'''
Created on 2019年1月3日

@author: User
'''

import psycopg2
import re
import pandas as pd
from BiddingKG.dl.common.Utils import *


#取编号名称数据到excel中
def getCodeNameToExcel():
    conn = psycopg2.connect(dbname="BiddingKG",user="postgres",password="postgres",host="192.168.2.101")
    cursor = conn.cursor()
    sql = " select A.id,A.content,B.projectcode,B.projectname from articles_processed A,project B where A.id=B.doc_id "
    
    cursor.execute(sql)
    
    rows = cursor.fetchall()
    
    list_doc_id = []
    list_codenameflag = []
    list_codename_before = []
    list_codename = []
    list_codename_after = []
    context_len = 10
    for row in rows:
        doc_id = row[0]
        contents = row[1]
        codes = row[2]
        names = row[3]
        for content in re.split("。",contents):
            for code in re.split("[;；]",codes):
                code = code.strip()
                if code=="":
                    continue
                for index in findAllIndex(code, content):
                    list_doc_id.append(doc_id)
                    list_codenameflag.append("code")
                    if index>context_len:
                        list_codename_before.append(content[index-context_len:index])
                    else:
                        list_codename_before.append(content[0:index])
                    list_codename.append(code)
                    if index<len(content)-context_len:
                        list_codename_after.append(content[index+len(code):index+len(code)+context_len])
                    else:
                        list_codename_after.append(content[index+len(code):len(content)])
                    #print(list_codename)
            for code in re.split("[;；]",names):
                code = code.strip()
                if code=="":
                    continue
                for index in findAllIndex(code, content):
                    list_doc_id.append(doc_id)
                    list_codenameflag.append("name")
                    if index>context_len:
                        list_codename_before.append(content[index-context_len:index])
                    else:
                        list_codename_before.append(content[0:index])
                    list_codename.append(code)
                    if index<len(content)-context_len:
                        list_codename_after.append(content[index+len(code):index+len(code)+context_len])
                    else:
                        list_codename_after.append(content[index+len(code):len(content)])
    for i in range(len(list_codename_after)):
        if list_codenameflag[i]=="code" and len(list_codename_after[i])>0 and re.search("\d",list_codename_after[i][0]) is not None:
            #print(i+1,list_codename[i]+list_codename_after[i])
            pass
        
    list_search = []
    for i in range(len(list_codename_after)):
        if list_codenameflag[i]=="name" and len(list_codename_after[i])>3:
            s = re.search("^项目|^施工|^监理|^[0-9A-Za-z一二三四五六七八九十]{1,2}标段|^系统|^设计|^服务",list_codename_after[i][0:4])
            if s is not None:
                list_search.append(list_codename[i]+s.group())
            else:
                list_search.append("")
        else:
            list_search.append("")
                    
    dataframe = pd.DataFrame({'id':list_doc_id,'flag':list_codenameflag,"before":list_codename_before,"value":list_codename,"after":list_codename_after,"search":list_search})
    columns = ['id','flag',"before","value","after","search"]
    dataframe.to_excel("handlabel/codenames.xls",index=False,header=False,encoding="utf8",columns=columns)
    
#保存结果到数据库
def persistenceToDB():
    
    conn = psycopg2.connect(dbname="BiddingKG",user="postgres",password="postgres",host="192.168.2.101")
    cursor = conn.cursor()
    
    file = "handlabel/codenames_tailCompleted.xls"
    pds = pd.read_excel(file)
    list_doc_id = pds['doc_id']
    list_flag = pds['flag']
    list_before = pds['before']
    list_codename = pds['codename']
    list_after = pds['after']
    list_codechange = pds['codechange']
    list_namechange = pds['namechange']
    
    dic = dict()
    for i in range(len(list_doc_id)):
        if list_doc_id[i] not in dic.keys():
            dic[list_doc_id[i]] = [set(),set()]
        if list_flag[i]=="code":
            if str(list_codechange[i])!="nan":
                dic[list_doc_id[i]][0].add(str(list_codechange[i]))
            else:
                if str(list_codename[i])!="nan":
                    dic[list_doc_id[i]][0].add(str(list_codename[i]))
        else:
            if str(list_namechange[i])!="nan":
                dic[list_doc_id[i]][1].add(str(list_namechange[i]))
            else:
                if str(list_codename[i])!="nan":
                    dic[list_doc_id[i]][1].add(str(list_codename[i]))
    for item in dic.keys():
        sql = " insert into project_relabel values('"+str(item)+"','"+";".join(list(dic[item][0]))+"','"+";".join(list(dic[item][1]))+"')"
        cursor.execute(sql)
    '''
    #确认无误之后替换
    update project set projectcode=(select projectcode from project_relabel A where A.doc_id=project.doc_id) where 
    exists(select 1 from project_relabel B where B.doc_id=project.doc_id and B.projectcode!='')
    
    update project set projectname=(select projectname from project_relabel A where A.doc_id=project.doc_id) where
    exists(select 1 from project_relabel B where B.doc_id=project.doc_id and B.projectname!='')
    '''
    conn.commit()
    conn.close()

if __name__=="__main__":
    #getCodeNameToExcel()
    persistenceToDB()