import os
from openai import OpenAI

# 初始化客戶端
client = OpenAI(
    base_url="https://ark.cn-beijing.volces.com/api/v3",
    api_key="<API_KEY>"  # 這里修改了一下，直接為變量賦值，輸入你的API_KEY即可
)

response = client.chat.completions.create(
    model="doubao-seed-1-6-250615",  #豆包的多模態(tài)處理模型。
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://wx3.sinaimg.cn/orj480/7ffa58d5ly1i44lggggjxj20k00zkjse.jpg"
                    },
                },
                {"type": "text", "text": "提取圖片中的文字信息，其它不要顯示。"},
            ],
        }
    ],
)

# 提取content內(nèi)容
content = response.choices[0].message.content
print("提取的內(nèi)容：")
print(content)

import os
import json
from tkinter import Tk, Label, Entry, filedialog, messagebox, StringVar, END, Button, LEFT,LabelFrame,Frame, Toplevel
from tkinter.scrolledtext import ScrolledText
from tkinter import PhotoImage, font
import threading
import base64
import pyperclip
from pathlib import Path
from openai import OpenAI
from PIL import Image, ImageTk

class ImageEssayEvaluator:
    def __init__(self, root):
        self.root = root
        self.root.title("圖片信息提取工具 | By Gordon")
        self.setup_ui()
        self.api = None
        self.output_dir = os.path.join(os.getcwd(), "識(shí)別結(jié)果")
        os.makedirs(self.output_dir, exist_ok=True)
        self.current_scale = 1.0  # Track current zoom scale
        self.image_files = []  # List to store multiple image paths
        self.current_image_index = 0  # Track current image index

        json_filename = "config.json"
        if not os.path.exists(json_filename):
            with open(json_filename, "w") as json_file:
                json.dump({"api_key": self.api}, json_file)
            print("JSON 文件已創(chuàng)建。")
        else:
            with open(json_filename, "r") as json_file:
                data = json.load(json_file)
            self.api = data.get("api")
        if self.api is None:
            self.show_settings()
        self.update_api_button_text()

    def update_api_button_text(self):
        """根據(jù)API狀態(tài)更新按鈕文本"""
        if self.api:
            self.api_button.config(text="已加載API")
        else:
            self.api_button.config(text="設(shè)置API")

    def setup_ui(self):
        self.info_var = StringVar()
        self.info_var.set("Tip: 請(qǐng)導(dǎo)入圖片，識(shí)別結(jié)果將顯示在下方文本框")
        Label(self.root, textvariable=self.info_var, font=("微軟雅黑", 12)).pack(pady=5)

        frame = Frame(self.root)
        frame.pack(padx=6)

        Label(frame, text="圖片路徑:", font=("微軟雅黑", 12)).pack(side='left')
        self.image_path_entry = Entry(frame, width=60, font=("微軟雅黑", 12))
        self.image_path_entry.pack(side=LEFT, padx=5, pady=5)

        self.api_button = Button(frame, text="設(shè)置API", font=("微軟雅黑", 12), command=self.show_settings)
        self.api_button.pack()

        # Create a frame for the image and text display
        main_frame = Frame(self.root)
        main_frame.pack(padx=5, pady=5)

        # Left Frame for Image Preview (no fixed height)
        #self.preview_frame = Frame(main_frame, width=500)
        self.preview_frame = LabelFrame(main_frame, text="  預(yù)覽圖片文件",font=("微軟雅黑", 12), width=400, height=460, padx=1, pady=1)
        self.preview_frame.grid(row=0, column=0, padx=10, pady=10)
        self.preview_frame.pack_propagate(False)

        # Image preview label
        self.image_label = Label(self.preview_frame)
        self.image_label.pack(padx=10, pady=10)

        # Frame for zoom and navigation buttons
        button_frame = Frame(self.preview_frame)
        button_frame.pack(side='bottom',pady=5)

