htmlurls.c

harvest是一个下载html网页得机器人

    if (strncasecmp(p, "url", 3) == 0) { /* there is URL */
      q += 3;
      while (isspace(*q)) q++;          /* skip spaces */
      if (*q == '=') { 			/* if there is URL = , skip it */
        q++;                            /* skip '=' */
        while (isspace(*q)) q++;          /* skip spaces */
	p = q;
      } /* else - this is HTML 4.0 conformant (?) refresh with (relative) URL starting with letters 'url' */
    }

    q = xstrdup(p);                         /* copy name */
    if ((p = strchr(q, '\"')) != NULL) {    /* terminate string on '"' */
      *p = '\0';
    } else {
      if ((p = strchr(q, '>')) != NULL)     /* terminate string on '>' */
      *p = '\0';
    }

    /* q should now contain the URL */

    /* Relative URLs are processed by various browsers, so enable them */
    if (base != (char *) NULL) {
      v = q;
      q = url_parse_relative(v, base);
      xfree(v);
    }

    if (q != (char *) NULL) {             /* Just in case... */
      add_buffer(urls, q, strlen(q));     /* Add URL to urls */
      add_buffer(urls, "\n", 1);
      xfree(q);
    }
    return;
}


/*
 *  process_framesrc() - Extracts the URL from the frame src tag.
 *
 *  DFM HACK - deals with:
 *      <FRAME SRC="url">
 *      <FRAME SRC = "url">
 *      <FRAME SRC = " url ">
 */
void process_framesrc(s)
     char *s;
{
    char *p, *q, *tmps, *v;

    /* Find the SRC in the FRAME */
    if ((tmps = strstr_icase(s, "src")) == NULL)
        return;

    /* Grab the URL from the SRC */
    if ((p = strchr(tmps, '=')) != NULL) {
        p++;                    /* skip '=' */
        while (isspace(*p) || (*p == '\"'))
            p++;                /* skip space '"'s */
        q = xstrdup(p);          /* copy URL */
        if ((p = strchr(q, '\"')) != NULL)      /* terminate string */
            *p = '\0';
        if ((p = strchr(q, ' ')) != NULL)       /* terminate string */
            *p = '\0';
        if (base != (char *) NULL) {
            v = q;
            q = url_parse_relative(v, base);
            xfree(v);
        }
        if (q != (char *) NULL) {
            add_buffer(urls, q, strlen(q));     /* Add URL to urls */
            add_buffer(urls, "\n", 1);
            xfree(q);
        }
        return;
    }
}

/*
 *  process_anchor() - Extracts the URL from the anchor href tag.
 *  ALSO DOES SO FROM AREA TAG FOR CLIENT SIDE IMAGE MAPS DFM
 *  Will process these anchors (HREF is case-insenstive):
 *
 *      <A HREF="url">
 *      <A HREF = "url">
 *      <A HREF = " url ">
 *      <A HREF='url'>
 *      <A HREF = 'url'>
 *      <A HREF = ' url '>
 *
 *
 */
void process_anchor(s)
     char *s;
{
    char *p, *q, *tmps, *v;
    int singlequote, doublequote;

    singlequote = doublequote = 0;

    /* Find the HREF in the anchor */
    if ((tmps = strstr_icase(s, "href")) == NULL)
	return;

    /* Grab the URL from the HREF */
    if ((p = strchr(tmps, '=')) != NULL) {
	p++;			/* skip '=' */
	while (isspace(*p) || (*p == '\"') || (*p == '\'')) {
	    if(*p ==  '\"')
		++doublequote;
	    if(*p == '\'')
		++singlequote;
	    p++;		/* skip space '"'s */
	}
	q = xstrdup(p);		/* copy URL */
	while ((p = strchr (q, '\n')) != NULL) {
	    strcpy(p, p+1);
	}
	if (((p = strchr(q, '\"')) != NULL && doublequote == 1) || (((p = strchr(q, '\'')) != NULL) && singlequote == 1))	/* terminate string */
	    *p = '\0';
	if ((p = strchr(q, ' ')) != NULL)	/* terminate string */
	    *p = '\0';
	if (base != (char *) NULL) {
	    v = q;
	    q = url_parse_relative(v, base);
	    xfree(v);
	}
	if (q != (char *) NULL) {
	    add_buffer(urls, q, strlen(q));	/* Add URL to urls */
	    add_buffer(urls, "\n", 1);
	    xfree(q);
	}
	return;
    }
}

