Merge pull request #68 from reinout/reinout-distutils-encoding-fix

Reinout distutils script encoding fix for #65

src/zc/buildout/easy_install.py

@@ -46,6 +46,8 @@ default_index_url = os.environ.get(
 logger = logging.getLogger('zc.buildout.easy_install')
 
 url_match = re.compile('[a-z0-9+.-]+://').match
+is_source_encoding_line = re.compile('coding[:=]\s*([-\w.]+)').search
+# Source encoding regex from http://www.python.org/dev/peps/pep-0263/
 
 is_win32 = sys.platform == 'win32'
 is_jython = sys.platform.startswith('java')
@@ -1046,12 +1048,18 @@ def _distutils_script(path, dest, script_content, initialization, rsetup):
     if not ('#!' in lines[0]) and ('python' in lines[0]):
         # The script doesn't follow distutil's rules.  Ignore it.
         return []
+    source_encoding_line = ''
     original_content = ''.join(lines[1:])
+    if (len(lines) > 1) and is_source_encoding_line(lines[1]):
+        # The second line contains a source encoding line. Copy it verbatim.
+        source_encoding_line = lines[1].rstrip()
+        original_content = ''.join(lines[2:])
 
     python = _safe_arg(sys.executable)
 
     contents = distutils_script_template % dict(
         python = python,
+        source_encoding_line = source_encoding_line,
         path = path,
         initialization = initialization,
         relative_paths_setup = rsetup,
@@ -1118,6 +1126,7 @@ if __name__ == '__main__':
 
 distutils_script_template = script_header + '''\
 
+%(source_encoding_line)s
 %(relative_paths_setup)s
 import sys
 sys.path[0:0] = [

src/zc/buildout/easy_install.txt

@@ -938,11 +938,13 @@ It also works for zipped eggs:
 
 Distutils copies the script files verbatim, apart from a line at the top that
 looks like ``#!/usr/bin/python``, which gets replaced by the actual python
-interpreter.  Buildout does the same, but additionally also adds the sys.path
-like for the console_scripts:
+interpreter. Buildout does the same, but additionally also adds the sys.path
+like for the console_scripts. Note that the second line in the scripts can
+contain a source encoding hint; buildout retains it.
 
     >>> cat(distbin, 'distutilsscript')
     #!/usr/local/bin/python2.7
+    # -*- coding: utf-8 -*-
     <BLANKLINE>
     import sys
     sys.path[0:0] = [

src/zc/buildout/tests.py

@@ -2899,6 +2899,7 @@ def create_sample_eggs(test, executable=sys.executable):
         write(
             tmp, 'distutilsscript',
             '#!/usr/bin/python\n'
+            '# -*- coding: utf-8 -*-\n'
             'import sys; sys.stdout.write("distutils!\\n")\n'
             )
         write(