TODO: dynamic buffer handling to never overflow

[EmilioSchi]

My solution works with small BUFFERSIZE value... but now i lose lastest bytes

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "html.h"

#define BUFFSIZE 2

int main(int argc, char **argv) {
    //TODO: dynamic buffer handling to never overflow
    FILE *f;
    HtmlDocument *doc;
    HtmlParseState *parse_state;

    char *buf = NULL; 
    buf = (char*) realloc (buf, (BUFFSIZE+1) * sizeof(char));
    buf[BUFFSIZE] = 0;

    const char *token = buf;


    if(argc != 2) {
        fprintf(stderr, "usage: html <file.html>\n");
        return 1;
    }

    size_t len = 0;
    size_t buffsize = BUFFSIZE;

    parse_state = html_parse_begin();
    if(!(f = fopen(argv[1], "r"))) {
        fprintf(stderr, "error: cannot open file %s\n", argv[1]);
        return 1;
    }

    int gain=0;
    int aftergain=0;
    int add=0;

    while(!feof(f)) {
        if (buffsize==0) {
                gain++;
                add=1;
                buf = (char*) realloc (buf, (gain+BUFFSIZE + 1) * sizeof(char));
                buf[gain+BUFFSIZE] = 0;
        } else {
            add = 0;
        }
        len = fread(buf+(BUFFSIZE+aftergain-buffsize), 1, add+buffsize, f);
        token = html_parse_stream(parse_state, buf + (BUFFSIZE+aftergain-buffsize), buf, len);
        buffsize = (token - buf);
        memmove(buf, token, BUFFSIZE+gain-buffsize);
        aftergain=gain;

    }

    doc = html_parse_end(parse_state);
    fclose(f);

    html_print_dom(doc);

    html_free_document(doc);

    free(buf);
    return 0;
}

Check this pls

[EmilioSchi]

Another solution without problems

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "html.h"

#define BUFFSIZE 3

int main(int argc, char **argv) {
    //TODO: dynamic buffer handling to never overflow
    FILE *f;
    HtmlDocument *doc;
    HtmlParseState *parse_state;

    char *buf = NULL; 
    buf = (char*) realloc (buf, (BUFFSIZE+1) * sizeof(char));
    buf[BUFFSIZE] = 0;

    const char *token = buf;

    if(argc != 2) {
        fprintf(stderr, "usage: html <file.html>\n");
        return 1;
    }

    size_t len = 0;
    size_t buffsize = BUFFSIZE;
    size_t plus_memory=0;
    size_t afterplus_memory=0;
    int plus_flag=0;

    parse_state = html_parse_begin();
    if(!(f = fopen(argv[1], "r"))) {
        fprintf(stderr, "error: cannot open file %s\n", argv[1]);
        return 1;
    }

    while(!feof(f)) {
        if (buffsize == 0) {
            plus_flag = 1;
            plus_memory++;

            buf = (char*) realloc (buf, (plus_memory+BUFFSIZE + 1) * sizeof(char));
            buf[plus_memory+BUFFSIZE] = 0;
        } else {
            plus_flag = 0;
        }

        len = fread(buf+(BUFFSIZE+afterplus_memory-buffsize), 1, plus_flag+buffsize, f);
        buf[BUFFSIZE+afterplus_memory-buffsize+len]=0;
        token = html_parse_stream(parse_state, buf + (BUFFSIZE+afterplus_memory-buffsize), buf, len);
        buffsize = (token - buf);
        memmove(buf, token, BUFFSIZE+plus_memory-buffsize);
        afterplus_memory=plus_memory;
    }

    doc = html_parse_end(parse_state);
    fclose(f);

    html_print_dom(doc);

    html_free_document(doc);

    free(buf);
    return 0;
}

[h4xxel]

Hi, could you submit this as a pull request so i can easily merge it?

[EmilioSchi]

Yes, Check this and merge it.
This is a clean solution. Easy to read ;)

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include "html.h"

#define BUFFSIZE    10

#define parser_balloc(BUFFER) \
    do {\
        BUFFER = (char*)realloc(BUFFER,\
         (parser_bplusmemory + BUFFSIZE + 1) * sizeof(char));\
        if (BUFFER == NULL) {\
            fprintf(stderr, "[PARSER] Error: memory not allocated.\n"); \
            return 1; /* TODO: Change it */ \
        } \
        BUFFER[parser_bplusmemory + BUFFSIZE] = 0;\
    } while (0)

int main (int argc, char* argv[]) {
    FILE *f;
    HtmlDocument *doc;
    HtmlParseState *parse_state;

    if(argc != 2) {
        fprintf(stderr, "usage: html <file.html>\n");
        return 1;
    }

    parse_state = html_parse_begin();

    if(!(f = fopen(argv[1], "r"))) {
        fprintf(stderr, "[PARSER] Error: cannot open file %s\n", argv[1]);
        return 1;
    }

    /* Buffer plus memory variables */
    size_t          parser_bplusmemory = 0;
    unsigned        plus_flag        = 0;
    size_t          read_bplusmemory     = 0;
    size_t          read             = 0;
    size_t          not_read         = 0;
    size_t          len_read         = 0;
    char*       buffer           = NULL;
    parser_balloc(buffer);
    size_t          token_miss       = BUFFSIZE;
    const char* token            = buffer;

    while (!feof(f)) {
        if (token_miss > 0) {
            plus_flag = 0;
        } else if (token_miss == 0){
            plus_flag = 1;
            parser_bplusmemory++;
            parser_balloc(buffer);
        } else {
            fprintf(stderr, "[PARSER] Error: token corrupt.\n");
            return 1;
        }
        read     = plus_flag + token_miss;
        not_read = BUFFSIZE + read_bplusmemory - token_miss;
        len_read = fread(buffer + (not_read), 1, read, f);
        buffer[not_read + len_read] = 0;
        token = html_parse_stream(parse_state, buffer + not_read, buffer, len_read);

        token_miss = (token - buffer);
        not_read = BUFFSIZE + parser_bplusmemory - token_miss;
        memmove(buffer, token, not_read);
        read_bplusmemory = parser_bplusmemory;
    }
    free(buffer);


    doc = html_parse_end(parse_state);

    fclose(f);

    html_print_dom(doc);
    html_free_document(doc);


    printf("\nMemory buffer end size: %zu\n", (BUFFSIZE+read_bplusmemory));

    return 0;
}