struct urlpos *
get_urls_css_file (const char *file, const char *url)
{
struct file_memory *fm;
struct map_context ctx;
/* Load the file. */
fm = read_file (file);
if (!fm)
{
logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
return NULL;
}
DEBUGP (("Loaded %s (size %s).\n", file, number_to_static_string (fm->length)));
ctx.text = fm->content;
ctx.head = ctx.tail = NULL;
ctx.base = NULL;
ctx.parent_base = url ? url : opt.base_href;
ctx.document_file = file;
ctx.nofollow = 0;
get_urls_css (&ctx, 0, fm->length);
read_file_free (fm);
return ctx.head;
}
struct robot_specs *
res_parse_from_file (const char *filename)
{
struct robot_specs *specs;
struct file_memory *fm = read_file (filename);
if (!fm)
{
logprintf (LOG_NOTQUIET, "Cannot open %s: %s",
filename, strerror (errno));
return NULL;
}
specs = res_parse (fm->content, fm->length);
read_file_free (fm);
return specs;
}
/* Change the links in one file. LINKS is a list of links in the
document, along with their positions and the desired direction of
the conversion. */
static void
convert_links (const char *file, struct urlpos *links)
{
struct file_memory *fm;
FILE *fp;
const char *p;
downloaded_file_t downloaded_file_return;
struct urlpos *link;
int to_url_count = 0, to_file_count = 0;
logprintf (LOG_VERBOSE, _("Converting %s... "), file);
{
/* First we do a "dry run": go through the list L and see whether
any URL needs to be converted in the first place. If not, just
leave the file alone. */
int dry_count = 0;
struct urlpos *dry;
for (dry = links; dry; dry = dry->next)
if (dry->convert != CO_NOCONVERT)
++dry_count;
if (!dry_count)
{
logputs (LOG_VERBOSE, _("nothing to do.\n"));
return;
}
}
fm = read_file (file);
if (!fm)
{
logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
file, strerror (errno));
return;
}
downloaded_file_return = downloaded_file (CHECK_FOR_FILE, file);
if (opt.backup_converted && downloaded_file_return)
write_backup_file (file, downloaded_file_return);
/* Before opening the file for writing, unlink the file. This is
important if the data in FM is mmaped. In such case, nulling the
file, which is what fopen() below does, would make us read all
zeroes from the mmaped region. */
if (unlink (file) < 0 && errno != ENOENT)
{
logprintf (LOG_NOTQUIET, _("Unable to delete %s: %s\n"),
quote (file), strerror (errno));
read_file_free (fm);
return;
}
/* Now open the file for writing. */
fp = fopen (file, "wb");
if (!fp)
{
logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
file, strerror (errno));
read_file_free (fm);
return;
}
/* Here we loop through all the URLs in file, replacing those of
them that are downloaded with relative references. */
p = fm->content;
for (link = links; link; link = link->next)
{
char *url_start = fm->content + link->pos;
if (link->pos >= fm->length)
{
DEBUGP (("Something strange is going on. Please investigate."));
break;
}
/* If the URL is not to be converted, skip it. */
if (link->convert == CO_NOCONVERT)
{
DEBUGP (("Skipping %s at position %d.\n", link->url->url, link->pos));
continue;
}
/* Echo the file contents, up to the offending URL's opening
quote, to the outfile. */
fwrite (p, 1, url_start - p, fp);
p = url_start;
switch (link->convert)
{
case CO_CONVERT_TO_RELATIVE:
/* Convert absolute URL to relative. */
{
char *newname = construct_relative (file, link->local_name);
char *quoted_newname = local_quote_string (newname);
if (link->link_css_p)
p = replace_plain (p, link->size, fp, quoted_newname);
else if (!link->link_refresh_p)
p = replace_attr (p, link->size, fp, quoted_newname);
else
p = replace_attr_refresh_hack (p, link->size, fp, quoted_newname,
link->refresh_timeout);
DEBUGP (("TO_RELATIVE: %s to %s at position %d in %s.\n",
link->url->url, newname, link->pos, file));
xfree (newname);
xfree (quoted_newname);
++to_file_count;
break;
}
case CO_CONVERT_TO_COMPLETE:
/* Convert the link to absolute URL. */
{
char *newlink = link->url->url;
char *quoted_newlink = html_quote_string (newlink);
if (link->link_css_p)
p = replace_plain (p, link->size, fp, quoted_newlink);
else if (!link->link_refresh_p)
p = replace_attr (p, link->size, fp, quoted_newlink);
else
p = replace_attr_refresh_hack (p, link->size, fp, quoted_newlink,
link->refresh_timeout);
DEBUGP (("TO_COMPLETE: <something> to %s at position %d in %s.\n",
newlink, link->pos, file));
xfree (quoted_newlink);
++to_url_count;
break;
}
case CO_NULLIFY_BASE:
/* Change the base href to "". */
p = replace_attr (p, link->size, fp, "");
break;
case CO_NOCONVERT:
abort ();
break;
}
}
/* Output the rest of the file. */
if (p - fm->content < fm->length)
fwrite (p, 1, fm->length - (p - fm->content), fp);
fclose (fp);
read_file_free (fm);
logprintf (LOG_VERBOSE, "%d-%d\n", to_file_count, to_url_count);
}
