software/cloudooo: set LANG to C.UTF-8

Libreoffice needs to run with an UTF-8 locale, otherwise it does not use proper
encoding for text files.
cloudooo sets a default locale as a fallback, but using en_US.UTF-8, which is
not always present (it is not present on capri testnodes). Anyway, this kind
of configuration is the job of slapos, so we prefer to have it done at slapos
level.

software/cloudooo/buildout.hash.cfg

@@ -18,4 +18,4 @@ md5sum = e986de01a57161b32425f1cd3ccac924
 
 [template-cloudooo-instance]
 filename = instance-cloudooo.cfg.in
-md5sum = 9e1a66cf18d7c30c14afeb66c20afb46
+md5sum = b83e2df7ed3a2af2bdcd0c323e189eb3

software/cloudooo/instance-cloudooo.cfg.in

@@ -162,6 +162,7 @@ environment =
   LD_LIBRARY_PATH = {{ parameter_dict['cairo'] }}/lib:{{ parameter_dict['cups'] }}/lib:{{ parameter_dict['cups'] }}/lib64:{{ parameter_dict['dbus'] }}/lib:{{ parameter_dict['dbus-glib'] }}/lib:{{ parameter_dict['file'] }}/lib:{{ parameter_dict['fontconfig'] }}/lib:{{ parameter_dict['freetype'] }}/lib:{{ parameter_dict['gcc'] }}/lib:{{ parameter_dict['gcc'] }}/lib64:{{ parameter_dict['glib'] }}/lib:{{ parameter_dict['glu'] }}/lib:{{ parameter_dict['libICE'] }}/lib:{{ parameter_dict['libSM'] }}/lib:{{ parameter_dict['libX11'] }}/lib:{{ parameter_dict['libXau'] }}/lib:{{ parameter_dict['libXdmcp'] }}/lib:{{ parameter_dict['libXext'] }}/lib:{{ parameter_dict['libXrender'] }}/lib:{{ parameter_dict['libexpat'] }}/lib:{{ parameter_dict['libffi'] }}/lib:{{ parameter_dict['libffi'] }}/lib64:{{ parameter_dict['libpng12'] }}/lib:{{ parameter_dict['libxcb'] }}/lib:{{ parameter_dict['mesa'] }}/lib:{{ parameter_dict['pixman'] }}/lib:{{ parameter_dict['xdamage'] }}/lib:{{ parameter_dict['xfixes'] }}/lib:{{ parameter_dict['zlib'] }}/lib
   FONTCONFIG_FILE = ${fontconfig-conf:rendered}
   PATH = ${binary-link:target-directory}
+  LANG = C.UTF-8
 mimetype_entry_addition =
 {% for entry in mimetype_entry_addition.splitlines() -%}
 {{ "  " ~ entry.strip() }}

software/cloudooo/test/test.py

@@ -26,6 +26,7 @@
 #
 ##############################################################################
 
+import codecs
 import os
 import json
 import six.moves.xmlrpc_client as xmlrpclib
@@ -37,12 +38,27 @@ import PyPDF2
 
 from slapos.testing.testcase import makeModuleSetUpAndTestCaseClass
 
-setUpModule, CloudOooTestCase = makeModuleSetUpAndTestCaseClass(
+setUpModule, _CloudOooTestCase = makeModuleSetUpAndTestCaseClass(
     os.path.abspath(
         os.path.join(os.path.dirname(__file__), '..', 'software.cfg')))
 
-# Cloudooo needs a lot of time before being available.
-CloudOooTestCase.instance_max_retry = 30
+
+class CloudOooTestCase(_CloudOooTestCase):
+  # Cloudooo needs a lot of time before being available.
+  instance_max_retry = 30
+
+  def setUp(self):
+    self.url = json.loads(
+        self.computer_partition.getConnectionParameterDict()["_"])['cloudooo']
+    # XXX ignore certificate errors
+    ssl_context = ssl.create_default_context()
+    ssl_context.check_hostname = False
+    ssl_context.verify_mode = ssl.CERT_NONE
+    self.server = xmlrpclib.ServerProxy(
+        self.url,
+        context=ssl_context,
+        allow_none=True,
+    )
 
 
 def normalizeFontName(font_name):
@@ -88,19 +104,6 @@ class HTMLtoPDFConversionFontTestMixin:
     """Convert the HTML source to pdf bytes.
     """
 
-  def setUp(self):
-    self.url = json.loads(
-        self.computer_partition.getConnectionParameterDict()["_"])['cloudooo']
-    # XXX ignore certificate errors
-    ssl_context = ssl.create_default_context()
-    ssl_context.check_hostname = False
-    ssl_context.verify_mode = ssl.CERT_NONE
-    self.server = xmlrpclib.ServerProxy(
-        self.url,
-        context=ssl_context,
-        allow_none=True,
-    )
-
   def test(self):
     actual_font_mapping_mapping = {}
     for font in self.expected_font_mapping:
@@ -237,3 +240,19 @@ class TestLibreoffice(HTMLtoPDFConversionFontTestMixin, CloudOooTestCase):
             'html',
             'pdf',
         ).encode())
+
+
+class TestLibreOfficeTextConversion(CloudOooTestCase):
+  __partition_reference__ = 'txt'
+
+  def test_html_to_text(self):
+    self.assertEqual(
+        base64.decodestring(
+            self.server.convertFile(
+                base64.encodestring(
+                    u'<html>héhé</html>'.encode('utf-8')).decode(),
+                'html',
+                'txt',
+            ).encode()),
+        codecs.BOM_UTF8 + b'h\xc3\xa9h\xc3\xa9\n',
+    )