Dynamic PEG for interpreted languages.
Nelze vybrat více než 25 témat Téma musí začínat písmenem nebo číslem, může obsahovat pomlčky („-“) a může být dlouhé až 35 znaků.
 

320 řádky
6.6 KiB

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
typedef struct Node Node;
typedef struct Exploration Exploration;
enum opcode {
STRING,
AND,
OR,
STAR,
PLUS,
CLASS,
QUESTION_MARK,
EXCLAMATION_MARK,
DOT
};
struct Exploration
{
int position;
char *text;
};
Exploration *mkExploration(char *text)
{
Exploration *exploration = calloc(1,sizeof(Exploration*)); //sizeof du type pointeur
exploration->position= 0;
exploration->text= text;
return exploration;
}
void advance(Exploration *exploration, int add)
{
exploration->position+= add;
}
int atEnd(Exploration *exploration) {
return exploration->text[exploration->position] == 0;
}
void setPosition(Exploration *exploration, int position)
{
exploration->position= position;
}
int getPosition(Exploration *exploration) {
return exploration->position;
}
int currentChar(Exploration *exploration)
{
return exploration->text[exploration->position];
}
char *currentText(Exploration *exploration)
{
return exploration->text + exploration->position;
}
struct String { enum opcode type; char *stringValue; int len;};
struct Class { enum opcode type; char *stringValue; int len;};
struct Node
{
enum opcode type;
union {
struct String STRING;
struct Class CLASS;
Node *children[2];
};
};
Node *mkNode(enum opcode type)
{
Node *node= calloc(1, sizeof(Node));
node->type= type;
return node;
}
Node *mkString(char *value)
{
Node *node= mkNode(STRING);
node->STRING.stringValue= value;
node->STRING.len=strlen(value);
return node;
}
Node *mkAnd(Node *node1, Node *node2)
{
Node *node= mkNode(AND);
node->children[0]= node1;
node->children[1]= node2;
return node;
}
Node *mkStar(Node *child)
{
Node *node= mkNode(STAR);
node->children[0]= child;
return node;
}
Node *mkPlus(Node *child)
{
Node *node= mkNode(PLUS);
node->children[0]= child;
return node;
}
Node *mkOr(Node *node1, Node *node2)
{
Node *node= mkNode(OR);
node->children[0]= node1;
node->children[1]= node2;
return node;
}
Node *mkClass(char* str)
{
Node *node= mkNode(CLASS);
node->STRING.stringValue= str;
node->STRING.len=strlen(str);
return node;
}
Node *mkQuestionMark(Node *child)
{
Node *node= mkNode(QUESTION_MARK);
node->children[0]= child;
return node;
}
Node *mkExclamationMark(Node *child)
{
Node *node= mkNode(EXCLAMATION_MARK);
node->children[0]= child;
return node;
}
Node *mkDot()
{
Node *node= mkNode(DOT);
return node;
}
Node *_checktype(Node *object, enum opcode type)
{
if (object->type == type) return object;
fprintf(stderr, "\naccesing type %i as if it were a %i\n", object->type, type);
exit(1);
return 0;
}
#define get(object, type, member) (_checktype(object, type)->type.member)
const char* getTypeName(enum opcode type)
{
switch (type)
{
case STAR: return "STAR";
case EXCLAMATION_MARK: return "NOT";
case DOT: return "DOT";
case QUESTION_MARK: return "QMARK";
case PLUS : return "PLUS";
case CLASS : return "CLASS";
default : printf("unexpected use");return 0;
abort();
}
}
void println(Node *node, int indent)
{
for(int i=0;i<indent;i++){
printf("----");
}
indent++;
switch (node->type) {
case STRING: printf("%.*s", get(node, STRING, len), get(node, STRING, stringValue)); return;
case AND: {
printf("AND\n");
println(node->children[0],indent);printf("\n");
println(node->children[1],indent);printf("\n");return;
}
case OR: {
printf("OR\n");
println(node->children[0],indent);printf("\n");
println(node->children[1],indent);printf("\n");return;
}
case DOT:{
printf("DOT\n");return;
}
case CLASS:{
printf("CLASS : Char in [%.*s]\n",get(node, CLASS, len),get(node, CLASS, stringValue));return;
}
default : {
printf("%s\n",getTypeName(node->type));
println(node->children[0],indent);printf("\n");return;
}
}
abort();
}
int execute(Node *node, Exploration *in)
{
switch (node->type) {
case STRING: {
if (strncmp(currentText(in), node->STRING.stringValue, node->STRING.len)) {
return 0;
}
advance(in, node->STRING.len);
return 1;
}
case AND: {
int pos= getPosition(in);
if (!execute(node->children[0], in)) //si il y a eu une erreur
{
return 0;
} //si ça s'est bien passé
if (!execute(node->children[1], in)) {
setPosition(in, pos);
return 0;
}
return 1;
}
case OR: {
if (execute(node->children[0], in)) {
return 1;
}
return execute(node->children[1], in);
}
case STAR: {
while (execute(node->children[0], in));
return 1;
}
case PLUS: {
if (!execute(node->children[0], in)) {
return 0;
}
while (execute(node->children[0], in));
return 1;
}
case CLASS: {
if (!currentChar(in)) {
return 0;
}
if (strchr(node->STRING.stringValue, currentChar(in))) {
advance(in, 1);
return 1;
}
return 0;
}
case QUESTION_MARK: {
execute(node->children[0], in);
return 1;
}
case EXCLAMATION_MARK: {
int pos= getPosition(in);
if (!execute(node->children[0], in)) {
return 1;
}
setPosition(in, pos);
return 0;
}
case DOT: {
if (atEnd(in)) {
return 0;
}
advance(in, 1);
return 1;
}
}
printf("this cannot happen\n");
abort();
}
int main(int argc, char **argv)
{
Exploration *in= mkExploration("aabbcc");
/*Node *program=
mkString("aabbcc");
Node *program=
mkAnd(mkString("aab"),
mkString("bcc"));*/
/*Node *program=
mkAnd(mkString("aa"),
mkAnd(mkString("bb"),
mkString("cc")));*/
/*Node *program=
mkOr(mkPlus(mkString("a")), // "a"*"b"*"c"*
mkAnd(mkStar(mkString("b")),
mkStar(mkString("c"))));*/
Node *program= mkStar(mkClass("abc"));
Node *program3=mkAnd(mkAnd(mkPlus(mkString("hello world")),mkExclamationMark(mkString("how are u world !?"))),mkStar(mkClass("abcde")));
if (!execute(program, in) || !atEnd(in))
{
printf("no match, current position : %i\n", getPosition(in));
} else
{
printf("match, current position : %i\n", getPosition(in));
}// 0 => no match, 1 => match
println(program3,0);
return 0;
}