implemented kwil parsing
parent
e518d1ae47
commit
cc159c2244
|
|
@ -0,0 +1,282 @@
|
||||||
|
#include "kwil_parse.h"
|
||||||
|
#include <stdio.h>
|
||||||
|
#include <stdlib.h>
|
||||||
|
#include <string.h>
|
||||||
|
#include <ctype.h>
|
||||||
|
|
||||||
|
int fpeek(FILE* file) {
|
||||||
|
int get = fgetc(file);
|
||||||
|
if(get != EOF) {
|
||||||
|
ungetc(get, file);
|
||||||
|
}
|
||||||
|
return get;
|
||||||
|
}
|
||||||
|
|
||||||
|
static
|
||||||
|
int read_line_without_whitespace(FILE* file, char* out_buf, int max_size) {
|
||||||
|
// int read this loop
|
||||||
|
int c = 0;
|
||||||
|
|
||||||
|
// initialize writer and end of buffer pointer
|
||||||
|
char* out_writer = out_buf;
|
||||||
|
char* out_end = out_buf + max_size;
|
||||||
|
|
||||||
|
do {
|
||||||
|
c = fgetc(file);
|
||||||
|
|
||||||
|
if(!isspace(c)) {
|
||||||
|
*out_writer = c;
|
||||||
|
++out_writer;
|
||||||
|
}
|
||||||
|
} while(!feof(file) && c != '\n' && out_writer < out_end);
|
||||||
|
|
||||||
|
if(out_writer < out_end) {
|
||||||
|
*out_writer = '\0';
|
||||||
|
}
|
||||||
|
|
||||||
|
return out_writer - out_buf;
|
||||||
|
}
|
||||||
|
|
||||||
|
static
|
||||||
|
int next_non_whitespace(FILE* file) {
|
||||||
|
int c = 0;
|
||||||
|
do {
|
||||||
|
c = fgetc(file);
|
||||||
|
} while(isspace(c));
|
||||||
|
|
||||||
|
return c;
|
||||||
|
}
|
||||||
|
|
||||||
|
static
|
||||||
|
int read_typename(FILE* file, char* out_buf, int max_size) {
|
||||||
|
int c = 0;
|
||||||
|
char* out_end = out_buf + max_size;
|
||||||
|
|
||||||
|
// initialze buffer to check for "struct"
|
||||||
|
char buffer[7];
|
||||||
|
buffer[6] = '\0';
|
||||||
|
|
||||||
|
for(int i = 0; i < 6; ++i) {
|
||||||
|
buffer[i] = fgetc(file);
|
||||||
|
}
|
||||||
|
|
||||||
|
// next line is not a compatible type declaration,
|
||||||
|
// write null terminator and return zero
|
||||||
|
if(strncmp(buffer, "struct", 7) != 0 && strncmp(buffer, "enum", 4) != 0) {
|
||||||
|
*out_buf = '\0';
|
||||||
|
return 0;
|
||||||
|
}
|
||||||
|
|
||||||
|
// ignore the whitespace after "struct"
|
||||||
|
c = next_non_whitespace(file);
|
||||||
|
|
||||||
|
// write all text until the next whitespace or '{' to the out buffer
|
||||||
|
char* out_writer = out_buf;
|
||||||
|
for(;(isalnum(c) || ispunct(c)) && c != '{' && out_writer < out_end; ++out_writer) {
|
||||||
|
*out_writer = c;
|
||||||
|
c = fgetc(file);
|
||||||
|
}
|
||||||
|
|
||||||
|
if(out_writer < out_end) {
|
||||||
|
*out_writer = '\0';
|
||||||
|
}
|
||||||
|
|
||||||
|
// read until the end of the line,
|
||||||
|
// setting the file pointer up for parsing the struct
|
||||||
|
while(c != '\n') {
|
||||||
|
c = fgetc(file);
|
||||||
|
}
|
||||||
|
|
||||||
|
return out_writer - out_buf;
|
||||||
|
}
|
||||||
|
|
||||||
|
void kwil_header_parse(struct kwil_header_t* self) {
|
||||||
|
printf("Reading top level type definitions for header \"%s\"\n", self->file_name);
|
||||||
|
|
||||||
|
// .h file associated with self
|
||||||
|
FILE* file = fopen(self->file_name, "r");
|
||||||
|
// should be able to store
|
||||||
|
// KWIL_ENUM() or KWIL_STRUCT() or KWIL_FIELD()
|
||||||
|
char line_buffer[48];
|
||||||
|
line_buffer[47] = '\0';
|
||||||
|
struct kwil_struct_t kwil_struct;
|
||||||
|
struct kwil_enum_t kwil_enum;
|
||||||
|
struct kwil_type_t type;
|
||||||
|
|
||||||
|
// read the entire file
|
||||||
|
while(!feof(file)) {
|
||||||
|
// search for a match with a kwil statemenent
|
||||||
|
read_line_without_whitespace(file, line_buffer, 47);
|
||||||
|
|
||||||
|
if(strncmp(line_buffer, "KWIL_STRUCT()", 48) == 0) {
|
||||||
|
// get the typename and update the user
|
||||||
|
read_typename(file, line_buffer, 47);
|
||||||
|
printf("\"%s\": {\n", line_buffer);
|
||||||
|
|
||||||
|
// line matches KWIL_STRUCT() instantiate a struct and pass control to struct parser
