]> git.madduck.net Git - etc/taskwarrior.git/blobdiff - tasklib/task.py

madduck's git repository

Every one of the projects in this repository is available at the canonical URL git://git.madduck.net/madduck/pub/<projectpath> — see each project's metadata for the exact URL.

All patches and comments are welcome. Please squash your changes to logical commits before using git-format-patch and git-send-email to patches@git.madduck.net. If you'd read over the Git project's submission guidelines and adhered to them, I'd be especially grateful.

SSH access, as well as push access can be individually arranged.

If you use my repositories frequently, consider adding the following snippet to ~/.gitconfig and using the third clone URL listed for each project:

[url "git://git.madduck.net/madduck/"]
  insteadOf = madduck:

Merge branch 'release/0.7.1'
[etc/taskwarrior.git] / tasklib / task.py
index 44c6965c1e3fdde7be836e48a4f8befb7f9a38c8..b6c3d09168ec6897749ade2c2d4ce26ceebf2b33 100644 (file)
@@ -24,26 +24,126 @@ class TaskWarriorException(Exception):
     pass
 
 
-class TaskResource(object):
+class SerializingObject(object):
+    """
+    Common ancestor for TaskResource & TaskFilter, since they both
+    need to serialize arguments.
+    """
+
+    def _deserialize(self, key, value):
+        hydrate_func = getattr(self, 'deserialize_{0}'.format(key),
+                               lambda x: x if x != '' else None)
+        return hydrate_func(value)
+
+    def _serialize(self, key, value):
+        dehydrate_func = getattr(self, 'serialize_{0}'.format(key),
+                                 lambda x: x if x is not None else '')
+        return dehydrate_func(value)
+
+    def timestamp_serializer(self, date):
+        if not date:
+            return None
+        return date.strftime(DATE_FORMAT)
+
+    def timestamp_deserializer(self, date_str):
+        if not date_str:
+            return None
+        return datetime.datetime.strptime(date_str, DATE_FORMAT)
+
+    def serialize_entry(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_entry(self, value):
+        return self.timestamp_deserializer(value)
+
+    def serialize_modified(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_modified(self, value):
+        return self.timestamp_deserializer(value)
+
+    def serialize_due(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_due(self, value):
+        return self.timestamp_deserializer(value)
+
+    def serialize_scheduled(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_scheduled(self, value):
+        return self.timestamp_deserializer(value)
+
+    def serialize_until(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_until(self, value):
+        return self.timestamp_deserializer(value)
+
+    def serialize_wait(self, value):
+        return self.timestamp_serializer(value)
+
+    def deserialize_wait(self, value):
+        return self.timestamp_deserializer(value)
+
+    def deserialize_annotations(self, data):
+        return [TaskAnnotation(self, d) for d in data] if data else []
+
+    def serialize_tags(self, tags):
+        return ','.join(tags) if tags else ''
+
+    def deserialize_tags(self, tags):
+        if isinstance(tags, basestring):
+            return tags.split(',') if tags else []
+        return tags
+
+    def serialize_depends(self, cur_dependencies):
+        # Return the list of uuids
+        return ','.join(task['uuid'] for task in cur_dependencies)
+
+    def deserialize_depends(self, raw_uuids):
+        raw_uuids = raw_uuids or ''  # Convert None to empty string
+        uuids = raw_uuids.split(',')
+        return set(self.warrior.tasks.get(uuid=uuid) for uuid in uuids if uuid)
+
+
+class TaskResource(SerializingObject):
     read_only_fields = []
 
     def _load_data(self, data):
         self._data = data
         # We need to use a copy for original data, so that changes
-        # are not propagated
+        # are not propagated. Shallow copy is alright, since data dict uses only
+        # primitive data types
         self._original_data = data.copy()
 
+    def _update_data(self, data, update_original=False):
+        """
+        Low level update of the internal _data dict. Data which are coming as
+        updates should already be serialized. If update_original is True, the
+        original_data dict is updated as well.
+        """
+
+        self._data.update(data)
+
+        if update_original:
+            self._original_data.update(data)
+
     def __getitem__(self, key):
-        hydrate_func = getattr(self, 'deserialize_{0}'.format(key),
-                               lambda x: x)
-        return hydrate_func(self._data.get(key))
+        # This is a workaround to make TaskResource non-iterable
+        # over simple index-based iteration
+        try:
+            int(key)
+            raise StopIteration
+        except ValueError:
+            pass
+
+        return self._deserialize(key, self._data.get(key))
 
     def __setitem__(self, key, value):
         if key in self.read_only_fields:
             raise RuntimeError('Field \'%s\' is read-only' % key)
-        dehydrate_func = getattr(self, 'serialize_{0}'.format(key),
-                                 lambda x: x)
-        self._data[key] = dehydrate_func(value)
+        self._data[key] = self._serialize(key, value)
 
     def __str__(self):
         s = six.text_type(self.__unicode__())
@@ -62,12 +162,6 @@ class TaskAnnotation(TaskResource):
         self.task = task
         self._load_data(data)
 
