FFmpeg截屏与虚拟实验硬件需求

import cv2
import easyocr
import numpy as np
import warnings
import subprocess
import time
import os
warnings.filterwarnings("ignore")

# ========== 全局配置 ==========
# 对话列表的大致区域（根据屏幕分辨率调整，不用精准）
LIST_X1, LIST_X2 = 30, 400    # 列表X范围
LIST_Y1, LIST_Y2 = 80, 1000   # 列表Y范围
OUTPUT_IMG_PATH = "/home/nick/workspace/RPA/real_time_ocr_verify.png"  # 验证图保存路径
SCREENSHOT_PATH = "/home/nick/workspace/RPA/real_time_screenshot.png"   # 实时截屏保存路径

# ========== 初始化EasyOCR ==========
reader = easyocr.Reader(
    lang_list=['ch_sim', 'en'],
    gpu=False,
    verbose=False,
    model_storage_directory="/home/nick/.EasyOCR/model"
)

def capture_screen():
    """实时截取整个屏幕（Linux系统）"""
    # 使用scrot命令截屏（需先安装：sudo apt install scrot）
    subprocess.run(["scrot", SCREENSHOT_PATH], check=True)
    print(f"✅ 已截取屏幕，保存至：{SCREENSHOT_PATH}")
    # 读取截屏图片
    img = cv2.imread(SCREENSHOT_PATH)
    if img is None:
        raise Exception("❌ 无法读取截屏图片")
    return img

def ocr_list_coordinates(img):
    """OCR识别列表项坐标，过滤无效结果"""
    # 裁剪列表区域
    crop_img = img[LIST_Y1:LIST_Y2, LIST_X1:LIST_X2]
    gray = cv2.cvtColor(crop_img, cv2.COLOR_BGR2GRAY)
    
    # OCR识别
    result = reader.readtext(gray, detail=1)
    
    # 过滤+计算坐标
    list_items = []
    img_verify = img.copy()  # 用于可视化标记
    for (bbox, text, conf) in result:
        text = text.strip()
        # 过滤规则：排除纯问号/高问号占比/过短文本
        if all(c == '?' for c in text) or len(text) < 2 or text.count('?')/len(text) > 0.5:
            continue
        # 计算屏幕绝对坐标（无需偏移，因为是实时截屏）
        center_x = int((bbox[0][0] + bbox[2][0])/2 + LIST_X1)
        center_y = int((bbox[0][1] + bbox[2][1])/2 + LIST_Y1)
        # 保存列表项信息（仅保留坐标，文字仅用于标记）
        list_items.append({
            "center": (center_x, center_y),
            "bbox": bbox  # 文本框坐标，用于拖拽选中
        })
        # 可视化标记：红色圆点+文本
        cv2.circle(img_verify, (center_x, center_y), 5, (0,0,255), -1)
        cv2.putText(img_verify, f"({center_x},{center_y})", 
                    (center_x+10, center_y), cv2.FONT_HERSHEY_SIMPLEX, 
                    0.4, (0,0,255), 1)
    
    # 保存验证图
    cv2.imwrite(OUTPUT_IMG_PATH, img_verify)
    print(f"✅ 验证图已保存至：{OUTPUT_IMG_PATH}")
    return list_items

if __name__ == "__main__":
    print("========================================")
    print("📸 按回车键触发实时截屏（请确保目标窗口在前台）")
    print("========================================")
    input()  # 等待用户按回车
    
    try:
        # 1. 实时截屏
        screen_img = capture_screen()
        # 2. OCR识别列表坐标
        list_items = ocr_list_coordinates(screen_img)
        # 3. 输出坐标结果
        print("\n✅ 识别到的列表项坐标（用于后续拷贝）：")
        print("=" * 50)
        for idx, item in enumerate(list_items, 1):
            print(f"列表项 {idx}：中心坐标 {item['center']}")
        print("=" * 50)
        print(f"📊 共识别到 {len(list_items)} 个列表项")
    except Exception as e:
        print(f"❌ 执行失败：{e}")

import subprocess
import time

def get_clipboard_text():
    """读取系统剪贴板内容（Linux）"""
    try:
        # 使用xclip读取剪贴板（需安装：sudo apt install xclip）
        result = subprocess.run(["xclip", "-o", "-selection", "clipboard"], 
                               capture_output=True, text=True, check=True)
        return result.stdout.strip()
    except Exception as e:
        print(f"❌ 读取剪贴板失败：{e}")
        return ""

