#!/usr/bin/env/python
# -*- coding:utf-8 -*-
import logging
from flask import Flask, render_template
from flask import request, redirect, Response
from flask_cors import *
from multiprocessing import Process, Queue
from server_tools import *
from six.moves import urllib
logger = logging.getLogger(__name__)
logger.setLevel(level=logging.INFO)
log_file = os.path.join(r'./logs', 'parse_log.txt') # 日志地址
handler = logging.FileHandler(log_file, mode='a', encoding='utf-8', delay=True)
handler.setLevel(logging.INFO)
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
handler.setFormatter(formatter)
# console = logging.StreamHandler()
# console.setLevel(logging.INFO)
# console.setFormatter(formatter)
logger.addHandler(handler)
# logger.addHandler(console)
app = Flask(__name__)
app.debug = True
CORS(app, supports_credentials=True)
'''
errcode=0成功
errcode=1失败
errmsg是具体的失败消息
'''
# 定义上传路径
UPLOAD_FOLDER = r'D:\zwj\word_uploads' # wordbin服务生成文件路径
if not os.path.isdir(UPLOAD_FOLDER):
os.makedirs(UPLOAD_FOLDER)
# --------------------------------------------------------------
@app.route('/')
def index():
return redirect("/all_subject_parse")
# 文件上传的HTML模板，这里没有额外去写html模板了，直接写在这里，方便点吧
@app.route('/all_subject_parse')
def upload():
return '''
欢迎使用全学科试题word解析
'''
def read(cache):
while True:
filename, filename_root, callback_url, flag, sid = cache.get(True) # 缓存
print("\n+++++++++++++++还有{}缓存解析+++++++++++++++\n".format(cache.qsize() - 1))
print("\n------callback_url===> {} -------\n".format(callback_url))
logger.info('开始解析文件信息：Get filename={}, filename_root={}, callback_url={}, is_cloud ={},sid ={} from queue\n'
.format(filename, filename_root, callback_url, flag, sid))
# 单线程解析
stime1 = time.time()
try:
res, wordbin_time, paper_type, img_upload_time = parse_word(filename, filename_root, flag, sid)
etime1 = time.time()
logger.info("----wordbin服务所占时间：{}".format(wordbin_time))
logger.info("----整个解析所占时间：{}".format(etime1 - stime1))
logger.info("----该份试卷的格式类型：{}".format(paper_type))
if img_upload_time:
logger.info("----图片上传时间：{}".format(img_upload_time))
except:
# print(traceback.print_exc())
print("--------have callback_url,but time out of parse_word-------")
res = {"time_out": 90,
"id": 0,
"errcode": 1,
"errmsgs": "解析超时，若文件太大，请将文件拆开、分两次上传！"}
# "errmsg": "word读取失败。请尝试：\n 1.请将word中图片改为嵌入式，然后上传。\n 2. 请将word题目拆分为多个word上传。\n 3.请将word中内容复制到新word上传。"
print("------------parse is fail---------------")
try:
headers = {'Content-Type': 'application/json', }
# print(json.dumps(res).encode("utf-8"))
# print(res)
r = requests.post(callback_url,
# json=res, # 可以,但是会进行转义
data=json.dumps(res).encode("utf-8"),
headers=headers)
# timeout=10
print("\n------post 回调地址===> {} -------\n".format(r.status_code))
print("\n------post 返回结果===> {} -------\n".format(r.text))
except TimeoutError:
print("回调超时")
except Exception as e:
print(e, "------回调出错")
if "items" in res and res["items"]:
if re.search(r"wid[/=](\d+)/?", str(callback_url)):
upload_id = re.search(r"wid[/=](\d+)/?", str(callback_url)).group(1)
save_fname = save_post_file(res, filename, upload_id) # 保存返回的解析结果
else:
save_fname = save_post_file(res, filename, '0000')
logger.info("\n----{}解析完保存的数据文件名：\n{}\n".format(os.path.basename(filename), save_fname))
# noinspection PyTypeChecker
@app.route('/all_subject_parse', methods=['POST'])
def do_upload():
"""
该接口函数专门用来接收word和参数，并保存到本地，将保存后的数据放入缓存中；
参数关键字为：mydata，callback_url,is_cloud,sid
# 文件上传，overwrite=True为覆盖原有的文件，是bottle的用法f.raw_filename
# 如果不加这参数，当服务器已存在同名文件时，将返回“IOError: File exists.”错误
:return:
"""
img_file_count = 0
if os.listdir(UPLOAD_FOLDER):
img_file_count = max([int(i) for i in os.listdir(UPLOAD_FOLDER)]) + 1
logger.info("==接收文件：request.POST.dict==>{}".format(request.form.to_dict())) # php传过来的文件
print("==request.POST.dict==>{}".format(request.form.to_dict()))
callback_url = request.form.get('callback_url', "")
flag = request.form.get('is_cloud', "0") # word中的图片是否上传ucloud
sid = request.form.get('sid', "0") # 学管端、教师端参数
# word文件接收
# 旧接口：整个word文件传过来，主要应用在云题库端上传
try:
upfile = request.files.get('mydata')
print("\n------接受文件名==> {}-------\n".format(os.path.basename(upfile.filename)))
if os.path.splitext(upfile.filename)[1] not in [".doc", ".docx"]:
return "only accept .doc .docx files"
