, RBRACE, 1, FLINE)) + return -1; + + tab->readonly = 1; + + return 0; +} + +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 int parse_array(context_t *ctx, toml_array_t *arr) { + if (eat_token(ctx, LBRACKET, 0, FLINE)) + return -1; + + for (;;) { + if (skip_newlines(ctx, 0)) + return -1; + + /* until ] */ + if (ctx->tok.tok == RBRACKET) + break; + + switch (ctx->tok.tok) { + case STRING: { + /* set array kind if this will be the first entry */ + if (arr->kind == 0) + arr->kind = 'v'; + else if (arr->kind != 'v') + arr->kind = 'm'; + + char *val = ctx->tok.ptr; + int vlen = ctx->tok.len; + + /* make a new value in array */ + toml_arritem_t *newval = create_value_in_array(ctx, arr); + if (!newval) + return e_outofmemory(ctx, FLINE); + + if (!(newval->val = STRNDUP(val, vlen))) + return e_outofmemory(ctx, FLINE); + + newval->valtype = valtype(newval->val); + + /* set array type if this is the first entry */ + if (arr->nitem == 1) + arr->type = newval->valtype; + else if (arr->type != newval->valtype) + arr->type = 'm'; /* mixed */ + + if (eat_token(ctx, STRING, 0, FLINE)) + return -1; + break; + } + + case LBRACKET: { /* [ [array], [array] ... ] */ + /* set the array kind if this will be the first entry */ + if (arr->kind == 0) + arr->kind = 'a'; + else if (arr->kind != 'a') + arr->kind = 'm'; + + toml_array_t *subarr = create_array_in_array(ctx, arr); + if (!subarr) + return -1; + if (parse_array(ctx, subarr)) + return -1; + break; + } + + case LBRACE: { /* [ {table}, {table} ... ] */ + /* set the array kind if this will be the first entry */ + if (arr->kind == 0) + arr->kind = 't'; + else if (arr->kind != 't') + arr->kind = 'm'; + + toml_table_t *subtab = create_table_in_array(ctx, arr); + if (!subtab) + return -1; + if (parse_inline_table(ctx, subtab)) + return -1; + break; + } + + default: + return e_syntax(ctx, ctx->tok.lineno, "syntax error"); + } + + if (skip_newlines(ctx, 0)) + return -1; + + /* on comma, continue to scan for next element */ + if (ctx->tok.tok == COMMA) { + if (eat_token(ctx, COMMA, 0, FLINE)) + return -1; + continue; + } + break; + } + + if (eat_token(ctx, RBRACKET, 1, FLINE)) + return -1; + return 0; +} + +/* handle lines like these: + key = "value" + key = [ array ] + key = { table } +*/ +static int parse_keyval(context_t *ctx, toml_table_t *tab) { + if (tab->readonly) { + return e_forbid(ctx, ctx->tok.lineno, + "cannot insert new entry into existing table"); + } + + token_t key = ctx->tok; + if (eat_token(ctx, STRING, 1, FLINE)) + return -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); + if (!subtabstr) + return -1; + + subtab = toml_table_in(tab, subtabstr); + xfree(subtabstr); + } + if (!subtab) { + subtab = create_keytable_in_table(ctx, tab, key); + if (!subtab) + return -1; + } + if (next_token(ctx, 1)) + return -1; + if (parse_keyval(ctx, subtab)) + return -1; + return 0; + } + + if (ctx->tok.tok != EQUAL) { + return e_syntax(ctx, ctx->tok.lineno, "missing ="); + } + + if (next_token(ctx, 0)) + return -1; + + switch (ctx->tok.tok) { + case STRING: { /* key = "value" */ + toml_keyval_t *keyval = create_keyval_in_table(ctx, tab, key); + if (!keyval) + return -1; + token_t val = ctx->tok; + + assert(keyval->val == 0); + if (!