Fix empty field parsing and resolve bison conflicts

[sa-parse] / src / csv.y

%{
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

extern int yylex();
extern int yylineno;
extern char* yytext;

void yyerror(const char* s);

/* Data structures for CSV representation */
typedef struct field_node {
    char* content;
    struct field_node* next;
} field_node_t;

typedef struct record_node {
    field_node_t* fields;
    struct record_node* next;
} record_node_t;

typedef struct csv_file {
    record_node_t* header;
    record_node_t* records;
} csv_file_t;

/* Global variables */
csv_file_t* csv_data = NULL;

/* Helper functions */
field_node_t* create_field(char* content);
record_node_t* create_record(field_node_t* fields);
field_node_t* append_field(field_node_t* list, field_node_t* new_field);
record_node_t* append_record(record_node_t* list, record_node_t* new_record);
void print_csv(csv_file_t* csv);
void free_csv(csv_file_t* csv);
%}

%union {
    char* str;
    struct field_node* field_list;
    struct record_node* record_node;
    struct csv_file* csv_file;
}

%token <str> FIELD_TOK
%token COMMA_TOK
%token CRLF_TOK

%type <field_list> field_list
%type <field_list> field_opt
%type <record_node> record
%type <record_node> consumed_record
%type <csv_file> file

%start file

%%

file:
    consumed_record
    {
        csv_data = malloc(sizeof(csv_file_t));
        if (!csv_data) YYERROR;
        csv_data->header = NULL;
        csv_data->records = $1;
        $$ = csv_data;
    }
    | file CRLF_TOK consumed_record
    {
        csv_data->records = append_record(csv_data->records, $3);
        $$ = csv_data;
    }
    ;

consumed_record:
    record
    {
        $$ = $1;
    }
    ;

record:
    field_list
    {
        $$ = create_record($1);
        if (!$$) YYERROR;
    }
    ;

field_list:
    field_opt
    {
        $$ = $1;
    }
    | field_list COMMA_TOK field_opt
    {
        $$ = append_field($1, $3);
    }
    ;

field_opt:
    FIELD_TOK
    {
        $$ = create_field($1);
        if (!$$) YYERROR;
    }
    | %empty
    {
        /* Empty field - create field with empty string */
        char* empty_content = strdup("");
        if (!empty_content) YYERROR;
        $$ = create_field(empty_content);
        if (!$$) YYERROR;
    }
    ;

%%

void yyerror(const char* s) {
    fprintf(stderr, "Parse error at line %d: %s\n", yylineno, s);
    if (yytext) {
        fprintf(stderr, "Near token: %s\n", yytext);
    }
}

field_node_t* create_field(char* content) {
    field_node_t* field = malloc(sizeof(field_node_t));
    if (!field) return NULL;
    /* Use the content directly without duplicating since lexer already allocated it */
    field->content = content;
    if (!field->content) {
        field->content = strdup("");
        if (!field->content) {
            free(field);
            return NULL;
        }
    }
    field->next = NULL;
    return field;
}

record_node_t* create_record(field_node_t* fields) {
    record_node_t* record = malloc(sizeof(record_node_t));
    if (!record) return NULL;
    record->fields = fields;
    record->next = NULL;
    return record;
}

field_node_t* append_field(field_node_t* list, field_node_t* new_field) {
    if (!list) return new_field;
    
    field_node_t* current = list;
    while (current->next) {
        current = current->next;
    }
    current->next = new_field;
    return list;
}

record_node_t* append_record(record_node_t* list, record_node_t* new_record) {
    if (!list) return new_record;
    
    record_node_t* current = list;
    while (current->next) {
        current = current->next;
    }
    current->next = new_record;
    return list;
}

void print_csv(csv_file_t* csv) {
    if (!csv) return;
    
    if (csv->header) {
        printf("Header: ");
        field_node_t* field = csv->header->fields;
        while (field) {
            printf("[%s]", field->content ? field->content : "");
            if (field->next) printf(", ");
            field = field->next;
        }
        printf("\n");
    }
    
    record_node_t* record = csv->records;
    int record_num = 1;
    while (record) {
        printf("Record %d: ", record_num++);
        field_node_t* field = record->fields;
        while (field) {
            printf("[%s]", field->content ? field->content : "");
            if (field->next) printf(", ");
            field = field->next;
        }
        printf("\n");
        record = record->next;
    }
}

void free_field_list(field_node_t* fields) {
    while (fields) {
        field_node_t* next = fields->next;
        free(fields->content);
        free(fields);
        fields = next;
    }
}

void free_record_list(record_node_t* records) {
    while (records) {
        record_node_t* next = records->next;
        free_field_list(records->fields);
        free(records);
        records = next;
    }
}

void free_csv(csv_file_t* csv) {
    if (!csv) return;
    
    if (csv->header) {
        free_field_list(csv->header->fields);
        free(csv->header);
    }
    
    free_record_list(csv->records);
    free(csv);
}

int main() {
    printf("CSV Parser - Enter CSV data (Ctrl+D to end):\n");
    
    int result = yyparse();
    
    if (result == 0) {
        printf("\nParsing successful!\n\n");
        print_csv(csv_data);
    } else {
        printf("\nParsing failed!\n");
    }
    
    free_csv(csv_data);
    return result;
}