gua2048
/
dupPRdetect
forked from zhangyu19/dupPRdetect
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

修复一些redis的小问题

This commit is contained in:
root 2021-11-11 19:06:34 +08:00
parent 8cd48f5000
commit 2f6f56e11a
3 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
app.py
View File

@ -120,4 +120,4 @@ def output():
if __name__ == '__main__': if __name__ == '__main__':
app.run(host="localhost", debug=True) app.run(host="0.0.0.0", debug=True)

5
detect_trustie.py
View File

@ -1,3 +1,4 @@
from re import T
import requests import requests
import json import json
import jieba import jieba
@ -10,7 +11,7 @@ import math
import redis import redis
token = "" token = ""
r = redis.StrictRedis(host="localhost", port=6379, db=0) r = redis.StrictRedis(host="localhost", port=6379, db=0, decode_responses=True)
def crawl_api(url): def crawl_api(url):
header = {"Authorization": "token %s" % token} header = {"Authorization": "token %s" % token}
@ -73,6 +74,7 @@ def crawl_all_pr(url):
def detect_text_file_dup(all_pr_info, project_key, pr_num): def detect_text_file_dup(all_pr_info, project_key, pr_num):
print(r.get(project_key + "_" + str(pr_num)))
current_pr_value = json.loads(r.get(project_key + "_" + str(pr_num))) current_pr_value = json.loads(r.get(project_key + "_" + str(pr_num)))
current_word = jieba.lcut(current_pr_value["title"] + " " + current_pr_value['description']) current_word = jieba.lcut(current_pr_value["title"] + " " + current_pr_value['description'])
current_word = [x.strip() for x in current_word if x.strip() != ''] current_word = [x.strip() for x in current_word if x.strip() != '']
@ -165,6 +167,7 @@ def main(repo, pr_num, url_head):
file = extract_filename(file_info) # 解析title file = extract_filename(file_info) # 解析title
value = {"title": title, "description": description, "file": file} value = {"title": title, "description": description, "file": file}
r.set(project_pr_key, json.dumps(value)) # 将需要缓存的数据写入redis数据库 r.set(project_pr_key, json.dumps(value)) # 将需要缓存的数据写入redis数据库
else: else:
pass # 跳过，下一步直接获取缓存数据 pass # 跳过，下一步直接获取缓存数据

2
templates/input.html
View File

@ -13,7 +13,7 @@
<!--Just the logo.. Don't mind this--> <!--Just the logo.. Don't mind this-->
<div> <div>
<form class="bs-example bs-example-form" style="width: 600px;" role="form" <form class="bs-example bs-example-form" style="width: 600px;" role="form"
action="http://localhost:5000/result" method="POST" name="input_url"> action="http://106.75.189.31:5000/result" method="POST" name="input_url">
<div class="input-group input-group-lg"> <div class="input-group input-group-lg">
<span class="input-group-addon" style="margin: 20px;"> PR URL </span> <span class="input-group-addon" style="margin: 20px;"> PR URL </span>
<input type="text" class="form-control" placeholder="URL of Source Pull Request" name="SrcPR" <input type="text" class="form-control" placeholder="URL of Source Pull Request" name="SrcPR"