Re("[\w\-]+")) + Spaces() +
Integer("dalign_query_end") +
Spaces() + AnyEol(),
Str("- ") + ToEol()) +
ToEol()) # blank line
domain_alignment = (domain_align_header +
Rep1(Opt(rf_line) +
domain_align_top +
domain_align_middle +
domain_align_bottom))
# //
record_end = Str("//") + AnyEol()
record = Std.record(Rep(sequence_info +
family_header +
(no_hit_line | Rep1(family_hit_line)) +
domain_header +
(no_hit_line | Rep1(domain_hit_line)) +
alignment_header +
(no_hit_line | Rep1(domain_alignment)) +
record_end
))
format = HeaderFooter("hmmpfam", {},
header, RecordReader.CountLines, (8,),
record, RecordReader.EndsWith, ("//\n",),
None, None, None)
# //
record_end = Martel.Group("record_end",
Martel.Str("//") +
Martel.Rep1(Martel.AnyEol()))
record = Std.record(Martel.Group("genbank_record",
locus_line + \
definition_block + \
accession_block + \
Martel.Opt(nid_line) + \
Martel.Opt(pid_line) + \
Martel.Opt(version_line) + \
Martel.Opt(db_source_block) + \
keywords_block + \
Martel.Opt(segment_line) + \
source_block + \
organism_block + \
Martel.Rep(reference) + \
Martel.Opt(primary) +\
Martel.Opt(comment_block) + \
features_line + \
feature_block + \
Martel.Alt(Martel.Opt(base_count_line) +
sequence_entry,
contig_block) + \
record_end))
# if you download a big mess of GenBank files, it'll have a header
# in that case you should be using 'ncbi_format' instead of the standard
# 'format'
header = Martel.Re("""\
#--- //
end = Martel.Group("END", Martel.Str("//") + Martel.AnyEol())
####################### put it all together
record = Std.record(
ID +
AC_block +
DT_created +
DT_seq_update +
DT_ann_update +
Martel.Opt(DE_block) +
Martel.Opt(GN_block) +
Martel.Opt(OS_block) +
Martel.Opt(OG_block) +
Martel.Opt(OC_block) +
Martel.Group("OX_block", Martel.NullOp()) +
Martel.Group("reference_block", Martel.Rep(reference)) +
comment +
Martel.Opt(DR_block) +
Martel.Opt(KW_block) +
Martel.Opt(feature_block) +
sequence +
end,
{"format": "swissprot/38"})
format_expression = Martel.Group("dataset", Martel.Rep1(record),
{"format": "swissprot/38"})
format = Martel.ParseRecords("dataset", {"format": "swissprot/38"},
# "|" them all together
ncbi_word = Std.dbxref(reduce(operator.or_, ids))
#ncbi_term = Assert(Re("[^ \R]+\|")) + \
ncbi_term = ncbi_word + Rep(Str("|") + ncbi_word)
# Anything else
generic_term = Std.dbxref(
Std.dbxref_dbid(UntilSep(sep = " "), {"dbname": "local"})
)
id_term = ncbi_term | generic_term
###########################################################
comment_lines = Rep(Str("#") + ToEol())
title = Str(">") + Std.description_line(id_term + UntilEol()) + AnyEol()
seqline = AssertNot(Str(">")) + Std.sequence(UntilEol()) + AnyEol()
# can get a sequence line without an Eol at the end of a file
seqline_nonewline = AssertNot(Str(">")) + Std.sequence(Word())
sequence = Std.sequence_block(Rep(seqline | seqline_nonewline))
record = Std.record(comment_lines + title + sequence + Rep(AnyEol()))
# define a format which reads records, but allows #-style comments in
# the FASTA file
format = HeaderFooter("dataset", {"format": "fasta"},
comment_lines, RecordReader.Until, (">",),
record, RecordReader.StartsWith, (">",),
comment_lines, RecordReader.Everything, ()
)
