get encoding from `locale.getdefaultlocale()`

Summary:This enables controlling the encoding chosen by infer via the usual environment
variables. For instance:

```
LC_ALL="C" infer ... # sets LOCALE to "ascii"
LC_ALL="en_US.UTF-8" infer ... # sets LOCALE to "UTF-8"
```

This gives an easy solution to #320: run `LC_ALL="en_US.UTF-8" infer ...`.
Right now the only solution is to edit the Python scripts by hand instead!

Reviewed By: jberdine

Differential Revision: D3207573

fb-gh-sync-id: 62d5b98
fbshipit-source-id: 62d5b98

infer/lib/python/infer

@@ -8,6 +8,7 @@ from __future__ import unicode_literals
 import argparse
 import imp
 import json
+import locale
 import logging
 import os
 import platform
@@ -138,6 +139,10 @@ def main():
         logging.info('PATH=%s', os.getenv('PATH'))
         logging.info('SHELL=%s', os.getenv('SHELL'))
         logging.info('PWD=%s', os.getenv('PWD'))
+        logging.info('output of locale.getdefaultlocale(): %s',
+                     str(locale.getdefaultlocale()))
+        logging.info('encoding we chose in the end: %s',
+                     config.CODESET)
 
         capture_exitcode = imported_module.gen_instance(args, cmd).capture()
         if capture_exitcode != os.EX_OK:

infer/lib/python/inferTraceBugs

@@ -345,7 +345,7 @@ def generate_html_report(args, reports):
     for bug in sel:
         bug_trace_path = path_of_bug_number(traces_dir, i)
         with codecs.open(bug_trace_path, 'w',
-                         encoding=config.LOCALE,
+                         encoding=config.CODESET,
                          errors='xmlcharrefreplace') as bug_trace_file:
             bug_trace_file.write(html_bug_trace(args, bug, i))
         i += 1
@@ -353,7 +353,7 @@ def generate_html_report(args, reports):
     remote_source_template = get_remote_source_template()
     bug_list_path = os.path.join(html_dir, 'index.html')
     with codecs.open(bug_list_path, 'w',
-                     encoding=config.LOCALE,
+                     encoding=config.CODESET,
                      errors='xmlcharrefreplace') as bug_list_file:
         bug_list_file.write(html_list_of_bugs(args,
                                               remote_source_template,

infer/lib/python/inferlib/config.py

@@ -14,13 +14,15 @@ import locale
 import os
 
 
-LOCALE = locale.getpreferredencoding()
-
+locale.setlocale(locale.LC_ALL, '')
+CODESET = locale.getlocale(locale.LC_CTYPE)[1]
+if CODESET is None:
+    CODESET = 'ascii'
 
 # this assumes that this file lives in infer/lib/python/infer/ and the binaries
 # are in infer/bin/
 INFER_PYTHON_DIRECTORY = os.path.dirname(os.path.realpath(__file__)
-                                         .decode(LOCALE))
+                                         .decode(CODESET))
 INFER_INFER_DIRECTORY = os.path.join(INFER_PYTHON_DIRECTORY,
                                      os.pardir, os.pardir, os.pardir)
 INFER_ROOT_DIRECTORY = os.path.join(INFER_INFER_DIRECTORY, os.pardir)
@@ -34,7 +36,7 @@ ANNOT_PROCESSOR_NAMES = 'com.facebook.infer.annotprocess.CollectSuppressWarnings
 WRAPPERS_DIRECTORY = os.path.join(LIB_DIRECTORY, 'wrappers')
 XCODE_WRAPPERS_DIRECTORY = os.path.join(LIB_DIRECTORY, 'xcode_wrappers')
 
