You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
senpy/senpy/extensions.py

417 lines
15 KiB
Python

"""
Main class for Senpy.
It orchestrates plugin (de)activation and analysis.
"""
from future import standard_library
standard_library.install_aliases()
from . import plugins, api
from .models import Error, AggregatedEvaluation
from .plugins import AnalysisPlugin
from .blueprints import api_blueprint, demo_blueprint, ns_blueprint
from threading import Thread
from functools import partial
import os
import copy
import errno
import logging
from . import gsitk_compat
logger = logging.getLogger(__name__)
class Senpy(object):
""" Default Senpy extension for Flask """
def __init__(self,
app=None,
plugin_folder=".",
data_folder=None,
default_plugins=False):
default_data = os.path.join(os.getcwd(), 'senpy_data')
self.data_folder = data_folder or os.environ.get('SENPY_DATA', default_data)
try:
os.makedirs(self.data_folder)
except OSError as e:
if e.errno == errno.EEXIST:
logger.debug('Data folder exists: {}'.format(self.data_folder))
else: # pragma: no cover
raise
self._default = None
self._plugins = {}
if plugin_folder:
self.add_folder(plugin_folder)
if default_plugins:
self.add_folder('plugins', from_root=True)
else:
# Add only conversion plugins
self.add_folder(os.path.join('plugins', 'postprocessing'),
from_root=True)
self.app = app
if app is not None:
self.init_app(app)
self._conversion_candidates = {}
def init_app(self, app):
""" Initialise a flask app to add plugins to its context """
"""
Note: I'm not particularly fond of adding self.app and app.senpy, but
I can't think of a better way to do it.
"""
app.senpy = self
# Use the newstyle teardown_appcontext if it's available,
# otherwise fall back to the request context
if hasattr(app, 'teardown_appcontext'):
app.teardown_appcontext(self.teardown)
else: # pragma: no cover
app.teardown_request(self.teardown)
app.register_blueprint(api_blueprint, url_prefix="/api")
app.register_blueprint(ns_blueprint, url_prefix="/ns")
app.register_blueprint(demo_blueprint, url_prefix="/")
def add_plugin(self, plugin):
self._plugins[plugin.name.lower()] = plugin
self._conversion_candidates = {}
def delete_plugin(self, plugin):
del self._plugins[plugin.name.lower()]
def plugins(self, plugin_type=None, is_activated=True, **kwargs):
""" Return the plugins registered for a given application. Filtered by criteria """
return sorted(plugins.pfilter(self._plugins,
plugin_type=plugin_type,
is_activated=is_activated,
**kwargs),
key=lambda x: x.id)
def get_plugin(self, name, default=None):
if name == 'default':
return self.default_plugin
elif name == 'conversion':
return None
if name.lower() in self._plugins:
return self._plugins[name.lower()]
results = self.plugins(id='endpoint:plugins/{}'.format(name.lower()),
plugin_type=None)
if results:
return results[0]
results = self.plugins(id=name,
plugin_type=None)
if results:
return results[0]
msg = ("Plugin not found: '{}'\n"
"Make sure it is ACTIVATED\n"
"Valid algorithms: {}").format(name,
self._plugins.keys())
raise Error(message=msg, status=404)
def get_plugins(self, name):
try:
name = name.split(',')
except AttributeError:
pass # Assume it is a tuple or a list
return tuple(self.get_plugin(n) for n in name)
def analysis_plugins(self, **kwargs):
""" Return only the analysis plugins that are active"""
candidates = self.plugins(**kwargs)
return list(plugins.pfilter(candidates, plugin_type=AnalysisPlugin))
def add_folder(self, folder, from_root=False):
""" Find plugins in this folder and add them to this instance """
if from_root:
folder = os.path.join(os.path.dirname(__file__), folder)
logger.debug("Adding folder: %s", folder)
if os.path.isdir(folder):
new_plugins = plugins.from_folder([folder],
data_folder=self.data_folder)
for plugin in new_plugins:
self.add_plugin(plugin)
else:
raise AttributeError("Not a folder or does not exist: %s", folder)
def _process(self, req, pending, done=None):
"""
Recursively process the entries with the first plugin in the list, and pass the results
to the rest of the plugins.
