166 lines
6.0 KiB
Python
166 lines
6.0 KiB
Python
#!/usr/bin/env python3
|
||
# -*- coding: utf-8 -*-
|
||
"""
|
||
測試修復後的翻譯映射邏輯
|
||
"""
|
||
|
||
import sys
|
||
import os
|
||
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
|
||
|
||
# 設定編碼
|
||
sys.stdout.reconfigure(encoding='utf-8')
|
||
|
||
from pathlib import Path
|
||
from app import create_app
|
||
|
||
def test_fixed_mapping_logic():
|
||
"""測試修復後的翻譯映射邏輯"""
|
||
|
||
print("=" * 80)
|
||
print("測試修復後的翻譯映射邏輯")
|
||
print("預期結果: 應該找到原始DIFY翻譯 (ROW 449)")
|
||
print("=" * 80)
|
||
|
||
app = create_app()
|
||
|
||
with app.app_context():
|
||
from sqlalchemy import text as sql_text
|
||
from app import db
|
||
from app.services.translation_service import ExcelParser
|
||
|
||
# 1. 取得Excel提取的D2文字
|
||
original_file = Path(r"C:\Users\EGG\WORK\data\user_scrip\TOOL\Document_translator_V2\uploads\98158984-f335-44f5-a0b4-88fb8ccd5d78") / "original_panjit_98158984.xlsx"
|
||
|
||
if not original_file.exists():
|
||
print("❌ 測試檔案不存在")
|
||
return
|
||
|
||
parser = ExcelParser(str(original_file))
|
||
segments = parser.extract_text_segments()
|
||
|
||
d2_extracted = None
|
||
for segment in segments:
|
||
if "WB inline" in segment:
|
||
d2_extracted = segment
|
||
break
|
||
|
||
if not d2_extracted:
|
||
print("❌ 沒有找到D2相關內容")
|
||
return
|
||
|
||
print(f"1. Excel提取的D2文字:")
|
||
print(f" {repr(d2_extracted)}")
|
||
|
||
# 2. 測試修復後的查詢邏輯
|
||
print(f"\n2. 測試修復後的查詢邏輯")
|
||
print("-" * 60)
|
||
|
||
target_language = 'ko'
|
||
|
||
# 精確匹配 (應該找到ROW 514)
|
||
print(f"步驟1: 精確匹配查詢")
|
||
result1 = db.session.execute(sql_text("""
|
||
SELECT id, translated_text, created_at
|
||
FROM dt_translation_cache
|
||
WHERE source_text = :text AND target_language = :lang
|
||
ORDER BY created_at DESC
|
||
LIMIT 1
|
||
"""), {'text': d2_extracted, 'lang': target_language})
|
||
|
||
row1 = result1.fetchone()
|
||
if row1:
|
||
print(f" ✅ 精確匹配找到: ROW {row1[0]} (時間: {row1[2]})")
|
||
print(f" 翻譯: {repr(row1[1][:40])}...")
|
||
else:
|
||
print(f" ❌ 精確匹配失敗")
|
||
|
||
# 標準化匹配 (應該找到ROW 449)
|
||
print(f"\n步驟2: 標準化匹配查詢")
|
||
normalized_text = d2_extracted.replace('\n', ' ').replace('\r', ' ').strip()
|
||
print(f" 標準化文字: {repr(normalized_text)}")
|
||
|
||
result2 = db.session.execute(sql_text("""
|
||
SELECT id, translated_text, created_at
|
||
FROM dt_translation_cache
|
||
WHERE REPLACE(REPLACE(TRIM(source_text), '\n', ' '), '\r', ' ') = :text
|
||
AND target_language = :lang
|
||
ORDER BY created_at ASC
|
||
LIMIT 1
|
||
"""), {'text': normalized_text, 'lang': target_language})
|
||
|
||
row2 = result2.fetchone()
|
||
if row2:
|
||
print(f" ✅ 標準化匹配找到: ROW {row2[0]} (時間: {row2[2]})")
|
||
print(f" 翻譯: {repr(row2[1][:40])}...")
|
||
|
||
if row2[0] == 449:
|
||
print(f" 🎯 太好了！找到原始DIFY翻譯 (ROW 449)")
|
||
else:
|
||
print(f" ⚠️ 不是原始DIFY翻譯")
|
||
else:
|
||
print(f" ❌ 標準化匹配也失敗")
|
||
|
||
# 3. 模擬完整映射邏輯
|
||
print(f"\n3. 模擬完整映射邏輯")
|
||
print("-" * 60)
|
||
|
||
# 模擬修復後的查詢邏輯
|
||
result = db.session.execute(sql_text("""
|
||
SELECT translated_text
|
||
FROM dt_translation_cache
|
||
WHERE source_text = :text AND target_language = :lang
|
||
ORDER BY created_at DESC
|
||
LIMIT 1
|
||
"""), {'text': d2_extracted, 'lang': target_language})
|
||
|
||
row = result.fetchone()
|
||
|
||
# 如果精確匹配失敗，嘗試標準化匹配
|
||
if not row:
|
||
normalized_text = d2_extracted.replace('\n', ' ').replace('\r', ' ').strip()
|
||
result = db.session.execute(sql_text("""
|
||
SELECT translated_text
|
||
FROM dt_translation_cache
|
||
WHERE REPLACE(REPLACE(TRIM(source_text), '\n', ' '), '\r', ' ') = :text
|
||
AND target_language = :lang
|
||
ORDER BY created_at ASC
|
||
LIMIT 1
|
||
"""), {'text': normalized_text, 'lang': target_language})
|
||
row = result.fetchone()
|
||
print(f" 使用標準化匹配")
|
||
else:
|
||
print(f" 使用精確匹配")
|
||
|
||
if row and row[0]:
|
||
print(f" ✅ 最終找到翻譯: {repr(row[0][:50])}...")
|
||
|
||
# 檢查這是否為原始DIFY翻譯的特徵
|
||
if "와이어 본딩" in row[0] or "처리 속도" in row[0]:
|
||
print(f" 🎯 這是原始DIFY翻譯！")
|
||
print(f" 特徵: 包含 '와이어 본딩' 或 '처리 속도'")
|
||
elif "연결" in row[0] and "단축" in row[0]:
|
||
print(f" ✋ 這是手動補充翻譯")
|
||
print(f" 特徵: 包含 '연결' 和 '단축'")
|
||
else:
|
||
print(f" ❓ 無法判斷翻譯來源")
|
||
else:
|
||
print(f" ❌ 最終也沒找到翻譯")
|
||
|
||
# 4. 建議下一步
|
||
print(f"\n4. 建議下一步")
|
||
print("-" * 60)
|
||
|
||
if row2 and row2[0] == 449:
|
||
print(f"✅ 修復成功！系統現在能找到原始DIFY翻譯")
|
||
print(f" 建議: 重新生成韓文翻譯檔案，應該會使用原始DIFY翻譯")
|
||
else:
|
||
print(f"⚠️ 修復不完全，還需要進一步調整")
|
||
print(f" 可能需要檢查SQL語法或邏輯")
|
||
|
||
print(f"\n" + "=" * 80)
|
||
print("修復後映射邏輯測試完成！")
|
||
print("=" * 80)
|
||
|
||
if __name__ == "__main__":
|
||
test_fixed_mapping_logic() |