-    def deserialize_entry(self, data):
-        return datetime.datetime.strptime(data, DATE_FORMAT) if data else None
-
-    def serialize_entry(self, date):
-        return date.strftime(DATE_FORMAT) if date else ''
-
     def remove(self):
         self.task.remove_annotation(self)
 
@@ -78,7 +172,7 @@ class TaskAnnotation(TaskResource):
 
 
 class Task(TaskResource):
-    read_only_fields = ['id', 'entry', 'urgency', 'uuid']
+    read_only_fields = ['id', 'entry', 'urgency', 'uuid', 'modified']
 
     class DoesNotExist(Exception):
         pass
@@ -102,26 +196,46 @@ class Task(TaskResource):
         """
         pass
 
-    def __init__(self, warrior, data={}, **kwargs):
+    def __init__(self, warrior, **kwargs):
         self.warrior = warrior
 
-        # We keep data for backwards compatibility
-        kwargs.update(data)
+        # Check that user is not able to set read-only value in __init__
+        for key in kwargs.keys():
+            if key in self.read_only_fields:
+                raise RuntimeError('Field \'%s\' is read-only' % key)
+
+        # We serialize the data in kwargs so that users of the library
+        # do not have to pass different data formats via __setitem__ and
+        # __init__ methods, that would be confusing
 
-        self._load_data(kwargs)
+        # Rather unfortunate syntax due to python2.6 comaptiblity
+        self._load_data(dict((key, self._serialize(key, value))
+                        for (key, value) in six.iteritems(kwargs)))
 
     def __unicode__(self):
         return self['description']
 
     def __eq__(self, other):
-        return self['uuid'] == other['uuid']
+        if self['uuid'] and other['uuid']:
+            # For saved Tasks, just define equality by equality of uuids
+            return self['uuid'] == other['uuid']
+        else:
+            # If the tasks are not saved, compare the actual instances
+            return id(self) == id(other)
+
 
     def __hash__(self):
-        return self['uuid'].__hash__()
+        if self['uuid']:
+            # For saved Tasks, just define equality by equality of uuids
+            return self['uuid'].__hash__()
+        else:
+            # If the tasks are not saved, return hash of instance id
+            return id(self).__hash__()
 
     @property
     def _modified_fields(self):
-        for key in self._data.keys():
+        writable_fields = set(self._data.keys()) - set(self.read_only_fields)
+        for key in writable_fields:
             if self._data.get(key) != self._original_data.get(key):
                 yield key
 
@@ -145,24 +259,42 @@ class Task(TaskResource):
     def saved(self):
         return self['uuid'] is not None or self['id'] is not None
 
-    def serialize_due(self, date):
-        return date.strftime(DATE_FORMAT)
+    def serialize_depends(self, cur_dependencies):
+        # Check that all the tasks are saved
+        for task in cur_dependencies:
+            if not task.saved:
+                raise Task.NotSaved('Task \'%s\' needs to be saved before '
+                                    'it can be set as dependency.' % task)
 
-    def deserialize_due(self, date_str):
-        if not date_str:
-            return None
-        return datetime.datetime.strptime(date_str, DATE_FORMAT)
+        return super(Task, self).serialize_depends(cur_dependencies)
 
-    def deserialize_annotations(self, data):
-        return [TaskAnnotation(self, d) for d in data] if data else []
+    def format_depends(self):
+        # We need to generate added and removed dependencies list,
+        # since Taskwarrior does not accept redefining dependencies.
 
-    def deserialize_tags(self, tags):
-        if isinstance(tags, basestring):
-            return tags.split(',') if tags else []
-        return tags
+        # This cannot be part of serialize_depends, since we need
+        # to keep a list of all depedencies in the _data dictionary,
+        # not just currently added/removed ones
 
-    def serialize_tags(self, tags):
-        return ','.join(tags) if tags else ''
+        old_dependencies_raw = self._original_data.get('depends','')
+        old_dependencies = self.deserialize_depends(old_dependencies_raw)
+
+        added = self['depends'] - old_dependencies
+        removed = old_dependencies - self['depends']
+
+        # Removed dependencies need to be prefixed with '-'
+        return 'depends:' + ','.join(
+                [t['uuid'] for t in added] +
+                ['-' + t['uuid'] for t in removed]
+            )
+
+    def format_description(self):
+        # Task version older than 2.4.0 ignores first word of the
+        # task description if description: prefix is used
+        if self.warrior.version < VERSION_2_4_0:
+            return self._data['description']
+        else:
+            return "description:'{0}'".format(self._data['description'] or '')
 
     def delete(self):
         if not self.saved:
@@ -174,9 +306,7 @@ class Task(TaskResource):
         if self.deleted:
             raise Task.DeletedTask("Task was already deleted")
 
