class ReuseSearch(ModelSearchAdapter):
model = Reuse
fuzzy = True
fields = (
'title^4',
'description^2',
'datasets.title',
)
facets = {
'tag': TermFacet('tags'),
'organization': ModelTermFacet('organization', Organization),
'owner': ModelTermFacet('owner', User),
'dataset': ModelTermFacet('dataset.id', Dataset),
'type': ReuseTypeFacet('type'),
'datasets': RangeFacet('metrics.datasets'),
'followers': RangeFacet('metrics.followers'),
'featured': BoolFacet('featured'),
'extra': ExtrasFacet('extras'),
'badge': TermFacet('badges', labelizer=reuse_badge_labelizer),
}
sorts = {
'title': Sort('title.raw'),
'created': Sort('created'),
'last_modified': Sort('last_modified'),
'datasets': Sort('metrics.datasets'),
'followers': Sort('metrics.followers'),
'views': Sort('metrics.views'),
}
mapping = {
'properties': {
'title': {
'type': 'string',
'analyzer': i18n_analyzer,
'fields': {
'raw': {
'type': 'string',
'index': 'not_analyzed'
}
}
},
'description': {
'type': 'string',
'analyzer': i18n_analyzer
},
'url': {
'type': 'string'
},
'organization': {
'type': 'string'
},
'owner': {
'type': 'string'
},
'type': {
'type': 'string'
},
'tags': {
'type': 'string',
'index_name': 'tag',
'index': 'not_analyzed'
},
'tag_suggest': {
'type': 'completion',
'index_analyzer': 'simple',
'search_analyzer': 'simple',
'payloads': False,
},
'badges': {
'type': 'string',
'index_name': 'badges',
'index': 'not_analyzed'
},
'created': {
'type': 'date',
'format': 'date_hour_minute_second'
},
'last_modified': {
'type': 'date',
'format': 'date_hour_minute_second'
},
'dataset': {
'type': 'object',
'properties': {
'id': {
'type': 'string'
},
'title': {
'type': 'string'
}
}
},
'metrics': metrics_mapping(Reuse),
'featured': {
'type': 'boolean'
},
'reuse_suggest': {
'type': 'completion',
'index_analyzer': 'simple',
'search_analyzer': 'simple',
'payloads': True,
},
'extras': {
'type': 'object',
'index_name': 'extra',
},
}
}
boosters = [
BoolBooster('featured', 1.1),
GaussDecay('metrics.datasets', max_datasets, decay=0.8),
GaussDecay('metrics.followers', max_followers, decay=0.8),
]
@classmethod
def is_indexable(cls, reuse):
return (reuse.deleted is None and len(reuse.datasets) > 0
and not reuse.private)
@classmethod
def serialize(cls, reuse):
"""By default use the ``to_dict`` method
and exclude ``_id``, ``_cls`` and ``owner`` fields.
