Langchain编程中常见Python库讲解【1】

本文罗列部分依赖的python库代码举例
strsimpy, pytz, python-magic-bin, python-decouple, pyreadline3, pyclipper, pathlib, mpmath, jieba, flatbuffers, filetype, faiss-cpu, brotli, zipp, wrapt, win32-setctime, websockets, watchdog, validators, urllib3, typing-extensions, tornado, toolz, toml, tenacity, tabulate, sympy, soupsieve, socksio, sniffio, smmap, six, simplejson
常见用法
依赖全文参考上一篇https://blog.csdn.net/Climbman/article/details/141509558

jieba 库示例（中文分词）：

import jieba

text = "我爱自然语言处理"
words = jieba.cut(text)
print(list(words))

validators 库示例（数据验证）：

import validators

email = "test@example.com"
is_valid_email = validators.email(email)
print(is_valid_email)

urllib3 库示例（发送 HTTP 请求）：

import urllib3

http = urllib3.PoolManager()
response = http.request('GET', 'https://www.example.com')
print(response.data)

toml 库示例（解析 TOML 格式文件）：

import toml

data = """
[tool.poetry]
name = "myproject"
version = "0.1.0"
"""
parsed_data = toml.loads(data)
print(parsed_data)

sympy 库示例（符号计算）：

import sympy

x = sympy.Symbol('x')
expression = x**2 + 2*x + 1
print(sympy.simplify(expression))

以下是更多库的简单示例代码：

tenacity 库示例（重试机制）：

import tenacity
import random

@tenacity.retry(wait=tenacity.wait_fixed(2), stop=tenacity.stop_after_attempt(3))
def some_function_with_possible_failure():
    random_number = random.randint(1, 10)
    if random_number < 5:
        raise ValueError("Random number is too small.")
    else:
        return "Success!"

try:
    result = some_function_with_possible_failure()
    print(result)
except Exception as e:
    print(f"Failed after retries: {e}")

tabulate 库示例（格式化表格输出）：

from tabulate import tabulate

data = [["Name", "Age"], ["Alice", 25], ["Bob", 30]]
print(tabulate(data))

socksio 库（假设是 socket 相关库）示例：

import socksio

sock = socksio.Socket()
sock.connect(("example.com", 80))
sock.send(b"GET / HTTP/1.1\r\nHost: example.com\r\n\r\n")
response = sock.recv(1024)
print(response)
sock.close()

sniffio 库（假设用于网络嗅探）示例：

import sniffio

async def some_async_function():
    current_async_backend = sniffio.current_async_library()
    print(f"Current async library: {current_async_backend}")

sniffio.run(some_async_function())

smmap 库（假设用于内存映射文件）示例：

import smmap

with open("large_file.txt", "r") as f:
    mm = smmap.mmap(f.fileno(), 0, access=mmap.ACCESS_READ)
    # 可以像处理字节串一样处理这个内存映射对象
    print(mm.read(10))
    mm.close()

six 库示例（用于 Python 2 和 3 的兼容性）：

from six import add_metaclass

class MyMeta(type):
    pass

@add_metaclass(MyMeta)
class MyClass:
    pass

print(MyClass)

simplejson 库示例（JSON 处理）：

import simplejson as json

data = {"name": "John", "age": 30}
json_data = json.dumps(data)
print(json_data)

parsed_data = json.loads(json_data)
print(parsed_data)

flatbuffers 库示例（高效的跨语言序列化库）：

假设你有一个定义好的 FlatBuffers schema 文件 MySchema.fbs，内容如下：

namespace my.namespace;

table Person {
  name: string;
  age: int;
}

生成 Python 绑定后，可以这样使用：

import flatbuffers
from my_package import Person

builder = flatbuffers.Builder(0)
name = builder.CreateString('Alice')
Person.PersonStart(builder)
Person.PersonAddName(builder, name)
Person.PersonAddAge(builder, 30)
person = Person.PersonEnd(builder)
builder.Finish(person)
buf = builder.Output()

parsed_person = Person.Person.GetRootAsPerson(buf)
print(parsed_person.Name())
print(parsed_person.Age())

filetype 库示例（文件类型检测）：

import filetype

kind = filetype.guess('image.jpg')
if kind is not None:
    print(f'File extension: {kind.extension}')
    print(f'File MIME type: {kind.mime}')

faiss-cpu 库示例（高效的相似性搜索库）：

import numpy as np
import faiss

d = 64  # 向量维度
nb = 1000  # 数据库大小
nq = 10  # 查询数量

np.random.seed(1234)
xb = np.random.random((nb, d)).astype('float32')
xq = np.random.random((nq, d)).astype('float32')

index = faiss.IndexFlatL2(d)
index.add(xb)

