From a492d67c21bb48adabbdb9d1d444a70477688bf6 Mon Sep 17 00:00:00 2001 From: Gabo Date: Mon, 7 Oct 2024 11:02:32 +0200 Subject: [PATCH] Add .xml to missing extension xmls --- setup.py | 4 ++-- src/app.py | 2 ++ 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/setup.py b/setup.py index a3ffebc..8d9dc5f 100644 --- a/setup.py +++ b/setup.py @@ -4,8 +4,8 @@ requirements_path = Path("requirements.txt") -dependency_links = [r for r in requirements_path.read_text().splitlines() if r.startswith("git+")] requirements = [r for r in requirements_path.read_text().splitlines() if not r.startswith("git+")] +dependency_links = [r for r in requirements_path.read_text().splitlines() if r.startswith("git+")] PROJECT_NAME = "pdf-document-layout-analysis" @@ -13,7 +13,7 @@ name=PROJECT_NAME, packages=["pdf_tokens_type_trainer", "pdf_features", "pdf_token_type_labels", "fast_trainer"], package_dir={"": "src"}, - version="0.11", + version="0.12", url="https://github.com/huridocs/pdf-document-layout-analysis", author="HURIDOCS", description="This tool is for PDF document layout analysis", diff --git a/src/app.py b/src/app.py index 6e8c27f..b795186 100755 --- a/src/app.py +++ b/src/app.py @@ -38,6 +38,7 @@ async def run(file: UploadFile = File(...), fast: bool = Form(False), extraction @app.post("/save_xml/{xml_file_name}") @catch_exceptions async def analyze_and_save_xml(file: UploadFile = File(...), xml_file_name: str | None = None, fast: bool = Form(False)): + xml_file_name = xml_file_name if xml_file_name.endswith(".xml") else f"{xml_file_name}.xml" if fast: return await run_in_threadpool(analyze_pdf_fast, file.file.read(), xml_file_name, "") return await run_in_threadpool(analyze_pdf, file.file.read(), xml_file_name, "") @@ -46,6 +47,7 @@ async def analyze_and_save_xml(file: UploadFile = File(...), xml_file_name: str @app.get("/get_xml/{xml_file_name}", response_class=PlainTextResponse) @catch_exceptions async def get_xml_by_name(xml_file_name: str): + xml_file_name = xml_file_name if xml_file_name.endswith(".xml") else f"{xml_file_name}.xml" return await run_in_threadpool(get_xml, xml_file_name)