ResourceCollection/flaticon/flaticon.py

# -*- coding: utf-8 -*-
# 共两个步骤, 1, 将目标图片的信息拉到数据库(标题, 所有img的url), 2, 从数据库中读取对应目标站点的所有未下载过的img的url, 下载到本地
# 需要安装psql, 并且 CREATE DATABASE collect; 运行会自动建表
import socket
import sys
import os
import time
import random
from concurrent.futures import ThreadPoolExecutor
import psycopg2

sys.path.append(os.path.join(os.path.abspath(__file__).split('ResourceCollection')[0] + 'ResourceCollection'))
import httpx
from playwright.sync_api import sync_playwright

target = 'flaticon'
step = 4  # 1 = 获取img链接, 2 = 下载图片, 3 = 1 + 2, 4 = 调试
local_proxy = 0
title_selector = '#pack-view__inner > section.pack-view__header > h1'  # 获取标题选择器
img_selector = '#pack-view__inner > section.search-result > ul > li:nth-child({}) > div > a > img'  # 获取图片的url
img_count_selector = '#pack-view__inner > section.pack-view__header > p'  # 获取图片总数选择器
not_find_page_selector = '#viewport > div.errorpage.e404 > h1'  # 当无法获取下一页时, 此选择器为最后一页

project_root = os.path.join(os.path.abspath(__file__).split('ResourceCollection')[0] + 'ResourceCollection')

#  获取局域网ip, 如果不是局域网, 则用公网连接数据库
s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
s.connect(('10.255.255.255', 1))
IP = s.getsockname()[0]
s.close()
if '192.168.100' not in IP:
    psql_params = {
        "host": "home.erhe.link",
        "port": 55434,
        "user": "psql",
        "password": "psql",
        "dbname": "collect"
    }
else:
    psql_params = {
        "host": "192.168.100.146",
        "port": 5434,
        "user": "psql",
        "password": "psql",
        "dbname": "collect"
    }


def open_browser(target_urls):
    # all_data = {}

    link_count = 1
    for target_url in target_urls:
        print(f'\n开始获取 {target_url} 数据, 当前链接是第 {link_count} 个, 共 {len(target_urls)} 个链接')
        link_count += 1
        pages = '/{}'
        urls = []
        title = ''  # 存放当前页面的title
        total_page_count = 0  # 合集总共页数
        with sync_playwright() as playwright:
            if local_proxy:
                browser = playwright.chromium.launch(
                    headless=True,
                    proxy={"server": "http://127.0.0.1:7890"}
                )
            else:
                browser = playwright.chromium.launch(headless=True)

            context = browser.new_context(viewport={'width': 1280, 'height': 700})
            page = context.new_page()

            img_sequence_num = 1
            for page_count in range(1, 999):
                try:
                    goto_url = target_url + pages.format(page_count)
                    page.goto(goto_url, timeout=8000)
                except Exception as e:
                    pass
                    # print(e)
                    # print(f'页面加载失败：url：{goto_url}')

                if page_count == 1:
                    # 获取title
                    page.wait_for_selector(title_selector, state="attached", timeout=10000)
                    title = page.query_selector(title_selector).inner_text()

                    img_count = page.query_selector(img_count_selector).inner_text()
                    img_count = int(img_count.split(' ')[0])

                    invalid_chars = ['<', '>', ':', '"', '/', '\\', '|', '?', '*', '.', '  ', 'Icon Pack ']
                    for char in invalid_chars:
                        title = title.replace(char, '')
                else:
                    try:
                        # 检查一下当前页面是不是 404
                        page.wait_for_selector(not_find_page_selector, state="attached", timeout=2000)
                        total_page_count = page_count - 1
                        break
                    except:
                        pass

                for i in range(1, img_count + 1):
                    # 选择所有的<a>标签
                    elements = page.query_selector_all(img_selector.format(i))

