ex-185 (jebene) added log_file flag to allow Bridget to specify the log

file destination

jacquard/command_validator.py

@@ -13,7 +13,6 @@
 """
 from __future__ import absolute_import
 
-from collections import defaultdict
 import errno
 import glob
 import os

jacquard/expand.py

@@ -108,6 +108,7 @@ def add_subparser(subparser):
     parser.add_argument("-v", "--verbose", action='store_true')
     parser.add_argument("-c", "--column_specification", help="Path to text file containing column regular expressions to be included in output file")
     parser.add_argument("--force", action='store_true', help="Overwrite contents of output directory")
+    parser.add_argument("--log_file", help="Log file destination")
 
 def _predict_output(args):
     return set([os.path.basename(args.output)])

jacquard/filter_hc_somatic.py

@@ -255,6 +255,7 @@ def add_subparser(subparser):
     parser.add_argument("output", help="Path to output directory. Will create if doesn't exist and will overwrite files in output directory as necessary")
     parser.add_argument("-v", "--verbose", action='store_true')
     parser.add_argument("--force", action='store_true', help="Overwrite contents of output directory")
+    parser.add_argument("--log_file", help="Log file destination")
 
 def _validate_arguments(args):
     input_dir = os.path.abspath(args.input)

jacquard/jacquard.py

@@ -163,7 +163,7 @@ def _dispatch(modules, arguments):
         command, args = _parse_command_line_args(modules, arguments)
         execution_context = _get_execution_context(command)
 
-        logger.initialize_logger(args.subparser_name)
+        logger.initialize_logger(args)
         logger.debug("Jacquard run begins")
         logger.debug("cwd|{}", os.getcwd())
         logger.debug("command|{}", " ".join(arguments))

jacquard/logger.py

@@ -6,13 +6,18 @@
 """
 #pylint: disable=invalid-name, global-statement
 from __future__ import print_function, absolute_import
+
 from datetime import datetime
+import errno
 import getpass
 import logging
 import os
 import socket
 import sys
 
+import jacquard.utils as utils
+
+
 WARNING_OCCURRED = False
 """Used to vary the Done message to emphasize upstream log warnings"""
 log_filename = None
@@ -25,11 +30,37 @@
 _logging_dict = {}
 _verbose = False
 
+def _makepath(path):
+    try:
+        os.makedirs(path)
+    except OSError as exc:
+        if exc.errno == errno.EEXIST and os.path.isdir(path):
+            pass
+        else: raise
+
+def _validate_log_file(log_file):
+    try:
+        _makepath(os.path.dirname(log_file))
+    except OSError:
+        raise utils.UsageError(("Jacquard cannot create specified log file "
+                                "[{}]. Review inputs and try again."), log_file)
+
+    try:
+        log = open(log_file, "w")
+        log.close()
+    except IOError:
+        raise utils.UsageError(("Jacquard cannot create specified log file "
+                               "[{}]. Review inputs and try again."), log_file)
 
-def initialize_logger(command, verbose=False):
+def initialize_logger(args, verbose=False):
     """Sets command name and formatting for subsequent calls to logger"""
+
     global log_filename
     log_filename = os.path.join(os.getcwd(), "jacquard.log")
+    if args.log_file:
+        _validate_log_file(args.log_file)
+        log_filename = args.log_file
+
     logging.basicConfig(format=_FILE_LOG_FORMAT,
                         level="DEBUG",
                         datefmt=_DATE_FORMAT,
@@ -43,7 +74,7 @@ def initialize_logger(command, verbose=False):
     _logging_dict = {'user': getpass.getuser(),
                      'host': socket.gethostname(),
                      'start_time': start_time,
-                     'tool': command}
+                     'tool': args.subparser_name}
 
 def error(message, *args):
     _print("ERROR", message, args)

jacquard/merge.py

@@ -37,7 +37,6 @@
 import jacquard.utils as utils
 from jacquard.vcf import FileWriter
 import jacquard.vcf as vcf
-import jacquard.variant_callers.variant_caller_factory as variant_caller_factory
 
 
 _DEFAULT_INCLUDED_FORMAT_TAGS = ["JQ_.*"]
@@ -411,6 +410,7 @@ def add_subparser(subparser):
     parser.add_argument("-v", "--verbose", action='store_true')
     parser.add_argument("--force", action='store_true', help="Overwrite contents of output directory")
     parser.add_argument("--include_format_tags", dest='tags', help="Comma-separated list of regexs for format tags to include in output. Defaults to all JQ tags.")
+    parser.add_argument("--log_file", help="Log file destination")
 
 def _predict_output(args):
     desired_output_files = set([os.path.basename(args.output)])

jacquard/summarize.py

@@ -75,6 +75,7 @@ def add_subparser(subparser):
     parser.add_argument("output", help="Path to output VCf")
     parser.add_argument("-v", "--verbose", action='store_true')
     parser.add_argument("--force", action='store_true', help="Overwrite contents of output directory")
+    parser.add_argument("--log_file", help="Log file destination")
 
 def report_prediction(args):
     return set([os.path.basename(args.output)])

jacquard/translate.py

@@ -234,6 +234,7 @@ def add_subparser(subparser):
     parser.add_argument("--force", action='store_true', help="Overwrite contents of output directory")
     parser.add_argument("--varscan_hc_filter_filename", help="Regex pattern that identifies optional VarScan high-confidence filter files. The VCF, high-confidence file pairs should share the same prefix. For example, given patientA.snp.vcf, patientA.indel.vcf, patientA.snp.fpfilter.pass, patientA.indel.fpfilter.pass, you could enable this option as varscan_hc_filter_filename='.fpfilter.pass$'")
     parser.add_argument("--allow_inconsistent_sample_sets", help="Allow inconsistent sample sets to be used")
+    parser.add_argument("--log_file", help="Path to log file destination. Defaults to current working directory if not specified.")
 
