Mercurial > repos > earlhaminst > ete

annotate ete_init_taxdb.py @ 11:2db72467da51 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 67e136d433c0d925db362342919305b46fdffecd"
author earlhaminst
date Thu, 10 Mar 2022 14:01:44 +0000
parents 03c10736e497
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
2
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
1 import optparse
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
2
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
3 import ete3.ncbi_taxonomy
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
4 from six.moves.urllib.request import urlretrieve
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
5
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
6 parser = optparse.OptionParser()
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
7 parser.add_option('-t', '--taxdump', dest='taxdump', default=None,
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
8 help='NCBI taxdump (tar.gz) will be downloaded if not given')
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
9 parser.add_option('-d', '--database', dest="database", default=None,
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
10 help='ETE sqlite data base to use (default: ~/.etetoolkit/taxa.sqlite)')
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
11 options, args = parser.parse_args()
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
12 if options.database is None:
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
13 parser.error("-d option must be specified")
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
14 if options.taxdump is not None:
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
15 taxdump = options.taxdump
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
16 else:
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
17 urlretrieve("http://ftp.ncbi.nih.gov/pub/taxonomy/taxdump.tar.gz", "taxdump.tar.gz")
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
18 taxdump = "taxdump.tar.gz"
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
19
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
20 # will remove a taxdump.tar.gz file at the end
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
21 # which will lead to an errmessage if not present
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
22 # if the tool is run on a taxdump in the current dir it will be
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
23 # deleted in the end
03c10736e497 planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
earlhaminst
parents:
diff changeset
24 ete3.ncbi_taxonomy.ncbiquery.update_db(dbfile=options.database, targz_file=taxdump)