def index_clusters():
""" Index clusters in elasticsearch
"""
es = get_es()
es.indices.put_mapping(
index=__index_name__, doc_type=__c_doc_type__, body={
__c_doc_type__: {
'properties': {
'keywords': {
"type": "string",
"position_increment_gap": 100
},
'items': {
"type": "string",
"position_increment_gap": 100
}
}
}
})
clusters = LoadClusters()
for query in clusters.get_queries()['queries']:
results = clusters.get_clusters(query)['children']
for result in results:
listings = list()
for child in result['children']:
listings.append(child['name'])
doc = {
'name': '{query}-{cluster}'.format(
query=query, cluster=result['name']),
'keywords': result['keywords'],
'items': listings
}
es.index(index=__index_name__, doc_type=__c_doc_type__, body=doc)
def get_news(day):
from es import get_es
""" 获取单独某一天的新闻 """
try:
es, index, doc_type = get_es('es_news_pub')
body = {
"query": {
"term": {
"datetime": day
}
}
}
res = es.search(index=index, doc_type=doc_type, body=body)
# 科技动态板块
technology_news = []
for r in res['hits']['hits']:
tmp = r['_source']
print(tmp)
if tmp.get("category") == "technology_news":
technology_news.append(tmp)
package = {
"technology_news": technology_news
}
return jsonify(
success=True,
data=package
)
except:
traceback.print_exc()
def main():
res = parse()
es, index, doc_type = get_es('es_news_kr36')
es1, index, doc_type = get_es('es_news_pub')
for item in res:
try:
es.index(index=index,
doc_type=doc_type,
body=item,
id=item['title'])
# 存入到pub中
es1.index(index=index,
doc_type=doc_type,
body=item,
id=item['title'])
except:
print(item['title'])
def facet():
es = get_es()
q = {
"query" : {
"match_all" : { }
},
"facets" : {
"text" : {
"terms" : {
"field" : "text",
"size" : 100
}
}
}
}
res = es.search(index='legal-index', body=q)['facets']
pp.pprint(res)
return res
def __init__(self, file_name):
self.es = get_es()
self.file_name = file_name
#!/usr/bin/env python
import re
import pprint
from es import get_es
import sys
from search.models import Phrase
import json
pp = pprint.PrettyPrinter(indent=4)
es = get_es()
def get_autocomplete(term):
s = {
"text-suggest" : {
"text" : term,
"completion" : {
"field" : "suggest"
}
}
}
res = es.suggest(index='autocomplete',body=s)
result = []
for r in res['text-suggest'][0]['options']:
r_dict = {
"id": r['text'].encode('utf-8'),
"label": r['text'].encode('utf-8'),
"value": r['text'].encode('utf-8')
}
result.append(r_dict)
pp.pprint(result)
return json.dumps(result)
def search(search_term,result,start_result=0):
es = get_es()
unquoted = dequoter(search_term).split('""')
auto_phrases = []
for u in unquoted:
auto_phrases.extend(phraser(u.rstrip()))
query_search_term = search_term
if len(auto_phrases) > 0:
query_search_term = search_term + ' ' + ' '.join(auto_phrases)
pp.pprint(query_search_term)
#phrase identifier
q = {
"fields" : ["title","urlAddress","text"],
"from" : start_result,
"size" : result,
"min_score": 0.1,
"query": {
"query_string": {
"query": query_search_term,
}
},
"highlight": {
"pre_tags" : ["<b>"],
"post_tags" : ["</b>"],
"fields": {
"text": {
"fragment_size" : 150,
"number_of_fragments": 4,
"no_match_size": 150,
"highlight_query": {
"query_string":{
"query":search_term.replace('"','')
}
}
}
}
},
"suggest" : {
"text" : search_term,
"simple_phrase" : {
"phrase" : {
"field" : "text",
"size" : 1,
"real_word_error_likelihood" : 0.95,
"max_errors" : 0.5,
"gram_size" : 2,
"direct_generator" : [ {
"field" : "text",
"suggest_mode" : "popular",
"min_word_len" : 1,
"min_doc_freq" : 0.01,
"max_term_freq" : 0.01
} ]
}
}
}
}
res = es.search(index="legal-index", body=q)
result = {}
try:
suggested =res['suggest']['simple_phrase'][0]['options'][0]['text'].encode('utf-8')
result.update({'suggestion': suggested})
result.update({'suggestion_urlencode': urllib.quote(suggested)})
pp.pprint(urllib.urlencode(suggested))
except:
pass
r = res['hits']['hits']
l = []
for re in r:
d = {"urlAddress" : re['fields']['urlAddress'],
#similar unicode bs
"title" : str(re['fields']['title'][3:-2].encode('utf-8').replace("\\n",'').replace("u'","").replace("' ","").replace("\\r","").replace("',","").replace("\\t","").replace('u"','').replace('\u2019','')),
"id" : re['_id'],
"score" : re['_score'],}
if d['title'] == '':
d['title'] = d['urlAddress']
try:
h_list = []
for h in re['highlight']['text']:
#really a hacked up fix to remove rubbish terms in the highlights
#thanks unicode not compatible python 2. thanks noob skills junyuan. thanks over encoding and ecoding of strings.
h = h.encode('utf-8','ignore').replace("\\n",'').replace("u'","").replace("' ","")\
.replace("\\r","").replace("',","").replace("\\t","").replace('u"','').replace('\u2019','')
if h[:2] in ['. ',') ',', ',"' ",")."]:
h = h[2:].strip()
h_list.append(h)
d.update({"highlight" : h_list})
except:
pass
l.append(d)
result_count = res['hits']['total']
if not result_count:
result_count = 0
pp.pprint('result count zero')
result.update({'result_list' : l})
result.update({'result_count' : result_count})
return result
def __init__(self, seed_title, quality=False):
self.seed = seed_title
self.quality = quality
self.es = get_es()
#!/usr/bin/env python
import re
import pprint
import sys
sys.path.append('/home/ec2-user/bblio/build/')
sys.path.append('/home/ec2-user/bblio/aws/')
import os
os.environ['DJANGO_SETTINGS_MODULE'] = 'Build.settings'
import time
from search.models import Document
import es
from io import StringIO, BytesIO
from lxml import etree
es = es.get_es()
_xpath = "body/descendant::*[not(self::script|self::link)]"
_index = "legal-index-html"
_doc_type = "legal-text-html"
def get_tree(text):
text = re.sub('<strong>','',text)
text = re.sub('</strong>','',text)
text = re.sub('\n','',text)
text = re.sub('\t','',text)
text = re.sub(r'\s+',' ',text)
parser = etree.HTMLParser()
tree = etree.parse(StringIO(text.encode('utf-8').decode('utf-8')), parser)
return tree
def get_body_html(text):
