6

Although it is often disabled, some web pages allow you to view index pages. I am not sure if that is the proper terminology (if not please say so), but an example is http://download.opensuse.org/. It shows you all of the files and "folders" for that web page.

I don't know if this is possible, but I think it would be great if konqueror allowed you to view these web pages using the normal file manager as if they were folders on a file system, allowing you to navigate up and down, view previews, copy files to your hard drive, and so on. It looks like the web page is effectively treated as a file system, so treating it as a real read-only file system seems natural. It may be infeasible, though.

Index pages are autogenerated html pages by server, the same as any other page online.

Your idea could be possible only if we write some regexps to get the data. Since most servers run on apache IIS and some other minor, it wount be a very hard task, but not native.

In generic, all links to files/folders could be found getting all links and excluding links with non-allowed characters for file name like ./

Source of http://download.opensuse.org/

Index of /



Index of /
Name Last modified Size debug/ 27-Nov-2008 11:13 -

distribution/ 19-Jun-2009 18:00 -
factory/ 18-Oct-2008 11:30 -
repositories/ 18-Jun-2009 13:44 -
source/ 27-Nov-2008 11:12 -
update/ 06-May-2009 01:35 -


Apache/2.2.11 (Linux/SUSE) Server at download.opensuse.org Port 80
MirrorBrain powered by Apache

Thanks for the info. I figured it would require some sort of interpreter, but I didn't know what sort or how difficult it would be.

this could be a perfect job for directoryindex:/ KIOslave

As someone pointed out, those directory pages are made by the server at the other end, and it's effectively just a webpage. Whilst a parsing program could be made (a KIO would be the easiest) it would only work if they are following the apache default (or ISS default / whichever other server defaults we supported) I think it would only end up being far more confusing for a user, with very little actual gain.

david_edmundson wrote:As someone pointed out, those directory pages are made by the server at the other end, and it's effectively just a webpage. Whilst a parsing program could be made (a KIO would be the easiest) it would only work if they are following the apache default (or ISS default / whichever other server defaults we supported) I think it would only end up being far more confusing for a user, with very little actual gain.

Well directory index has links ( tags) So to find all links isn't a hard task. Then it should remove links that contains illegal characters like ? in url.


Or improved idea

Under "File" add a view as directory option
Phrase web page, find all links and display them all as folders. Links that looks like "regular" links (no # ? etc in it)/"non regular"/ is from same/another domain display in different colors.

When you would like to copy a "folder" a menu cames up, so you define a filters: depth to go, include associated files (for html files), get files from same domain only.

This would be more generic solution Not only for index pages, but also for websvn's or any other web file source (like for serves, that came up in the end with list/table of (many) files to download one by one, like my university resource)

@Lukas: Very good idea! Konqueror should have option called show links as files. In this mode Konqueror gets any links from page and using some information to grouping it. We can also generate preview, but it isn't very good idea - some files can be huge(we can of course get only some first bytes).

It would be great to go on osnews.pl and tell show page as directory. Konqueror will show four group: main, news, news candidate and categories. In main we could have any link from body root. We can also have week article and best information. In news we have all news from this page. In news candidate we could have any news candidate. And in categories we should have Apple, etc.

The problem is to handling folders in existing KIO File.

I would seen this option realized in Nautilus style. When we went into directory with index.xhtml/index.html file, it will show up top pane: View web page from this directory. I think, it will also great to read additional options from .directory file - this option will be placed in special menu and user will be warned about malware problem.