-        self.warrior.execute_command([self['uuid'], 'delete'], config_override={
-            'confirmation': 'no',
-        })
+        self.warrior.execute_command([self['uuid'], 'delete'])
 
         # Refresh the status again, so that we have updated info stored
         self.refresh(only_fields=['status'])
@@ -229,7 +359,7 @@ class Task(TaskResource):
 
     def remove_annotation(self, annotation):
         if not self.saved:
-            raise Task.NotSaved("Task needs to be saved to add annotation")
+            raise Task.NotSaved("Task needs to be saved to remove annotation")
 
         if isinstance(annotation, TaskAnnotation):
             annotation = annotation['description']
@@ -241,12 +371,15 @@ class Task(TaskResource):
         args = []
 
         def add_field(field):
-            # Task version older than 2.4.0 ignores first word of the
-            # task description if description: prefix is used
-            if self.warrior.version < VERSION_2_4_0 and field == 'description':
-                args.append(self._data[field])
-            else:
-                args.append('{0}:{1}'.format(field, self._data[field]))
+            # Add the output of format_field method to args list (defaults to
+            # field:'value')
+            format_default = lambda k: "{0}:{1}".format(k,
+                                           "'{0}'".format(self._data[k])
+                                           if self._data[k] is not None
+                                           else '')
+            format_func = getattr(self, 'format_{0}'.format(field),
+                                  lambda: format_default(field))
+            args.append(format_func())
 
         # If we're modifying saved task, simply pass on all modified fields
         if self.saved:
@@ -274,17 +407,12 @@ class Task(TaskResource):
         if only_fields:
             to_update = dict(
                 [(k, new_data.get(k)) for k in only_fields])
-            self._data.update(to_update)
-            self._original_data.update(to_update)
+            self._update_data(to_update, update_original=True)
         else:
-            self._data = new_data
-            # We need to create a clone for original_data though
-            # Shallow copy is alright, since data dict uses only
-            # primitive data types
-            self._original_data = new_data.copy()
+            self._load_data(new_data)
 
 
-class TaskFilter(object):
+class TaskFilter(SerializingObject):
     """
     A set of parameters to filter the task list with.
     """
@@ -300,14 +428,23 @@ class TaskFilter(object):
 
         # Replace the value with empty string, since that is the
         # convention in TW for empty values
-        value = value if value is not None else ''
+        attribute_key = key.split('.')[0]
+        value = self._serialize(attribute_key, value)
 
         # If we are filtering by uuid:, do not use uuid keyword
         # due to TW-1452 bug
         if key == 'uuid':
             self.filter_params.insert(0, value)
         else:
-            self.filter_params.append('{0}:{1}'.format(key, value))
+            # Surround value with aphostrophes unless it's a empty string
+            value = "'%s'" % value if value else ''
+
+            # We enforce equality match by using 'is' (or 'none') modifier
+            # Without using this syntax, filter fails due to TW-1479
+            modifier = '.is' if value else '.none'
+            key = key + modifier if '.' not in key else key
+
+            self.filter_params.append("{0}:{1}".format(key, value))
 
     def get_filter_params(self):
         return [f for f in self.filter_params if f]
@@ -435,6 +572,8 @@ class TaskWarrior(object):
             os.makedirs(data_location)
         self.config = {
             'data.location': os.path.expanduser(data_location),
+            'confirmation': 'no',
+            'dependency.confirmation': 'no', # See TW-1483 or taskrc man page
         }
         self.tasks = TaskQuerySet(self)
         self.version = self._get_version()
@@ -478,7 +617,9 @@ class TaskWarrior(object):
             if line:
                 data = line.strip(',')
                 try:
-                    tasks.append(Task(self, json.loads(data)))
+                    filtered_task = Task(self)
+                    filtered_task._load_data(json.loads(data))
+                    tasks.append(filtered_task)
                 except ValueError:
                     raise TaskWarriorException('Invalid JSON: %s' % data)
         return tasks
@@ -490,6 +631,4 @@ class TaskWarrior(object):
         })
 
     def undo(self):
-        self.execute_command(['undo'], config_override={
-            'confirmation': 'no',
-        })
+        self.execute_command(['undo'])