def copy_list_item_text(item_center, drag_offset=50):
    """
    拖拽选中列表项文字并拷贝
    :param item_center: 列表项中心坐标 (x,y)
    :param drag_offset: 拖拽宽度（左右各偏移50像素）
    """
    x, y = item_center
    # 1. 移动鼠标到列表项左侧（开始拖拽的位置）
    subprocess.run(["xdotool", "mousemove", str(x - drag_offset), str(y)])
    time.sleep(0.1)  # 等待鼠标移动到位
    # 2. 按住左键 → 拖拽到右侧 → 释放左键（选中文字）
    subprocess.run(["xdotool", "mousedown", "1"])  # 按住左键
    time.sleep(0.1)
    subprocess.run(["xdotool", "mousemove", str(x + drag_offset), str(y)])  # 拖拽
    time.sleep(0.1)
    subprocess.run(["xdotool", "mouseup", "1"])  # 释放左键
    # 3. 按下Ctrl+C拷贝
    subprocess.run(["xdotool", "key", "ctrl+c"])
    time.sleep(0.1)
    # 4. 读取剪贴板内容
    copied_text = get_clipboard_text()
    return copied_text

if __name__ == "__main__":
    # ========== 第一步：手动输入（或从step1读取）列表项坐标 ==========
    # 示例：从step1的输出中复制一个列表项坐标
    list_items = [
        (192, 258),  # 2026年人民币换美元新规解读
        (134, 448),  # 蔡文胜介绍
        (173, 524)   # 《西江月》词牌韵律表
    ]
    
    # ========== 第二步：逐个拷贝列表项文字 ==========
    print("========================================")
    print("📋 开始拷贝列表项文字（确保目标窗口在前台）")
    print("========================================")
    time.sleep(2)  # 预留2秒切换到目标窗口
    
    for idx, center in enumerate(list_items, 1):
        print(f"\n🔹 拷贝列表项 {idx}（坐标 {center}）：")
        copied_text = copy_list_item_text(center)
        if copied_text:
            print(f"✅ 拷贝结果：{copied_text}")
        else:
            print(f"❌ 拷贝失败")

import cv2
import easyocr
import numpy as np
import warnings
import subprocess
import time
import os
warnings.filterwarnings("ignore")

# ========== 全局配置（根据你的屏幕/列表调整） ==========
# 对话列表的大致区域（不用精准，OCR会自动识别）
LIST_X1, LIST_X2 = 30, 400    # 列表X范围
LIST_Y1, LIST_Y2 = 80, 1000   # 列表Y范围
OUTPUT_IMG_PATH = "/home/nick/workspace/RPA/real_time_ocr_verify.png"
SCREENSHOT_PATH = "/home/nick/workspace/RPA/real_time_screenshot.png"
DRAG_OFFSET = 100  # 拖拽宽度（覆盖文字长度）
COPY_DELAY = 3     # 拷贝前的窗口切换时间（秒）

# ========== 初始化EasyOCR ==========
reader = easyocr.Reader(
    lang_list=['ch_sim', 'en'],
    gpu=False,
    verbose=False,
    model_storage_directory="/home/nick/.EasyOCR/model"
)

# ========== 函数1：实时截屏 ==========
def capture_screen():
    """截取整个屏幕并返回图片"""
    subprocess.run(["scrot", SCREENSHOT_PATH], check=True, capture_output=True)
    img = cv2.imread(SCREENSHOT_PATH)
    if img is None:
        raise Exception("❌ 无法读取截屏图片")
    print(f"✅ 已截取屏幕：{SCREENSHOT_PATH}")
    return img

# ========== 函数2：OCR识别列表坐标 ==========
def ocr_list_coordinates(img):
    """识别列表项中心坐标，返回坐标列表+带标记的验证图"""
    # 裁剪列表区域
    crop_img = img[LIST_Y1:LIST_Y2, LIST_X1:LIST_X2]
    gray = cv2.cvtColor(crop_img, cv2.COLOR_BGR2GRAY)
    
    # OCR识别
    result = reader.readtext(gray, detail=1)
    
    # 过滤无效结果，提取坐标
    list_coords = []
    img_verify = img.copy()
    for (bbox, text, conf) in result:
        text = text.strip()
        # 过滤规则：排除纯问号/高问号占比/过短文本
        if all(c == '?' for c in text) or len(text) < 2 or text.count('?')/len(text) > 0.5:
            continue
        # 计算屏幕绝对坐标
        center_x = int((bbox[0][0] + bbox[2][0])/2 + LIST_X1)
        center_y = int((bbox[0][1] + bbox[2][1])/2 + LIST_Y1)
        list_coords.append((center_x, center_y))
        # 可视化标记
        cv2.circle(img_verify, (center_x, center_y), 5, (0,0,255), -1)
        cv2.putText(img_verify, f"({center_x},{center_y})", 
                    (center_x+10, center_y), cv2.FONT_HERSHEY_SIMPLEX, 
                    0.4, (0,0,255), 1)
    
    # 保存验证图
    cv2.imwrite(OUTPUT_IMG_PATH, img_verify)
    print(f"✅ 验证图已保存：{OUTPUT_IMG_PATH}")
    print(f"✅ 识别到 {len(list_coords)} 个列表项")
    return list_coords

