用Python和Flask打造属于自己的API

1827 字

9 分钟

用Python和Flask打造属于自己的API

2022-04-16

Tech

/

Python

/

Flask

/

API

/

Host

先上api链接：https://api.ninym.top

文档链接：https://ninym.top

快速开始#

环境安装#

首先既然是用python+flask，这两个东西肯定要装好的嘛~ Python的安装我就不讲了，主要说下flask

很简单，安装命令就一行

Linux#

1
pip3 install flask

Windows#

1
pip install flask

因为我这里已经安装过了，所以写的是Requirement already satisfied，如果没有安装过的话会进入安装状态

# Hello World#

既然要用flask框架，那就首先要引用，在python中，我们可以使用import语句来引用外部模块

1
from flask import Flask

当然可以在后面加上as xxx给它取个别名（特别是对于那些名字很长的modules，下面会讲）

接下来我们创建我们的第一个flask程序

1
from flask import Flask
2

3
app = Flask(__name__)
4
@app.route('/', methods=['GET'])
5
def parser():
6
    return '<h1>Hello World!</h1>'
7

8
if __name__ == '__main__':
9
    app.run(host='0.0.0.0', port=8080, debug=False)

然后我们直接运行命令python <file>.py就可以运行我们的服务器了（Linux请使用python3 <file>.py，下面不再赘述，只讲Windows，Linux用户请自行替换）

然后我们在浏览器访问我们的网站，在这里可以使用127.0.0.1:8080访问，因为服务器就是部署在本机器上面的，打开后就会显示Hello World了

注：图标是我的浏览器本地缓存，正常情况下网站是没有图标的（因为没有设置）

太棒啦，你现在已经学会创建Flask程序啦！

我的Flask程序#

网易云音乐下载#

我在玩这个东西的过程中，主要是想把我的网易云音乐的API给重制一下（之前是用JavaScript写的烂代码，而且不会自动重命名，就想把这个功能完善一下），所以在我住院的期间，我就开始干起了这个东西。

首先先新建一个utils文件夹，里面放各种自制模块，然后新建了NeteaseCloudMusic.py来写我的这部分的代码

1
import requests as r
2
import bs4
3
import os
4
import re
5

6

7
def NeteaseDownload(id, ContentType):
8
    if ContentType == 'attachment':
9
        song, author = FileInfo(id)
10
        # If the song id doesn't map to a song, this will return.
11
        if song == None and author == None:
12
            return False, 'Not Found', 'Not Found'
13
        base = 'https://music.163.com/song/media/outer/url?id='
14
        file = './cache/' + id + '.mp3'
15
        filename = id+'.mp3'
16
        if os.path.exists(file):
17
            return filename, song, author
18
        try:
19
            stream = r.get(base + id, timeout=30)
20
        except TimeoutError:
21
            return False, None, None
22
        if 'u-errlg u-errlg-404' in stream.text:    # Some VIP songs cannot be downloaded
23
            print(
24
                '[NETEASEMUSICDOWNLOAD] Failed while getting the resources and information of a song.')
25
            return False, 404, 404
26
        print('[NETEASEMUSICDOWNLOAD] Getting song {} - {}, returned status_code {}'.format(
27
            author, song, stream.status_code))
28
        with open(file, 'wb') as f:
29
            f.write(stream.content)
30
        return filename, song, author
31
    else:
32
        base = 'https://music.163.com/song/media/outer/url?id='
33
        song, author = FileInfo(id)
34
        if song == None and author == None:
35
            return {'code': -2, 'msg': 'It seems that id {} is not a valid song id.'.format(id)}
36
        try:
37
            stream = r.get(base + id, timeout=30)
38
        except TimeoutError:
39
            return {'code': -3, 'msg': 'Timed out. Please try again later.'}
40
        if stream.url == "https://music.163.com/404":
41
            return {'code': 500, 'link': None, 'name': song, 'author': author, 'msg': 'Cannot fetch download link.'}
42
        return {'code': 200, 'link': stream.url, 'name': song, 'author': author, 'msg': 'Success'}
43

