Plurk paste

"""
PDF 關鍵字擷取 (繁體中文, Tkinter)
版本：自由文字選取 + 一鍵複製版
功能：
- Windows7 / Windows11 可執行 (需安裝 Tesseract 與 Poppler)
- 將 PDF 轉圖片後以 Tesseract (chi_tra) OCR
- 使用者可新增/刪除關鍵字（不寫檔）
- 結果改用 Text 顯示，可自由拖曳選取部分文字
- 可用 Ctrl+C 或按「複製選取文字」按鈕複製內容
"""

import os
import tkinter as tk
from tkinter import filedialog, messagebox, simpledialog
from PIL import Image
from pdf2image import convert_from_path
import subprocess
import tempfile
import shutil
import threading

# ------------------ 使用前請調整這三個路徑 ------------------
TESSERACT_PATH = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
TESSDATA_ROOT = r"C:\Program Files\Tesseract-OCR"
POPPLER_PATH = r"" # 例如 r"C:\poppler-xx\Library\bin"
# ----------------------------------------------------------

# 預設關鍵字
keywords = ["UR", "糞便培養", "糞", "潛血", "毒品"]

class KeywordFinderApp:
def __init__(self, master):
self.master = master
master.title("PDF 關鍵字擷取工具 (繁體中文，自由選取＋一鍵複製)")
master.geometry("820x620")

self.current_pdf_path = None

# 上方：檔案操作
frame_top = tk.Frame(master, padx=8, pady=8)
frame_top.pack(fill='x')

self.btn_open = tk.Button(frame_top, text="① 開啟 PDF 檔案", command=self.open_pdf, width=18)
self.btn_open.pack(side='left', padx=6)

self.btn_search = tk.Button(frame_top, text="② 開始搜尋關鍵字", command=self.start_search_thread, width=18, bg='#ffdddd')
self.btn_search.pack(side='left', padx=6)

self.label_pdf = tk.Label(frame_top, text="目前未開啟檔案", anchor='w')
self.label_pdf.pack(side='left', padx=12)

# 關鍵字管理
frame_kw = tk.Frame(master, padx=8, pady=6)
frame_kw.pack(fill='x')

tk.Label(frame_kw, text="當前關鍵字：").pack(side='left')
self.var_kw = tk.StringVar(value=", ".join(keywords))
self.lbl_kw = tk.Label(frame_kw, textvariable=self.var_kw, fg='blue')
self.lbl_kw.pack(side='left', padx=6)

tk.Button(frame_kw, text="新增關鍵字", command=self.add_keyword, width=12).pack(side='right', padx=6)
tk.Button(frame_kw, text="刪除關鍵字", command=self.delete_keyword, width=12).pack(side='right', padx=6)

# 中下方：結果顯示 + 一鍵複製
frame_results = tk.Frame(master, padx=8, pady=6)
frame_results.pack(fill='both', expand=True)

tk.Label(frame_results, text="搜尋結果（可自由拖曳選取文字或使用Ctrl+C）").pack(anchor='w')

frame_text = tk.Frame(frame_results)
frame_text.pack(fill='both', expand=True)

scrollbar = tk.Scrollbar(frame_text)
scrollbar.pack(side='right', fill='y')

# ✅ 改用 Text 元件顯示結果
self.textbox = tk.Text(frame_text, wrap='word', yscrollcommand=scrollbar.set)
self.textbox.pack(fill='both', expand=True)
scrollbar.config(command=self.textbox.yview)
self.textbox.config(state='disabled')

# ✅ 新增「複製選取文字」按鈕
tk.Button(frame_results, text="複製選取文字", command=self.copy_selected_text, width=18, bg='#ddf').pack(anchor='e', pady=(4, 0))

# 狀態列
self.status_var = tk.StringVar(value="準備就緒")
tk.Label(master, textvariable=self.status_var, anchor='w').pack(fill='x', padx=8, pady=(0,8))

# ------------------ 關鍵字管理 ------------------
def update_kw_label(self):
self.var_kw.set(", ".join(keywords))

def add_keyword(self):
s = simpledialog.askstring("新增關鍵字", "請輸入關鍵字：", parent=self.master)
if s:
s = s.strip()
if s and s not in keywords:
keywords.append(s)
self.update_kw_label()

def delete_keyword(self):
s = simpledialog.askstring("刪除關鍵字", "請輸入要刪除的關鍵字：", parent=self.master)
if s:
s = s.strip()
if s in keywords:
keywords.remove(s)
self.update_kw_label()
else:
messagebox.showinfo("提示", "關鍵字不在目前列表中。")

