Ejemplo n.º 1
0
    def get_by_ids(self, ids, **params):
        if not ids:
            return _ESDocs()

        _raise_on_empty = params.pop('_raise_on_empty', False)
        fields = params.pop('_fields', [])

        _limit = params.pop('_limit', len(ids))
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        _start, _limit = process_limit(_start, _page, _limit)

        docs = []
        for _id in ids:
            docs.append(
                dict(_index=self.index_name,
                     _type=self.src2type(_id['_type']),
                     _id=_id['_id']))

        params = dict(body=dict(docs=docs))
        if fields:
            fields_params = process_fields_param(fields)
            params.update(fields_params)

        documents = _ESDocs()
        documents._nefertari_meta = dict(
            start=_start,
            fields=fields,
        )

        try:
            data = self.api.mget(**params)
        except IndexNotFoundException:
            if _raise_on_empty:
                raise JHTTPNotFound(
                    '{}({}) resource not found (Index does not exist)'.format(
                        self.doc_type, params))
            documents._nefertari_meta.update(total=0)
            return documents

        for found_doc in data['docs']:
            try:
                output_doc = found_doc['_source']
                output_doc['_type'] = found_doc['_type']
            except KeyError:
                msg = "ES: '%s(%s)' resource not found" % (found_doc['_type'],
                                                           found_doc['_id'])
                if _raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.error(msg)
                    continue

            documents.append(dict2obj(dictset(output_doc)))

        documents._nefertari_meta.update(total=len(documents), )

        return documents
Ejemplo n.º 2
0
    def get_by_ids(self, ids, **params):
        if not ids:
            return _ESDocs()

        __raise_on_empty = params.pop('__raise_on_empty', False)
        fields = params.pop('_fields', [])

        _limit = params.pop('_limit', len(ids))
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        _start, _limit = process_limit(_start, _page, _limit)

        docs = []
        for _id in ids:
            docs.append(
                dict(
                    _index=self.index_name,
                    _type=self.src2type(_id['_type']),
                    _id=_id['_id']
                )
            )

        params = dict(
            body=dict(docs=docs)
        )
        if fields:
            params['fields'] = fields

        data = ES.api.mget(**params)
        documents = _ESDocs()

        for _d in data['docs']:
            try:
                _d = _d['fields'] if fields else _d['_source']
            except KeyError:
                msg = "ES: '%s(%s)' resource not found" % (
                    _d['_type'], _d['_id'])
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.error(msg)
                    continue

            documents.append(dict2obj(dictset(_d)))

        documents._nefertari_meta = dict(
            total=len(documents),
            start=_start,
            fields=fields,
        )

        return documents
Ejemplo n.º 3
0
    def get_by_ids(self, ids, **params):
        if not ids:
            return _ESDocs()

        __raise_on_empty = params.pop("__raise_on_empty", False)
        fields = params.pop("_fields", [])

        _limit = params.pop("_limit", len(ids))
        _page = params.pop("_page", None)
        _start = params.pop("_start", None)
        _start, _limit = process_limit(_start, _page, _limit)

        docs = []
        for _id in ids:
            docs.append(dict(_index=self.index_name, _type=self.src2type(_id["_type"]), _id=_id["_id"]))

        params = dict(body=dict(docs=docs))
        if fields:
            fields_params = process_fields_param(fields)
            params.update(fields_params)

        documents = _ESDocs()
        documents._nefertari_meta = dict(start=_start, fields=fields)

        try:
            data = self.api.mget(**params)
        except IndexNotFoundException:
            if __raise_on_empty:
                raise JHTTPNotFound("{}({}) resource not found (Index does not exist)".format(self.doc_type, params))
            documents._nefertari_meta.update(total=0)
            return documents

        for found_doc in data["docs"]:
            try:
                output_doc = found_doc["_source"]
                output_doc["_type"] = found_doc["_type"]
            except KeyError:
                msg = "ES: '%s(%s)' resource not found" % (found_doc["_type"], found_doc["_id"])
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.error(msg)
                    continue

            documents.append(dict2obj(dictset(output_doc)))

        documents._nefertari_meta.update(total=len(documents))

        return documents
Ejemplo n.º 4
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(
            index=self.index_name,
            doc_type=self.doc_type
        )
        _raw_terms = params.pop('q', '')

        if 'body' not in params:
            query_string = build_qs(params.remove(RESERVED_PARAMS), _raw_terms)
            if query_string:
                _params['body'] = {
                    'query': {
                        'query_string': {
                            'query': query_string
                        }
                    }
                }
            else:
                _params['body'] = {"query": {"match_all": {}}}
        else:
            _params['body'] = params['body']

        if '_limit' not in params:
            params['_limit'] = self.api.count()['count']

        _params['from_'], _params['size'] = process_limit(
            params.get('_start', None),
            params.get('_page', None),
            params['_limit'])

