/* Conversion of links to local files.
- Copyright (C) 2003, 2004, 2005, 2006, 2007,
- 2008 Free Software Foundation, Inc.
+ Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011
+ Free Software Foundation, Inc.
This file is part of GNU Wget.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
-#ifdef HAVE_UNISTD_H
-# include <unistd.h>
-#endif /* HAVE_UNISTD_H */
+#include <unistd.h>
#include <errno.h>
#include <assert.h>
#include "convert.h"
#include "hash.h"
#include "ptimer.h"
#include "res.h"
+#include "html-url.h"
+#include "css-url.h"
+#include "iri.h"
static struct hash_table *dl_file_url_map;
struct hash_table *dl_url_file_map;
-/* Set of HTML files downloaded in this Wget run, used for link
+/* Set of HTML/CSS files downloaded in this Wget run, used for link
conversion after Wget is done. */
struct hash_table *downloaded_html_set;
+struct hash_table *downloaded_css_set;
static void convert_links (const char *, struct urlpos *);
-/* This function is called when the retrieval is done to convert the
- links that have been downloaded. It has to be called at the end of
- the retrieval, because only then does Wget know conclusively which
- URLs have been downloaded, and which not, so it can tell which
- direction to convert to.
-
- The "direction" means that the URLs to the files that have been
- downloaded get converted to the relative URL which will point to
- that file. And the other URLs get converted to the remote URL on
- the server.
-
- All the downloaded HTMLs are kept in downloaded_html_files, and
- downloaded URLs in urls_downloaded. All the information is
- extracted from these two lists. */
-void
-convert_all_links (void)
+static void
+convert_links_in_hashtable (struct hash_table *downloaded_set,
+ int is_css,
+ int *file_count)
{
int i;
- double secs;
- int file_count = 0;
-
- struct ptimer *timer = ptimer_new ();
int cnt;
char **file_array;
cnt = 0;
- if (downloaded_html_set)
- cnt = hash_table_count (downloaded_html_set);
+ if (downloaded_set)
+ cnt = hash_table_count (downloaded_set);
if (cnt == 0)
- goto cleanup;
+ return;
file_array = alloca_array (char *, cnt);
- string_set_to_array (downloaded_html_set, file_array);
+ string_set_to_array (downloaded_set, file_array);
for (i = 0; i < cnt; i++)
{
char *url;
char *file = file_array[i];
- /* Determine the URL of the HTML file. get_urls_html will need
+ /* Determine the URL of the file. get_urls_{html,css} will need
it. */
url = hash_table_get (dl_file_url_map, file);
if (!url)
DEBUGP (("Scanning %s (from %s)\n", file, url));
- /* Parse the HTML file... */
- urls = get_urls_html (file, url, NULL);
+ /* Parse the file... */
+ urls = is_css ? get_urls_css_file (file, url) :
+ get_urls_html (file, url, NULL, NULL);
/* We don't respect meta_disallow_follow here because, even if
the file is not followed, we might still want to convert the
for (cur_url = urls; cur_url; cur_url = cur_url->next)
{
char *local_name;
- struct url *u = cur_url->url;
+ struct url *u;
+ struct iri *pi;
if (cur_url->link_base_p)
{
/* We decide the direction of conversion according to whether
a URL was downloaded. Downloaded URLs will be converted
ABS2REL, whereas non-downloaded will be converted REL2ABS. */
+
+ pi = iri_new ();
+ set_uri_encoding (pi, opt.locale, true);
+
+ u = url_parse (cur_url->url->url, NULL, pi, true);
+ if (!u)
+ continue;
+
local_name = hash_table_get (dl_url_file_map, u->url);
/* Decide on the conversion type. */
cur_url->local_name = NULL;
DEBUGP (("will convert url %s to complete\n", u->url));
}
+
+ url_free (u);
+ iri_free (pi);
}
/* Convert the links in the file. */
convert_links (file, urls);
- ++file_count;
+ ++*file_count;
/* Free the data. */
free_urlpos (urls);
}
+}
+
+/* This function is called when the retrieval is done to convert the
+ links that have been downloaded. It has to be called at the end of
+ the retrieval, because only then does Wget know conclusively which
+ URLs have been downloaded, and which not, so it can tell which
+ direction to convert to.
+
+ The "direction" means that the URLs to the files that have been
+ downloaded get converted to the relative URL which will point to
+ that file. And the other URLs get converted to the remote URL on
+ the server.
