为了更好的理解某些单词的意思 ， 需要定位它最恰当使用的上下文， 所以 把几千部文学作品放进elasticsearch, 搜索这个单词， 就可以找到这个上下文

article.py

import requests
import os
import json
import sys
import pdb
import subprocess

headers = {
        "Content-Type": "application/json"
    }

BASE_URL = "http://192.168.2.113:9200/article"
def post_to_es(title, article):

    res = requests.post(BASE_URL+"/_doc", headers=headers, data=json.dumps({
        "filename": title,
        "content": article.decode("gbk", "ignore"),
    }))
    print res.text


def query(word):
    filenames = []
    print word
    ret = requests.get(BASE_URL+"/_search", params={
        "q": '"%s"' % word,
    }, headers=headers).json()
    for hit in ret["hits"]["hits"]:
        filename =  hit["_source"]["filename"]
        print hit["_score"], filename
        filenames.append(filename)
    print filenames[0]
    p = subprocess.Popen([
        'C:\\Program Files (x86)\\Notepad++\\notepad++.exe', filenames[0]])
    p.wait()


def create_index():
    content = {
        "mappings": {
            "_doc": {
                "properties": {
                    "filename": {
                        "type": "text"
                    },
                    "content": {
                        "type": "text"
                    }
                }
            }
        }
    }
    res = requests.put(BASE_URL, data=json.dumps(content), headers=headers)
    print res.text

def send_docs():
    for base, dirs, files in os.walk("."):
        for file in files:
            if file.lower().endswith(".txt"):
                fullpath = os.path.join(base, file)
                print fullpath
                post_to_es(fullpath, open(fullpath).read())

if __name__ == "__main__":
    
    arg1 = sys.argv[1]
    if arg1 == "index":
        create_index()
    elif arg1 == "docs":
        send_docs()
    elif arg1 == "del":
        requests.delete(BASE_URL)
    elif arg1 == "q":
        query(sys.argv[2])
    else:
        print "unknown"