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')
            search_fields.reverse()
            search_fields = [s + '^' + str(i) for i, s in
                             enumerate(search_fields, 1)]
            current_qs = _params['body']['query']['query_string']
            if isinstance(current_qs, str):
                _params['body']['query']['query_string'] = {'query': current_qs}
            _params['body']['query']['query_string']['fields'] = search_fields

        return _params
Ejemplo n.º 5
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(index=self.index_name, doc_type=self.doc_type)
        _raw_terms = params.pop('q', '')

        if 'body' not in params:
            query_string = build_qs(params.remove(RESERVED_PARAMS), _raw_terms)
            if query_string:
                _params['body'] = {
                    'query': {
                        'query_string': {
                            'query': query_string
                        }
                    }
                }
            else:
                _params['body'] = {"query": {"match_all": {}}}
        else:
            _params['body'] = params['body']

        if '_limit' not in params:
            params['_limit'] = self.api.count()['count']

        _params['from_'], _params['size'] = process_limit(
            params.get('_start', None), params.get('_page', None),
            params['_limit'])

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')
            search_fields.reverse()
            search_fields = [
                s + '^' + str(i) for i, s in enumerate(search_fields, 1)
            ]
            current_qs = _params['body']['query']['query_string']
            if isinstance(current_qs, str):
                _params['body']['query']['query_string'] = {
                    'query': current_qs
                }
            _params['body']['query']['query_string']['fields'] = search_fields

        return _params
Ejemplo n.º 6
0
    def get_by_ids(self, ids, **params):
        if not ids:
            return _ESDocs()

        __raise_on_empty = params.pop('__raise_on_empty', False)
        fields = params.pop('_fields', [])

        _limit = params.pop('_limit', len(ids))
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        _start, _limit = process_limit(_start, _page, _limit)

        docs = []
        for _id in ids:
            docs.append(
                dict(_index=self.index_name,
                     _type=self.src2type(_id['_type']),
                     _id=_id['_id']))

        params = dict(body=dict(docs=docs))
        if fields:
            params['fields'] = fields

        data = ES.api.mget(**params)
        documents = _ESDocs()

        for _d in data['docs']:
            try:
                _d = _d['fields'] if fields else _d['_source']
            except KeyError:
                msg = "ES: '%s(%s)' resource not found" % (_d['_type'],
                                                           _d['_id'])
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.error(msg)
                    continue

            documents.append(dict2obj(dictset(_d)))

        documents._nefertari_meta = dict(
            total=len(documents),
            start=_start,
            fields=fields,
        )

        return documents
Ejemplo n.º 7
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(
            index=self.index_name,
            doc_type=self.doc_type
        )

        if 'body' not in params:
            query_string = build_qs(
                params.remove(RESERVED),
                params.get('_raw_terms', ''))
            if query_string:
                _params['body'] = {
                    'query': {
                        'query_string': {
                            'query': query_string
                        }
                    }
                }
            else:
                _params['body'] = {"query": {"match_all": {}}}

        if '_limit' not in params:
            raise JHTTPBadRequest('Missing _limit')

        _params['from_'], _params['size'] = process_limit(
            params.get('_start', None),
            params.get('_page', None),
            params['_limit'])

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')
            search_fields.reverse()
            search_fields = [s + '^' + str(i) for i, s in
                             enumerate(search_fields, 1)]
            _params['body']['query']['query_string']['fields'] = search_fields

        return _params
Ejemplo n.º 8
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(index=self.index_name, doc_type=self.doc_type)

        if 'body' not in params:
            query_string = build_qs(params.remove(RESERVED),
                                    params.get('_raw_terms', ''))
            if query_string:
                _params['body'] = {
                    'query': {
                        'query_string': {
                            'query': query_string
                        }
                    }
                }
                print query_string
            else:
                _params['body'] = {"query": {"match_all": {}}}

        if '_limit' in params:
            _params['from_'], _params['size'] = process_limit(
                params.get('_start', None), params.get('_page', None),
                params['_limit'])

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')
            search_fields.reverse()
            search_fields = [
                s + '^' + str(i) for i, s in enumerate(search_fields, 1)
            ]
            _params['body']['query']['query_string']['fields'] = search_fields

        return _params
Ejemplo n.º 9
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(index=self.index_name, doc_type=self.doc_type)
        _raw_terms = params.pop("q", "")

        if "body" not in params:
            query_string = build_qs(params.remove(RESERVED_PARAMS), _raw_terms)
            if query_string:
                _params["body"] = {"query": {"query_string": {"query": query_string}}}
            else:
                _params["body"] = {"query": {"match_all": {}}}
        else:
            _params["body"] = params["body"]

        if "_limit" not in params:
            params["_limit"] = self.api.count()["count"]

