archeus/src/std/parser.c

#include "arc/std/parser.h"
#include "arc/std/bool.h"
#include "arc/std/errno.h"
#include "arc/std/lexer.h"
#include "arc/std/vector.h"
#include <stdint.h>
#include <stdlib.h>
#include <string.h>

struct ARC_Parser {
    ARC_Array language;

    ARC_Lexer *lexer;

    void                     *data;
    void                     *userData;
    ARC_ParserData_CreateFn  *createDataFn;
    ARC_ParserData_DestroyFn *destroyDataFn;

};

void ARC_Parser_Create(ARC_Parser **parser, ARC_Array *language, ARC_Parser_InitLexerRulesFn initLexerRulesFn, ARC_ParserData_CreateFn *createDataFn, ARC_ParserData_DestroyFn *destroyDataFn, void *userData){
    *parser = (ARC_Parser *)malloc(sizeof(ARC_Parser));

    //set the language size to 0 and data to NULL in case the language is NULL
    (*parser)->language.size = 0;
    (*parser)->language.data = NULL;

    //if the language exists, copy the language
    if(language != NULL){
        (*parser)->language.size = language->size;
        (*parser)->language.data = malloc(sizeof(ARC_ParserTag) * language->size);

        for(uint32_t index = 0; index < language->size; index++){
            ARC_ParserTag *languageTag = ((ARC_ParserTag *)language->data) + index;
            ARC_ParserTag *currentTag  = ((ARC_ParserTag *)(*parser)->language.data) + index;

            //copy the language tag into the current tag
            currentTag->tagId            = languageTag->tagId;
            currentTag->tokensOrTagsSize = languageTag->tokensOrTagsSize;

            //create place to store tokens or tags
            currentTag->tokensOrTags = (uint32_t **)malloc(sizeof(uint32_t *) * languageTag->tokensOrTagsSize);

            //copy each or section into the tokensOrTags
            for(uint32_t orIndex = 0; orIndex < languageTag->tokensOrTagsSize; orIndex++){
                currentTag->tokensOrTags[orIndex] = (uint32_t *)malloc(sizeof(uint32_t) * (languageTag->tokensOrTags[orIndex][0] + 1));

                for(uint32_t tokenOrTagIndex = 0; tokenOrTagIndex < languageTag->tokensOrTags[orIndex][0] + 1; tokenOrTagIndex++){
                    currentTag->tokensOrTags[orIndex][tokenOrTagIndex] = languageTag->tokensOrTags[orIndex][tokenOrTagIndex];
                }
            }
        }
    }

    //create the lexer
    ARC_Lexer_Create(&((*parser)->lexer));

    //register instructions to the lexer
    initLexerRulesFn(((*parser)->lexer));

    //set the data to null (the parse function is what creates it)
    (*parser)->data = NULL;

    //set the userData for the create callback function
    (*parser)->userData = userData;

    //init the create function callback with null, then copy the callback if it exists
    (*parser)->createDataFn = NULL;
    if(createDataFn != NULL){
        (*parser)->createDataFn = (ARC_ParserData_CreateFn *)malloc(sizeof(ARC_ParserData_CreateFn));
        *((*parser)->createDataFn) = *createDataFn;
    }

    //init the destroy function callback with null, then copy the callback if it exists
    (*parser)->destroyDataFn = NULL;
    if(createDataFn != NULL){
        (*parser)->destroyDataFn = (ARC_ParserData_DestroyFn *)malloc(sizeof(ARC_ParserData_DestroyFn));
        *((*parser)->destroyDataFn) = *destroyDataFn;
    }
}

void ARC_Parser_CreateFromVector(ARC_Parser **parser, ARC_Vector *language, ARC_Parser_InitLexerRulesFn initLexerRulesFn, ARC_ParserData_CreateFn *createDataFn, ARC_ParserData_DestroyFn *destroyDataFn, void *userData){
    //creates the variables to copy the vector into
    const uint32_t languageSize = ARC_Vector_GetSize(language);
    ARC_ParserTag languageArray[languageSize];

    //copy the language from a vector into an array
    for(uint32_t index = 0; index < languageSize; index++){
        languageArray[index] = *(ARC_ParserTag *)ARC_Vector_Get(language, index);
    }

    //set the vector data as an ARC_Array
    ARC_Array languageAsArray = {
        languageSize,
        languageArray
    };

