From 6dacc267d1f44f73b15e4a5de09c7bd4e5a8fd0c Mon Sep 17 00:00:00 2001
From: Panciera <michael.panciera@amedrhcwrair001.amed.ds.army.mil>
Date: Thu, 18 Aug 2016 15:45:03 -0400
Subject: [PATCH 1/2] initial commit of genbank script

---
 bio_bits/genbank.py  | 45 ++++++++++++++++++++++++++++++++++++++++++++
 requirements-pip.txt |  1 +
 setup.py             |  1 +
 3 files changed, 47 insertions(+)
 create mode 100644 bio_bits/genbank.py

diff --git a/bio_bits/genbank.py b/bio_bits/genbank.py
new file mode 100644
index 0000000..56f826b
--- /dev/null
+++ b/bio_bits/genbank.py
@@ -0,0 +1,45 @@
+'''
+Usage:
+  genbank <infile>
+'''
+from Bio import SeqIO
+import dateparser
+import sys
+from docopt import docopt
+
+def rec_qualifier(rec, key):
+  for f in rec.features:
+    if f.qualifiers:
+      if 'collection_date' in f.qualifiers:
+        return f.qualifiers[key]
+
+def fix_date(d):
+    date = dateparser.parse(d)
+    return date.strftime('%Y-%m-%d')
+
+def rec_to_fasta(rec):
+  raw_collection_date = rec_qualifier(rec, 'collection_date')[0]
+  country = rec_qualifier(rec, 'country')[0]
+  date = dateparser.parse(raw_collection_date)
+  date = date.strftime('%Y-%m-%d')
+  rec.description = "|{}|{}".format(country, date)
+  return rec
+
+  # acc = rec.name
+  #header = "{}|{}|{}".format(acc, country, date)
+  #seq = str(rec.seq)
+  #return rec.format("fasta")
+def run(infile):
+  with open(infile) as input:
+    gb = SeqIO.parse(input, 'genbank')
+    results = map(rec_to_fasta, gb)
+    SeqIO.write(results, sys.stdout, 'fasta')
+
+def main():
+  raw_args = docopt(__doc__, version='Version 1.0')
+  run(raw_args['<infile>'])
+  sys.exit(0)
+
+
+# ['17-Feb-2010']
+# rec = degen.id_to_record("KJ627355")
diff --git a/requirements-pip.txt b/requirements-pip.txt
index 21186bf..8fea4e1 100644
--- a/requirements-pip.txt
+++ b/requirements-pip.txt
@@ -1,5 +1,6 @@
 argparse
 docopt
+dateparser
 future
 schema
 sh
diff --git a/setup.py b/setup.py
index 46e5069..109f230 100644
--- a/setup.py
+++ b/setup.py
@@ -27,6 +27,7 @@
             'degen = bio_bits.degen:main',
             'plot_muts = bio_bits.plot_muts:main',
             'fasta = bio_bits.fasta:main',
+            'genbank = bio_bits.genbank:main',
             #'sequence_concat = bio_bits.sequence_concat:main',
             #'sequence_files_concat = bio_bits.sequence_files_concat:main',
             #'sequence_split = bio_bits_old.sequence_split:main',

From bd36057645925e7691b8d86139c3cd7c77a13c53 Mon Sep 17 00:00:00 2001
From: Mike Panciera <michael.panciera.work@gmail.com>
Date: Thu, 18 Aug 2016 16:30:29 -0400
Subject: [PATCH 2/2] Update CHANGELOG.rst

---
 CHANGELOG.rst | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/CHANGELOG.rst b/CHANGELOG.rst
index c3d9b6d..db97202 100644
--- a/CHANGELOG.rst
+++ b/CHANGELOG.rst
@@ -2,6 +2,11 @@
 CHANGELOG
 =========
 
+Version 1.4.0
+-------------
+
+* added genbank downloader script
+
 Version 1.3.2
 -------------