diff ete_init_taxdb.py @ 2:03c10736e497 draft

planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 91b634b8f9b131045bbbbf43cc8edbea59ac686b-dirty
author earlhaminst
date Tue, 07 Nov 2017 11:45:13 -0500
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/ete_init_taxdb.py	Tue Nov 07 11:45:13 2017 -0500
@@ -0,0 +1,24 @@
+import optparse
+
+import ete3.ncbi_taxonomy
+from six.moves.urllib.request import urlretrieve
+
+parser = optparse.OptionParser()
+parser.add_option('-t', '--taxdump', dest='taxdump', default=None,
+                  help='NCBI taxdump (tar.gz) will be downloaded if not given')
+parser.add_option('-d', '--database', dest="database", default=None,
+                  help='ETE sqlite data base to use (default: ~/.etetoolkit/taxa.sqlite)')
+options, args = parser.parse_args()
+if options.database is None:
+    parser.error("-d option must be specified")
+if options.taxdump is not None:
+    taxdump = options.taxdump
+else:
+    urlretrieve("http://ftp.ncbi.nih.gov/pub/taxonomy/taxdump.tar.gz", "taxdump.tar.gz")
+    taxdump = "taxdump.tar.gz"
+
+# will remove a taxdump.tar.gz file at the end
+# which will lead to an errmessage if not present
+# if the tool is run on a taxdump in the current dir it will be
+# deleted in the end
+ete3.ncbi_taxonomy.ncbiquery.update_db(dbfile=options.database, targz_file=taxdump)