    //create the parser
    ARC_Parser_Create(parser, &languageAsArray, initLexerRulesFn, createDataFn, destroyDataFn, userData);
}

void ARC_Parser_CreateFromString(ARC_Parser **parser, ARC_String *languageString, ARC_Parser_InitLexerRulesFn initLexerRulesFn){
}

void ARC_Parser_Destroy(ARC_Parser *parser){
    //clear all the copied token or tags from memory
    for(uint32_t index = 0; index < parser->language.size; index++){
        ARC_ParserTag *currentTag  = ((ARC_ParserTag *)parser->language.data) + index;

        //free the orIndex vlues
        for(uint32_t orIndex = 0; orIndex < currentTag->tokensOrTagsSize; orIndex++){
            free(currentTag->tokensOrTags[orIndex]);
        }

        //free the tokens or tags
        free(currentTag->tokensOrTags);
    }

    //free the creation function callback
    if(parser->createDataFn != NULL){
        free(parser->createDataFn);
    }

    //free the data and the deletion function callback
    if(parser->destroyDataFn != NULL){
        (*(parser->destroyDataFn))(parser->data, parser->userData);
        free(parser->destroyDataFn);
    }

    //clear the copied language from memory
    free(parser->language.data);

    ARC_Lexer_Destroy(parser->lexer);

    free(parser);
}

//private creation function for ARC_ParserTagToken type
//note: token and tagTokens will be set to null, it is safe to create tagTokens outside of this as the destructor will clean it up
void ARC_ParserTagToken_Create(ARC_ParserTagToken **tagToken, uint32_t id){
    *tagToken = (ARC_ParserTagToken *)malloc(sizeof(ARC_ParserTagToken));

    (*tagToken)->id        = id;
    (*tagToken)->token     = NULL;
    (*tagToken)->tagTokens = NULL;
}

//private destroy function for ARC_ParserTagToken type
void ARC_ParserTagToken_Destroy(ARC_ParserTagToken *tagToken){
    //destroy the tag token vector
    if(tagToken->tagTokens != NULL){
        ARC_Vector_Destroy(tagToken->tagTokens);
    }

    free(tagToken);
}

//private function to cleanup a parser tag token from a vector
void ARC_ParserTagToken_VectorDestroyDataFn(void *data){
    //we can just use the destroy function with casted data
    ARC_ParserTagToken_Destroy((ARC_ParserTagToken *)data);
}

//private recusive function to parse a tag
ARC_Bool ARC_Parser_ParseTag(ARC_Parser *parser, ARC_ParserTagToken *tagToken, uint32_t *lexerIndex){
    //get the current tag
    ARC_ParserTag *tag = NULL;
    for(uint32_t index = 0; index < parser->language.size; index++){
        ARC_ParserTag *foundTag = ((ARC_ParserTag *)parser->language.data) + index;
        if(foundTag->tagId == tagToken->id){
            tag = foundTag;
            break;
        }
    }

    //if the tag was not found can't do much, so throw an error
    if(tag == NULL){
        arc_errno = ARC_ERRNO_NULL;
        ARC_DEBUG_LOG_ERROR_WITH_VARIABLES("ARC_Parser_ParseTag(parser, subdata, tagId), could not find tag with id: %u", tagToken->id);
        return ARC_False;
    }

    //create a vector of tag token to use if a rule is validated, a comparison function is not needed as it will be iterated through
    ARC_Vector_DestroyDataFn destroyTokenTagFn = ARC_ParserTagToken_VectorDestroyDataFn;
    ARC_Vector_Create(&(tagToken->tagTokens), NULL, &destroyTokenTagFn);

    //loop through each or section of the tags and tokens
    for(uint32_t orIndex = 0; orIndex < tag->tokensOrTagsSize; orIndex++){
        //reset the tag tokens for each or index
        ARC_Vector_Clear(tagToken->tagTokens);

        //loop through each token or tag to check if the lexed data matches
        uint32_t lexerCheckIndex = *lexerIndex;
        ARC_Bool foundRule = ARC_True;
        for(uint32_t tokenOrTagIndex = 1; tokenOrTagIndex < tag->tokensOrTags[orIndex][0] + 1; tokenOrTagIndex++){
            //get next tag id to check
            uint32_t nextTagId = tag->tokensOrTags[orIndex][tokenOrTagIndex];

            //check if it is lambda (can return safely)
            if(nextTagId == ARC_PARSER_TAG_LAMBDA){
                break;
            }

            //if the value isn't a token it is a tag, so recurs if it isn't a token
            ARC_Bool isToken = ARC_Lexer_IsTokenId(parser->lexer, nextTagId);
            if(isToken == ARC_False){
                //create a temporary lexer index in case the rule does not exist
                uint32_t tempLexerCheckIndex = lexerCheckIndex;