-DEFAULT_INFER_OUT = os.path.join(os.getcwd().decode(LOCALE), 'infer-out')
+DEFAULT_INFER_OUT = os.path.join(os.getcwd().decode(CODESET), 'infer-out')
 CSV_PERF_FILENAME = 'performances.csv'
 STATS_FILENAME = 'stats.json'
 PROC_STATS_FILENAME = 'proc_stats.json'

infer/lib/python/inferlib/issues.py

@@ -269,11 +269,12 @@ def print_and_save_errors(json_report, bugs_out, xml_out):
     plain_out = _text_of_report_list(errors,
                                      formatter=colorize.PLAIN_FORMATTER)
     with codecs.open(bugs_out, 'w',
-                     encoding=config.LOCALE, errors='replace') as file_out:
+                     encoding=config.CODESET, errors='replace') as file_out:
         file_out.write(plain_out)
     if xml_out is not None:
         with codecs.open(xml_out, 'w',
-                         encoding=config.LOCALE, errors='replace') as file_out:
+                         encoding=config.CODESET,
+                         errors='replace') as file_out:
             file_out.write(_pmd_xml_of_issues(errors))
 
 
@@ -323,7 +324,7 @@ def _pmd_xml_of_issues(issues):
         violation.text = issue[JSON_INDEX_QUALIFIER]
         file_node.append(violation)
         root.append(file_node)
-    return etree.tostring(root, pretty_print=True, encoding=config.LOCALE)
+    return etree.tostring(root, pretty_print=True, encoding=config.CODESET)
 
 
 def _sort_and_uniq_rows(l):

infer/lib/python/inferlib/source.py

@@ -44,7 +44,7 @@ class Indenter(str):
 
     def add(self, x):
         if type(x) != unicode:
-            x = x.decode(config.LOCALE)
+            x = x.decode(config.CODESET)
         lines = x.splitlines()
         indent = self.indent_get()
         lines = [indent + l for l in lines]
@@ -55,7 +55,7 @@ class Indenter(str):
         return self.text
 
     def __str__(self):
-        return unicode(self).encode(config.LOCALE)
+        return unicode(self).encode(config.CODESET)
 
 
 def build_source_context(source_name, mode, report_line):
@@ -67,7 +67,7 @@ def build_source_context(source_name, mode, report_line):
     line_number = 1
     excerpt = ''
     with codecs.open(source_name, 'r',
-                     encoding=config.LOCALE, errors="replace") as source_file:
+                     encoding=config.CODESET, errors="replace") as source_file:
         # avoid going past the end of the file
         for line in source_file:
             if start_line <= line_number <= end_line:

infer/lib/python/inferlib/utils.py

@@ -55,7 +55,7 @@ if 'check_output' not in dir(subprocess):
 def locale_csv_reader(iterable, dialect='excel', **kwargs):
     rows = csv.reader(iterable, dialect=dialect, **kwargs)
     for row in rows:
-        yield [unicode(cell, config.LOCALE) for cell in row]
+        yield [unicode(cell, config.CODESET) for cell in row]
 
 
 def configure_logging(args):
@@ -89,8 +89,8 @@ def get_cmd_in_bin_dir(binary_name):
 
 def load_json_from_path(path, errors='replace'):
     with codecs.open(path, 'r',
-                     encoding=config.LOCALE, errors=errors) as file_in:
-        return json.load(file_in, encoding=config.LOCALE)
+                     encoding=config.CODESET, errors=errors) as file_in:
+        return json.load(file_in, encoding=config.CODESET)
 
 
 def dump_json_to_path(
@@ -99,10 +99,10 @@ def dump_json_to_path(
         cls=None,
         indent=2,  # customized
         separators=None,
-        encoding=config.LOCALE,  # customized
+        encoding=config.CODESET,  # customized
         default=None, sort_keys=False, **kw):
     with codecs.open(path, 'w',
-                     encoding=config.LOCALE, errors='replace') as file_out:
+                     encoding=config.CODESET, errors='replace') as file_out:
         json.dump(data, file_out, skipkeys=skipkeys, ensure_ascii=ensure_ascii,
                   check_circular=check_circular, allow_nan=allow_nan, cls=cls,
                   indent=indent, separators=separators, encoding=encoding,
@@ -288,11 +288,11 @@ def get_plural(_str, count):
 
 
 def decode(s, errors='replace'):
-    return s.decode(encoding=config.LOCALE, errors=errors)
+    return s.decode(encoding=config.CODESET, errors=errors)
 
 
 def encode(u, errors='replace'):
-    return u.encode(encoding=config.LOCALE, errors=errors)
+    return u.encode(encoding=config.CODESET, errors=errors)
 
 
 def stdout(s, errors='replace'):