python 操作solr索引数据
#-*- coding: UTF-8 -*-import csv
import os
import codecs
def ReadCSV(filename):
if os.path.exists(filename):
with open(filename, 'r') as f:
reader = csv.reader(f)
list = []
for item in reader:
list.append(item)
return list
#################################################
#coding=utf-8
import json
import requests
import os
import time
from os import walk
import CSVOP
from datetime import datetime
import pysolr
import math
class SolrClientObj:
def test(self):
data = {"add": {"doc": {"id": "100001", "*字段名*": u"我是一个大好人"}}}
params = {"boost": 1.0, "overwrite": "true", "commitWithin": 1000}
url = 'http://127.0.0.1:8983/solr/mycore/update?wt=json'
headers = {"Content-Type": "application/json"}
r = requests.post(url, json=data, params=params, headers=headers)
print r.text
def pysolr_Index_Data_FromCSV(self, csvfile,url='http://127.0.0.1:8983/solr/mycore/'):
'''
从CSV文件中读取数据，并索引到solr中
:param csvfile: csv文件，包括完整路径
:return:
'''
list = CSVOP.ReadCSV(csvfile)
index = 0
listdocs = []
for item in list:
if index > 0:# 第一行是标题
doc = {}
try:
doc['title'] = item.decode('GB2312')
doc['link'] = item
# doc['date'] = item
doc['source'] = item.decode('GB2312')
doc['keyword'] = item.decode('GB2312')
listdocs.append(doc)
except Exception,e:
print e.message
index += 1
solr = pysolr.Solr(url, timeout=10)
result = solr.add(listdocs)
print result
def Index_Data_FromCSV(self, csvfile):
'''
从CSV文件中读取数据，并索引到solr中
:param csvfile: csv文件，包括完整路径
:return:
'''
list = CSVOP.ReadCSV(csvfile)
index = 0
doc = {}
params = {"boost": 1.0, "overwrite": "true", "commitWithin": 1000}
url = 'http://127.0.0.1:8983/solr/mycore/update?wt=json'
headers = {"Content-Type": "application/json"}
for item in list:
if index > 0:# 第一行是标题
try:
doc['title'] = item.decode('GB2312')
doc['link'] = item
# doc['date'] = item
doc['source'] = item.decode('GB2312')
doc['keyword'] = item.decode('GB2312')
data = {"add": {"doc": doc}}
r = requests.post(url, json=data, params=params, headers=headers)
print r.text
except Exception,e:
print e.message
print index
index += 1
def Index_data(self):
solr = pysolr.Solr('http://127.0.0.1:8983/solr/mycore/', timeout=10)
# How you'd index data.
result = solr.add([
{
"id": "doc_1",
"title": "A test document",
},
{
"id": "doc_2",
"title": "The Banana: Tasty or Dangerous?",
},
])
print result
def search_data(self,where='视频',url='http://127.0.0.1:8983/solr/mycore/'):
solr = pysolr.Solr(url, timeout=10)
dict = {'start':10,'rows': 30,'fl':'title,keyword,source,link'}
result = solr.search('title:视频',**dict)
# result = solr.search('title:视频')
# print result.raw_response['response']['numFound']
for item in result:
print 'keyword: %s'% item['keyword']
print 'title: %s'% item['title']
print 'source: %s'% item['source']
print 'link: %s'% item['link']
print ' '
def delete_index_data(self,where,url='http://127.0.0.1:8983/solr/mycore/'):
'''
删除索引
:param where: 删除的条件
:param url: url
:return:
'''
solr = pysolr.Solr(url, timeout=10)
# solr.delete(id=where) #id='id1':删除id为“id1”的索引
result = solr.delete(q=where) #q='*:*'删除所有索引
print result
obj = SolrClientObj()
# obj.delete_index_data('*:*') #删除所有索引
# obj.Index_data()
# obj.search_data()
# obj.delete_index_data('doc_1')
obj.search_data('视频')
# csvfile = 'D:/work/Solr/other/exportExcels/2017-07-07_info.csv'
# obj.pysolr_Index_Data_FromCSV(csvfile)
页:
[1]