        # Zoom buttons (Up and Down) and Navigation buttons (Previous and Next)
        self.zoom_in_button = Button(button_frame, text="放大", font=("微軟雅黑", 12), command=self.zoom_in)
        self.zoom_in_button.pack(side=LEFT, padx=5)
        self.zoom_out_button = Button(button_frame, text="縮小", font=("微軟雅黑", 12), command=self.zoom_out)
        self.zoom_out_button.pack(side=LEFT, padx=5)
        self.prev_button = Button(button_frame, text="上一張", font=("微軟雅黑", 12), command=self.previous_image)
        self.prev_button.pack(side=LEFT, padx=5)
        self.next_button = Button(button_frame, text="下一張", font=("微軟雅黑", 12), command=self.next_image)
        self.next_button.pack(side=LEFT, padx=5)

        # Right Frame for Text Display
        #self.text_frame = Frame(main_frame, width=400, height=300)
        self.text_frame = LabelFrame(main_frame, text="  提取信息顯示 ", font=("微軟雅黑", 12), width=400, height=300, padx=1, pady=1)
        self.text_frame.grid(row=0, column=1, padx=10, pady=10)
        # Text display area
        self.text_display = ScrolledText(self.text_frame, width=50, height=20, font=("微軟雅黑", 12))
        self.text_display.pack(padx=5, pady=5)
        button_frame = Frame(self.root)
        button_frame.pack(expand=True)
        # Other buttons
        Button(button_frame, text="導(dǎo)入圖片", font=("微軟雅黑", 12), command=self.load_image).pack(side="left", padx=90, pady=10)
        Button(button_frame, text="提取信息", font=("微軟雅黑", 12), command=self.start_evaluation).pack(side="left", padx=60, pady=10)
        Button(button_frame, text="打開(kāi)目錄", font=("微軟雅黑", 12), command=self.open_output_dir).pack(side="left", padx=60, pady=10)
        Button(button_frame, text="復(fù)制信息", font=("微軟雅黑", 12), command=self.copy_info).pack(side="left", padx=60, pady=10)

    def show_settings(self):
        self.settings_window = Toplevel(self.root)
        self.settings_window.attributes('-topmost', True)
        self.settings_window.title("豆包 API設(shè)置")

        Label(self.settings_window, text="請(qǐng)把kimi的API放在這里,使用ctrl+V:").pack(pady=5)
        self.api_var = StringVar()
        self.entry = Entry(self.settings_window, textvariable=self.api_var, width=30, font=("微軟雅黑", 12))
        self.entry.pack()
        confirm_button = Button(self.settings_window, text="確認(rèn)", command=lambda: self.apply_settings())
        confirm_button.pack(pady=10)

        screen_width = self.settings_window.winfo_screenwidth()
        screen_height = self.settings_window.winfo_screenheight()
        self.settings_window.update_idletasks()
        window_width = self.settings_window.winfo_width()
        window_height = self.settings_window.winfo_height()

        x_position = (screen_width - window_width) // 2
        y_position = (screen_height - window_height) // 2
        self.settings_window.geometry(f"{window_width}x{window_height}+{x_position}+{y_position}")

    def apply_settings(self):
        new_time = self.api_var.get()
        self.api = new_time.strip()
        data = {'api': self.api}
        with open('config.json', 'w+') as f:
            json.dump(data, f, indent=4)
        self.settings_window.destroy()

    def copy_info(self):
        pyperclip.copy(self.text_display.get(1.0, END))
    def load_image(self): 
        if len(str(self.api)) < 10:
            messagebox.showwarning("警告！", "API設(shè)置不正確或者沒(méi)有")
            self.show_settings()
            return

        # 選擇導(dǎo)入方式
        choice = messagebox.askquestion("選擇導(dǎo)入方式", "是：加載多個(gè)圖片文件\n否：加載整個(gè)文件夾中的圖片")

        image_paths = []

        if choice == 'yes':
            # 導(dǎo)入多個(gè)圖片文件
            file_paths = filedialog.askopenfilenames(
                title="選擇圖片",
                filetypes=[("圖片文件", "*.jpg;*.png;*.jpeg;*.bmp")]
            )
            if file_paths:
                image_paths = [fp.replace("\\", "/") for fp in file_paths]
                # 顯示第一張圖片路徑到 entry
                self.image_path_entry.delete(0, END)
                self.image_path_entry.insert(0, image_paths[0])

