Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

playlist.xml now draws it's data from the common model in httprequests.lua output data is the same for backwards compatibility
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

instead of sending the uri as name to input_item_NewExt, send NULL and the core will deal with it (fix #5378)
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

This should help on Win64

Close #3148

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Some speed/seek functions are now gone, because they don't really make
sense for an audio playback.
Volume functions are now in, because this is useful

This reverts commit 0c16b5d0.

The same effect can already be obtained by setting CONTRIB_DIR to "yes"

Signed-off-by: Rafaël Carré <funman@videolan.org>

Instead of ignoring the URI anchor (if present), the input now parses
it for start and stop seek points. This is non-standard - anchors are
normally names found in HTML document. But at least, it does not
interfere with parsing real-world URLs (see the bug #5022 for example).

Also contrary to stream output chain, the hash symbol should be free of
shell quoting issues when used this.

Unfortunately, there is currently no way to determine whether the
MRL scheme uses file paths (e.g. file, dir, vdr, zip...) or not, so
we always try.

Anyway, the folder browsing menu item should not be visible at all when
the item is not a local file. So there is no need to deal with that
failure care here.

I don't know when "raw_device" was used, but it must have been removed
a loooooooong time ago.

This reverts commit fb9bbdc9.

This reverts commit 06108030.