"""
done = done or []
if not pending:
return req
analysis = pending[0]
results = analysis.run(req)
results.activities.append(analysis)
done += analysis
return self._process(results, pending[1:], done)
def install_deps(self):
logger.info('Installing dependencies')
# If a plugin is activated, its dependencies should already be installed
# Otherwise, it would've failed to activate.
plugins.install_deps(*self.plugins(is_activated=False))
def analyse(self, request, analyses=None):
"""
Main method that analyses a request, either from CLI or HTTP.
It takes a processed request, provided by the user, as returned
by api.parse_call().
"""
if not self.plugins():
raise Error(
status=404,
message=("No plugins found."
" Please install one."))
if analyses is None:
plugins = self.get_plugins(request.parameters['algorithm'])
analyses = api.parse_analyses(request.parameters, plugins)
logger.debug("analysing request: {}".format(request))
results = self._process(request, analyses)
logger.debug("Got analysis result: {}".format(results))
results = self.postprocess(results, analyses)
logger.debug("Returning post-processed result: {}".format(results))
return results
def convert_emotions(self, resp, analyses):
"""
Conversion of all emotions in a response **in place**.
In addition to converting from one model to another, it has
to include the conversion plugin to the analysis list.
Needless to say, this is far from an elegant solution, but it works.
@todo refactor and clean up
"""
logger.debug("Converting emotions")
if 'parameters' not in resp:
logger.debug("NO PARAMETERS")
return resp
params = resp['parameters']
toModel = params.get('emotion-model', None)
if not toModel:
logger.debug("NO tomodel PARAMETER")
return resp
logger.debug('Asked for model: {}'.format(toModel))
output = params.get('conversion', None)
newentries = []
done = []
for i in resp.entries:
if output == "full":
newemotions = copy.deepcopy(i.emotions)
else:
newemotions = []
for j in i.emotions:
activity = j['prov:wasGeneratedBy']
act = resp.activity(activity)
if not act:
raise Error('Could not find the emotion model for {}'.format(activity))
fromModel = act.plugin['onyx:usesEmotionModel']
if toModel == fromModel:
continue
candidate = self._conversion_candidate(fromModel, toModel)
if not candidate:
e = Error(('No conversion plugin found for: '
'{} -> {}'.format(fromModel, toModel)),
status=404)
e.original_response = resp
e.parameters = params
raise e
analysis = candidate.activity(params)
done.append(analysis)
for k in candidate.convert(j, fromModel, toModel, params):
k.prov__wasGeneratedBy = analysis.id
if output == 'nested':
k.prov__wasDerivedFrom = j
newemotions.append(k)
i.emotions = newemotions
newentries.append(i)
resp.entries = newentries
return resp
def _conversion_candidate(self, fromModel, toModel):
if not self._conversion_candidates:
candidates = {}
for conv in self.plugins(plugin_type=plugins.EmotionConversion):
for pair in conv.onyx__doesConversion:
logging.debug(pair)
key = (pair['onyx:conversionFrom'], pair['onyx:conversionTo'])
if key not in candidates:
candidates[key] = []
candidates[key].append(conv)
self._conversion_candidates = candidates
key = (fromModel, toModel)
if key not in self._conversion_candidates:
return None
return self._conversion_candidates[key][0]
def postprocess(self, response, analyses):
'''
Transform the results from the analysis plugins.
It has some pre-defined post-processing like emotion conversion,
and it also allows plugins to auto-select themselves.
'''
response = self.convert_emotions(response, analyses)
for plug in self.plugins(plugin_type=plugins.PostProcessing):
if plug.check(response, response.activities):
activity = plug.activity(response.parameters)
response = plug.process(response, activity)
return response
def _get_datasets(self, request):
if not self.datasets:
raise Error(
status=404,
message=("No datasets found."