                //create tag token for if the rule works
                ARC_ParserTagToken *nextTagToken;
                ARC_ParserTagToken_Create(&nextTagToken, nextTagId);

                //check if the tag works if not break to continue checking next or index
                foundRule = ARC_Parser_ParseTag(parser, nextTagToken, &tempLexerCheckIndex);
                if(foundRule == ARC_False){
                    //clean up the tag token
                    ARC_ParserTagToken_Destroy(nextTagToken);
                    break;
                }

                //add the tag token because rule was found
                ARC_Vector_Add(tagToken->tagTokens, nextTagToken);

                //increase the lexer check index as a recursed rule was found, and continue checking
                lexerCheckIndex = tempLexerCheckIndex;

                continue;
            }

            //check if there is another token that can be used
            if(lexerCheckIndex >= ARC_Lexer_GetTokensSize(parser->lexer)){
                //out of tokens to the current or does not work, so break
                foundRule = ARC_False;
                break;
            }

            //get the next token in the lexer and increment the lexers index
            ARC_LexerToken *token = ARC_Lexer_GetToken(parser->lexer, lexerCheckIndex);
            lexerCheckIndex++;

            //if the token rule does not match the current token in the current or statement the token rule could not be found for the current or index so break
            if(token->rule != tag->tokensOrTags[orIndex][tokenOrTagIndex]){
                foundRule = ARC_False;
                break;
            }

            //the rule was a match so create a tag token to store the token in
            ARC_ParserTagToken *nextTagToken;
            ARC_ParserTagToken_Create(&nextTagToken, nextTagId);
            nextTagToken->token = token;

            //add the token to the tag tokens
            ARC_Vector_Add(tagToken->tagTokens, nextTagToken);
        }

        //if the rule is found we don't need to check anymore so we can return out
        if(foundRule == ARC_True){
            *lexerIndex = lexerCheckIndex;

            //cleanup
            return ARC_True;
        }
    }

    //cleanup
    ARC_Vector_Destroy(tagToken->tagTokens);
    tagToken->tagTokens = NULL;

    //no rule was found, so return false
    return ARC_False;
}

void ARC_Parser_Parse(ARC_Parser *parser, ARC_String **data){
    //make sure the parser has a language
    if(parser->language.size == 0){
        ARC_DEBUG_LOG_ERROR("ARC_Parser_Parse(parser, data), no parser language defined");
        return;
    }

    //lex the subdata
    ARC_Lexer_LexString(parser->lexer, data);
    if(arc_errno){
        ARC_DEBUG_LOG_ERROR("ARC_Parser_Parse(parser, data), could not lex the given data");
        ARC_Lexer_Clear(parser->lexer);
        return;
    }

    //set the lexer index to start and get the first tag
    uint32_t lexerIndex = 0;
    ARC_ParserTag *startTag = parser->language.data;

    //setup a tag token that will be passed to the creation callback on success
    ARC_ParserTagToken *tagToken;
    ARC_ParserTagToken_Create(&tagToken, startTag->tagId);

    //TODO: handle error checks for if parsing fails
    //recursivly parse from the inital start tag
    ARC_Bool parsed = ARC_Parser_ParseTag(parser, tagToken, &lexerIndex);
    ARC_Bool allTokensParsed = lexerIndex == ARC_Lexer_GetTokensSize(parser->lexer);

    //create the data if the creation callback exists
    if(parser->createDataFn != NULL){
        (*(parser->createDataFn))(&(parser->data), tagToken, parser->userData);
    }

    //cleanup
    ARC_ParserTagToken_Destroy(tagToken);
    ARC_Lexer_Clear(parser->lexer);

    //error if anything went wrong
    if(parsed == ARC_False || allTokensParsed == ARC_False || arc_errno){
        arc_errno = ARC_ERRNO_DATA;
        ARC_DEBUG_LOG_ERROR_WITH_VARIABLES("ARC_Parser_Parse(parser, data), could not parse the given data at lexer index: %u", lexerIndex);
        return;
    }
}

void ARC_Parser_ParseFile(ARC_Parser *parser, ARC_String *path){

}

void ARC_Parser_ClearData(ARC_Parser *parser){
    //check if that data exists and the destructor exists to make sure they can be run
    if(parser->data != NULL && parser->destroyDataFn != NULL){
        (*(parser->destroyDataFn))(parser->data, parser->userData);
    }

    //TODO: might want to error here
    parser->data = NULL;
}

void *ARC_Parser_GetData(ARC_Parser *parser){
    return parser->data;
}