        _params["from_"], _params["size"] = process_limit(
            params.get("_start", None), params.get("_page", None), params["_limit"]
        )

        if "_sort" in params:
            _params["sort"] = apply_sort(params["_sort"])

        if "_fields" in params:
            _params["fields"] = params["_fields"]

        if "_search_fields" in params:
            search_fields = params["_search_fields"].split(",")
            search_fields.reverse()
            search_fields = [s + "^" + str(i) for i, s in enumerate(search_fields, 1)]
            current_qs = _params["body"]["query"]["query_string"]
            if isinstance(current_qs, str):
                _params["body"]["query"]["query_string"] = {"query": current_qs}
            _params["body"]["query"]["query_string"]["fields"] = search_fields

        return _params
Ejemplo n.º 10
0
    def get_by_ids(self, ids, **params):
        if not ids:
            return _ESDocs()

        _raise_on_empty = params.pop('_raise_on_empty', False)
        fields = params.pop('_fields', [])

        _limit = params.pop('_limit', len(ids))
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        _start, _limit = process_limit(_start, _page, _limit)

        docs = []
        for _id in ids:
            docs.append(
                dict(
                    _index=self.index_name,
                    _type=self.src2type(_id['_type']),
                    _id=_id['_id']
                )
            )

        params = dict(
            body=dict(docs=docs)
        )
        if fields:
            fields_params = process_fields_param(fields)
            params.update(fields_params)

        documents = _ESDocs()
        documents._nefertari_meta = dict(
            start=_start,
            fields=fields,
        )

        try:
            data = self.api.mget(**params)
        except IndexNotFoundException:
            if _raise_on_empty:
                raise JHTTPNotFound(
                    '{}({}) resource not found (Index does not exist)'.format(
                        self.doc_type, params))
            documents._nefertari_meta.update(total=0)
            return documents

        for found_doc in data['docs']:
            try:
                output_doc = found_doc['_source']
                output_doc['_type'] = found_doc['_type']
            except KeyError:
                msg = "ES: '%s(%s)' resource not found" % (
                    found_doc['_type'], found_doc['_id'])
                if _raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.error(msg)
                    continue

            documents.append(dict2proxy(dictset(output_doc), ES.document_proxy.get_document_proxies_by_type(found_doc['_type'])))

        documents._nefertari_meta.update(
            total=len(documents),
        )

        return documents
Ejemplo n.º 11
0
    def get_collection(cls, **params):
        """
        params may include '_limit', '_page', '_sort', '_fields'
        returns paginated and sorted query set
        raises JHTTPBadRequest for bad values in params
        """
        params.pop('__confirmation', False)
        __strict = params.pop('__strict', True)

        _sort = _split(params.pop('_sort', []))
        _fields = _split(params.pop('_fields', []))
        _limit = params.pop('_limit', None)
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)

        _count = '_count' in params; params.pop('_count', None)
        _explain = '_explain' in params; params.pop('_explain', None)
        __raise_on_empty = params.pop('__raise_on_empty', False)

        query_set = cls.objects

        # Remove any __ legacy instructions from this point on
        params = dictset(filter(lambda item: not item[0].startswith('__'), params.items()))

        if __strict:
            _check_fields = [f.strip('-+') for f in params.keys() + _fields + _sort]
            cls.check_fields_allowed(_check_fields)
        else:
            params = cls.filter_fields(params)

        process_lists(params)
        process_bools(params)

        #if param is _all then remove it
        params.pop_by_values('_all')

        try:
            query_set = query_set(**params)
            _total = query_set.count()
            if _count:
                return _total

            if _limit is None:
                raise JHTTPBadRequest('Missing _limit')

            _start, _limit = process_limit(_start, _page, _limit)

            # Filtering by fields has to be the first thing to do on the query_set!
            query_set = cls.apply_fields(query_set, _fields)
            query_set = cls.apply_sort(query_set, _sort)
            query_set = query_set[_start:_start+_limit]

            if not query_set.count():
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.debug(msg)

        except (mongo.ValidationError, mongo.InvalidQueryError) as e:
            raise JHTTPBadRequest(str(e), extra={'data': e})

        if _explain:
            return query_set.explain()

        log.debug('get_collection.query_set: %s(%s)',
                  cls.__name__, query_set._query)

        query_set._nefertari_meta = dict(
            total=_total,
            start=_start,
            fields=_fields)

        return query_set
Ejemplo n.º 12
0
    def get_collection(cls, **params):
        """ Query collection and return results.

        Notes:
        *   Before validating that only model fields are present in params,
            reserved params, query params and all params starting with
            double underscore are dropped.
        *   Params which have value "_all" are dropped.
        *   When ``_count`` param is used, objects count is returned
            before applying offset and limit.