(keyval->val = STRNDUP(val.ptr, val.len))) + return e_outofmemory(ctx, FLINE); + + if (next_token(ctx, 1)) + return -1; + + return 0; + } + + case LBRACKET: { /* key = [ array ] */ + toml_array_t *arr = create_keyarray_in_table(ctx, tab, key, 0); + if (!arr) + return -1; + if (parse_array(ctx, arr)) + return -1; + return 0; + } + + case LBRACE: { /* key = { table } */ + toml_table_t *nxttab = create_keytable_in_table(ctx, tab, key); + if (!nxttab) + return -1; + if (parse_inline_table(ctx, nxttab)) + return -1; + return 0; + } + + default: + return e_syntax(ctx, ctx->tok.lineno, "syntax error"); + } + return 0; +} + +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 int 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) + return e_syntax(ctx, lineno, + "table path is too deep; max allowed is 10."); + + if (ctx->tok.tok != STRING) + return e_syntax(ctx, lineno, "invalid or missing key"); + + char *key = normalize_key(ctx, ctx->tok); + if (!key) + return -1; + ctx->tpath.tok[ctx->tpath.top] = ctx->tok; + ctx->tpath.key[ctx->tpath.top] = key; + ctx->tpath.top++; + + if (next_token(ctx, 1)) + return -1; + + if (ctx->tok.tok == RBRACKET) + break; + + if (ctx->tok.tok != DOT) + return e_syntax(ctx, lineno, "invalid key"); + + if (next_token(ctx, 1)) + return -1; + } + + if (ctx->tpath.top <= 0) + return e_syntax(ctx, lineno, "empty table selector"); + + return 0; +} + +/* Walk tabpath from the root, and create new tables on the way. + * Sets ctx->curtab to the final table. + */ +static int 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') + return e_internal(ctx, FLINE); + + if (nextarr->nitem == 0) + return e_internal(ctx, FLINE); + + nexttab = nextarr->item[nextarr->nitem - 1].tab; + break; + + case 'v': + return e_keyexists(ctx, ctx->tpath.tok[i].lineno); + + default: { /* Not found. Let's create an implicit table. */ + int n = curtab->ntab; + toml_table_t **base = + (toml_table_t **)expand_ptrarr((void **)curtab->tab, n); + if (0 == base) + return e_outofmemory(ctx, FLINE); + + curtab->tab = base; + + if (0 == (base[n] = (toml_table_t *)CALLOC(1, sizeof(*base[n])))) + return e_outofmemory(ctx, FLINE); + + if (0 == (base[n]->key = STRDUP(key))) + return e_outofmemory(ctx, FLINE); + + nexttab = curtab->tab[curtab->ntab++]; + + /* tabs created by walk_tabpath are considered implicit */ + nexttab->implicit = true; + } break; + } + + /* switch to next tab */ + curtab = nexttab; + } + + /* save it */ + ctx->curtab = curtab; + + return 0; +} + +/* handle lines like [x.y.z] or [[x.y.z]] */ +static int 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 [[ */ + if (eat_token(ctx, LBRACKET, 1, FLINE)) + return -1; + if (llb) { + assert(ctx->tok.tok == LBRACKET); + if (eat_token(ctx, LBRACKET, 1, FLINE)) + return -1; + } + + if (fill_tabpath(ctx)) + return -1; + + /* 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 */ + if (walk_tabpath(ctx)) + return -1; + + if (!llb) { + /* [x.y.z] -> create z = {} in x.y */ + toml_table_t *curtab = create_keytable_in_table(ctx, ctx->curtab, z); + if (!curtab) + return -1; + ctx->curtab = curtab; + } else { + /* [[x.y.z]] -> create z = [] in x.y */ + toml_array_t *arr = 0; + { + char *zstr = normalize_key(ctx, z); + if (!zstr) + return -1; + arr = toml_array_in(ctx->curtab, zstr); + xfree(zstr); + } + if (!arr) { + arr = create_keyarray_in_table(ctx, ctx->curtab, z, 't'); + if (!arr) + return -1; + } + if (arr->kind != 't') + return e_syntax(ctx, z.lineno, "array mismatch"); + + /* add to z[] */ + toml_table_t *dest; + { + toml_table_t *t = create_table_in_array(ctx, arr); + if (!t) + return -1; + + if (0 == (t->key = STRDUP("__anon__"))) + return e_outofmemory(ctx, FLINE); + + dest = t; + } + + ctx->curtab = dest; + } + + if (ctx->tok.tok != RBRACKET) { + return e_syntax(ctx, ctx->tok.lineno, "expects ]"); + } + if (llb) { + if (!