commit 548804f4b49a06aa9033bea2dd292f8a12997b82 Author: Clyne Sullivan Date: Tue Jan 23 08:17:07 2018 -0500 initial commit diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..6a2f0f7 --- /dev/null +++ b/.gitignore @@ -0,0 +1,2 @@ +shell +*.o diff --git a/Makefile b/Makefile new file mode 100644 index 0000000..82b6431 --- /dev/null +++ b/Makefile @@ -0,0 +1,7 @@ +CC = gcc -m32 + +CFLAGS = -Wall -Wextra -I. -ggdb -fno-builtin + +all: + $(CC) $(CFLAGS) -c parser.c + $(CC) $(CFLAGS) shell.c parser.o -o shell diff --git a/README.md b/README.md new file mode 100644 index 0000000..385650c --- /dev/null +++ b/README.md @@ -0,0 +1,6 @@ +# interpreter +This project aims to provide a very minimal scripting language for embedded systems. Many other languages already exist, such as Lua, Tcl, or BASIC; however, most implementations require certain system calls like a read() and write(), expecting a filesystem. This interpreter aims to be as independent as possible: parsing script from strings at a time, having minimal built-in functions (so the user can define their own prints and such), and only requiring a few library functions. + +To use this program with your own device, only two functions are truly needed: malloc, and strcmp. + +This project is still in heavy development, so don't expect much. Right now only function calls are supported, without variable name expansion. To include it in your own project, just link in parser.o and use the header files. diff --git a/parser.c b/parser.c new file mode 100644 index 0000000..ba1ce2a --- /dev/null +++ b/parser.c @@ -0,0 +1,205 @@ +#include + +#include +#include +#include + +static const char *interpreter_operators = "=("; + +uint8_t isalpha(char c) +{ + return (c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z'); +} + +uint8_t isnum(char c) +{ + return (c >= '0' && c <= '9'); +} + +uint8_t isname(char c) +{ + return isalpha(c) || isnum(c); +} + +uint8_t isspace(char c) +{ + return (c == ' ' || c == '\t' || c == '\n'); +} + +uint8_t isoper(char c) +{ + for (uint8_t i = 0; i < sizeof(interpreter_operators); i++) { + if (c == interpreter_operators[i]) + return 1; + } + + return 0; +} + +void interpreter_init(interpreter *interp) +{ + interp->status = READY; + interp->vcount = 0; + interp->vars = (variable *)calloc(32, sizeof(variable)); + interp->names = (char **)calloc(32, sizeof(char *)); + interp->stack = (stack_t *)calloc(64, sizeof(stack_t)); +} + +void interpreter_define_value(interpreter *interp, const char *name, int32_t value) +{ + interp->names[interp->vcount] = (char *)name; + interp->vars[interp->vcount].nameidx = interp->vcount; + interp->vars[interp->vcount].type = VALUE; + interp->vars[interp->vcount].value = (uint32_t)value; + interp->vcount++; +} + +void interpreter_define_cfunc(interpreter *interp, const char *name, func_t addr) +{ + interp->names[interp->vcount] = (char *)name; + interp->vars[interp->vcount].nameidx = interp->vcount; + interp->vars[interp->vcount].type = CFUNCTION; + interp->vars[interp->vcount].value = (uint32_t)addr; + interp->vcount++; +} + +int32_t interpreter_get_value(interpreter *interp, const char *name) +{ + for (uint16_t i = 0; i < interp->vcount; i++) { + if (!strcmp(interp->names[i], name)) + return (int32_t)interp->vars[i].value; + } + + return 0; +} + +/** + * doline section + */ + +bool namencmp(const char *name, const char *s) +{ + uint16_t i; + for (i = 0; name[i] == s[i] && s[i] != '\0'; i++); + return (name[i] == '\0'); +} + +uint16_t spacecount(const char *s) +{ + uint16_t i; + for (i = 0; isspace(s[i]); i++); + return i; +} + +char *copystr(const char *s) +{ + uint16_t len = 0; + while (s[len++] != '\n'); + char *buf = (char *)malloc(len); + for (uint16_t i = 0; i < len; i++) + buf[i] = s[i]; + return buf; +} + +char *copysubstr(const char *s, int end) +{ + char *buf = (char *)malloc(end); + for (uint16_t i = 0; i < end; i++) + buf[i] = s[i]; + return buf; +} + +int interpreter_doline(interpreter *interp, const char *line) +{ + variable *bits[16]; + uint16_t offset = 0, boffset = 0; + + // check for var/func set or usage +getvar: + for (uint16_t i = 0; i < interp->vcount; i++) { + if (namencmp(interp->names[i], line)) { + bits[boffset++] = &interp->vars[i]; + // get past name + for (uint16_t j = 0; interp->names[i][j] != '\0'; j++, offset++); + break; + } + } + + // defining new variable + if (boffset == 0) { + uint16_t end; + for (end = 0; isname(line[end]); end++); + interpreter_define_value(interp, copysubstr(line, end), 0); + goto getvar; // try again + } + + // skip whitespace + offset += spacecount(line + offset); + + if (boffset == 0 && line[offset] != '=') + return -1; // variable not found + + // find operator + if (line[offset] == '\0') { + // print value + return -99; + } else if (line[offset] == '=') { + return -23; + // assignment/expression + //offset++; + //offset += spacecount(line + offset); + //if (boffset > 0) + // bits[boffset]->value = (uint32_t)copystr(line + offset); + } else if (line[offset] == '(') { + // function call + offset++; + if (bits[0]->type != FUNCTION && bits[0]->type != CFUNCTION) + return -2; + offset += spacecount(line + offset); + + // collect arg offsets + uint16_t offsets[8]; + uint8_t ooffset = 0; + while (line[offset] != ')' && line[offset] != '\0') { + offsets[ooffset] = offset; + offset += spacecount(line + offset); + + uint8_t isvn = 1; + do { + if (line[offset] == ' ' || line[offset] == '\t') { + offset += spacecount(line + offset); + isvn = 0; + } + + if (line[offset] == ',') { + offset++; + ooffset++; + break; + } else if (line[offset] == ')') { + ooffset++; + break; + } else if (isvn == 0) { + return -3; + } + } while (++offset); + } + + // populate stack + for (uint8_t i = 0; i < ooffset; i++) { + uint16_t j; + for (j = offsets[i]; line[j] != ' ' && line[j] != '\t' && + line[j] != ',' && line[j] != ')'; j++); + j -= offsets[i]; + interp->stack[i] = (char *)malloc(j); + for (uint16_t k = 0; k < j; k++) + ((char *)interp->stack[i])[k] = line[offsets[i] + k]; + } + + ((func_t)bits[0]->value)(interp->stack); + } else { + return -2; // invalid operation + } + + return 0; +} + diff --git a/parser.h b/parser.h new file mode 100644 index 0000000..c6a2b17 --- /dev/null +++ b/parser.h @@ -0,0 +1,31 @@ +#ifndef PARSER_H_ +#define PARSER_H_ + +#include + +typedef void *stack_t; + +typedef struct { + uint16_t status; + uint16_t vcount; + variable *vars; + char **names; + stack_t *stack; +} interpreter; + +enum status { + READY = 0 +}; + +typedef void (*func_t)(stack_t *); + +void interpreter_init(interpreter *); + +void interpreter_define_value(interpreter *, const char *, int32_t); +void interpreter_define_cfunc(interpreter *, const char *, func_t); + +int32_t interpreter_get_value(interpreter *, const char *); + +int interpreter_doline(interpreter *, const char *); + +#endif // PARSER_H_ diff --git a/shell.c b/shell.c new file mode 100644 index 0000000..b4bfd63 --- /dev/null +++ b/shell.c @@ -0,0 +1,23 @@ +#include + +#include + +void test(stack_t *stack) +{ + printf("%s\n", stack[0]); +} + +int main(int argc, char *argv[]) +{ + interpreter interp; + + interpreter_init(&interp); + interpreter_define_value(&interp, "answer", 42); + interpreter_define_cfunc(&interp, "test", test); + + if (argc == 2) { + printf("%d\n", interpreter_doline(&interp, argv[1])); + } + + return 0; +} diff --git a/variable.h b/variable.h new file mode 100644 index 0000000..fd84a6e --- /dev/null +++ b/variable.h @@ -0,0 +1,24 @@ +#ifndef TOKEN_H_ +#define TOKEN_H_ + +#include + +typedef struct { + uint16_t nameidx; + uint8_t type; + uint8_t info; + uint32_t value; +} variable; + +#define INFO_ARGS(x) ((x) & 0x07) +#define INFO_RET (1 << 3) + +enum vartype { + VALUE = 0, + VARIABLE, + OPERATOR, + FUNCTION, + CFUNCTION +}; + +#endif // TOKEN_H_