|
||||||
|
kwil_struct_parse(&kwil_struct, file);
|
||||||
|
|
||||||
|
// move struct definition into a type struct and store in self
|
||||||
|
kwil_type_from_struct(&type, &kwil_struct, line_buffer);
|
||||||
|
kwil_header_add_type(self, &type);
|
||||||
|
} else if(strncmp(line_buffer, "KWIL_ENUM()", 48) == 0) {
|
||||||
|
//read typename and update the user
|
||||||
|
read_typename(file, line_buffer, 47);
|
||||||
|
printf("\"%s\": [\n", line_buffer);
|
||||||
|
|
||||||
|
// line matches KWIL_ENUM() instantiate an enum and pass control to the enum parser
|
||||||
|
kwil_enum_parse(&kwil_enum, file);
|
||||||
|
|
||||||
|
// move enum definition into a type struct and register with header
|
||||||
|
kwil_type_from_enum(&type, &kwil_enum, line_buffer);
|
||||||
|
kwil_header_add_type(self, &type);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// close file after finishing
|
||||||
|
fclose(file);
|
||||||
|
}
|
||||||
|
|
||||||
|
void kwil_struct_parse(struct kwil_struct_t *self, FILE *file) {
|
||||||
|
kwil_init_struct(self);
|
||||||
|
// create a line buffer with a null terminator
|
||||||
|
// this should always be treated as a 47-character-long array
|
||||||
|
// this way the string will always be valid
|
||||||
|
char line_buffer[48];
|
||||||
|
line_buffer[47] = '\0';
|
||||||
|
|
||||||
|
struct kwil_field_t field;
|
||||||
|
|
||||||
|
do {
|
||||||
|
// read the next line
|
||||||
|
read_line_without_whitespace(file, line_buffer, 47);
|
||||||
|
if(strncmp("KWIL_FIELD()", line_buffer, 12) == 0) {
|
||||||
|
kwil_field_parse(&field, file);
|
||||||
|
kwil_struct_add_field(self, &field);
|
||||||
|
printf("\t\"%s\": \"%s%s\",\n", field.name_str, field.type_str, field.array_length > 0 ? "*" : "");
|
||||||
|
}
|
||||||
|
} while(!feof(file) && line_buffer[0] != '}');
|
||||||
|
|
||||||
|
// print an end of line to separate structs when declaring fields
|
||||||
|
printf("}\n");
|
||||||
|
}
|
||||||
|
|
||||||
|
static
|
||||||
|
int kwil_field_parse_split(FILE* file, char* line_buffer,
|
||||||
|
char** o_words, int* o_word_lengths, int* o_word_count,
|
||||||
|
int* o_is_array) {
|
||||||
|
// last character read from file
|
||||||
|
int c = 0;
|
||||||
|
char* line_buffer_writer = line_buffer;
|
||||||
|
size_t word_index = 1;
|
||||||
|
// the "start" of the line is considered a whitespace character, this way leading whitespace is ignored
|
||||||
|
int last_was_space = 1;
|
||||||
|
|
||||||
|
do {
|
||||||
|
c = fgetc(file);
|
||||||
|
if(c == '*') {
|
||||||
|
*o_is_array = 2;
|
||||||
|
} else if(isalnum(c) || c == '_' || c == ']') {
|
||||||
|
// write non-whitespace characters to the line buffer
|
||||||
|
*line_buffer_writer = c;
|
||||||
|
++line_buffer_writer;
|
||||||
|
last_was_space = 0;
|
||||||
|
} else if(c == '[') {
|
||||||
|
*o_is_array = 1;
|
||||||
|
// store the start of the next word
|
||||||
|
o_words[word_index] = line_buffer_writer + 1;
|
||||||
|
// get the length of the previous word
|
||||||
|
o_word_lengths[word_index-1] = o_words[word_index] - o_words[word_index-1] - 1;
|
||||||
|
++word_index;
|
||||||
|
// add "\0[" or "\0]" to the line buffer
|
||||||
|
line_buffer_writer[0] = '\0';
|
||||||
|
line_buffer_writer[1] = c;
|
||||||
|
line_buffer_writer += 2;
|
||||||
|
} else if(last_was_space == 0) {
|
||||||
|
// write a null terminator at the end of the last word
|
||||||
|
*line_buffer_writer = '\0';
|
||||||
|
last_was_space = 1;
|
||||||
|
// store the start of the next word
|
||||||
|
o_words[word_index] = line_buffer_writer + 1;
|
||||||
|
// get the length of the previous word
|
||||||
|
o_word_lengths[word_index-1] = o_words[word_index] - o_words[word_index-1] - 1;
|
||||||
|
++word_index;
|
||||||
|
++line_buffer_writer;
|
||||||
|
}
|
||||||
|