# ------------------ 開啟 PDF ------------------
def open_pdf(self):
path = filedialog.askopenfilename(title="選擇 PDF 檔案", filetypes=[("PDF 檔案", "*.pdf")])
if path:
self.current_pdf_path = path
self.label_pdf.config(text=f"目前檔案：{os.path.basename(path)}")
self._set_text(f"已載入檔案：{os.path.basename(path)}。按「開始搜尋關鍵字」。")
self.status_var.set("已載入檔案")
else:
self.status_var.set("未選擇檔案")

# ------------------ 背景執行搜尋 ------------------
def start_search_thread(self):
if not self.current_pdf_path:
messagebox.showwarning("警告", "請先開啟一個 PDF 檔案。")
return
self.btn_open.config(state='disabled')
self.btn_search.config(state='disabled')
self._set_text("開始處理...（請稍候）")
self.status_var.set("處理中")
threading.Thread(target=self.start_search, daemon=True).start()

def start_search(self):
if not os.path.isfile(TESSERACT_PATH):
self._show_error_and_finish(f"找不到 tesseract：{TESSERACT_PATH}")
return

tessdata_guess = TESSDATA_ROOT
if os.path.isdir(os.path.join(TESSDATA_ROOT, "tessdata")):
tessdata_guess = os.path.join(TESSDATA_ROOT, "tessdata")
os.environ["TESSDATA_PREFIX"] = tessdata_guess

chi_path = os.path.join(tessdata_guess, "chi_tra.traineddata")
if not os.path.isfile(chi_path):
self._show_error_and_finish(f"找不到 chi_tra.traineddata：{chi_path}")
return

temp_dir = tempfile.mkdtemp(prefix="pdf2img_")
found_count = 0

try:
convert_kwargs = {}
if POPPLER_PATH:
convert_kwargs['poppler_path'] = POPPLER_PATH

pages = convert_from_path(self.current_pdf_path, **convert_kwargs)
self._set_text("") # 清空顯示

for idx, page in enumerate(pages):
self._append_text(f"處理第 {idx+1} 頁...\n")
img_path = os.path.join(temp_dir, f"page_{idx}.png")
page.save(img_path, "PNG")

cmd = [
TESSERACT_PATH,
img_path,
"stdout",
"-l", "chi_tra",
"--psm", "6",
"--tessdata-dir", tessdata_guess
]
try:
proc = subprocess.run(cmd, capture_output=True, text=True, encoding='utf-8', errors='replace', check=True)
text = proc.stdout or ""
except Exception as e:
self._append_text(f"[錯誤] 第 {idx+1} 頁 OCR 失敗：{e}\n")
continue

cleaned = "\n".join([s.strip() for s in text.splitlines() if s.strip()])
for line in cleaned.splitlines():
for kw in keywords:
if kw.lower() in line.lower():
result = f"[第{idx+1}頁/關鍵字:{kw}] {line}"
self._append_text(result + "\n")
found_count += 1
break

if found_count == 0:
self._append_text("\n===== 搜尋完成：未找到任何關鍵字 =====\n")
else:
self._append_text(f"\n===== 搜尋完成：共找到 {found_count} 筆結果 =====\n")
self.status_var.set("完成")
finally:
shutil.rmtree(temp_dir, ignore_errors=True)
self._finish_buttons()

# ------------------ 一鍵複製功能 ------------------
def copy_selected_text(self):
try:
selected_text = self.textbox.selection_get()
self.master.clipboard_clear()
self.master.clipboard_append(selected_text)
self.master.update()
messagebox.showinfo("已複製", "選取文字已複製到剪貼簿。")
except tk.TclError:
messagebox.showwarning("提示", "請先選取要複製的文字。")

# ------------------ 輔助函式 ------------------
def _set_text(self, content):
self.textbox.config(state='normal')
self.textbox.delete(1.0, tk.END)
self.textbox.insert(tk.END, content + "\n")
self.textbox.config(state='disabled')
self.textbox.see(tk.END)
self.master.update()

def _append_text(self, content):
self.textbox.config(state='normal')
self.textbox.insert(tk.END, content)
self.textbox.config(state='disabled')
self.textbox.see(tk.END)
self.master.update()

def _finish_buttons(self):
self.btn_open.config(state='normal')
self.btn_search.config(state='normal')

def _show_error_and_finish(self, msg):
messagebox.showerror("錯誤", msg)
self.status_var.set("錯誤")
self._finish_buttons()

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