        elif choice == 'no':
            # 導(dǎo)入整個(gè)文件夾中的圖片
            folder_path = filedialog.askdirectory(title="選擇文件夾")
            if folder_path:
                folder_path = folder_path.replace("\\", "/")  # 替換為統(tǒng)一的路徑分隔符
                for fname in os.listdir(folder_path):
                    if fname.lower().endswith(('.jpg', '.jpeg', '.png', '.bmp')):
                        image_path = os.path.join(folder_path, fname).replace("\\", "/")
                        image_paths.append(image_path)
                image_paths.sort()

                # 設(shè)置 entry 顯示文件夾路徑
                self.image_path_entry.delete(0, END)
                self.image_path_entry.insert(0, folder_path)

        # 若成功導(dǎo)入圖片路徑
        if image_paths:
            self.image_files = image_paths
            self.current_image_index = 0
            self.current_scale = 1.0  # Reset zoom

            # 顯示狀態(tài)信息
            self.info_var.set(
                f"已加載圖片: {os.path.basename(self.image_files[self.current_image_index])} "
                f"({self.current_image_index + 1}/{len(self.image_files)})"
            )
            self.display_image(self.image_files[self.current_image_index])
            self.update_navigation_buttons()
        else:
            self.info_var.set("未選擇任何圖片")

    def update_navigation_buttons(self):
        """Enable or disable navigation buttons based on image list and current index"""
        if len(self.image_files) > 1:
            self.prev_button.config(state="normal" if self.current_image_index > 0 else "disabled")
            self.next_button.config(state="normal" if self.current_image_index < len(self.image_files) - 1 else "disabled")
        else:
            self.prev_button.config(state="disabled")
            self.next_button.config(state="disabled")

    def previous_image(self):
        if self.current_image_index > 0:
            self.current_image_index -= 1
            self.image_path_entry.delete(0, END)
            self.image_path_entry.insert(0, self.image_files[self.current_image_index])
            self.info_var.set(f"已加載圖片: {os.path.basename(self.image_files[self.current_image_index])} ({self.current_image_index + 1}/{len(self.image_files)})")
            self.current_scale = 1.0  # Reset zoom scale
            self.display_image(self.image_files[self.current_image_index])
            self.update_navigation_buttons()

    def next_image(self):
        if self.current_image_index < len(self.image_files) - 1:
            self.current_image_index += 1
            self.image_path_entry.delete(0, END)
            self.image_path_entry.insert(0, self.image_files[self.current_image_index])
            self.info_var.set(f"已加載圖片: {os.path.basename(self.image_files[self.current_image_index])} ({self.current_image_index + 1}/{len(self.image_files)})")
            self.current_scale = 1.0  # Reset zoom scale
            self.display_image(self.image_files[self.current_image_index])
            self.update_navigation_buttons()

    def display_image(self, image_path):
        # Open the image and display it as preview
        img = Image.open(image_path)
        # Estimate text area height (15 lines of font "微軟雅黑" size 12)
        text_font = font.Font(family="微軟雅黑", size=12)
        line_height = text_font.metrics("linespace")
        target_height = int(line_height * 15 * self.current_scale)
        # Calculate width based on aspect ratio
        original_width, original_height = img.size
        target_width = int(target_height * original_width / original_height)
        img = img.resize((target_width, target_height), Image.Resampling.LANCZOS)
        self.image_preview = ImageTk.PhotoImage(img)
        self.image_label.config(image=self.image_preview)

    def zoom_in(self):
        # Increase image size by 10%
        self.current_scale *= 1.1
        self.update_image()

    def zoom_out(self):
        # Decrease image size by 10%
        self.current_scale *= 0.9
        self.update_image()

    def update_image(self):
        if not self.image_files or self.current_image_index >= len(self.image_files):
            messagebox.showwarning("警告", "請(qǐng)先加載圖片")
            return

        image_path = self.image_files[self.current_image_index]

        try:
            img = Image.open(image_path)