                    # 遍历所有<a>标签，提取href属性
                    for element in elements:
                        src = element.get_attribute('src')
                        if src:
                            src = src.replace('/128/', '/512/')
                            suffix = src.split('.')[-1]
                            sequence = str(img_sequence_num).zfill(3)
                            urls.append({
                                'url': src,
                                'file_title': title,
                                'serial': sequence,
                                'img': f'{title}_{sequence}',
                                'suffix': suffix
                            })
                            img_sequence_num += 1
                            break

            print(f'所有图片URL已获取。总页数: 共 {total_page_count} 页, 总共图片 {len(urls)},  正在写入数据库...')

            page.close()
            browser.close()

            # all_data[title] = urls  # 所有数据爬取完成再存
            save_data({title: urls})
        print(f'{title} 已保存')

    # 获取所有 url 数据之后, 存数据库
    # return all_data


def download_img(load_data, target_file_path):
    print('正在下载图片')
    with ThreadPoolExecutor(max_workers=4) as executor:
        executor.map(single_img_download,
                     [(index, data, load_data, target_file_path) for index, data in enumerate(load_data)])


def single_img_download(args):
    index, data, load_data, target_file_path = args
    # 连接数据库, 准备反写下载状态
    conn = psycopg2.connect(**psql_params)
    cursor = conn.cursor()

    # 如果img文件存在, 即已经下载过, 直接跳过
    id = data['id']
    name = data['name']
    target_site = data['target_site'],
    file_title = data['file_title'].replace(' ', '_')
    set_name = data['set_name']
    serial = str(data['serial']).zfill(3)
    image_suffix = data['image_suffix']
    img_url = data['img_url']

    # 查看每个合集的文件夹是否存在, 不存在就创建
    title_file_path = os.path.join(target_file_path, file_title)
    if not os.path.exists(title_file_path):
        os.mkdir(title_file_path)

    img_name = f'{file_title}_{serial}.{image_suffix}'  # 图片文件名
    img_file_path = os.path.join(str(title_file_path), img_name)  # 图片完整路径

    if os.path.exists(img_file_path):
        #  当此 img 已存在本地时, 在 psql 将数据库状态改为已下载
        query = f"UPDATE {target} SET download_state = %s WHERE id = %s"
        cursor.execute(query, (True, id))
        conn.commit()
        print(f'图片 {img_file_path} 已存在。继续！')
        return

    retry = 8
    while retry:
        try:
            resp = httpx.get(img_url, headers={
                "user-agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36"
            })
            with open(img_file_path, 'wb') as f:
                f.write(resp.content)

            #  下载成功后, 在 psql 将数据库状态改为已下载
            query = f"UPDATE {target} SET download_state = %s WHERE id = %s"
            cursor.execute(query, (True, id))
            conn.commit()

            # 算一下进度
            rate = index / len(load_data) * 100
            print(f'已下载：{img_name}, 当前第 {index + 1} 个, 共 {len(load_data)} 个, 已下载 {rate:.2f}%')
            time.sleep(random.uniform(1, 2))
            break
        except Exception as e:
            print(f'下载图片失败：{img_name}。错误：{e}  重试: {retry}')
            retry -= 1
            time.sleep(random.uniform(3, 5))

    conn.close()


def save_data(data_item):
    conn = psycopg2.connect(**psql_params)
    cursor = conn.cursor()
    for k, v in data_item.items():
        for data in v:
            # 检查img_url是否重复
            cursor.execute("SELECT img_url FROM flaticon WHERE img_url = %s", (data['url'],))
            if cursor.fetchone() is None:
                # 插入数据
                cursor.execute("""
                    INSERT INTO flaticon (name, target_site, file_title, set_name, serial, download_state, image_suffix, img_url)
                    VALUES (%s, %s, %s, %s, %s, %s, %s, %s)
                """, (
                    None,
                    target,
                    data['file_title'],
                    None,
                    data['serial'],
                    False,
                    data['suffix'],
                    data['url']
                ))
                conn.commit()
                # print(f"数据 {data['url']} 保存成功")
            else:
                print(f"数据 {data['url']} 已存在，跳过")

