Leon/docutranslate
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

修改为导出的markdownu压缩包不含顶层目录

Browse Source
This commit is contained in:
xunbu
2025-07-03 17:30:31 +08:00
parent 8cef4451b2
commit b53a1d6005
2 changed files with 18 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
docutranslate/translater.py
View File

@@ -109,6 +109,10 @@ class FileTranslater:
return cached_result return cached_result
if document.suffix in [".md", ".txt"]: if document.suffix in [".md", ".txt"]:
return document.filebytes.decode("utf-8") return document.filebytes.decode("utf-8")
if document.suffix in ['zip']:
#寻找zip内的filename
filename=find_markdown_in_zip(document.filebytes)
return embed_inline_image_from_zip(document.filebytes,filename)
translater_logger.info("正在转化为markdown") translater_logger.info("正在转化为markdown")
if self.convert_engin == "docling": if self.convert_engin == "docling":
if artifact is None: if artifact is None:
@@ -289,18 +293,20 @@ class FileTranslater:
def save_as_markdown(self, filename: str | Path | None = None, output_dir: str | Path = "./output", embeded=True): def save_as_markdown(self, filename: str | Path | None = None, output_dir: str | Path = "./output", embeded=True):
if isinstance(filename, str): if isinstance(filename, str):
filename = Path(filename) filename = Path(filename)
if isinstance(output_dir, str):
output_dir = Path(output_dir)
if filename is None: if filename is None:
filename = Path(f"{self.document.stem}.md") filename = Path(f"{self.document.stem}.md")
# 确保输出目录存在 # 确保输出目录存在
output_dir.mkdir(parents=True, exist_ok=True) if isinstance(output_dir, str):
output_dir = Path(output_dir)
if embeded: if embeded:
full_name = output_dir / filename output_dir.mkdir(parents=True, exist_ok=True)
full_name = output_dir / filename.name
with open(full_name, "w") as file: with open(full_name, "w") as file:
file.write(self.export_to_markdown()) file.write(self.export_to_markdown())
translater_logger.info(f"文件已写入{full_name.resolve()}") translater_logger.info(f"文件已写入{full_name.resolve()}")
else: else:
output_dir=output_dir/filename.stem
output_dir.mkdir(parents=True, exist_ok=True)
with zipfile.ZipFile(io.BytesIO(self.export_to_unembed_markdown())) as zip_ref: with zipfile.ZipFile(io.BytesIO(self.export_to_unembed_markdown())) as zip_ref:
zip_ref.extractall(output_dir) zip_ref.extractall(output_dir)
return self return self
@@ -316,7 +322,7 @@ class FileTranslater:
if filename is None: if filename is None:
filename = Path(f"{self.document.stem}.md") filename = Path(f"{self.document.stem}.md")
self._markdown_format() self._markdown_format()
return unembed_base64_images_to_zip(self.markdown, folder_name=str(filename.stem), markdown_name=str(filename)) return unembed_base64_images_to_zip(self.markdown, markdown_name=filename.name)
def save_as_html(self, filename: str | Path | None = None, output_dir: str | Path = "./output"): def save_as_html(self, filename: str | Path | None = None, output_dir: str | Path = "./output"):
if isinstance(filename, str): if isinstance(filename, str):

17
docutranslate/utils/markdown_utils.py
View File

@@ -179,11 +179,9 @@ def embed_inline_image_from_zip(zip_bytes: bytes, filename_in_zip: str, encoding
return modified_md_content return modified_md_content
def unembed_base64_images_to_zip(markdown:str,folder_name:str,markdown_name:str,image_folder_name="images")->bytes: def unembed_base64_images_to_zip(markdown:str,markdown_name:str,image_folder_name="images")->bytes:
with tempfile.TemporaryDirectory() as temp_dir: with tempfile.TemporaryDirectory() as temp_dir:
subfolder = os.path.join(temp_dir, folder_name)#所有的操作都在这个subfolder里进行 image_folder=os.path.join(temp_dir,image_folder_name)
os.makedirs(subfolder, exist_ok=True)
image_folder=os.path.join(subfolder,image_folder_name)
os.makedirs(image_folder,exist_ok=True) os.makedirs(image_folder,exist_ok=True)
pattern=r"!\[(.*?)\]\(data:(.*?);.*base64,(.*)\)" pattern=r"!\[(.*?)\]\(data:(.*?);.*base64,(.*)\)"
def unembed_base64_images(match:re.Match)->str: def unembed_base64_images(match:re.Match)->str:
@@ -197,16 +195,15 @@ def unembed_base64_images_to_zip(markdown:str,folder_name:str,markdown_name:str,
f.write(base64.b64decode(b64data)) f.write(base64.b64decode(b64data))
return f"![{match.group(1)}]({url})" return f"![{match.group(1)}]({url})"
modified_md_content = re.sub(pattern, unembed_base64_images,markdown) modified_md_content = re.sub(pattern, unembed_base64_images,markdown)
with open(os.path.join(subfolder,f"{markdown_name}"),"w") as f: with open(os.path.join(temp_dir,f"{markdown_name}"),"w") as f:
f.write(modified_md_content) f.write(modified_md_content)
zip_buffer=io.BytesIO() zip_buffer=io.BytesIO()
folder_path=Path(subfolder) folder_path=Path(temp_dir)
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zipf: with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zipf:
for file in folder_path.rglob('*'): for file in folder_path.rglob('*'):
if file.is_file(): if file.is_file():
zipf.write(file, file.relative_to(folder_path.parent)) zipf.write(file, file.relative_to(folder_path))
return zip_buffer.getvalue()
return zip_buffer.read()
def clean_markdown_math_block(markdown): def clean_markdown_math_block(markdown):
"""清除公式块的多余空格字符""" """清除公式块的多余空格字符"""
@@ -222,5 +219,5 @@ def clean_markdown_math_block(markdown):
if __name__ == '__main__': if __name__ == '__main__':
with open(r"C:\Users\jxgm\Desktop\translate\docutranslate\tests\files\test7.md",'r') as f: with open(r"C:\Users\jxgm\Desktop\translate\docutranslate\tests\files\test7.md",'r') as f:
markdown=f.read() markdown=f.read()
print(unembed_base64_images_to_zip(markdown)) print(unembed_base64_images_to_zip(markdown,"markdown.md"))