Switch to using preferred encoding from locale module instead of system default for creating byte strings.

Author: dan-blanchard

drmaa/const.py

@@ -25,6 +25,11 @@
 
 from __future__ import absolute_import, print_function, unicode_literals
 
+import locale
+
+# Encoding to use for passing strings to C library
+ENCODING = locale.getpreferredencoding()
+
 # drmaa_get_attribute()
 ATTR_BUFFER = 1024
 

drmaa/helpers.py

@@ -31,7 +31,7 @@
 from ctypes import (byref, c_uint, create_string_buffer, POINTER, pointer,
                     sizeof)
 
-from drmaa.const import ATTR_BUFFER, NO_MORE_ELEMENTS
+from drmaa.const import ATTR_BUFFER, ENCODING, NO_MORE_ELEMENTS
 from drmaa.errors import error_buffer
 from drmaa.wrappers import (drmaa_attr_names_t, drmaa_attr_values_t,
                             drmaa_get_attribute, drmaa_get_attribute_names,
@@ -61,10 +61,10 @@ class BoolConverter(object):
 
     def __init__(self, true=b'y', false=b'n'):
         if isinstance(true, str):
-            true = true.encode()
+            true = true.encode(ENCODING)
         self.true = true
         if isinstance(false, str):
-            false = false.encode()
+            false = false.encode(ENCODING)
         self.false = false
 
     def to_drmaa(self, value):
@@ -104,7 +104,8 @@ def __get__(self, *args):
 
 Version = namedtuple("Version", "major minor")
 if sys.version_info < (3, 0):
-    Version.__str__ = lambda x: "{0}.{1}".format(x.major, x.minor).encode()
+    Version.__str__ = lambda x: "{0}.{1}".format(x.major,
+                                                 x.minor).encode(ENCODING)
 else:
     Version.__str__ = lambda x: "{0}.{1}".format(x.major, x.minor)
 
@@ -139,15 +140,15 @@ def __init__(self, name, type_converter=None):
            implementation. See BoolConverter for an example.
         """
         if isinstance(name, str):
-            name = name.encode()
+            name = name.encode(ENCODING)
         self.name = name
         self.converter = type_converter
 
     def __set__(self, instance, value):
         if self.converter:
             v = self.converter.to_drmaa(value)
         elif isinstance(value, str):
-            v = value.encode()
+            v = value.encode(ENCODING)
         else:
             v = value
         c(drmaa_set_attribute, instance, self.name, v)
@@ -174,7 +175,7 @@ class VectorAttribute(object):
 
     def __init__(self, name):
         if isinstance(name, str):
-            name = name.encode()
+            name = name.encode(ENCODING)
         self.name = name
 
     def __set__(self, instance, value):
@@ -195,11 +196,12 @@ class DictAttribute(object):
 
     def __init__(self, name):
         if isinstance(name, str):
-            name = name.encode()
+            name = name.encode(ENCODING)
         self.name = name
 
     def __set__(self, instance, value):
-        v = ["{0}={1}".format(k, v).encode() for (k, v) in value.items()]
+        v = ["{0}={1}".format(k, v).encode(ENCODING) for (k, v) in
+             value.items()]
         c(drmaa_set_vector_attribute, instance, self.name,
           string_vector(v))
 
@@ -296,7 +298,7 @@ def string_vector(v):
     vlen = len(v)
     values = (STRING * (vlen + 1))()
     for i, el in enumerate(v):
-        values[i] = STRING(el.encode() if isinstance(el, str) else el)
+        values[i] = STRING(el.encode(ENCODING) if isinstance(el, str) else el)
     values[vlen] = STRING()
     return values
 

drmaa/session.py

@@ -27,14 +27,15 @@
 from ctypes import byref, c_int, create_string_buffer, pointer, POINTER, sizeof
 
