Merge pull request #376 from FNNDSC/tail-logs-keep-nonascii

Fix truncation conditional length, keep non-ascii chars

chris_backend/plugininstances/services/manager.py

@@ -70,8 +70,6 @@
 
 class PluginInstanceManager(object):
 
-    __NONASCII = re.compile(r'[^\x00-\x7F]')
-
     def __init__(self, plugin_instance):
 
         self.c_plugin_inst = plugin_instance
@@ -672,8 +670,8 @@ def _register_output_files(self, filenames):
                 total_size += plg_inst_file.fname.size
         self.c_plugin_inst.size += total_size
 
-    @classmethod
-    def get_job_status_summary(cls, d_response=None):
+    @staticmethod
+    def get_job_status_summary(d_response=None):
         """
         Get a job status summary JSON string from pfcon response.
         """
@@ -705,16 +703,8 @@ def get_job_status_summary(cls, d_response=None):
             d_jobStatusSummary['compute']['return']['job_status'] = d_c['status']
             logs = d_jobStatusSummary['compute']['return']['job_logs'] = d_c['logs']
 
-            if len(logs) > 3000:
-                # Workaround which fixes https://github.com/FNNDSC/ChRIS_ultron_backEnd/issues/364
-                # Better solution discussed in https://github.com/FNNDSC/ChRIS_ultron_backEnd/issues/374
-                #
-                # JSON serialization causes string to be longer than you might expect
-                # because of escape sequences, e.g. json.dumps('\n') -> '"\\n"'
-
-                # replace non-ASCII characters, such as emojis and Chinese characters
-                logs_ascii = cls.__NONASCII.sub('?', logs)
-
-                # truncate logs, assuming worst case where every character needs to be escaped
-                d_jobStatusSummary['compute']['return']['job_logs'] = logs_ascii[-1800:]
-        return json.dumps(d_jobStatusSummary)
+            # truncate logs, assuming worst case where every character needs to be escaped
+            if len(logs) > 1800:
+                d_jobStatusSummary['compute']['return']['job_logs'] = logs[-1800:]
+        # PostgreSQL allows UTF-8, supports emojis, chinese, etc.
+        return json.dumps(d_jobStatusSummary, ensure_ascii=False)