# ========== 函数3：读取剪贴板 ==========
def get_clipboard_text():
    """读取系统剪贴板内容"""
    try:
        result = subprocess.run(["xclip", "-o", "-selection", "clipboard"], 
                               capture_output=True, text=True, check=True)
        return result.stdout.strip()
    except:
        return ""

# ========== 函数4：拖拽选中+拷贝文字 ==========
def copy_list_item_text(center_coords):
    """批量拷贝列表项文字"""
    print(f"\n========================================")
    print(f"📋 {COPY_DELAY}秒后开始拷贝，请立刻切回目标窗口！")
    print(f"========================================")
    time.sleep(COPY_DELAY)
    
    copied_results = []
    for idx, (x, y) in enumerate(center_coords, 1):
        print(f"\n🔹 处理列表项 {idx}（坐标：{x},{y}）：")
        # 1. 移动到文字左侧
        subprocess.run(["xdotool", "mousemove", str(x - DRAG_OFFSET), str(y)])
        time.sleep(0.2)
        # 2. 拖拽选中文字
        subprocess.run(["xdotool", "mousedown", "1"])
        time.sleep(0.2)
        subprocess.run(["xdotool", "mousemove", str(x + DRAG_OFFSET), str(y)])
        time.sleep(0.2)
        subprocess.run(["xdotool", "mouseup", "1"])
        time.sleep(0.2)
        # 3. Ctrl+C拷贝
        subprocess.run(["xdotool", "key", "ctrl+c"])
        time.sleep(0.2)
        # 4. 读取并保存结果
        text = get_clipboard_text()
        copied_results.append({
            "index": idx,
            "coords": (x, y),
            "text": text
        })
        # 输出结果
        if text:
            print(f"✅ 拷贝成功：{text}")
        else:
            print(f"❌ 拷贝失败")
    
    return copied_results

# ========== 主函数：串联所有操作 ==========
def main():
    try:
        # 步骤1：等待用户触发截屏
        print("========================================")
        print("📸 按回车键触发实时截屏（确保目标窗口在前台）")
        print("========================================")
        input()
        
        # 步骤2：实时截屏
        screen_img = capture_screen()
        
        # 步骤3：OCR识别列表坐标
        list_coords = ocr_list_coordinates(screen_img)
        if not list_coords:
            print("❌ 未识别到任何列表项")
            return
        
        # 步骤4：自动拷贝所有列表项文字
        copy_list_item_text(list_coords)
        
        print("\n🎉 所有操作完成！验证图可查看坐标标记，拷贝结果已输出。")
        
    except Exception as e:
        print(f"\n❌ 执行失败：{str(e)}")

if __name__ == "__main__":
    main()

import subprocess
import time

def get_clipboard_text():
    """读取系统剪贴板内容（Linux）"""
    try:
        result = subprocess.run(["xclip", "-o", "-selection", "clipboard"], 
                               capture_output=True, text=True, check=True)
        return result.stdout.strip()
    except Exception as e:
        return ""

def copy_list_item_text(item_center, drag_offset=100):
    """
    优化版：拖拽选中 + Ctrl+C 拷贝
    :param item_center: 实时OCR识别的中心坐标 (x,y)
    :param drag_offset: 拖拽宽度（放大到100，确保覆盖文字）
    """
    x, y = item_center
    # 1. 移动到文字左侧
    subprocess.run(["xdotool", "mousemove", str(x - drag_offset), str(y)])
    time.sleep(0.2)
    # 2. 按住左键拖拽到右侧
    subprocess.run(["xdotool", "mousedown", "1"])
    time.sleep(0.2)
    subprocess.run(["xdotool", "mousemove", str(x + drag_offset), str(y)])
    time.sleep(0.2)
    subprocess.run(["xdotool", "mouseup", "1"])
    time.sleep(0.2)
    # 3. Ctrl+C 拷贝
    subprocess.run(["xdotool", "key", "ctrl+c"])
    time.sleep(0.2)
    # 4. 读取剪贴板
    return get_clipboard_text()

if __name__ == "__main__":
    # ========== 关键：用实时OCR输出的新坐标！ ==========
    list_items = [
        (199, 144),   # 列表项1（实时OCR结果）
        (199, 197),   # 列表项2
        (173, 234),   # 列表项3
        (259, 257),   # 列表项4
    ]
    
    print("========================================")
    print("📋 3秒后开始拷贝，请立刻切回浏览器窗口！")
    print("========================================")
    time.sleep(3)  # 延长到3秒，确保切回窗口
    
    for idx, center in enumerate(list_items, 1):
        print(f"\n🔹 拷贝列表项 {idx}（坐标 {center}）：")
        copied_text = copy_list_item_text(center)
        if copied_text:
            print(f"✅ 拷贝成功：{copied_text}")
        else:
            print(f"❌ 拷贝失败，请检查坐标/窗口焦点")