            # Estimate text area height (15 lines of font "微軟雅黑" size 12)
            text_font = font.Font(family="微軟雅黑", size=12)
            line_height = text_font.metrics("linespace")
            target_height = int(line_height * 15 * self.current_scale)

            # Calculate width based on aspect ratio
            original_width, original_height = img.size
            target_width = int(target_height * original_width / original_height)

            img = img.resize((target_width, target_height), Image.Resampling.LANCZOS)
            self.image_preview = ImageTk.PhotoImage(img)
            self.image_label.config(image=self.image_preview)

            # 更新路徑欄顯示當(dāng)前圖片（可選）
            self.image_path_entry.delete(0, END)
            self.image_path_entry.insert(0, image_path)

        except Exception as e:
            messagebox.showerror("錯(cuò)誤", f"無(wú)法加載圖片：{e}")


    def start_evaluation(self):
        threading.Thread(target=self.evaluate_essay).start()
    def evaluate_essay(self):
        if not self.image_files:
            messagebox.showerror("錯(cuò)誤", "請(qǐng)先導(dǎo)入圖片文件或文件夾！")
            return

        self.info_var.set("正在提取內(nèi)容，請(qǐng)稍候...")

        try:
            for file in self.image_files:
                result = self.chat_doubao(file)
                base_name = os.path.splitext(os.path.basename(file))[0]
                output_path = os.path.join(self.output_dir, f"{base_name}.txt")
                with open(output_path, "a+", encoding="utf-8") as f:
                    f.write(result)

            self.info_var.set("提取完成！")
        except Exception as e:
            messagebox.showerror("錯(cuò)誤", f"提取失敗: {e}")
            self.info_var.set("提取失敗，請(qǐng)重試")

    def get_order(self):
        with open("order.txt","r",encoding="utf-8") as f:
            order = f.read().strip()
        return order
    def image_to_base64(self,image_path):
        # 獲取圖片文件的MIME類型
        ext = os.path.splitext(image_path)[1].lower()
        mime_type = f"image/{ext[1:]}" if ext in ['.jpg', '.jpeg', '.png', '.gif'] else "image/jpeg"
        
        with open(image_path, "rb") as image_file:
            # 讀取文件內(nèi)容并進(jìn)行Base64編碼
            base64_data = base64.b64encode(image_file.read()).decode('utf-8')
            # 返回完整的data URI格式
            return f"data:{mime_type};base64,{base64_data}"
    def chat_doubao(self,local_image_path):
        # 初始化客戶端
        client = OpenAI(
            base_url="https://ark.cn-beijing.volces.com/api/v3",
            api_key="<YOUR API KEY>"  # 這里要輸入自己的API
        )

        try:
            # 轉(zhuǎn)換本地圖片為Base64編碼
            image_data = self.image_to_base64(local_image_path)
            
            # 調(diào)用API處理圖片
            response = client.chat.completions.create(
                model="doubao-seed-1-6-250615",
                messages=[
                    {
                        "role": "user",
                        "content": [
                            {
                                "type": "image_url",
                                "image_url": {
                                    "url": image_data  # 使用Base64編碼的本地圖片數(shù)據(jù)
                                },
                            },
                            {"type": "text", "text": self.get_order()},
                        ],
                    }
                ]
            )

            # 提取并打印內(nèi)容
            content = response.choices[0].message.content
            self.text_display.insert(END, f"{content}\n")
            self.text_display.see(END)
            # 更新預(yù)覽圖片顯示
            self.display_image(local_image_path)

            # 更新頂部路徑欄顯示
            self.image_path_entry.delete(0, END)
            self.image_path_entry.insert(0, local_image_path)

        except FileNotFoundError:
            self.text_display.insert('1.0',f"錯(cuò)誤：找不到圖片文件 '{local_image_path}'")
        except Exception as e:
            print(f"處理過(guò)程中發(fā)生錯(cuò)誤：{str(e)}")
        return content

    def open_output_dir(self):
        os.startfile(self.output_dir)


if __name__ == "__main__":
    root = Tk()
    app = ImageEssayEvaluator(root)
    root.mainloop()