 from drmaa.const import (BLOCK_EMAIL, DEADLINE_TIME, DURATION_HLIMIT,
-                         DURATION_SLIMIT, ERROR_PATH, INPUT_PATH, JOB_CATEGORY,
-                         JOB_IDS_SESSION_ALL, JOB_IDS_SESSION_ANY, JOB_NAME,
-                         JobState, JobControlAction, JobSubmissionState,
-                         JOIN_FILES, JS_STATE, NATIVE_SPECIFICATION,
-                         OUTPUT_PATH, REMOTE_COMMAND, SIGNAL_BUFFER, START_TIME,
-                         status_to_string, string_to_control_action,
-                         TIMEOUT_NO_WAIT, TIMEOUT_WAIT_FOREVER, TRANSFER_FILES,
-                         V_ARGV, V_EMAIL, V_ENV, WCT_HLIMIT, WCT_SLIMIT, WD)
+                         DURATION_SLIMIT, ENCODING, ERROR_PATH, INPUT_PATH,
+                         JOB_CATEGORY, JOB_IDS_SESSION_ALL, JOB_IDS_SESSION_ANY,
+                         JOB_NAME, JobState, JobControlAction,
+                         JobSubmissionState, JOIN_FILES, JS_STATE,
+                         NATIVE_SPECIFICATION, OUTPUT_PATH, REMOTE_COMMAND,
+                         SIGNAL_BUFFER, START_TIME, status_to_string,
+                         string_to_control_action, TIMEOUT_NO_WAIT,
+                         TIMEOUT_WAIT_FOREVER, TRANSFER_FILES, V_ARGV, V_EMAIL,
+                         V_ENV, WCT_HLIMIT, WCT_SLIMIT, WD)
 from drmaa.helpers import (adapt_rusage, Attribute, attribute_names_iterator,
                            BoolConverter, c, DictAttribute, IntConverter,
                            run_bulk_job, SessionStringAttribute,
@@ -376,7 +377,7 @@ def control(jobId, operation):
         or jobs submitted via native utilities.
         """
         if isinstance(jobId, str):
-            jobId = jobId.encode()
+            jobId = jobId.encode(ENCODING)
         c(drmaa_control, jobId, string_to_control_action(operation))
 
     # takes string list, num value and boolean, no return value
@@ -404,9 +405,9 @@ def synchronize(jobIds, timeout=-1, dispose=False):
             data record at the end of the synchronize() call. If set to
             False, the data record will be left for future access via the
             wait() method. It is the responsibility of the application to
-            make sure that either `synchronize()` or `wait()`is called for 
+            make sure that either `synchronize()` or `wait()`is called for
             every job. Not doing so creates a memory leak. Note that calling
-            synchronize() with dispose set to true flushes all accounting 
+            synchronize() with dispose set to true flushes all accounting
             information for all jobs in the list.
 
         To avoid thread race conditions in multithreaded applications, the
@@ -465,7 +466,7 @@ def wait(jobId, timeout=-1):
         jid_out = create_string_buffer(128)
         rusage = pointer(POINTER(drmaa_attr_values_t)())
         if isinstance(jobId, str):
-            jobId = jobId.encode()
+            jobId = jobId.encode(ENCODING)
         c(drmaa_wait, jobId, jid_out, sizeof(jid_out), byref(stat), timeout,
           rusage)
         res_usage = adapt_rusage(rusage)
@@ -513,7 +514,7 @@ def jobStatus(jobId):
         """
         status = c_int()
         if isinstance(jobId, str):
-            jobId = jobId.encode()
+            jobId = jobId.encode(ENCODING)
         c(drmaa_job_ps, jobId, byref(status))
         return status_to_string(status.value)
 

drmaa/wrappers.py

@@ -29,6 +29,7 @@
                     POINTER, RTLD_GLOBAL, sizeof, Structure)
 from ctypes.util import find_library
 
+from drmaa.const import ENCODING
 from drmaa.errors import error_check, error_buffer
 
 
@@ -68,7 +69,7 @@
 
 def py_drmaa_init(contact=None):
     if isinstance(contact, str):
-        contact = contact.encode()
+        contact = contact.encode(ENCODING)
     return _lib.drmaa_init(contact, error_buffer, sizeof(error_buffer))
 
 _lib.drmaa_exit.argtypes = [c_char_p, c_size_t]