From d3a3f09ad1d49ce6accda00c805735bdf8b1ffca Mon Sep 17 00:00:00 2001
From: amercader <amercadero@gmail.com>
Date: Thu, 11 Jun 2015 10:19:07 +0100
Subject: [PATCH] [#127] Use site user on the CKAN harvester

To avoid having to create a 'harvest' sysadmin explicitly. It will still
be used if present, but if not the site user will be used. You can also
define to user to use via a config option.
---
 ckanext/harvest/harvesters/base.py | 48 +++++++++++++++++++++++++++---
 1 file changed, 44 insertions(+), 4 deletions(-)

diff --git a/ckanext/harvest/harvesters/base.py b/ckanext/harvest/harvesters/base.py
index b4db6ce..41a5061 100644
--- a/ckanext/harvest/harvesters/base.py
+++ b/ckanext/harvest/harvesters/base.py
@@ -4,6 +4,7 @@ import uuid
 
 from sqlalchemy.sql import update,and_, bindparam
 from sqlalchemy.exc import InvalidRequestError
+from pylons import config
 
 from ckan import plugins as p
 from ckan import model
@@ -38,6 +39,8 @@ class HarvesterBase(SingletonPlugin):
 
     config = None
 
+    _user_name = None
+
     def _gen_new_name(self, title):
         '''
         Creates a URL friendly name from a title
@@ -80,6 +83,46 @@ class HarvesterBase(SingletonPlugin):
             log_message = '{0}, line {1}'.format(message,line) if line else message
             log.debug(log_message)
 
+    def _get_user_name(self):
+        '''
+        Returns the name of the user that will perform the harvesting actions
+        (deleting, updating and creating datasets)
+
+        By default this will be the old 'harvest' user to maintain
+        compatibility. If not present, the internal site admin user will be
+        used. This is the recommended setting, but if necessary it can be
+        overridden with the `ckanext.harvest.user_name` config option:
+
+           ckanext.harvest.user_name = harvest
+
+        '''
+        if self._user_name:
+            return self._user_name
+
+        config_user_name = config.get('ckanext.harvest.user_name')
+        if config_user_name:
+            self._user_name = config_user_name
+            return self._user_name
+
+        context = {'model': model,
+                   'ignore_auth': True,
+                   }
+
+        # Check if 'harvest' user exists and if is a sysadmin
+        try:
+            user_harvest = p.toolkit.get_action('user_show')(
+                context, {'id': 'harvest'})
+            if user_harvest['sysadmin']:
+                self._user_name = 'harvest'
+                return self._user_name
+        except p.toolkit.ObjectNotFound:
+            pass
+
+        context['defer_commit'] = True  # See ckan/ckan#1714
+        self._site_user = p.toolkit.get_action('get_site_user')(context, {})
+        self._user_name = self._site_user['name']
+
+        return self._user_name
 
     def _create_harvest_objects(self, remote_ids, harvest_job):
         '''
@@ -135,13 +178,10 @@ class HarvesterBase(SingletonPlugin):
                     api_version = int(self.config.get('api_version', 2))
                 except ValueError:
                     raise ValueError('api_version must be an integer')
-
-                #TODO: use site user when available
-                user_name = self.config.get('user', u'harvest')
             else:
                 api_version = 2
-                user_name = u'harvest'
 
+            user_name = self._get_user_name()
             context = {
                 'model': model,
                 'session': Session,