data.config

# this file contains lists of corpus texts. To fully load a corpus cpname,
# use 'make cpname'. To download only a specific file named fname of corpus cpname,
# use 'make cpname/fname'.

# http://www.data-compression.info/Corpora/CanterburyCorpus/index.html
CANTERBURY = \
	canterbury/alice29.txt \
	canterbury/asyoulik.txt \
	canterbury/cp.html \
	canterbury/fields.c \
	canterbury/grammar.lsp \
	canterbury/kennedy.xls \
	canterbury/lcet10.txt \
	canterbury/plrabn12.txt \
	canterbury/ptt5 \
	canterbury/sum \
	canterbury/xargs.1

# http://www.data-compression.info/Corpora/CanterburyCorpus/index.html
LARGECANTERBURY = \
	largecanterbury/bible.txt \
	largecanterbury/E.coli \
	largecanterbury/world192.txt

# http://sun.aei.polsl.pl/~sdeor/index.php?page=silesia
SILESIA = \
	silesia/dickens \
	silesia/mozilla \
	silesia/mr \
	silesia/nci \
	silesia/ooffice \
	silesia/osdb \
	silesia/reymont \
	silesia/samba \
	silesia/sao \
	silesia/webster \
	silesia/xml \
	silesia/x-ray

# http://pizzachili.dcc.uchile.cl/texts.html
PIZZACHILI = \
	pizzachili/sources \
	pizzachili/pitches \
	pizzachili/proteins \
	pizzachili/dna \
	pizzachili/english \
	pizzachili/dblp.xml

# http://pizzachili.dcc.uchile.cl/repcorpus.html (only REAL texts)
REPETITIVE = \
	repetitive/Escherichia_Coli \
	repetitive/cere \
	repetitive/coreutils \
	repetitive/einstein.de.txt \
	repetitive/einstein.en.txt \
	repetitive/influenza \
	repetitive/kernel \
	repetitive/para \
	repetitive/world_leaders

# http://hgdownload.soe.ucsc.edu/downloads.html (only a selection)
GENOMES = \
	genomes/hg38 \
	genomes/mm10 \
	genomes/rn6

# all files in one macro
ALLTESTDATA = $(CANTERBURY) $(LARGECANTERBURY) $(SILESIA) $(PIZZACHILI) $(REPETITIVE) $(GENOMES)