/*
 *  read_file() - Reads the file fp into memory and returns a pointer to it.
 */
Buffer *read_file(fp)
     FILE *fp;
{
    static Buffer *b;
    char buf[BUFSIZ];
    int nread;

    b = create_buffer(BUFSIZ);

    while ((nread = fread(buf, 1, BUFSIZ, fp)) > 0)
	add_buffer(b, buf, nread);

    return (b);
}

void process_base(buf)
     char *buf;
{
    char *t = NULL;
    char *p = NULL;
    char *q = NULL;

    /* Find the HREF in the anchor */
    if ((t = strstr_icase(buf, "href")) == (char *) NULL)
	return;

    /* Grab the URL from the HREF */
    if ((p = strchr(t, '=')) != NULL) {
	p++;			/* skip '=' */
	while (isspace(*p) || (*p == '\"'))
	    p++;		/* skip space '"'s */
	q = xstrdup(p);		/* copy URL */
	if ((p = strchr(q, '\"')) != NULL)	/* terminate string */
	    *p = '\0';
	if ((p = strchr(q, ' ')) != NULL)	/* terminate string */
	    *p = '\0';
    }
    if (q != (char *) NULL) {
	xfree(base);
	base = xstrdup(q);
    }
}

void process_node(mp)
     struct mark_up *mp;
{
    if ((mp->type == M_BASE) &&
	(mp->start != NULL) && (strlen(mp->start) > 5))
	process_base(mp->start);

    if ((mp->type == M_ANCHOR) &&
	(mp->start != NULL) && (strlen(mp->start) > 5))
	process_anchor(mp->start);

	/*DFM hack for FRAMES*/
    if ((mp->type == M_FRAME) &&
        (mp->start != NULL) && (strlen(mp->start) > 5))
        process_framesrc(mp->start);

        /*DFM hack for AREA (client side Image Map*/
    if ((mp->type == M_AREA) &&
        (mp->start != NULL) && (strlen(mp->start) > 5))
        process_anchor(mp->start);

    if ((mp->type == M_META) &&
        (mp->start != NULL) && (strlen(mp->start) > 5)) {
            process_metarobots(mp->start);
            process_metapull(mp->start);	/* HS */
        }
}


static void free_struct_markup(x)
     struct mark_up *x;
{
    if (x->text)
	free(x->text);
    if (x->start)
	free(x->start);
    if (x->end)
	free(x->end);
    free(x);
}

int main(argc, argv)
     int argc;
     char *argv[];
{
    struct mark_up *HTMLParse();
    struct mark_up *mp = NULL;
    struct mark_up *walker = NULL;
    struct mark_up *t = NULL;
    Buffer *b = NULL;
    FILE *fp = NULL;
    FILE *logfp = NULL;

    if (getenv("HARVEST_GATHERER_LOGFILE") != (char *) NULL)
	logfp = fopen(getenv("HARVEST_GATHERER_LOGFILE"), "a+");
    if (logfp == (FILE *) NULL)
	logfp = stderr;

    init_log3("HTMLurls", logfp, stderr);
    debug_init();
    for (argc--, argv++; argc > 0 && **argv == '-'; argc--, argv++) {
	if (!strncmp(*argv, "-D", 2)) {
	    debug_flag(*argv);
	} else if (!strcmp(*argv, "--base-url")) {
	    argc--;
	    argv++;
	    if (argc < 1)
		usage();
	    base = xstrdup(*argv);
	}
    }

    if (argc < 1)
	usage();


    /* Parse the HTML file */
    if ((fp = fopen(*argv, "r")) == NULL) {
	log_errno(*argv);
	exit(1);
    }
#if 0 /* kjl/7mar2002 */
    if (getenv("ENUMERATOR_URL"))
	Url = xstrdup(getenv("ENUMERATOR_URL"));
    if (Url == (char *) NULL)
	Url = xstrdup(*argv);
#endif

    b = read_file(fp);
    fclose(fp);
    mp = HTMLParse(NULL, b->data);
    free_buffer(b);

    urls = create_buffer(BUFSIZ);

    /* Extract important information from the parsed HTML */
    for (walker = mp; walker != NULL;
	t = walker, walker = walker->next, free_struct_markup(t))
	process_node(walker);

    if (RobotsFollow) fwrite(urls->data, 1, urls->length, stdout);
    free_buffer(urls);
    if (RobotsIndex)
      exit(0);
    else
      exit(99); /* Pick a number, any number */
}