D, I = index.search(xq, 5)  # 返回每个查询向量的最近 5 个邻居的距离和索引
print(D)
print(I)

brotli 库示例（压缩和解压缩数据）：

import brotli

data = b'Hello, world!'
compressed_data = brotli.compress(data)
decompressed_data = brotli.decompress(compressed_data)
print(decompressed_data)

zipp 库示例（操作 ZIP 归档文件）：

import zipp

archive = zipp.Path('archive.zip')
archive.write_text('file1.txt', 'Content of file 1')
archive.write_text('file2.txt', 'Content of file 2')

for file in archive.iterdir():
    print(file.name)

wrapt 库示例（装饰器相关库）：

import wrapt

@wrapt.decorator
def my_decorator(wrapped, instance, args, kwargs):
    print('Before function call')
    result = wrapped(*args, **kwargs)
    print('After function call')
    return result

@my_decorator
def my_function():
    print('Inside function')

my_function()

win32-setctime 库示例（在 Windows 上设置文件创建时间）：

import win32file
import win32con
import os

file_path = 'test_file.txt'
with open(file_path, 'w') as f:
    f.write('Some content')

creation_time = 1629459200  # 时间戳
handle = win32file.CreateFile(file_path, win32con.GENERIC_WRITE, 0, None, win32con.OPEN_EXISTING, 0, None)
win32file.SetFileTime(handle, creation_time, None, None)
handle.close()

websockets 库示例（WebSockets 通信）：

import asyncio
import websockets

async def echo(websocket, path):
    async for message in websocket:
        await websocket.send(message)

start_server = websockets.serve(echo, 'localhost', 8765)

asyncio.get_event_loop().run_until_complete(start_server)
asyncio.get_event_loop().run_forever()

watchdog 库示例（文件系统监控）：

import time
from watchdog.observers import Observer
from watchdog.events import FileSystemEventHandler

class MyHandler(FileSystemEventHandler):
    def on_created(self, event):
        print(f'File created: {event.src_path}')

    def on_modified(self, event):
        print(f'File modified: {event.src_path}')

if __name__ == "__main__":
    event_handler = MyHandler()
    observer = Observer()
    observer.schedule(event_handler, path='.', recursive=False)
    observer.start()

    try:
        while True:
            time.sleep(1)
    except KeyboardInterrupt:
        observer.stop()

    observer.join()

strsimpy 库（假设是字符串相似度计算库）示例：

# 假设 strsimpy 有一个函数 calculate_similarity 用于计算字符串相似度
from strsimpy import calculate_similarity

str1 = "hello world"
str2 = "hi world"
similarity = calculate_similarity(str1, str2)
print(similarity)

pytz 库示例（时区处理）：

import pytz
from datetime import datetime

tz = pytz.timezone('America/New_York')
now = datetime.now(tz)
print(now)

python-magic-bin 库示例（文件类型识别）：

import magic

file_path = 'test.txt'
file_type = magic.from_file(file_path)
print(file_type)

python-decouple 库示例（配置文件读取）：

假设你有一个配置文件 .env，内容如下：

MY_VARIABLE=some_value

from decouple import config

my_variable = config('MY_VARIABLE')
print(my_variable)

pyreadline3 库示例（增强命令行交互）：

import pyreadline3

# 假设可以用它来实现更好的命令行输入提示
line = pyreadline3.Completer().complete('print(', 0)
print(line)

pyclipper 库示例（多边形裁剪库）：

from pyclipper import PyclipperOffset

coordinates = [(0, 0), (5, 0), (5, 5), (0, 5)]
clipper_offset = PyclipperOffset()
offset_coordinates = clipper_offset.Execute(coordinates, 2)
print(offset_coordinates)

pathlib 库示例（路径操作）：

from pathlib import Path

file_path = Path('/home/user/documents/test.txt')
if file_path.exists():
    print(file_path.name)
else:
    print('File does not exist.')

mpmath 库示例（高精度数学计算）：

from mpmath import mp

mp.dps = 50
pi = mp.pi
print(pi)

toolz 库示例（函数式编程工具）：

from toolz import compose, partial

def add(a, b):
    return a + b

add_five = partial(add, b=5)
double_result = compose(lambda x: x * 2, add_five)

print(double_result(3))

tornado 库示例（异步网络框架）：

import tornado.ioloop
import tornado.web

class MainHandler(tornado.web.RequestHandler):
    def get(self):
        self.write("Hello, Tornado!")

def make_app():
    return tornado.web.Application([
        (r"/", MainHandler),
    ])

if __name__ == "__main__":
    app = make_app()
    app.listen(8888)
    tornado.ioloop.IOLoop.current().start()