" Please verify DatasetManager"))
datasets_name = request.parameters.get('dataset', None).split(',')
for dataset in datasets_name:
if dataset not in self.datasets:
logger.debug(("The dataset '{}' is not valid\n"
"Valid datasets: {}").format(
dataset, self.datasets.keys()))
raise Error(
status=404,
message="The dataset '{}' is not valid".format(dataset))
dm = gsitk_compat.DatasetManager()
datasets = dm.prepare_datasets(datasets_name)
return datasets
@property
def datasets(self):
self._dataset_list = {}
dm = gsitk_compat.DatasetManager()
for item in dm.get_datasets():
for key in item:
if key in self._dataset_list:
continue
properties = item[key]
properties['@id'] = key
self._dataset_list[key] = properties
return self._dataset_list
def evaluate(self, params):
logger.debug("evaluating request: {}".format(params))
results = AggregatedEvaluation()
results.parameters = params
datasets = self._get_datasets(results)
plugs = []
for plugname in params['algorithm']:
plugs = self.get_plugins(plugname)
for plug in plugs:
if not isinstance(plug, plugins.Evaluable):
raise Exception('Plugin {} can not be evaluated', plug.id)
for eval in plugins.evaluate(plugs, datasets):
results.evaluations.append(eval)
if 'with-parameters' not in results.parameters:
del results.parameters
logger.debug("Returning evaluation result: {}".format(results))
return results
@property
def default_plugin(self):
if not self._default or not self._default.is_activated:
candidates = self.analysis_plugins()
if len(candidates) > 0:
self._default = candidates[0]
else:
self._default = None
logger.debug("Default: {}".format(self._default))
return self._default
@default_plugin.setter
def default_plugin(self, value):
if isinstance(value, plugins.Plugin):
if not value.is_activated:
raise AttributeError('The default plugin has to be activated.')
self._default = value
else:
self._default = self._plugins[value.lower()]
def activate_all(self, sync=True, allow_fail=False):
ps = []
for plug in self._plugins.keys():
try:
self.activate_plugin(plug, sync=sync)
except Exception as ex:
if not allow_fail:
raise
logger.error('Could not activate {}: {}'.format(plug, ex))
return ps
def deactivate_all(self, sync=True):
ps = []
for plug in self._plugins.keys():
ps.append(self.deactivate_plugin(plug, sync=sync))
return ps
def _activate(self, plugin):
success = False
with plugin._lock:
if plugin.is_activated:
return
plugin._activate()
msg = "Plugin activated: {}".format(plugin.name)
logger.info(msg)
success = plugin.is_activated
return success
def activate_plugin(self, plugin_name, sync=True):
plugin_name = plugin_name.lower()
if plugin_name not in self._plugins:
raise Error(
message="Plugin not found: {}".format(plugin_name), status=404)
plugin = self._plugins[plugin_name]
logger.info("Activating plugin: {}".format(plugin.name))
if sync or not getattr(plugin, 'async', True) or getattr(
plugin, 'sync', False):
return self._activate(plugin)
else:
th = Thread(target=partial(self._activate, plugin))
th.start()
return th
def _deactivate(self, plugin):
with plugin._lock:
if not plugin.is_activated:
return
plugin._deactivate()
logger.info("Plugin deactivated: {}".format(plugin.name))
def deactivate_plugin(self, plugin_name, sync=True):
plugin_name = plugin_name.lower()
if plugin_name not in self._plugins:
raise Error(
message="Plugin not found: {}".format(plugin_name), status=404)
plugin = self._plugins[plugin_name]
if sync or not getattr(plugin, 'async', True) or not getattr(
plugin, 'sync', False):
plugin._deactivate()
else:
th = Thread(target=plugin.deactivate)
th.start()
return th
def teardown(self, exception):
pass