"""
return {
'title':
reuse.title,
'description':
reuse.description,
'url':
reuse.url,
'organization':
(str(reuse.organization.id) if reuse.organization else None),
'owner':
str(reuse.owner.id) if reuse.owner else None,
'type':
reuse.type,
'tags':
reuse.tags,
'tag_suggest':
reuse.tags,
'badges': [badge.kind for badge in reuse.badges],
'created':
reuse.created_at.strftime('%Y-%m-%dT%H:%M:%S'),
'last_modified':
reuse.last_modified.strftime('%Y-%m-%dT%H:%M:%S'),
'dataset': [{
'id': str(d.id),
'title': d.title
} for d in reuse.datasets if isinstance(d, Dataset)],
'metrics':
reuse.metrics,
'featured':
reuse.featured,
'extras':
reuse.extras,
'reuse_suggest': {
'input': cls.completer_tokenize(reuse.title) + [reuse.id],
'output': str(reuse.id),
'payload': {
'title': reuse.title,
'slug': reuse.slug,
'image_url': reuse.image(40),
},
},
}
class ReuseSearch(ModelSearchAdapter):
model = Reuse
fuzzy = True
class Meta:
doc_type = 'Reuse'
title = String(analyzer=i18n_analyzer,
fields={'raw': String(index='not_analyzed')})
description = String(analyzer=i18n_analyzer)
url = String(index='not_analyzed')
organization = String(index='not_analyzed')
owner = String(index='not_analyzed')
type = String(index='not_analyzed')
tags = String(index='not_analyzed',
fields={'i18n': String(index='not_analyzed')})
badges = String(index='not_analyzed')
topic = String(index='not_analyzed')
tag_suggest = Completion(analyzer=simple,
search_analyzer=simple,
payloads=False)
datasets = Object(properties={
'id': String(index='not_analyzed'),
'title': String(),
})
created = Date(format='date_hour_minute_second')
last_modified = Date(format='date_hour_minute_second')
metrics = Reuse.__search_metrics__
featured = Boolean()
reuse_suggest = Completion(analyzer=simple,
search_analyzer=simple,
payloads=True)
extras = Object()
facets = {
'tag':
TermsFacet(field='tags'),
'organization':
ModelTermsFacet(field='organization', model=Organization),
'owner':
ModelTermsFacet(field='owner', model=User),
'dataset':
ModelTermsFacet(field='dataset.id', model=Dataset),
'type':
TermsFacet(field='type', labelizer=reuse_type_labelizer),
'datasets':
RangeFacet(field='metrics.datasets',
ranges=[('none', (None, 1)), ('few', (1, 5)),
('many', (5, None))],
labels={
'none': _('No datasets'),
'few': _('Few datasets'),
'many': _('Many datasets'),
}),
'followers':
RangeFacet(field='metrics.followers',
ranges=[('none', (None, 1)), ('few', (1, 5)),
('many', (5, None))],
labels={
'none': _('No followers'),
'few': _('Few followers'),
'many': _('Many followers'),
}),
'badge':
TermsFacet(field='badges', labelizer=reuse_badge_labelizer),
'featured':
BoolFacet(field='featured'),
'topic':
TermsFacet(field='topic', labelizer=reuse_topic_labelizer),
}
sorts = {
'title': 'title.raw',
'created': 'created',
'last_modified': 'last_modified',
'datasets': 'metrics.datasets',
'followers': 'metrics.followers',
'views': 'metrics.views',
}
boosters = [
BoolBooster('featured', lazy('featured_boost')),
GaussDecay('metrics.datasets',
max_datasets,
decay=lazy('datasets_decay')),
GaussDecay('metrics.followers',
max_followers,
decay=lazy('followers_decay')),
]
@classmethod
def is_indexable(cls, reuse):
return (reuse.deleted is None and len(reuse.datasets) > 0
and not reuse.private)
@classmethod
def serialize(cls, reuse):
"""By default use the ``to_dict`` method
and exclude ``_id``, ``_cls`` and ``owner`` fields.
"""
datasets = Dataset.objects(id__in=[r.id for r in reuse.datasets])
datasets = list(datasets.only('id', 'title').no_dereference())
organization = None
owner = None
if reuse.organization:
organization = Organization.objects(
id=reuse.organization.id).first()
elif reuse.owner:
owner = User.objects(id=reuse.owner.id).first()
return {
'title': reuse.title,
'description': reuse.description,
'url': reuse.url,
'organization': str(organization.id) if organization else None,
'owner': str(owner.id) if owner else None,
'type': reuse.type,
'topic': reuse.topic,
'tags': reuse.tags,
'tag_suggest': reuse.tags,
'badges': [badge.kind for badge in reuse.badges],
'created': to_iso_datetime(reuse.created_at),
'last_modified': to_iso_datetime(reuse.last_modified),
'dataset': [{
'id': str(d.id),
'title': d.title
} for d in datasets],
'metrics': reuse.metrics,
'featured': reuse.featured,
'extras': reuse.extras,
'reuse_suggest': {
'input': cls.completer_tokenize(reuse.title) + [reuse.id],
'output': str(reuse.id),
'payload': {
'title': reuse.title,
'slug': reuse.slug,
'image_url': reuse.image(500, external=True),
},
},
}