} while(!feof(file) && c != ';');
|
||||||
|
|
||||||
|
*o_word_count = word_index - 1;
|
||||||
|
|
||||||
|
// consume the line to set up for the next field
|
||||||
|
while(c != '\n') {
|
||||||
|
c = fgetc(file);
|
||||||
|
}
|
||||||
|
|
||||||
|
return line_buffer_writer - line_buffer;
|
||||||
|
}
|
||||||
|
|
||||||
|
void kwil_field_parse(struct kwil_field_t* self, FILE* file) {
|
||||||
|
// 47 character long temp buffer for reading from file
|
||||||
|
char line_buffer[48];
|
||||||
|
// array of pointers into line_buffer that point at the starts of words
|
||||||
|
char* words[5];
|
||||||
|
// array of word lengths matched to start_ptrs
|
||||||
|
int word_lengths[5];
|
||||||
|
|
||||||
|
// initialize these as 0x0
|
||||||
|
memset(words, '\0', sizeof(words));
|
||||||
|
memset(word_lengths, '\0', sizeof(word_lengths));
|
||||||
|
// the start of the first word matches the start of the line buffer
|
||||||
|
words[0] = line_buffer;
|
||||||
|
|
||||||
|
// set the terminal element to null
|
||||||
|
line_buffer[47] = '\0';
|
||||||
|
|
||||||
|
// Set to true if this field is a pointer or array
|
||||||
|
// Both dynamic and static sized arrays are supported,
|
||||||
|
// though dynamic arrays will need to be marked with their minimum length
|
||||||
|
int is_array = 0;
|
||||||
|
// length of the entire buffer
|
||||||
|
int word_count;
|
||||||
|
int length = kwil_field_parse_split(file, line_buffer,
|
||||||
|
words, word_lengths, &word_count, &is_array);
|
||||||
|
line_buffer[length - 1] = '\0';
|
||||||
|
|
||||||
|
// scan for the length of static arrays
|
||||||
|
int array_len = 0;
|
||||||
|
if(is_array == 1) {
|
||||||
|
sscanf(words[word_count-1], "[%d]",
|
||||||
|
&array_len);
|
||||||
|
word_count -= 1;
|
||||||
|
}
|
||||||
|
|
||||||
|
// get the typename from the split declaration
|
||||||
|
char* type = words[0];
|
||||||
|
if(strncmp("struct", type, 7) == 0 || strncmp("enum", type, 5) == 0) {
|
||||||
|
words[0][word_lengths[0]] = ' ';
|
||||||
|
}
|
||||||
|
|
||||||
|
kwil_init_field(self, words[word_count - 1], words[0]);
|
||||||
|
self->array_length = array_len;
|
||||||
|
self->array_dynamic = is_array == 2;
|
||||||
|
}
|
||||||
|
|
||||||
|
void kwil_enum_parse(struct kwil_enum_t* self, FILE* file) {
|
||||||
|
kwil_init_enum(self);
|
||||||
|
|
||||||
|
char value_buffer[24];
|
||||||
|
int length;
|
||||||
|
value_buffer[23] = '\0';
|
||||||
|
|
||||||
|
length = read_line_without_whitespace(file, value_buffer, 23);
|
||||||
|
while(!feof(file) && value_buffer[0] != '}') {
|
||||||
|
if(value_buffer[length-1] == ',') {
|
||||||
|
value_buffer[length-1] = '\0';
|
||||||
|
}
|
||||||
|
kwil_enum_add_value(self, value_buffer);
|
||||||
|
length = read_line_without_whitespace(file, value_buffer, 23);
|
||||||
|
printf("\t\"%s\",\n", self->enum_values[self->enum_values_len-1].name);
|
||||||
|
}
|
||||||
|
|
||||||
|
printf("]\n\n");
|
||||||
|
}
|
||||||
|
|
@ -0,0 +1,17 @@
|
||||||
|
#ifndef _kwil_parse_H
|
||||||
|
#define _kwil_parse_H
|
||||||
|
|
||||||
|
#include <stdio.h>
|
||||||
|
#include "kwil_def.h"
|
||||||
|
|
||||||
|
// Open a header file and parse it for it's kwil types
|
||||||
|
extern void kwil_header_parse(struct kwil_header_t* self);
|
||||||
|
// Read a file until the end of struct definition.
|
||||||
|
extern void kwil_struct_parse(struct kwil_struct_t* self, FILE* file);
|
||||||
|
// Parse a field without checking custom types.
|
||||||
|
// Custom types will not be checked yet as they may not be known to kwel yet.
|
||||||
|
extern void kwil_field_parse(struct kwil_field_t* self, FILE* file);
|
||||||
|
// Parse an enum and it's possible values.
|
||||||
|
extern void kwil_enum_parse(struct kwil_enum_t* self, FILE* file);
|
||||||
|
|
||||||
|
#endif // !_kwil_parse_H
|
||||||
Loading…
Reference in New Issue