        :param bool _strict: If True ``params`` are validated to contain
            only fields defined on model, exception is raised if invalid
            fields are present. When False - invalid fields are dropped.
            Defaults to ``True``.
        :param bool _item_request: Indicates whether it is a single item
            request or not. When True and DataError happens on DB request,
            JHTTPNotFound is raised. JHTTPBadRequest is raised when False.
            Defaults to ``False``.
        :param list _sort: Field names to sort results by. If field name
            is prefixed with "-" it is used for "descending" sorting.
            Otherwise "ascending" sorting is performed by that field.
            Defaults to an empty list in which case sorting is not
            performed.
        :param list _fields: Names of fields which should be included
            or excluded from results. Fields to excluded should be
            prefixed with "-". Defaults to an empty list in which
            case all fields are returned.
        :param int _limit: Number of results per page. Defaults
            to None in which case all results are returned.
        :param int _page: Number of page. In conjunction with
            ``_limit`` is used to calculate results offset. Defaults to
            None in which case it is ignored. Params ``_page`` and
            ``_start` are mutually exclusive.
        :param int _start: Results offset. If provided ``_limit`` and
            ``_page`` params are ignored when calculating offset. Defaults
            to None. Params ``_page`` and ``_start`` are mutually
            exclusive. If not offset-related params are provided, offset
            equals to 0.
        :param Query query_set: Existing queryset. If provided, all queries
            are applied to it instead of creating new queryset. Defaults
            to None.
        :param _count: When provided, only results number is returned as
            integer.
        :param _explain: When provided, query performed(SQL) is returned
            as a string instead of query results.
        :param bool _raise_on_empty: When True JHTTPNotFound is raised
            if query returned no results. Defaults to False in which case
            error is just logged and empty query results are returned.

        :returns: Query results as ``sqlalchemy.orm.query.Query`` instance.
            May be sorted, offset, limited.
        :returns: Dict of {'field_name': fieldval}, when ``_fields`` param
            is provided.
        :returns: Number of query results as an int when ``_count`` param
            is provided.
        :returns: String representing query ran when ``_explain`` param
            is provided.

        :raises JHTTPNotFound: When ``_raise_on_empty=True`` and no
            results found.
        :raises JHTTPNotFound: When ``_item_request=True`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query. Latter exception is raised when querying DB with
            an identifier of a wrong type. E.g. when querying Int field
            with a string.
        :raises JHTTPBadRequest: When ``_item_request=False`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query.
        :raises JHTTPBadRequest: When ``sqlalchemy.exc.InvalidRequestError``
            or ``sqlalchemy.exc.IntegrityError`` errors happen during DB
            query.
        """
        log.debug('Get collection: {}, {}'.format(cls.__name__, params))
        params.pop('__confirmation', False)
        _strict = params.pop('_strict', True)
        _item_request = params.pop('_item_request', False)

        _sort = _split(params.pop('_sort', []))
        _fields = _split(params.pop('_fields', []))
        _limit = params.pop('_limit', None)
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        query_set = params.pop('query_set', None)

        _count = '_count' in params
        params.pop('_count', None)
        _explain = '_explain' in params
        params.pop('_explain', None)
        _raise_on_empty = params.pop('_raise_on_empty', False)

        if query_set is None:
            query_set = Session().query(cls)

        # Remove any __ legacy instructions from this point on
        params = dictset({
            key: val
            for key, val in params.items() if not key.startswith('__')
        })

        iterables_exprs, params = cls._pop_iterables(params)

        params = drop_reserved_params(params)
        if _strict:
            _check_fields = [
                f.strip('-+') for f in list(params.keys()) + _fields + _sort
            ]
            cls.check_fields_allowed(_check_fields)
        else:
            params = cls.filter_fields(params)

        process_lists(params)
        process_bools(params)

        # If param is _all then remove it
        params.pop_by_values('_all')

        try:

            query_set = query_set.filter_by(**params)

            # Apply filtering by iterable expressions
            for expr in iterables_exprs:
                query_set = query_set.from_self().filter(expr)

            _total = query_set.count()
            if _count:
                return _total

            # Filtering by fields has to be the first thing to do on
            # the query_set!
            query_set = cls.apply_fields(query_set, _fields)
            query_set = cls.apply_sort(query_set, _sort)

            if _limit is not None:
                _start, _limit = process_limit(_start, _page, _limit)
                query_set = query_set.offset(_start).limit(_limit)

            if not query_set.count():
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                if _raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.debug(msg)

        except DataError as ex:
            if _item_request:
                msg = "'{}({})' resource not found".format(
                    cls.__name__, params)
                raise JHTTPNotFound(msg, explanation=ex.message)
            else:
                raise JHTTPBadRequest(str(ex), extra={'data': ex})
        except (InvalidRequestError, ) as ex:
            raise JHTTPBadRequest(str(ex), extra={'data': ex})

        query_sql = str(query_set).replace('\n', '')
        if _explain:
            return query_sql

        log.debug('get_collection.query_set: %s (%s)', cls.__name__, query_sql)

        if _fields:
            query_set = cls.add_field_names(query_set, _fields)

        query_set._nefertari_meta = dict(total=_total,
                                         start=_start,
                                         fields=_fields)
        return query_set
Ejemplo n.º 13
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(
            index=self.index_name,
            doc_type=self.doc_type
        )
        _raw_terms = params.pop('q', '')

        if 'body' not in params and 'es_q' not in params:
            analyzed_terms = apply_analyzer(params, self.doc_type, engine.get_document_cls)

