# $Id$
#
# This is the playlist file, for specifying iTunes-style
# "Smart Playlists".
# 
# The syntax is as follows:
#
# "Playlist Name" { expression }
#
# An expression consists of:
#
# ID3-tag-name operator operand
#
# Where valid ID3-tag-names are:
#   Artist (string)
#   Album (string)
#   Genre (string) 
#   Path (string) -- full path to song, including filename
#   Composer (string)
#   Orchestra (string)
#   Conductor (string)
#   Grouping (string) -- I don't even know what this is...
#   Comment (string)
#   Type (string) -- "mp3 audio file", "m4a audio file", etc.
#   Year (int)
#   BPM (int)
#   Bitrate (int)
#   Date (date)
#
# Valid operators include:
#   is, includes (string), matches (string)
#   >, <, <=, >=, = (int)
#   after, before (date)
#
# the "is" operator must exactly match the tag, while
# the "includes" operator matches a substring and
# "matches" operator makes trivial wildcard match using
# "?" and "*" as wildcards for any character and any
# string accordingly.  All matches are case-insensitive.
#
# Valid operands include:
#   "string value" (string)
#   integer (int)
#
# Multiple expressions can be anded or ored together,
# using the keywords OR and AND (or || and &&).
# The unary not operator is also supported using the
# keyword NOT (or !)
#
# Examples:
#
# "techno" {
#    genre includes "techno" ||
#    artist includes "zombie"
# }
#
# This would match songs by "Rob Zombie" or "White Zombie",
# as well as those with a genre of "Techno-Industrial" or
# "Trance/Techno", for example.
#
# "AAC Files" {
#   path includes ".m4a" ||
#   path includes ".m4p"
# }
#
# This would match all m4a and m4p files -- i.e. iTunes-ripped aac files
# or songs downloaded from iTMS.
#
# "Orchestral Music" {
#   Orchestra !IS "" ||
#   Conductor !IS ""
# }
#
# This would match anything with *anything* entered as a
# orchestra or conductor... this would probably include any
# orchestral music.  Kind of ugly, but works!
#
#
# DATES
#
# Dates are kind of funky.  The "date" of a file is when it
# was created on the file system, or the date that it was first
# entered into the database, whichever is earlier.  The date of 
# a file can be matched with the terms "before" or "after".
#
# One example of a valid date is a date in yyyy-mm-dd format:
#
# "Files added after January 1, 2004" {
#   date after 2004-01-01
# }
#
# There are also some special date keywords:
#  "today", "yesterday", "last week", "last month", "last year"
# 
# A valid date can also be made by appling an interval to a 
# date.  As an example, a valid date might be:
#
# 3 weeks before today
# or
# 3 weeks ago   
#
# You can combine these, of course.
#
# "3 weeks before today" is the same as "2 weeks before last week" 
# or "1 week after last month" or "21 days before today" or
# "20 days before yesterday" or "7 days after last month".  You get
# the idea.
# 
# Note that the playlists are only generated at the time that mt-daapd
# starts... so while the dates will be accurate at start time, they
# may become inaccurate with time.  Yes, I know... it's on my list.  :)
#
# So, examples:
#
# "Recently Added MP3s" {
#   date after last month AND file includes ".mp3"
# } 
#
# This matches only mp3 files added in the last 30 days.
#
# 
# SUMMARY
#
# I expect that this language will grow over time.  If you want
# to hack on it, see src/lexer.l, src/parser.y, and src/playlist.c
#
# If there is something missing you particularly want, let me
# (rpedde@users.sourceforge.net) know!
#

"60's Music" {
    Year >= 1960 && Year < 1970
}

"Recently Added" {
    Date after 2 weeks ago
}

"Non-DRMed Music" {
    path not includes ".m4p"
}

"AAC Files" {
    path includes ".m4p" ||
    path includes ".m4a" ||
    path includes ".aac"
}