Fix pastebin when captured output contains non-ascii characters

Fix pytest-dev#1219

Author: nicoddemus

CHANGELOG

@@ -9,6 +9,9 @@
 - fix #1198: ``--pastebin`` option now works on Python 3. Thanks
   Mehdy Khoshnoody for the PR.
 
+- fix #1219: ``--pastebin`` now works correctly when captured output contains
+  non-ascii characters. Thanks Bruno Oliveira for the PR.
+
 - fix #1204: another error when collecting with a nasty __getattr__().
   Thanks Florian Bruhin for the PR.
 

_pytest/pastebin.py

@@ -13,17 +13,21 @@ def pytest_addoption(parser):
 
 @pytest.hookimpl(trylast=True)
 def pytest_configure(config):
+    import py
     if config.option.pastebin == "all":
         tr = config.pluginmanager.getplugin('terminalreporter')
         # if no terminal reporter plugin is present, nothing we can do here;
         # this can happen when this function executes in a slave node
         # when using pytest-xdist, for example
         if tr is not None:
-            config._pastebinfile = tempfile.TemporaryFile('w+')
+            # pastebin file will be utf-8 encoded binary file
+            config._pastebinfile = tempfile.TemporaryFile('w+b')
             oldwrite = tr._tw.write
             def tee_write(s, **kwargs):
                 oldwrite(s, **kwargs)
-                config._pastebinfile.write(str(s))
+                if py.builtin._istext(s):
+                    s = s.encode('utf-8')
+                config._pastebinfile.write(s)
             tr._tw.write = tee_write
 
 def pytest_unconfigure(config):
@@ -45,7 +49,7 @@ def create_new_paste(contents):
     """
     Creates a new paste using bpaste.net service.
 
-    :contents: paste contents
+    :contents: paste contents as utf-8 encoded bytes
     :returns: url to the pasted contents
     """
     import re
@@ -61,8 +65,8 @@ def create_new_paste(contents):
         'expiry': '1week',
     }
     url = 'https://bpaste.net'
-    response = urlopen(url, data=urlencode(params).encode()).read()
-    m = re.search(r'href="/raw/(\w+)"', response.decode())
+    response = urlopen(url, data=urlencode(params).encode('ascii')).read()
+    m = re.search(r'href="/raw/(\w+)"', response.decode('utf-8'))
     if m:
         return '%s/show/%s' % (url, m.group(1))
     else:

testing/test_pastebin.py

@@ -1,3 +1,4 @@
+# encoding: utf-8
 import sys
 import pytest
 
@@ -27,6 +28,7 @@ def test_skip():
         assert reprec.countoutcomes() == [1,1,1]
 
     def test_all(self, testdir, pastebinlist):
+        from _pytest.pytester import LineMatcher
         testpath = testdir.makepyfile("""
             import pytest
             def test_pass():
@@ -39,9 +41,34 @@ def test_skip():
         reprec = testdir.inline_run(testpath, "--pastebin=all", '-v')
         assert reprec.countoutcomes() == [1,1,1]
         assert len(pastebinlist) == 1
-        s = pastebinlist[0]
-        for x in 'test_fail test_skip test_pass'.split():
-            assert x in s
+        contents = pastebinlist[0].decode('utf-8')
+        matcher = LineMatcher(contents.splitlines())
+        matcher.fnmatch_lines([
+            '*test_pass PASSED*',
+            '*test_fail FAILED*',
+            '*test_skip SKIPPED*',
+            '*== 1 failed, 1 passed, 1 skipped in *'
+        ])
+
+    def test_non_ascii_paste_text(self, testdir):
+        """Make sure that text which contains non-ascii characters is pasted
+        correctly. See #1219.
+        """
+        testdir.makepyfile(test_unicode="""
+            # encoding: utf-8
+            def test():
+                assert '☺' == 1
+        """)
+        result = testdir.runpytest('--pastebin=all')
+        if sys.version_info[0] == 3:
+            expected_msg = "*assert '☺' == 1*"
+        else:
+            expected_msg = "*assert '\\xe2\\x98\\xba' == 1*"
+        result.stdout.fnmatch_lines([
+            expected_msg,
+            "*== 1 failed in *",
+            '*Sending information to Paste Service*',
+        ])
 
 
 class TestPaste:
@@ -74,7 +101,7 @@ def read(self):
         return calls
 
     def test_create_new_paste(self, pastebin, mocked_urlopen):
-        result = pastebin.create_new_paste('full-paste-contents')
+        result = pastebin.create_new_paste(b'full-paste-contents')
         assert result == 'https://bpaste.net/show/3c0c6750bd'
         assert len(mocked_urlopen) == 1
         url, data = mocked_urlopen[0]