            query_string = self.build_qs(params.remove(RESERVED_PARAMS), _raw_terms)

            query = {'must': []}

            if query_string:
                query['must'].append({'query_string': {'query': query_string}})

            if analyzed_terms:
                query['must'].append(analyzed_terms)

            if query['must']:
                _params['body'] = {'query': {'bool': query}}
            else:
                _params['body'] = {'query': {'match_all': {}}}

        if 'body' in params:
            raise JHTTPUnprocessableEntity('Illegal parameter "body"')

        if '_limit' not in params:
            params['_limit'] = self.api.count(index=self.index_name)['count']
        _params['from_'], _params['size'] = process_limit(
            params.get('_start', None),
            params.get('_page', None),
            params['_limit'])

        if 'es_q' in params:
            _params['body'] = {}

            try:
                _params['body']['query'] = compile_es_query(params)
            except Exception as exc:
                log.exception('es_q parsing error: {exc}'.format(exc=exc))
                raise JHTTPBadRequest('Bad query string for {params}'
                                        .format(
                                                params=_params['body']['query']['query_string']['query']))

            log.debug('Parsed ES request body {body}'.format(body=_params['body']['query']))

        if '_sort' in params and self.proxy:
            params['_sort'] = substitute_nested_terms(params['_sort'], self.proxy.substitutions)

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            params['_fields'] = self.add_nested_fields(params['_fields'], ',')
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')

            search_fields.reverse()

            # Substitute search fields and add ^index
            for index, search_field in enumerate(search_fields):
                sf_terms = search_field.split('.')

                if self.proxy is not None:
                    if len(sf_terms) > 0 and sf_terms[0] in self.proxy.substitutions:
                        sf_terms[0] += "_nested"
                        search_field = '.'.join(sf_terms)

                search_fields[index] = search_field + '^' + str(index + 1)

            must_query = _params['body']['query']['bool']['must']
            query_string = {}

            for query_item in must_query:
                if 'query_string' in query_item:
                    query_string = query_item
                    break

            current_qs = query_string.get('query_string', None)

            if current_qs:
                query_string['query_string']['fields'] = search_fields

        return _params
Ejemplo n.º 14
0
    def get_collection(cls, **params):
        """ Query collection and return results.

        Notes:
        *   Before validating that only model fields are present in params,
            reserved params, query params and all params starting with
            double underscore are dropped.
        *   Params which have value "_all" are dropped.
        *   When ``_count`` param is used, objects count is returned
            before applying offset and limit.

        :param bool __strict: If True ``params`` are validated to contain
            only fields defined on model, exception is raised if invalid
            fields are present. When False - invalid fields are dropped.
            Defaults to ``True``.
        :param bool _item_request: Indicates whether it is a single item
            request or not. When True and DataError happens on DB request,
            JHTTPNotFound is raised. JHTTPBadRequest is raised when False.
            Defaults to ``False``.
        :param list _sort: Field names to sort results by. If field name
            is prefixed with "-" it is used for "descending" sorting.
            Otherwise "ascending" sorting is performed by that field.
            Defaults to an empty list in which case sorting is not
            performed.
        :param list _fields: Names of fields which should be included
            or excluded from results. Fields to excluded should be
            prefixed with "-". Defaults to an empty list in which
            case all fields are returned.
        :param int _limit: Number of results per page. Defaults
            to None in which case all results are returned.
        :param int _page: Number of page. In conjunction with
            ``_limit`` is used to calculate results offset. Defaults to
            None in which case it is ignored. Params ``_page`` and
            ``_start` are mutually exclusive.
        :param int _start: Results offset. If provided ``_limit`` and
            ``_page`` params are ignored when calculating offset. Defaults
            to None. Params ``_page`` and ``_start`` are mutually
            exclusive. If not offset-related params are provided, offset
            equals to 0.
        :param Query query_set: Existing queryset. If provided, all queries
            are applied to it instead of creating new queryset. Defaults
            to None.
        :param _count: When provided, only results number is returned as
            integer.
        :param _explain: When provided, query performed(SQL) is returned
            as a string instead of query results.
        :param bool __raise_on_empty: When True JHTTPNotFound is raised
            if query returned no results. Defaults to False in which case
            error is just logged and empty query results are returned.

        :returns: Query results as ``sqlalchemy.orm.query.Query`` instance.
            May be sorted, offset, limited.
        :returns: Dict of {'field_name': fieldval}, when ``_fields`` param
            is provided.
        :returns: Number of query results as an int when ``_count`` param
            is provided.
        :returns: String representing query ran when ``_explain`` param
            is provided.