    # 关闭数据库连接
    cursor.close()
    conn.close()


def load_data():
    # 连接数据库
    conn = psycopg2.connect(**psql_params)
    cursor = conn.cursor()

    # 查询download_state为false的所有数据
    query = f"SELECT * FROM {target} WHERE download_state = %s order by id asc"

    load_data_list = []

    try:
        # 执行查询
        cursor.execute(query, (False,))
        # 获取查询结果
        rows = cursor.fetchall()

        # 打印结果
        for row in rows:
            load_data_list.append(
                {
                    'id': row[0],
                    'name': row[1],
                    'target_site': row[2],
                    'file_title': row[3],
                    'set_name': row[4],
                    'serial': row[5],
                    'download_state': row[6],
                    'image_suffix': row[7],
                    'img_url': row[8]
                }
            )
    except psycopg2.Error as e:
        print(f"Database error: {e}")
    finally:
        # 关闭数据库连接
        cursor.close()
        conn.close()

    if load_data_list:
        return load_data_list
    else:
        print("没有需要下载的数据。")
        exit(0)


def check_psql():
    # 连接数据库
    try:
        conn = psycopg2.connect(**psql_params)
    except Exception as e:
        print(f"无法连接到数据库：{e}")
        exit(1)

    # 创建cursor对象
    cur = conn.cursor()

    cur.execute("SELECT EXISTS(SELECT FROM pg_catalog.pg_tables WHERE schemaname = 'public' AND tablename = %s)",
                (target,))
    exist = cur.fetchone()[0]

    if not exist:
        # 如果不存在，则创建表
        cur.execute(f"""
        CREATE TABLE {target} (
            id SERIAL PRIMARY KEY,
            name VARCHAR(255),
            target_site VARCHAR(255),
            file_title VARCHAR(255),
            set_name VARCHAR(255),
            serial INT,
            download_state BOOLEAN,
            image_suffix VARCHAR(50),
            img_url VARCHAR(255)
        );
        """)
        print(f"表 '{target}' 创建成功。")

    # 提交事务
    conn.commit()

    # 关闭cursor和连接
    cur.close()
    conn.close()


def check_local_downloads_dir():
    # 查看一下是否存在 downloads 文件夹, 不存在就创建一个
    download_file_path = os.path.join(str(project_root), 'downloads')
    if not os.path.exists(download_file_path):
        os.mkdir(download_file_path)
    target_file_path = os.path.join(download_file_path, target)
    if not os.path.exists(target_file_path):
        os.mkdir(target_file_path)
    return target_file_path


def check_target_url_txt():
    txt_file_name = 'target_link.txt'
    if not os.path.exists(txt_file_name):
        with open(txt_file_name, 'w') as file:
            file.write('')
        print('需要在 target_link.txt 中填写目标链接')
        exit(0)
    else:
        with open('target_link.txt', 'r') as f:
            targets = [target.strip() for target in f.readlines()]
            if not targets:
                print('在 target_link.txt 中未找到目标链接')
                exit(0)

    return targets


if __name__ == "__main__":
    # 检查数据库
    check_psql()

    if step == 1:
        targets = check_target_url_txt()
        open_browser(targets)
    elif step == 2:
        # 开始读取数据
        load_data = load_data()

        # 开始下载 img
        target_file_path = check_local_downloads_dir()
        download_img(load_data, target_file_path)
        print('下载完成, 程序退出')
    elif step == 3:
        targets = check_target_url_txt()
        #  保存 img 合集链接
        open_browser(targets)

        # 开始读取数据
        load_data = load_data()

        # 开始下载 img
        target_file_path = check_local_downloads_dir()
        download_img(load_data, target_file_path)
        print('下载完成, 程序退出')
    elif step == 4:
        #  调试
        pass
    else:
        pass