44

45
def FileInfo(id):   # Get the information of the song, including name and author
46
    TargetWebInfo = r.get('https://music.163.com/song?id=' + id)
47
    html = TargetWebInfo.content
48
    bf = bs4.BeautifulSoup(html, "lxml")
49
    try:
50
        song = str(list(bf.find_all('em', class_="f-ff2"))[0])
51
        song = song.replace('<em class="f-ff2">', '').replace('</em>', '')
52
        ForbiddenCharacters = [('\\', ' '), ('/', ' '), (':', '：'),
53
                               ('*', ' '), ('?', '？'), ('<', ' '), ('>', ' '), ('|', '丨')]
54
        for i in ForbiddenCharacters:
55
            song = song.replace(i[0], i[1])
56
    except:
57
        song = None
58
    try:
59
        author = str(list(bf.find_all('a', class_='s-fc7'))[1])
60
        ReplaceLink = re.findall(r'<a class="s-fc7" href=".+">', author)
61
        for i in ReplaceLink:
62
            author = author.replace(i, '')
63
        author = author.replace('</a>', '')
64
        if author == '${escape(x.beRepliedUser.nickname)}':
65
            author = None
66
    except:
67
        author = None
68
    return song, author

首先因为涉及到网络，所以就引入了一个requests模块（精简版的utllib），然后把我原来在JavaScript写的逻辑搬过来，就形成了NeteaseDownload这个函数，那就剩下重命名没搞定。因为手动解析网易云音乐的网页很麻烦，所以就用了一个叫做beautifulsoup4的模块来帮助我解析。在这里直接搜索<em>元素，限定class为f-ff2，这样就可以找到歌曲名；在用re寻找<a>标签下class为<s-fc7>的内容，这些元素和class都是通过分析网易云音乐网站的源代码获得的。当然当用户输入了非法的id的时候，在网易云上就无法获取到歌曲信息，这时候通过搜索作者中内容为${escape(x.beRepliedUser.nickname)}的情况就可以判断，向用户返回非法id的信息

接着在主程序中注册网易云音乐的解析器，给网易云音乐的API设定路径即可

1
from utils.NeteaseCloudMusic import NeteaseDownload
2
@app.route('/<query>', methods=['GET']) # First path handler
3
def parser(query):
4
    paths = ['song']  # All requests paths
5
    path = query.split('/')
6
    parameter = path[0]
7
    if parameter not in paths:   # When the path not exists, this will return 404
8
        abort(404)
9
    if parameter == 'song':
10
        id = request.args.get('id')
11
        ContentType = request.args.get('type')
12
        return NeteaseHandler(id, ContentType)

Github RELEASE下载次数饼状图#

这个用到了matplotlib.pyplot这个模块，上面说这种模块名字很长的就可以给它命个别名，所以在引入的时候就给它改了个名字

1
import matplotlib.pyplot as plt
2
import matplotlib # 本体，在后面颜色的设定用到了

然后就是疯狂地写代码，Github的下载次数是有API可以获取的（不过有QPS就是了）

从Github获取次数后进行json解析，把RELEASE版本信息和次数提取出来，接着放到图片里就可以了，总之就是写成了下面这样

1
import requests as r
2
import matplotlib.pyplot as plt
3
import matplotlib
4
import numpy as np
5
import json
6
import flask
7

8
ReleaseBase = 'https://api.github.com/repos/'
9
domain = 'dev.api.ninym.top'
10

11
def ghParser(operation, author, repo, ContentType):
12
    operations = ['release']
13
    Error404 = {
14
        'code': 404,
15
        'msg': 'Invalid path gh/{}'.format(operation)
16
    }
17
    Error501 = {
18
        'code': 501,
19
        'msg': 'It seems that this repo has no any release'.format(operation)
20
    }
21
    if operation not in operations: return json.dump(Error404)
22
    if operation == 'release':
23
        release = Release(author,repo,ContentType)
24
        if release.isEmpty:
25
            return json.dumps(Error501)
26
        return release.Draw()
27

28
class Release():
29
    def __init__(self, author, repo, ContentType):
30
        self.author = author
31
        self.repo = repo
32
        self.info = json.loads(r.get(ReleaseBase+self.author+'/'+self.repo+'/releases').text)
33
        self.ContentType = ContentType
34
        if len(self.info) == 0: self.isEmpty = True
35
        else: self.isEmpty = False
36