        :raises JHTTPNotFound: When ``__raise_on_empty=True`` and no
            results found.
        :raises JHTTPNotFound: When ``_item_request=True`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query. Latter exception is raised when querying DB with
            an identifier of a wrong type. E.g. when querying Int field
            with a string.
        :raises JHTTPBadRequest: When ``_item_request=False`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query.
        :raises JHTTPBadRequest: When ``sqlalchemy.exc.InvalidRequestError``
            or ``sqlalchemy.exc.IntegrityError`` errors happen during DB
            query.
        """
        log.debug('Get collection: {}, {}'.format(cls.__name__, params))
        params.pop('__confirmation', False)
        __strict = params.pop('__strict', True)
        _item_request = params.pop('_item_request', False)

        _sort = _split(params.pop('_sort', []))
        _fields = _split(params.pop('_fields', []))
        _limit = params.pop('_limit', None)
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        query_set = params.pop('query_set', None)

        _count = '_count' in params
        params.pop('_count', None)
        _explain = '_explain' in params
        params.pop('_explain', None)
        __raise_on_empty = params.pop('__raise_on_empty', False)

        if query_set is None:
            query_set = Session().query(cls)

        # Remove any __ legacy instructions from this point on
        params = dictset({
            key: val for key, val in params.items()
            if not key.startswith('__')
        })

        iterables_exprs, params = cls._pop_iterables(params)

        params = drop_reserved_params(params)
        if __strict:
            _check_fields = [
                f.strip('-+') for f in list(params.keys()) + _fields + _sort]
            cls.check_fields_allowed(_check_fields)
        else:
            params = cls.filter_fields(params)

        process_lists(params)
        process_bools(params)

        # If param is _all then remove it
        params.pop_by_values('_all')

        try:

            query_set = query_set.filter_by(**params)

            # Apply filtering by iterable expressions
            for expr in iterables_exprs:
                query_set = query_set.from_self().filter(expr)

            _total = query_set.count()
            if _count:
                return _total

            # Filtering by fields has to be the first thing to do on
            # the query_set!
            query_set = cls.apply_fields(query_set, _fields)
            query_set = cls.apply_sort(query_set, _sort)

            if _limit is not None:
                _start, _limit = process_limit(_start, _page, _limit)
                query_set = query_set.offset(_start).limit(_limit)

            if not query_set.count():
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.debug(msg)

        except DataError as ex:
            if _item_request:
                msg = "'{}({})' resource not found".format(
                    cls.__name__, params)
                raise JHTTPNotFound(msg, explanation=ex.message)
            else:
                raise JHTTPBadRequest(str(ex), extra={'data': ex})
        except (InvalidRequestError,) as ex:
            raise JHTTPBadRequest(str(ex), extra={'data': ex})

        query_sql = str(query_set).replace('\n', '')
        if _explain:
            return query_sql

        log.debug('get_collection.query_set: %s (%s)', cls.__name__, query_sql)

        if _fields:
            query_set = cls.add_field_names(query_set, _fields)

        query_set._nefertari_meta = dict(
            total=_total,
            start=_start,
            fields=_fields)
        return query_set
Ejemplo n.º 15
0
    def get_collection(cls, **params):
        """
        Params may include '_limit', '_page', '_sort', '_fields'.
        Returns paginated and sorted query set.
        Raises JHTTPBadRequest for bad values in params.
        """
        log.debug('Get collection: {}, {}'.format(cls.__name__, params))
        params.pop('__confirmation', False)
        _strict = params.pop('_strict', True)
        _item_request = params.pop('_item_request', False)

        _sort = _split(params.pop('_sort', []))
        _fields = _split(params.pop('_fields', []))
        _limit = params.pop('_limit', None)
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)
        query_set = params.pop('query_set', None)

        _count = '_count' in params
        params.pop('_count', None)
        _explain = '_explain' in params
        params.pop('_explain', None)
        _raise_on_empty = params.pop('_raise_on_empty', False)

        if query_set is None:
            query_set = cls.objects

        # Remove any __ legacy instructions from this point on
        params = dictset({
            key: val for key, val in params.items()
            if not key.startswith('__')
        })

        params = drop_reserved_params(params)
        if _strict:
            _check_fields = [
                f.strip('-+') for f in list(params.keys()) + _fields + _sort]
            cls.check_fields_allowed(_check_fields)
        else:
            params = cls.filter_fields(params)

        process_lists(params)
        process_bools(params)

        # If param is _all then remove it
        params.pop_by_values('_all')

        try:
            query_set = query_set(**params)
            _total = query_set.count()
            if _count:
                return _total

            # Filtering by fields has to be the first thing to do on the
            # query_set!
            query_set = cls.apply_fields(query_set, _fields)
            query_set = cls.apply_sort(query_set, _sort)

            if _limit is not None:
                _start, _limit = process_limit(_start, _page, _limit)
                query_set = query_set[_start:_start+_limit]

            if not query_set.count():
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                if _raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.debug(msg)

        except mongo.ValidationError as ex:
            if _item_request:
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                raise JHTTPNotFound(msg, explanation=ex.message)
            else:
                raise JHTTPBadRequest(str(ex), extra={'data': ex})
        except mongo.InvalidQueryError as ex:
            raise JHTTPBadRequest(str(ex), extra={'data': ex})

        if _explain:
            return query_set.explain()

        log.debug('get_collection.query_set: %s(%s)',
                  cls.__name__, query_set._query)

        query_set._nefertari_meta = dict(
            total=_total,
            start=_start,
            fields=_fields)

        return query_set
Ejemplo n.º 16
0
    def build_search_params(self, params):
        params = dictset(params)