(ctx->tok.ptr + 1 < ctx->stop && ctx->tok.ptr[1] == ']')) { + return e_syntax(ctx, ctx->tok.lineno, "expects ]]"); + } + if (eat_token(ctx, RBRACKET, 1, FLINE)) + return -1; + } + + if (eat_token(ctx, RBRACKET, 1, FLINE)) + return -1; + + if (ctx->tok.tok != NEWLINE) + return e_syntax(ctx, ctx->tok.lineno, "extra chars after ] or ]]"); + + return 0; +} + +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)))) { + e_outofmemory(&ctx, FLINE); + // Do not goto fail, root table not set up yet + return 0; + } + + // set root as default table + ctx.curtab = ctx.root; + + /* Scan forward until EOF */ + for (token_t tok = ctx.tok; !tok.eof; tok = ctx.tok) { + switch (tok.tok) { + + case NEWLINE: + if (next_token(&ctx, 1)) + goto fail; + break; + + case STRING: + if (parse_keyval(&ctx, ctx.curtab)) + goto fail; + + if (ctx.tok.tok != NEWLINE) { + e_syntax(&ctx, ctx.tok.lineno, "extra chars after value"); + goto fail; + } + + if (eat_token(&ctx, NEWLINE, 1, FLINE)) + goto fail; + break; + + case LBRACKET: /* [ x.y.z ] or [[ x.y.z ]] */ + if (parse_select(&ctx)) + goto fail; + break; + + default: + e_syntax(&ctx, tok.lineno, "syntax error"); + goto fail; + } + } + + /* success */ + for (int i = 0; i < ctx.tpath.top; i++) + xfree(ctx.tpath.key[i]); + return ctx.root; + +fail: + // 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; +} + +toml_table_t *toml_parse_file(FILE *fp, char *errbuf, int errbufsz) { + int bufsz = 0; + char *buf = 0; + int off = 0; + + /* read from fp into buf */ + while (!feof(fp)) { + + if (off == bufsz) { + int xsz = bufsz + 1000; + char *x = expand(buf, bufsz, xsz); + if (!x) { + snprintf(errbuf, errbufsz, "out of memory"); + xfree(buf); + return 0; + } + buf = x; + bufsz = xsz; + } + + 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 */ + if (off == bufsz) { + int xsz = bufsz + 1; + char *x = expand(buf, bufsz, xsz); + if (!x) { + snprintf(errbuf, errbufsz, "out of memory"); + xfree(buf); + return 0; + } + buf = x; + bufsz = xsz; + } + buf[off] = 0; + + /* 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); + const int n = p->nitem; + for (int i = 0; i < n; i++) { + toml_arritem_t *a = &p->item[i]; + if (a->val) + xfree(a->val); + else if (a->arr) + xfree_arr(a->arr); + else if (a->tab) + xfree_tab(a->tab); + } + xfree(p->item); + 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 void set_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; +} + +static void set_eof(context_t *ctx, int lineno) { + set_token(ctx, NEWLINE, lineno, ctx->stop, 0); + ctx->tok.eof = 1; +} + +/* 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 int scan_string(context_t *ctx, char *p, int lineno, int dotisspecial) { + char *orig = p; + if (0 == strncmp(p, "'''", 3)) { + char *q = p + 3; + + while (1) { + q = strstr(q, "'''"); + if (0 == q) { + return e_syntax(ctx, lineno, "unterminated triple-s-quote"); + } + while (q[3] == '\'') + q++; + break; + } + + set_token(ctx, STRING, lineno, orig, q + 3 - orig); + return 0; + } + + if (0 == strncmp(p, "\"\"\"", 3)) { + char *q = p + 3; + + while (1) { + q = strstr(q, "\"\"\""); + if (0 == q) { + return e_syntax(ctx, lineno, "unterminated triple-d-quote"); + } + if (q[-1] == '\\') { + q++; + continue; + } + while (q[3] == '\"') + q++; + break; + } + + // the string is [p+3, q-1] + + int hexreq = 0; /* #hex required */ + int escape = 0; + for (p += 3; p < q; 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 */ + return e_syntax(ctx, lineno, "bad escape char"); + } + if (hexreq) { + hexreq--; + if (strchr("0123456789ABCDEF", *p)) + continue; + return e_syntax(ctx, lineno, "expect hex char"); + } + if (*p == '\\') { + escape = 1; + continue; + } + } + if (escape) + return e_syntax(ctx, lineno, "expect an escape char"); + if (hexreq) + return e_syntax(ctx, lineno, "expected more hex char"); + + set_token(ctx, STRING, lineno, orig, q + 3 - orig); + return 0; + } + + if ('\'' == *p) { + for (p++; *p && *p != '\n' && *p != '\''; p++) + ; + if (*p != '\'') { + return e_syntax(ctx, lineno, "unterminated s-quote"); + } + + set_token(ctx, STRING, lineno, orig, p + 1 - orig); + return 0; + } + + 