file_root1 = os.path.join(UPLOAD_FOLDER, str(img_file_count)) # 每份word保存的文件夹路径
if not os.path.isdir(file_root1):
os.makedirs(file_root1)
new_filename = str(img_file_count) + os.path.splitext(upfile.filename)[1]
filename = os.path.join(file_root1, new_filename) # 文件的绝对路径, 不会含有汉字
upfile.save(filename) # 保存到当地
except:
# 新接口：mydata传过来的是文件地址(json格式)，需另外下载
# j_data = json.loads(request.json.get('mydata'))
up_para = request.form.get('mydata')
print("up_para:", up_para)
upfile_url = json.loads(up_para)[0]["file_url"] # word文件的绝对路径
print("文件地址：", upfile_url)
print("\n------接受文件名==> {}-------\n".format(os.path.basename(upfile_url)))
if os.path.splitext(upfile_url)[1] not in [".doc", ".docx"]:
return "only accept .doc .docx files"
file_root1 = os.path.join(UPLOAD_FOLDER, str(img_file_count)) # 每份word保存的文件夹路径
if not os.path.isdir(file_root1):
os.makedirs(file_root1)
# 根据url地址下载word文件
new_filename = str(img_file_count) + os.path.splitext(upfile_url)[1]
filename = os.path.join(file_root1, new_filename) # 文件的绝对路径, 不会含有汉字
try:
urllib.request.urlretrieve(upfile_url, filename) # 下载到指定位置
except:
return "word文件下载失败，请重新上传"
print('------word Successfully downloaded-----')
# ----------------------------------------------------------------------
# 开一个进程对filename 进行解析
if callback_url:
cache_file.put([filename, file_root1, callback_url, flag, sid])
print("-----当前缓存还有{}-----".format(int(cache_file.qsize()) - 1))
return json.dumps({"errcode": 0, "errmsg": "OK", "docsbefore": int(cache_file.qsize()) - 1},
ensure_ascii=False).encode("utf-8")
else: # 没有回调地址时，简易网页上传
stime1 = time.time()
try:
res, wordbin_time, paper_type, img_upload_time = parse_word(filename, None)
etime1 = time.time()
print("------------简易接口上传,解析成功---------------")
logger.info("\n------------简易接口上传,解析成功---------------")
logger.info("\n开始解析文件：{}".format(filename))
logger.info("\n----wordbin服务所占时间：{}".format(wordbin_time))
logger.info("\n----整个解析所占时间：{}".format(etime1 - stime1))
logger.info("\n----该份试卷的格式类型：{}".format(paper_type))
except:
print("------------简易接口上传,解析失败---------------")
res = {"time_out": 3,
"errcode": 1,
"errmsg": "word读取失败。请尝试：\n 1.请确定试卷版面格式是否为常见的试卷排版格式,删除不规范格式;\n"
"2.请将word中内容复制到新word上传;\n 3. 请将word题目拆分为多个word上传。\n 4.请检查图片格式,将word中图片改为嵌入式后再上传。"}
if "items" in res and res["items"]:
save_fname = save_post_file(res, filename, '0011')
print("\n----{}解析完保存的数据文件名：\n{}".format(os.path.basename(filename), save_fname))
new_res = []
for r in res["items"]:
dd = [r["item_id"], r["type"], r["stem"],
r["key"], r["analysis"],r["text_errmsgs"]]
if "options" in r:
r["options"] = [chr(65+k)+'、'+ opt for k, opt in enumerate(r["options"])]
dd[2] += "
===================
【选项】
" \
+ ";
".join(r["options"]).replace("
;", ";")
new_res.append(dd)
import pandas as pd
resdf = pd.DataFrame(new_res)
resdf.columns = ["题号", "题型", "题干", "答案", "解析","错误信息"]
pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', None)
pd.set_option('expand_frame_repr', True) # 允许换行
pd.set_option('display.width', -1)
pd.set_option('display.max_colwidth', -1) # 自动最大列宽
# 如何将图片显示出来:图片属性格式要符合前端显示标准
return render_template('res.html',
items=[resdf.to_html(header=True,
index=False,
escape=False,
justify="center")],
res_json=json.dumps(res, ensure_ascii=False)
)
else:
# resp = Response_headers(json.dumps(res, ensure_ascii=False).encode("utf-8")) # , indent=4
save_fname = save_post_file(res, filename, '0011')
print("\n----{}解析完保存的数据文件名：\n{}".format(os.path.basename(filename), save_fname))
resp = json.dumps(res, ensure_ascii=False)
return resp
@app.route('/ser_static/', methods=["GET"])
def ser_static(file_path): # endpoint的位置是函数接口名，不能用static，与flask内部变量重名
"""
将本地图片供php调用，上传Ucloud时该函数不使用
:param file_path: 图片的本地绝对路径
:return:
"""
# filepath = request.args.get('q') # 针对/**/=？类型
file = os.path.join(UPLOAD_FOLDER, file_path)
if "image" in file:
resize_img(file, file)
# resp = Response(base64.b64encode(open(file, 'rb').read()).decode()) # 仍无法显示图片
resp = Response(open(file, 'rb'))
return resp
if __name__ == "__main__":
# 解析文件的队列
cache_file = Queue()
start_word2html_app(kill_mathtype=True)
# 读取解析进程
pr_parse = Process(target=read, args=(cache_file,))
pr_parse.start()
print("parse pid:", pr_parse.pid)
app.run(host="0.0.0.0", port=11088, threaded=True, debug=True)