        _params = dict(
            index=self.index_name,
            doc_type=self.doc_type
        )
        _raw_terms = params.pop('q', '')

        if 'body' not in params:
            query_string = self.build_qs(params.remove(RESERVED_PARAMS), _raw_terms)
            if query_string:
                _params['body'] = {
                    'query': {
                        'query_string': {
                            'query': query_string
                        }
                    }
                }
            else:
                _params['body'] = {"query": {"match_all": {}}}
        else:
            _params['body'] = params['body']
        if '_limit' not in params:
            params['_limit'] = self.api.count(index=self.index_name)['count']

        _params['from_'], _params['size'] = process_limit(
            params.get('_start', None),
            params.get('_page', None),
            params['_limit'])

        if '_nested' in params:
            if 'query' in _params['body']:
                old_query = _params['body']['query']
                _params['body']['query'] = {'bool': {'must': []}}
                nested = dict()

                nested['nested'] = self.build_nested_query(params)
                _params['body']['query']['bool']['must'].append(nested)
                _params['body']['query']['bool']['must'].append(old_query)

        if '_sort' in params and self.proxy:
            params['_sort'] = substitute_nested_terms(params['_sort'], self.proxy.substitutions)

        if '_sort' in params:
            _params['sort'] = apply_sort(params['_sort'])

        if '_fields' in params:
            params['_fields'] = self.add_nested_fields(params['_fields'], ',')
            _params['fields'] = params['_fields']

        if '_search_fields' in params:
            search_fields = params['_search_fields'].split(',')

            search_fields.reverse()

            # Substitute search fields and add ^index
            for index, search_field in enumerate(search_fields):
                sf_terms = search_field.split('.')

                if self.proxy is not None:
                    if len(sf_terms) > 0 and sf_terms[0] in self.proxy.substitutions:
                        sf_terms[0] += "_nested"
                        search_field = '.'.join(sf_terms)

                search_fields[index] = search_field + '^' + str(index + 1)

            current_qs = _params['body']['query']['query_string']

            if isinstance(current_qs, str):
                _params['body']['query']['query_string'] = {'query': current_qs}
            _params['body']['query']['query_string']['fields'] = search_fields
        return _params
Ejemplo n.º 17
0
    def get_collection(cls, _count=False, _strict=True, _sort=None,
                       _fields=None, _limit=None, _page=None, _start=None,
                       _query_set=None, _item_request=False, _explain=None,
                       _search_fields=None, q=None, **params):
        """ Query collection and return results.

        Notes:
        *   Before validating that only model fields are present in params,
            reserved params, query params and all params starting with
            double underscore are dropped.
        *   Params which have value "_all" are dropped.
        *   When ``_count`` param is used, objects count is returned
            before applying offset and limit.

        :param bool _strict: If True ``params`` are validated to contain
            only fields defined on model, exception is raised if invalid
            fields are present. When False - invalid fields are dropped.
            Defaults to ``True``.
        :param list _sort: Field names to sort results by. If field name
            is prefixed with "-" it is used for "descending" sorting.
            Otherwise "ascending" sorting is performed by that field.
            Defaults to an empty list in which case sorting is not
            performed.
        :param list _fields: Names of fields which should be included
            or excluded from results. Fields to excluded should be
            prefixed with "-". Defaults to an empty list in which
            case all fields are returned.
        :param int _limit: Number of results per page. Defaults
            to None in which case all results are returned.
        :param int _page: Number of page. In conjunction with
            ``_limit`` is used to calculate results offset. Defaults to
            None in which case it is ignored. Params ``_page`` and
            ``_start` are mutually exclusive.
        :param int _start: Results offset. If provided ``_limit`` and
            ``_page`` params are ignored when calculating offset. Defaults
            to None. Params ``_page`` and ``_start`` are mutually
            exclusive. If not offset-related params are provided, offset
            equals to 0.
        :param Query _query_set: Existing queryset. If provided, all queries
            are applied to it instead of creating new queryset. Defaults
            to None.
        :param bool _item_request: Indicates whether it is a single item
            request or not. When True and DataError happens on DB request,
            JHTTPNotFound is raised. JHTTPBadRequest is raised when False.
            Defaults to ``False``.
        :param _count: When provided, only results number is returned as
            integer.
        :param _explain: When provided, query performed(SQL) is returned
            as a string instead of query results.
        :param bool _raise_on_empty: When True JHTTPNotFound is raised
            if query returned no results. Defaults to False in which case
            error is just logged and empty query results are returned.
        :param q: Query string to perform full-text search with.
        :param _search_fields: Coma-separated list of field names to use
            with full-text search(q param) to limit fields which are
            searched.

