# header file for published Mandarin lexicon.

# Each line of the lexicon contains the following seven fields
# in the following order.  The first field contains a name, the second
# contains a regular expression that should match every instance of that
# field.  Fields are tab separated, here as well as in the lexicon.

# For full documentation of the lexicon, please read the file ma_lex.doc!

# Usage of this file:
#  simple syntax check:
#	parse-lex -s ma_lex.hdr ma_lex.v03
#  output full "parse" of fields:
#	parse-lex ma_lex.hdr ma_lex.v03
# The "parse-lex" tool is available via FTP from ftp.ldc.upenn.edu

let $mchar=([¡-þ]{2})

HEADWORD	(T)?$mchar+

PINYIN	[a-z:]+[1-5]( [a-z:]+[1-5])*

TONE	[1-5]( [1-5])*

# pronunciation

PRON	[a-zA-Z@&%>]+( [a-zA-Z@&%>]+)*

# number of occurrences in Xinhua newswire

XINHUA	[0-9]+

# number of occurrences in training transcripts

TRAIN	[0-9]+

# part-of-speech tagging
let $pos=(acronym|adj|adj_r|adv|adv_r|class|class_r|conj|for_name|interj|name|name_affix|noun|number|onom|part|part_struc|part_asp|part_final|phrase|prep|pro|surname|surname_affix|verb|verb_r|number_class|name_seg|affix)

POS	$pos(/$pos)*/?