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; + } + return e_syntax(ctx, lineno, "bad escape char"); + } + if (hexreq) { + hexreq--; + if (strchr("0123456789ABCDEF", *p)) + continue; + return e_syntax(ctx, lineno, "expect hex char"); + } + if (*p == '\\') { + escape = 1; + continue; + } + if (*p == '\'') { + if (p[1] == '\'' && p[2] == '\'') { + return e_syntax(ctx, lineno, "triple-s-quote inside string lit"); + } + continue; + } + if (*p == '\n') + break; + if (*p == '"') + break; + } + if (*p != '"') { + return e_syntax(ctx, lineno, "unterminated quote"); + } + + set_token(ctx, STRING, lineno, orig, p + 1 - orig); + return 0; + } + + /* check for timestamp without quotes */ + if (0 == scan_date(p, 0, 0, 0) || 0 == scan_time(p, 0, 0, 0)) { + // forward thru the timestamp + p += strspn(p, "0123456789.:+-Tt Zz"); + // squeeze out any spaces at end of string + for (; p[-1] == ' '; p--) + ; + // tokenize + set_token(ctx, STRING, lineno, orig, p - orig); + return 0; + } + + /* 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; + } + + set_token(ctx, STRING, lineno, orig, p - orig); + return 0; +} + +static int 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 == '.') { + set_token(ctx, DOT, lineno, p, 1); + return 0; + } + + switch (*p) { + case ',': + set_token(ctx, COMMA, lineno, p, 1); + return 0; + case '=': + set_token(ctx, EQUAL, lineno, p, 1); + return 0; + case '{': + set_token(ctx, LBRACE, lineno, p, 1); + return 0; + case '}': + set_token(ctx, RBRACE, lineno, p, 1); + return 0; + case '[': + set_token(ctx, LBRACKET, lineno, p, 1); + return 0; + case ']': + set_token(ctx, RBRACKET, lineno, p, 1); + return 0; + case '\n': + set_token(ctx, NEWLINE, lineno, p, 1); + return 0; + case '\r': + case ' ': + case '\t': + /* ignore white spaces */ + p++; + continue; + } + + return scan_string(ctx, p, lineno, dotisspecial); + } + + set_eof(ctx, lineno); + return 0; +} + +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; +} + +int toml_key_exists(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 1; + } + for (i = 0; i < tab->narr; i++) { + if (0 == strcmp(key, tab->arr[i]->key)) + return 1; + } + for (i = 0; i < tab->ntab; i++) { + if (0 == strcmp(key, tab->tab[i]->key)) + return 1; + } + 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) { + return (0 <= idx && idx < arr->nitem) ? arr->item[idx].val : 0; +} + +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->nitem == 0) + return 0; + + return arr->type; +} + +int toml_array_nelem(const toml_array_t *arr) { return arr->nitem; } + +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) { + return (0 <= idx && idx < arr->nitem) ? arr->item[idx].arr : 0; +} + +toml_table_t *toml_table_at(const toml_array_t *arr, int idx) { + return (0 <= idx && idx < arr->nitem) ? arr->item[idx].tab : 0; +} + +static int parse_millisec(const char *p, const char **endp); + +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 != '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 == '.') { + p++; /* skip '.' */ + const char *qq; + *millisec = parse_millisec(p, &qq); + 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++; + if (ch == '_') { + // disallow '__' + if (s[0] == '_') + return -1; + // numbers cannot end with '_' + if (s[0] == '\0') + return -1; + continue; /* skip _ */ + } + *p++ = ch; + } + + // if not at end-of-string or we ran out of buffer ... + if (*s || p == q) + 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; + + /* decimal point, if used, must be surrounded by at least one digit on each + * side */ + { + char *dot = strchr(s, '.'); + if (dot) { + if (dot == s || !isdigit(dot[-1]) || !isdigit(dot[1])) + 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++; + if (ch == '_') { + // disallow '__' + if (s[0] == '_') + return -1; + // disallow last char '_' + if (s[0] == 0) + return -1; + continue; /* skip _ */ + } + *p++ = ch; + } + if (*s || p == q) + return -1; /* reached end of string or buffer is full? */ + + /* 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; +} + +toml_datum_t toml_string_at(const toml_array_t *arr, int idx) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtos(toml_raw_at(arr, idx), &ret.u.s)); + return ret; +} + +toml_datum_t toml_bool_at(const toml_array_t *arr, int idx) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtob(toml_raw_at(arr, idx), &ret.u.b)); + return ret; +} + +toml_datum_t toml_int_at(const toml_array_t *arr, int idx) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtoi(toml_raw_at(arr, idx), &ret.u.i)); + return ret; +} + +toml_datum_t toml_double_at(const toml_array_t *arr, int idx) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtod(toml_raw_at(arr, idx), &ret.u.d)); + return ret; +} + +toml_datum_t toml_timestamp_at(const toml_array_t *arr, int idx) { + toml_timestamp_t ts; + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtots(toml_raw_at(arr, idx), &ts)); + if (ret.ok) { + ret.ok = !!(ret.u.ts = MALLOC(sizeof(*ret.u.ts))); + if (ret.ok) { + *ret.u.ts = ts; + if (ret.u.ts->year) + ret.u.ts->year = &ret.u.ts->__buffer.year; + if (ret.u.ts->month) + ret.u.ts->month = &ret.u.ts->__buffer.month; + if (ret.u.ts->day) + ret.u.ts->day = &ret.u.ts->__buffer.day; + if (ret.u.ts->hour) + ret.u.ts->hour = &ret.u.ts->__buffer.hour; + if (ret.u.ts->minute) + ret.u.ts->minute = &ret.u.ts->__buffer.minute; + if (ret.u.ts->second) + ret.u.ts->second = &ret.u.ts->__buffer.second; + if (ret.u.ts->millisec) + ret.u.ts->millisec = &ret.u.ts->__buffer.millisec; + if (ret.u.ts->z) + ret.u.ts->z = ret.u.ts->__buffer.z; + } + } + return ret; +} + +toml_datum_t toml_string_in(const toml_table_t *arr, const char *key) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + toml_raw_t raw = toml_raw_in(arr, key); + if (raw) { + ret.ok = (0 == toml_rtos(raw, &ret.u.s)); + } + return ret; +} + +toml_datum_t toml_bool_in(const toml_table_t *arr, const char *key) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtob(toml_raw_in(arr, key), &ret.u.b)); + return ret; +} + +toml_datum_t toml_int_in(const toml_table_t *arr, const char *key) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtoi(toml_raw_in(arr, key), &ret.u.i)); + return ret; +} + +toml_datum_t toml_double_in(const toml_table_t *arr, const char *key) { + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtod(toml_raw_in(arr, key), &ret.u.d)); + return ret; +} + +toml_datum_t toml_timestamp_in(const toml_table_t *arr, const char *key) { + toml_timestamp_t ts; + toml_datum_t ret; + memset(&ret, 0, sizeof(ret)); + ret.ok = (0 == toml_rtots(toml_raw_in(arr, key), &ts)); + if (ret.ok) { + ret.ok = !!(ret.u.ts = MALLOC(sizeof(*ret.u.ts))); + if (ret.ok) { + *ret.u.ts = ts; + if (ret.u.ts->year) + ret.u.ts->year = &ret.u.ts->__buffer.year; + if (ret.u.ts->month) + ret.u.ts->month = &ret.u.ts->__buffer.month; + if (ret.u.ts->day) + ret.u.ts->day = &ret.u.ts->__buffer.day; + if (ret.u.ts->hour) + ret.u.ts->hour = &ret.u.ts->__buffer.hour; + if (ret.u.ts->minute) + ret.u.ts->minute = &ret.u.ts->__buffer.minute; + if (ret.u.ts->second) + ret.u.ts->second = &ret.u.ts->__buffer.second; + if (ret.u.ts->millisec) + ret.u.ts->millisec = &ret.u.ts->__buffer.millisec; + if (ret.u.ts->z) + ret.u.ts->z = ret.u.ts->__buffer.z; + } + } + return ret; +} + +static int parse_millisec(const char *p, const char **endp) { + int ret = 0; + int unit = 100; /* unit in millisec */ + for (; '0' <= *p && *p <= '9'; p++, unit /= 10) { + ret += (*p - '0') * unit; + } + *endp = p; + return ret; +} |