Merge pull request #41 from pbashyal-nmdp/master

Specify a path for temp files

Author: mmaiers-nmdp

README.rst

@@ -7,7 +7,7 @@ py-ard
         :target: https://pypi.python.org/pypi/py-ard
 
 .. image:: https://readthedocs.org/projects/pyars/badge/?version=latest
-        :target: https://pyars.readthedocs.io/en/latest/?badge=latest
+        :target: https://pyard.readthedocs.io/en/latest/?badge=latest
         :alt: Documentation Status
 
 
@@ -22,45 +22,48 @@ Install from source
 
 .. code-block::
 
-	python3 -m venv venv
-	source venv/bin/activate
+    python3 -m venv venv
+    source venv/bin/activate
 
-	python setup.py install
+    python setup.py install
 
 Install from PyPi
 -----------------
 
 .. code-block::
 
-	pip install py-ard
+    pip install py-ard
 
 
 Example
 -------
 
 .. code-block:: python3
 
-	from pyard import ARD
+    from pyard import ARD
 
-	# Initialize ARD object
-	ard = ARD('3290')
+    # Initialize ARD object
+    ard = ARD('3290')
 
-	# Initialize with latest DB
-	ard = ARD()
+    # You can specify a data directory for temp files
+    # ard = ard = ARD('3290', data_dir='/tmp/py-ard')
 
-	allele = "A*01:01:01"
+    # Initialize with latest DB
+    ard = ARD()
 
-	ard.redux(allele, 'G')
-	# >> 'A*01:01:01G'
+    allele = "A*01:01:01"
 
-	ard.redux(allele, 'lg')
-	# >> 'A*01:01g'
+    ard.redux(allele, 'G')
+    # >> 'A*01:01:01G'
 
-	ard.redux(allele, 'lgx')
-	# 'A*01:01'
+    ard.redux(allele, 'lg')
+    # >> 'A*01:01g'
 
-	ard_gl = ard.redux_gl("A*01:01/A*01:01N+A*02:AB^B*07:02+B*07:AB", "G")
-	# >>> ard_gl
-	# 'B*07:02:01G+B*07:02:01G^A*01:01:01G+A*02:01:01G/A*02:02'
+    ard.redux(allele, 'lgx')
+    # 'A*01:01'
+
+    ard_gl = ard.redux_gl("A*01:01/A*01:01N+A*02:AB^B*07:02+B*07:AB", "G")
+    # >>> ard_gl
+    # 'B*07:02:01G+B*07:02:01G^A*01:01:01G+A*02:01:01G/A*02:02'
 
 

pyard/pyard.py

@@ -35,6 +35,7 @@
 from functools import partial
 from typing import Dict
 import logging
+import pathlib
 
 ismac = lambda x: True if re.search(":\D+", x) else False
 
@@ -79,7 +80,8 @@ class ARD(object):
     def __init__(self, dbversion: str='Latest',
                  download_mac: bool=True,
                  verbose: bool=False,
-                 remove_invalid: bool=True):
+                 remove_invalid: bool=True,
+                 data_dir: str=None):
         """
         ARD -
         :param dbversion:
@@ -117,14 +119,20 @@ def __init__(self, dbversion: str='Latest',
 
         # List of expression characters
         expre_chars = ['N', 'Q', 'L', 'S']
-        data_dir = os.path.dirname(__file__)
-        ars_url = 'https://raw.githubusercontent.com/ANHIG/IMGTHLA/' \
-                  + dbversion + '/wmda/hla_nom_g.txt'
+
+        # Set data directory where all the downloaded files will go
+        if data_dir is None:
+            data_dir = os.path.dirname(__file__)
+        else:
+            pathlib.Path(data_dir).mkdir(exist_ok=True)
+
+        imgt_hla_url = 'https://raw.githubusercontent.com/ANHIG/IMGTHLA/'
+        ars_url = imgt_hla_url + dbversion + '/wmda/hla_nom_g.txt'
         ars_file = data_dir + '/hla_nom_g.' + str(dbversion) + ".txt"
         allele_file = data_dir + '/AlleleList.' + str(dbversion) + ".txt"
         mac_file = data_dir + "/mac.txt"
         mac_pickle = data_dir + "/mac.pickle"
-        broad_file = data_dir + "/dna_relshp.csv"
+        broad_file = os.path.dirname(__file__) + "/dna_relshp.csv"
         #print("mac_file:", mac_file)
 
         allele_url = "https://raw.githubusercontent.com/ANHIG/IMGTHLA/" \

pyard/smart_sort.py

@@ -1,7 +1,7 @@
 import functools
 import re
 
-expr_regex = re.compile('[NQLS]')
+expr_regex = re.compile('[NQLSGg]')
 
 @functools.lru_cache(maxsize=None)
 def smart_sort_comparator(a1, a2):