        :returns: Query results as ``elasticsearch_dsl.XXX`` instance.
            May be sorted, offset, limited.
        :returns: Dict of {'field_name': fieldval}, when ``_fields`` param
            is provided.
        :returns: Number of query results as an int when ``_count`` param
            is provided.

        :raises JHTTPNotFound: When ``_raise_on_empty=True`` and no
            results found.
        :raises JHTTPNotFound: When ``_item_request=True`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query. Latter exception is raised when querying DB with
            an identifier of a wrong type. E.g. when querying Int field
            with a string.
        :raises JHTTPBadRequest: When ``_item_request=False`` and
            ``sqlalchemy.exc.DataError`` exception is raised during DB
            query.
        :raises JHTTPBadRequest: When ``sqlalchemy.exc.InvalidRequestError``
            or ``sqlalchemy.exc.IntegrityError`` errors happen during DB
            query.
        """
        # see if the items are cached
        pk_field = cls.pk_field()
        if (list(params.keys()) == [pk_field] and _count==False
            and _strict==True and _sort==None and _fields==None
            and _limit==None and _page==None and _start==None
            and _query_set==None and _item_request==False and _explain==None
            and _search_fields==None and q==None):
            ids = params[pk_field]
            if not isinstance(ids, (list, tuple)):
                ids = [ids]
            results = []
            for id in ids:
                if not id in cls._cache:
                    break
                results.append(cls._cache[id])
            else:
                return results

        search_obj = cls.search()

        if _limit is not None:
            _start, limit = process_limit(_start, _page, _limit)
            search_obj = search_obj.extra(from_=_start, size=limit)

        if _fields:
            include, exclude = process_fields(_fields)
            if _strict:
                _validate_fields(cls, include + exclude)
            # XXX partial fields support isn't yet released. for now
            # we just use fields, later we'll add support for excluded fields
            search_obj = search_obj.fields(include)

        if params:
            params = _cleaned_query_params(cls, params, _strict)
            params = _restructure_params(cls, params)
            if params:
                search_obj = search_obj.filter('terms', **params)

        if q is not None:
            query_kw = {'query': q}
            if _search_fields is not None:
                query_kw['fields'] = _search_fields.split(',')
            search_obj = search_obj.query('query_string', **query_kw)

        if _count:
            return search_obj.count()

        if _explain:
            return search_obj.to_dict()

        if _sort:
            sort_fields = split_strip(_sort)
            if _strict:
                _validate_fields(
                    cls,
                    [f[1:] if f.startswith('-') else f for f in sort_fields]
                    )
            search_obj = search_obj.sort(*sort_fields)

        hits = search_obj.execute().hits
        hits._nefertari_meta = dict(
            total=hits.total,
            start=_start,
            fields=_fields
            )
        return hits
Ejemplo n.º 18
0
    def get_collection(cls, **params):
        """
        params may include '_limit', '_page', '_sort', '_fields'
        returns paginated and sorted query set
        raises JHTTPBadRequest for bad values in params
        """
        params.pop('__confirmation', False)
        __strict = params.pop('__strict', True)

        _sort = _split(params.pop('_sort', []))
        _fields = _split(params.pop('_fields', []))
        _limit = params.pop('_limit', None)
        _page = params.pop('_page', None)
        _start = params.pop('_start', None)

        _count = '_count' in params
        params.pop('_count', None)
        _explain = '_explain' in params
        params.pop('_explain', None)
        __raise_on_empty = params.pop('__raise_on_empty', False)

        query_set = cls.objects

        # Remove any __ legacy instructions from this point on
        params = dictset(
            filter(lambda item: not item[0].startswith('__'), params.items()))

        if __strict:
            _check_fields = [
                f.strip('-+') for f in params.keys() + _fields + _sort
            ]
            cls.check_fields_allowed(_check_fields)
        else:
            params = cls.filter_fields(params)

        process_lists(params)
        process_bools(params)

        #if param is _all then remove it
        params.pop_by_values('_all')

        try:
            query_set = query_set(**params)
            _total = query_set.count()
            if _count:
                return _total

            if _limit is None:
                raise JHTTPBadRequest('Missing _limit')

            _start, _limit = process_limit(_start, _page, _limit)

            # Filtering by fields has to be the first thing to do on the query_set!
            query_set = cls.apply_fields(query_set, _fields)
            query_set = cls.apply_sort(query_set, _sort)
            query_set = query_set[_start:_start + _limit]

            if not query_set.count():
                msg = "'%s(%s)' resource not found" % (cls.__name__, params)
                if __raise_on_empty:
                    raise JHTTPNotFound(msg)
                else:
                    log.debug(msg)

        except (mongo.ValidationError, mongo.InvalidQueryError) as e:
            raise JHTTPBadRequest(str(e), extra={'data': e})

        if _explain:
            return query_set.explain()

        log.debug('get_collection.query_set: %s(%s)', cls.__name__,
                  query_set._query)

        query_set._nefertari_meta = dict(total=_total,
                                         start=_start,
                                         fields=_fields)

        return query_set