Skip to content

Added DPI parameter to control PDF resolution. #53

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Embracex1998 wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Added DPI parameter to control PDF resolution. #53

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5e8fb5b
change the pdf_images dpi
Embracex1998 Sep 20, 2024
bf51dae
change the pdf_images dpi
Embracex1998 Sep 20, 2024
b71bb7d
change the pdf_images dpi
Embracex1998 Sep 26, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 10 additions & 7 deletions gptpdf/parse.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,8 +130,7 @@ def _parse_rects(page: fitz.Page) -> List[Tuple[float, float, float, float]]:

return [rect.bounds for rect in merged_rects]


def _parse_pdf_to_images(pdf_path: str, output_dir: str = './') -> List[Tuple[str, List[str]]]:
def _parse_pdf_to_images(pdf_path: str,dpi:int,output_dir: str = './') -> List[Tuple[str, List[str]]]:
"""
Parse PDF to images and save to output_dir.
"""
Expand All @@ -143,11 +142,15 @@ def _parse_pdf_to_images(pdf_path: str, output_dir: str = './') -> List[Tuple[st
logging.info(f'parse page: {page_index}')
rect_images = []
rects = _parse_rects(page)
if not os.path.exists(output_dir):
os.makedirs(output_dir)
for index, rect in enumerate(rects):
fitz_rect = fitz.Rect(rect)
# 保存页面为图片
pix = page.get_pixmap(clip=fitz_rect, matrix=fitz.Matrix(4, 4))
pix = page.get_pixmap(clip=fitz_rect, matrix=fitz.Matrix(dpi/72, dpi/72))
name = f'{page_index}_{index}.png'


pix.save(os.path.join(output_dir, name))
rect_images.append(name)
# # 在页面上绘制红色矩形
Expand All @@ -164,15 +167,14 @@ def _parse_pdf_to_images(pdf_path: str, output_dir: str = './') -> List[Tuple[st
page.draw_rect(text_rect, color=(1, 1, 1), fill=(1, 1, 1))
# 插入带有白色背景的文字
page.insert_text((text_x, text_y), name, fontsize=10, color=(1, 0, 0))
page_image_with_rects = page.get_pixmap(matrix=fitz.Matrix(3, 3))
page_image_with_rects = page.get_pixmap(matrix=fitz.Matrix(dpi/72, dpi/72))
page_image = os.path.join(output_dir, f'{page_index}.png')
page_image_with_rects.save(page_image)
image_infos.append((page_image, rect_images))

pdf_document.close()
return image_infos


def _gpt_parse_images(
image_infos: List[Tuple[str, List[str]]],
prompt_dict: Optional[Dict] = None,
Expand Down Expand Up @@ -239,7 +241,7 @@ def _process_page(index: int, image_info: Tuple[str, List[str]]) -> Tuple[int, s

return '\n\n'.join(contents)


#增加了PDF图片解析的dpi参数，默认值设为200，根据需求设置300-1000可以满足清晰度需求
def parse_pdf(
pdf_path: str,
output_dir: str = './',
Expand All @@ -249,6 +251,7 @@ def parse_pdf(
model: str = 'gpt-4o',
verbose: bool = False,
gpt_worker: int = 1,
dpi:int = 200,
**args
) -> Tuple[str, List[str]]:
"""
Expand All @@ -257,7 +260,7 @@ def parse_pdf(
if not os.path.exists(output_dir):
os.makedirs(output_dir)

image_infos = _parse_pdf_to_images(pdf_path, output_dir=output_dir)
image_infos = _parse_pdf_to_images(pdf_path, output_dir=output_dir,dpi=dpi)
content = _gpt_parse_images(
image_infos=image_infos,
output_dir=output_dir,
Expand Down
5 changes: 3 additions & 2 deletions test/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@ def test_use_api_key():
api_key = os.getenv('OPENAI_API_KEY')
base_url = os.getenv('OPENAI_API_BASE')
# Manually provide OPENAI_API_KEY and OPEN_API_BASE
content, image_paths = parse_pdf(pdf_path, output_dir=output_dir, api_key=api_key, base_url=base_url, model='gpt-4o', gpt_worker=6)
content, image_paths = parse_pdf(pdf_path, output_dir=output_dir, api_key=api_key, base_url=base_url, model='gpt-4o', gpt_worker=6,dpi=1000)
print(content)
print(image_paths)
# also output_dir/output.md is generated
Expand Down Expand Up @@ -50,7 +50,8 @@ def test_qwen_vl_max():
base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1"
# Refer to: https://help.aliyun.com/zh/dashscope/developer-reference/compatibility-of-openai-with-dashscope
model = 'qwen-vl-max'
content, image_paths = parse_pdf(pdf_path, output_dir=output_dir, api_key=api_key, base_url=base_url, model=model, verbose=True, temperature=0.5, max_tokens=1000, top_p=0.9, frequency_penalty=1)
#新增了dpi参数，默认值为200
content, image_paths = parse_pdf(pdf_path, output_dir=output_dir, api_key=api_key, base_url=base_url, model=model, verbose=True, temperature=0.5, max_tokens=1000, top_p=0.9, frequency_penalty=1,dpi=1000)
print(content)
print(image_paths)

Expand Down