 #TODO (cgates): This module is both a command and also manipulates VcfRecords
 # like a caller. This is the only body of code that does both these things.

test/command_validator_test.py

@@ -393,7 +393,7 @@ def test_create_temp_working_dir_createsIfNotExists(self):
             command_validator._create_temp_working_dir(None, args)
             self.assertTrue(os.path.isdir(tmp_dir))
 
-    ## cgates: This test is a good idea, but doesn't work reliably on Windows 7.
+    ##TODO: cgates: This test is a good idea, but doesn't work reliably on Windows 7.
     ## Can we remove it with a clean conscience? What's a better way?
     def xtest_create_temp_working_dir_doesNotExistCannotCreateTmp_directoryGivenAsOutput(self):
         with TempDirectory() as output_dir:

test/logger_test.py

@@ -1,9 +1,12 @@
 #pylint: disable=line-too-long, invalid-name, too-many-public-methods
+from StringIO import StringIO
+from argparse import Namespace
 from datetime import datetime
 import os
+import shutil
 import sys
-from StringIO import StringIO
 import unittest
+
 import jacquard.logger as logger
 
 
@@ -20,15 +23,41 @@ def tearDown(self):
         except OSError:
             pass
 
-    def test_initialize_logger(self):
+    def test_initialize_logger_defaultFilename(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
         self.assertEquals(['host', 'tool', 'start_time', 'user'],
                           logger._logging_dict.keys())
+        self.assertEquals("jacquard.log", os.path.basename(logger.log_filename))
+
+    def test_initialize_logger_suppliedFilename(self):
+        tool = "foo"
+        log_filename = "tmp/log.foo"
+        args = Namespace(subparser_name=tool,
+                         log_file=log_filename)
+        logger.initialize_logger(args)
+        self.assertEquals(['host', 'tool', 'start_time', 'user'],
+                          logger._logging_dict.keys())
+        self.assertEquals(log_filename, logger.log_filename)
+
+    def test_initialize_logger_createsParentDirs(self):
+        try:
+            tool = "foo"
+            log_filename = "tmp/log.foo"
+            args = Namespace(subparser_name=tool,
+                             log_file=log_filename)
+            logger.initialize_logger(args)
+            self.assertTrue(os.path.isdir(os.path.dirname(log_filename)))
+        finally:
+            shutil.rmtree(os.path.dirname(log_filename))
 
     def test_error(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
         logger.error("bar")
 
         root_logger = logger.logging.getLogger()
@@ -43,7 +72,9 @@ def test_error(self):
 
     def test_warning(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
         self.assertFalse(logger.WARNING_OCCURRED)
         logger.warning("bar")
         self.assertTrue(logger.WARNING_OCCURRED)
@@ -57,7 +88,9 @@ def test_warning(self):
 
     def test_info(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
         current_time = datetime.now()
 
         logger.info("bar")
@@ -78,7 +111,9 @@ def test_info(self):
 
     def test_info_message_args(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
         current_time = datetime.now()
 
         logger.info("bar {}/{}", "1", "2")
@@ -99,7 +134,9 @@ def test_info_message_args(self):
 
     def test_debug(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
 
         logger.debug("bar")
 
@@ -110,7 +147,9 @@ def test_debug(self):
 
     def test_debug_consoleWhenVerbose(self):
         tool = "foo"
-        logger.initialize_logger(tool, True)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args, True)
         logger.debug("bar")
         root_logger = logger.logging.getLogger()
 
@@ -123,7 +162,9 @@ def test_debug_consoleWhenVerbose(self):
 
     def test_debug_noConsoleWhenNotVerbose(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
 
         logger.debug("bar")
 
@@ -134,7 +175,9 @@ def test_debug_noConsoleWhenNotVerbose(self):
 
     def test_noExceptionOnMalformedMessage(self):
         tool = "foo"
-        logger.initialize_logger(tool)
+        args = Namespace(subparser_name=tool,
+                         log_file=None)
+        logger.initialize_logger(args)
 
         logger.info("bar {}/{}/{}", "1", "2")
 

test/mock_module.py

@@ -23,6 +23,7 @@ def add_subparser(subparser):
     parser.add_argument("input", help="foo")
     parser.add_argument("output", help="foo")
     parser.add_argument("--force", action='store_true', help="foo")
+    parser.add_argument("--log_file", help="foo")
 
 def execute(args, execution_context):
     global execute_called