def merge_suggestions(request): # TODO: test
req = request.json
entity_type_primary = req.get("entity_type_primary") or "Transcription"
entity_type_secondary = req.get("entity_type_secondary") or "Translation"
threshold = req["threshold"] or 0.2
levenstein = req["levenstein"] or 1
client_id = req["client_id"]
object_id = req["object_id"]
lexes = list(
DBSession.query(LexicalEntry)
.filter_by(parent_client_id=client_id, parent_object_id=object_id, marked_for_deletion=False)
.all()
)
if not lexes:
return json.dumps([])
# first_persp = json.loads(lexes[0].additional_metadata)['came_from']
lexes_1 = [o.track(False) for o in lexes]
remove_deleted(lexes_1)
lexes_2 = list(lexes_1)
def parse_response(elem):
words = filter(
lambda x: x["entity_type"] == entity_type_primary and not x["marked_for_deletion"], elem["contains"]
)
words = map(lambda x: x["content"], words)
trans = filter(
lambda x: x["entity_type"] == entity_type_secondary and not x["marked_for_deletion"], elem["contains"]
)
trans = map(lambda x: x["content"], trans)
tuples_res = [
(i_word, i_trans, (elem["client_id"], elem["object_id"])) for i_word in words for i_trans in trans
]
return tuples_res
tuples_1 = [parse_response(i) for i in lexes_1]
tuples_1 = [item for sublist in tuples_1 for item in sublist]
tuples_2 = [parse_response(i) for i in lexes_2]
tuples_2 = [item for sublist in tuples_2 for item in sublist]
def get_dict(elem):
return {
"suggestion": [
{"lexical_entry_client_id": elem[0][0], "lexical_entry_object_id": elem[0][1]},
{"lexical_entry_client_id": elem[1][0], "lexical_entry_object_id": elem[1][1]},
],
"confidence": elem[2],
}
if (not tuples_1) or (not tuples_2):
return {}
results = [get_dict(i) for i in mergeDicts(tuples_1, tuples_2, float(threshold), int(levenstein))]
results = sorted(results, key=lambda k: k["confidence"])
return results
def merge_suggestions_old(request):
subreq = Request.blank('/dictionary/' + request.matchdict.get('dictionary_client_id_1') + '/' +
request.matchdict.get('dictionary_object_id_1') + '/perspective/' +
request.matchdict.get('perspective_client_id_1') + '/' +
request.matchdict.get('perspective_object_id_1') + '/all')
subreq.method = 'GET'
response_1 = request.invoke_subrequest(subreq).json
subreq = Request.blank('/dictionary/' + request.matchdict.get('dictionary_client_id_2') + '/' +
request.matchdict.get('dictionary_object_id_2') + '/perspective/' +
request.matchdict.get('perspective_client_id_2') + '/' +
request.matchdict.get('perspective_object_id_2') + '/all')
subreq.method = 'GET'
response_2 = request.invoke_subrequest(subreq).json
#entity_type_primary = 'Word'
#entity_type_secondary = 'Transcription'
#threshold = 0.2
#levenstein = 2
entity_type_primary = request.matchdict.get('entity_type_primary')
entity_type_secondary = request.matchdict.get('entity_type_secondary')
threshold = request.matchdict.get('threshold')
levenstein = request.matchdict.get('levenstein')
def parse_response(elem):
words = filter(lambda x: x['entity_type'] == entity_type_primary and not x['marked_for_deletion'], elem['contains'])
words = map(lambda x: x['content'], words)
trans = filter(lambda x: x['entity_type'] == entity_type_secondary and not x['marked_for_deletion'], elem['contains'])
trans = map(lambda x: x['content'], trans)
tuples_res = [(i_word, i_trans, (elem['client_id'], elem['object_id'])) for i_word in words for i_trans in trans]
return tuples_res
tuples_1 = [parse_response(i) for i in response_1['lexical_entries']]
tuples_1 = [item for sublist in tuples_1 for item in sublist]
tuples_2 = [parse_response(i) for i in response_2['lexical_entries']]
tuples_2 = [item for sublist in tuples_2 for item in sublist]
def get_dict(elem):
return {'suggestion': [
{'lexical_entry_client_id': elem[0][0], 'lexical_entry_object_id': elem[0][1]},
{'lexical_entry_client_id': elem[1][0], 'lexical_entry_object_id': elem[1][1]}
], 'confidence': elem[2]}
results = [get_dict(i) for i in mergeDicts(tuples_1, tuples_2, float(threshold), int(levenstein))]
return json.dumps(results)
def merge_suggestions_old(request):
subreq = Request.blank(
"/dictionary/"
+ request.matchdict.get("dictionary_client_id_1")
+ "/"
+ request.matchdict.get("dictionary_object_id_1")
+ "/perspective/"
+ request.matchdict.get("perspective_client_id_1")
+ "/"
+ request.matchdict.get("perspective_object_id_1")
+ "/all"
)
subreq.method = "GET"
response_1 = request.invoke_subrequest(subreq).json
subreq = Request.blank(
"/dictionary/"
+ request.matchdict.get("dictionary_client_id_2")
+ "/"
+ request.matchdict.get("dictionary_object_id_2")
+ "/perspective/"
+ request.matchdict.get("perspective_client_id_2")
+ "/"
+ request.matchdict.get("perspective_object_id_2")
+ "/all"
)
subreq.method = "GET"
response_2 = request.invoke_subrequest(subreq).json
# entity_type_primary = 'Word'
# entity_type_secondary = 'Transcription'
# threshold = 0.2
# levenstein = 2
entity_type_primary = request.matchdict.get("entity_type_primary")
entity_type_secondary = request.matchdict.get("entity_type_secondary")
threshold = request.matchdict.get("threshold")
levenstein = request.matchdict.get("levenstein")
def parse_response(elem):
words = filter(
lambda x: x["entity_type"] == entity_type_primary and not x["marked_for_deletion"], elem["contains"]
)
words = map(lambda x: x["content"], words)
trans = filter(
lambda x: x["entity_type"] == entity_type_secondary and not x["marked_for_deletion"], elem["contains"]
)
trans = map(lambda x: x["content"], trans)
tuples_res = [
(i_word, i_trans, (elem["client_id"], elem["object_id"])) for i_word in words for i_trans in trans
]
return tuples_res
tuples_1 = [parse_response(i) for i in response_1["lexical_entries"]]
tuples_1 = [item for sublist in tuples_1 for item in sublist]
tuples_2 = [parse_response(i) for i in response_2["lexical_entries"]]
tuples_2 = [item for sublist in tuples_2 for item in sublist]
def get_dict(elem):
return {
"suggestion": [
{"lexical_entry_client_id": elem[0][0], "lexical_entry_object_id": elem[0][1]},
{"lexical_entry_client_id": elem[1][0], "lexical_entry_object_id": elem[1][1]},
],
"confidence": elem[2],
}
results = [get_dict(i) for i in mergeDicts(tuples_1, tuples_2, float(threshold), int(levenstein))]
return json.dumps(results)