WIN32 does have some kind of symbolic links (e.g. in NTFS), but the
readlink() function is not available.  Disable symlink checking for
now.

Don't use fixed stack buffers.

Don't use fixed stack buffers.

If we want to check whether a file is a directory, use GLib's
g_file_test() instead of stat().

Make the event_pipe (formerly main_notify) send/receive a set of
events, with a callback for each one.

The default event PIPE_EVENT_SIGNAL does not have a callback.  It
is still there for waking up the main thread, when it is waiting for
the player thread.

Continuing the previous patch.

We are going to migrate away from the concept of notifying the main
thread.  There should be events sent to it instead.  This patch starts
a series to implement that.

Only include headers which are really needed.  os_compat.h aimed to
make MPD easily portable, but was never actually made portable.

When there are no archive plugins, we do not need the archive API at
all.  Drop all its overhead.

MPD 0.13 and older followed all symbolic links.  Although this can be
a security problem (as it has always been), 0.14 should offer the same
default behaviour as 0.13.

The configuration options "follow_outside_symlinks" and
"follow_inside_symlinks" let the user control whether MPD should
follow symbolic links in the music directory.

[mk: converted variables to "bool"; moved configuration to
update_global_init()]

Those two functions are called when MPD starts and exits.  It allows
the update library to perform global initialization and
deinitialization.  The implementations are currently empty.

Renamed the local variable "basename", which shadows the POSIX
function basename().

skip_symlinks() expects an UTF-8 encoded file name, but
updateDirectory() passed ent->d_name (in file system encoding) to it.
Convert it to UTF-8 first.

Nearly all mapper functions can fail and will then return NULL.  Add
checks to all callers.

GLib's g_path_get_basename() is much more reliable than
mpd_basename().  The latter could be tricked into an assertion
failure.

Skip only the special directory entries "." and "..", don't skip all
other "hidden" files.

"idle" waits until something noteworthy happens on the server,
e.g. song change, playlist modified, database updated.  This allows
clients to keep up to date without polling.

Don't follow relative symlinks which point into the music directory.
This allows you to organize music with symbolic links, without MPD
managing separate copies of each song.

The mapper library maps directory and song objects to file system
paths.  With this central library, the code mixture in path.c should
be cleaned up, and we will be able to add neat features like aliasing.

By always creating the parent directory, we can use delete_name_in()
without further lookups.  The parents which may non exist will be
pruned later.  An update request for a non-existing or empty directory
should be quite unusual, so this doesn't add any measurable overhead.

In order to optimize buffer usage, pass only the base file name to
updateInDirectory().  This way, updateInDirectory() may choose when to
allocate a larger buffer for the full path.

delete_name_in() is similar to delete_path(), but it does not need to
look up the parent directory.

Previously only updates with subdirectories being
specified could be queued.  No harm in queueing
full updates.

There is only once update thread at a time.  Make the "modified" flag
global and remove the return values of most functions.  Propagating an
error is only useful for updateDirectory(), since updateInDirectory()
will delete failed subdirectories.

When a song file was not modified, MPD printed the debug message "not
a directory or music", because the first "if" branch did not return.

When the update queue is full, directory_update_init() did not free
the path argument.

Since the return value cannot be -1 anymore, we can make it unsigned.

The documentation for directory_update_init() was incorrect: a job ID
must be positive, not non-negative.  If the update queue is full and
no job was created, it makes more sense to return 0 instead of -1,
because it is more consistent with the return value of isUpdatingDB().

When the update task is idle, there is no need to check for deleted
songs.  Return early from reap_update_task().

pthread_join() expects a "pointer to a pointer" parameter, but it got
a "pointer to an enum".  On AMD64, an enum is smaller than a pointer,
leading to a buffer overflow.