+
+ All the downloaded HTMLs are kept in downloaded_html_files, and
+ downloaded URLs in urls_downloaded. All the information is
+ extracted from these two lists. */
+
+void
+convert_all_links (void)
+{
+ double secs;
+ int file_count = 0;
+
+ struct ptimer *timer = ptimer_new ();
+
+ convert_links_in_hashtable (downloaded_html_set, 0, &file_count);
+ convert_links_in_hashtable (downloaded_css_set, 1, &file_count);
secs = ptimer_measure (timer);
logprintf (LOG_VERBOSE, _("Converted %d files in %s seconds.\n"),
file_count, print_decimal (secs));
-cleanup:
+
ptimer_destroy (timer);
}
static void write_backup_file (const char *, downloaded_file_t);
+static const char *replace_plain (const char*, int, FILE*, const char *);
static const char *replace_attr (const char *, int, FILE *, const char *);
static const char *replace_attr_refresh_hack (const char *, int, FILE *,
const char *, int);
-static char *local_quote_string (const char *);
+static char *local_quote_string (const char *, bool);
static char *construct_relative (const char *, const char *);
-/* Change the links in one HTML file. LINKS is a list of links in the
+/* 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
}
}
- fm = read_file (file);
+ fm = wget_read_file (file);
if (!fm)
{
logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
{
logprintf (LOG_NOTQUIET, _("Unable to delete %s: %s\n"),
quote (file), strerror (errno));
- read_file_free (fm);
+ wget_read_file_free (fm);
return;
}
/* Now open the file for writing. */
{
logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
file, strerror (errno));
- read_file_free (fm);
+ wget_read_file_free (fm);
return;
}
/* Convert absolute URL to relative. */
{
char *newname = construct_relative (file, link->local_name);
- char *quoted_newname = local_quote_string (newname);
+ char *quoted_newname = local_quote_string (newname,
+ link->link_css_p);
- if (!link->link_refresh_p)
+ 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,
char *newlink = link->url->url;
char *quoted_newlink = html_quote_string (newlink);
- if (!link->link_refresh_p)
+ if (link->link_css_p)
+ p = replace_plain (p, link->size, fp, 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,
if (p - fm->content < fm->length)
fwrite (p, 1, fm->length - (p - fm->content), fp);
fclose (fp);
- read_file_free (fm);
+ wget_read_file_free (fm);
logprintf (LOG_VERBOSE, "%d-%d\n", to_file_count, to_url_count);
}
/* Rather than just writing over the original .html file with the
converted version, save the former to *.orig. Note we only do
this for files we've _successfully_ downloaded, so we don't
- clobber .orig files sitting around from previous invocations. */
+ clobber .orig files sitting around from previous invocations.
+ On VMS, use "_orig" instead of ".orig". See "wget.h". */
/* Construct the backup filename as the original name plus ".orig". */
size_t filename_len = strlen (file);
char* filename_plus_orig_suffix;
+ /* TODO: hack this to work with css files */
if (downloaded_file_return == FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED)
{
/* Just write "orig" over "html". We need to do it this way
else /* downloaded_file_return == FILE_DOWNLOADED_NORMALLY */
{
/* Append ".orig" to the name. */
- filename_plus_orig_suffix = alloca (filename_len + sizeof (".orig"));
+ filename_plus_orig_suffix = alloca (filename_len + sizeof (ORIG_SFX));
strcpy (filename_plus_orig_suffix, file);
- strcpy (filename_plus_orig_suffix + filename_len, ".orig");
+ strcpy (filename_plus_orig_suffix + filename_len, ORIG_SFX);
}
if (!converted_files)
static bool find_fragment (const char *, int, const char **, const char **);
+/* Replace a string with NEW_TEXT. Ignore quoting. */
+static const char *
+replace_plain (const char *p, int size, FILE *fp, const char *new_text)
+{
+ fputs (new_text, fp);
+ p += size;
+ return p;
+}
+
/* Replace an attribute's original text with NEW_TEXT. */
static const char *
We quote ? as %3F to avoid passing part of the file name as the
parameter when browsing the converted file through HTTP. However,
- it is safe to do this only when `--html-extension' is turned on.
+ it is safe to do this only when `--adjust-extension' is turned on.
This is because converting "index.html?foo=bar" to
"index.html%3Ffoo=bar" would break local browsing, as the latter
isn't even recognized as an HTML file! However, converting
"index.html?foo=bar.html" to "index.html%3Ffoo=bar.html" should be
safe for both local and HTTP-served browsing.
- We always quote "#" as "%23" and "%" as "%25" because those
- characters have special meanings in URLs. */
+ We always quote "#" as "%23", "%" as "%25" and ";" as "%3B"
+ because those characters have special meanings in URLs. */
static char *
-local_quote_string (const char *file)
+local_quote_string (const char *file, bool no_html_quote)
{
const char *from;
char *newname, *to;
- char *any = strpbrk (file, "?#%");
+ char *any = strpbrk (file, "?#%;");
if (!any)
- return html_quote_string (file);
+ return no_html_quote ? strdup (file) : html_quote_string (file);
/* Allocate space assuming the worst-case scenario, each character
having to be quoted. */
*to++ = '2';
*to++ = '3';
break;
+ case ';':
+ *to++ = '%';
+ *to++ = '3';
+ *to++ = 'B';
+ break;
case '?':
- if (opt.html_extension)
+ if (opt.adjust_extension)
{
*to++ = '%';
*to++ = '3';
}
*to = '\0';
- return html_quote_string (newname);
+ return no_html_quote ? strdup (newname) : html_quote_string (newname);
}
\f
/* Book-keeping code for dl_file_url_map, dl_url_file_map,
/* Register that FILE is an HTML file that has been downloaded. */
void
-register_html (const char *url, const char *file)
+register_html (const char *file)
{
if (!downloaded_html_set)
downloaded_html_set = make_string_hash_table (0);
string_set_add (downloaded_html_set, file);
}
+/* Register that FILE is a CSS file that has been downloaded. */
+
+void
+register_css (const char *file)
+{
+ if (!downloaded_css_set)
+ downloaded_css_set = make_string_hash_table (0);
+ string_set_add (downloaded_css_set, file);
+}
+
static void downloaded_files_free (void);
/* Cleanup the data structures associated with this file. */
However, our hash tables only accept pointers for keys and values.
So when we need a pointer, we use the address of a
downloaded_file_t variable of static storage. */
-
+
static downloaded_file_t *
downloaded_mode_to_ptr (downloaded_file_t mode)
{
projects / wget / blobdiff