mirror of
https://github.com/netsurf-browser/netsurf
synced 2024-11-24 15:29:45 +03:00
d36758ad97
If "easy" charset conversion fails, return a copy of the original string. svn path=/trunk/netsurf/; revision=10580
115 lines
2.4 KiB
C
115 lines
2.4 KiB
C
/*
|
|
* Copyright 2009 Rene W. Olsen <ac@rebels.com>
|
|
* Copyright 2009 Stephen Fellner <sf.amiga@gmail.com>
|
|
* Copyright 2009 Chris Young <chris@unsatisfactorysoftware.co.uk>
|
|
*
|
|
* This file is part of NetSurf, http://www.netsurf-browser.org/
|
|
*
|
|
* NetSurf is free software; you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License as published by
|
|
* the Free Software Foundation; version 2 of the License.
|
|
*
|
|
* NetSurf is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
* GNU General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public License
|
|
* along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
|
|
#include <stdio.h>
|
|
#include <ctype.h>
|
|
#include <string.h>
|
|
|
|
#include "proto/exec.h"
|
|
|
|
#include "urlhistory.h"
|
|
|
|
#include "content/urldb.h"
|
|
#include "desktop/options.h"
|
|
|
|
static struct List PageList;
|
|
|
|
void URLHistory_Init( void )
|
|
{
|
|
// Initialise page list
|
|
NewList( &PageList );
|
|
}
|
|
|
|
|
|
void URLHistory_Free( void )
|
|
{
|
|
struct Node *node;
|
|
|
|
while(( node = RemHead( &PageList )))
|
|
{
|
|
if( node->ln_Name) FreeVec( node->ln_Name );
|
|
FreeVec( node );
|
|
}
|
|
}
|
|
|
|
|
|
void URLHistory_ClearList( void )
|
|
{
|
|
struct Node *node;
|
|
|
|
while(( node = RemHead( &PageList )))
|
|
{
|
|
if( node->ln_Name) FreeVec( node->ln_Name );
|
|
FreeVec( node );
|
|
}
|
|
}
|
|
|
|
|
|
struct List * URLHistory_GetList( void )
|
|
{
|
|
return &PageList;
|
|
}
|
|
|
|
static bool URLHistoryFound(const char *url, const struct url_data *data)
|
|
{
|
|
struct Node *node;
|
|
|
|
/* skip non-visted pages */
|
|
if(data->visits <= 0) return true;
|
|
|
|
/* skip this URL if it is already in the list */
|
|
if(URLHistory_FindPage(url)) return true;
|
|
|
|
node = AllocVec( sizeof( struct Node ), MEMF_SHARED|MEMF_CLEAR );
|
|
|
|
if ( node )
|
|
{
|
|
STRPTR urladd = (STRPTR) AllocVec( strlen ( url ) + 1, MEMF_SHARED|MEMF_CLEAR );
|
|
if ( urladd )
|
|
{
|
|
strcpy(urladd, url);
|
|
node->ln_Name = urladd;
|
|
AddTail( &PageList, node );
|
|
}
|
|
else
|
|
{
|
|
FreeVec(node);
|
|
}
|
|
}
|
|
return true;
|
|
}
|
|
|
|
struct Node * URLHistory_FindPage( const char *urlString )
|
|
{
|
|
return FindName(&PageList,urlString);
|
|
}
|
|
|
|
|
|
void URLHistory_AddPage( const char * urlString )
|
|
{
|
|
if(!option_url_suggestion) return;
|
|
|
|
// Only search if length > 0
|
|
if( strlen( urlString ) > 0 )
|
|
{
|
|
urldb_iterate_partial(urlString, URLHistoryFound);
|
|
}
|
|
}
|