37
    def getInfo(self):  # For debugging
38
        return self.info
39

40
    def makeDimensionList(self):
41
        self.ReleaseLabel = []
42
        self.ReleaseDownloads = []
43
        for i in self.info:
44
            Downloads = 0
45
            TagName = i['tag_name']
46
            assets = i['assets']
47
            if len(assets) == 0:
48
                self.ReleaseLabel = None
49
                self.ReleaseDownloads = None
50
            else:
51
                for j in assets:
52
                    Downloads += j['download_count']
53
                self.ReleaseLabel.append(TagName)
54
                self.ReleaseDownloads.append(Downloads)
55

56

57
    def Draw(self):
58
        self.makeDimensionList()
59
        plt.figure(figsize=(10,10)) # Make the circle to be a formal one
60
        matplotlib.rcParams.update({'font.size': 20})   # Change the font size of title
61
        if self.ReleaseDownloads == None and self.ReleaseLabel == None:
62
            values = [100]
63
            plt.title('No Any Downloadable Asset from {}/{}'.format(self.author,self.repo))
64
            r = np.linspace(106,142,1,dtype=np.uint8)
65
            g = np.linspace(103,140,1,dtype=np.uint8)
66
            b = np.linspace(232,216,1,dtype=np.uint8)
67
            colors = ['#'+'{:0>2}{:0>2}{:0>2}'.format(str(hex(r[i])),str(hex(g[i])),str(hex(b[i]))).replace('0x','') for i in range(len(r))]
68
            plt.pie(values,colors=colors)
69
            plt.savefig('./cache/{}-{}.png'.format(self.author,self.repo))
70
            if self.ContentType == 'pic': return flask.send_from_directory('./cache/', '{}-{}.png'.format(self.author,self.repo), as_attachment=False, download_name='{} - {}.png'.format(self.author, self.repo))
71
            else:
72
                dt = {
73
                    'author': self.author,
74
                    'repo': self.repo,
75
                    'labels': self.ReleaseLabel,
76
                    'downloads': self.ReleaseDownloads,
77
                    'pic': 'https://{}/cache/{}-{}.png'.format(domain,self.author,self.repo),
78
                    'remark': 'The access to the pic will be removed when the service is redeployed or by the operation of administrator'
79
                }
80
                return json.dumps(dt)
81
        explode = [0.01] * len(self.ReleaseLabel)
82
        r = np.linspace(106,142,len(self.ReleaseDownloads),dtype=np.uint8)
83
        g = np.linspace(103,140,len(self.ReleaseDownloads),dtype=np.uint8)
84
        b = np.linspace(232,216,len(self.ReleaseDownloads),dtype=np.uint8)
85
        colors = ['#'+'{:0>2}{:0>2}{:0>2}'.format(str(hex(r[i])),str(hex(g[i])),str(hex(b[i]))).replace('0x','') for i in range(len(r))]
86
        plt.pie(self.ReleaseDownloads,explode=explode,labels=self.ReleaseLabel,autopct='%1.1f%%',colors=colors)
87
        plt.title('Downloads of {}/{}'.format(self.author,self.repo))
88
        plt.savefig('./cache/{}-{}.png'.format(self.author,self.repo))
89
        if self.ContentType == 'pic': return flask.send_from_directory('./cache/', '{}-{}.png'.format(self.author,self.repo), as_attachment=False, download_name='{} - {}.png'.format(self.author, self.repo))
90
        else:
91
            dt = {
92
                'author': self.author,
93
                'repo': self.repo,
94
                'labels': self.ReleaseLabel,
95
                'downloads': self.ReleaseDownloads,
96
                'pic': 'https://{}/cache/{}-{}.png'.format(domain,self.author,self.repo),
97
                'remark': 'The access to the pic will be removed when the service is redeployed or by the operation of administrator'
98
            }
99
            return json.dumps(dt)

然后还是一样在主程序中注册路径即可

1
from utils.Github import ghParser
2

3
@app.route('/gh/<operation>', methods=['GET'])   # Github Handler
4
def ghHandler(operation):
5
    Analytics(request)
6
    author = request.args.get('author')
7
    repo = request.args.get('repo')
8
    ContentType = request.args.get('type')
9
    if ContentType != 'pic' and ContentType != 'json':
10
        ContentType = 'pic'
11
    return ghParser(operation, author, repo, ContentType)