vlmparse 0.1.0__py3-none-any.whl → 0.1.3__py3-none-any.whl

vlmparse/st_viewer/st_viewer.py

@@ -0,0 +1,80 @@
+import argparse
+import subprocess
+import sys
+
+import streamlit as st
+from streamlit import runtime
+
+from vlmparse.data_model.document import Document
+from vlmparse.st_viewer.fs_nav import file_selector
+
+st.set_page_config(layout="wide")
+
+
+@st.cache_resource
+def get_doc(file_path):
+    return Document.from_zip(file_path)
+
+
+def render_sidebar_controls(doc, file_path):
+    """Render sidebar controls and return settings."""
+    return {
+        "page_no": st.number_input("Page", 0, len(doc.pages) - 1, 0),
+        "plot_layouts": st.checkbox("Plot layouts", value=True),
+    }
+
+
+def run_streamlit(folder: str) -> None:
+    with st.sidebar:
+        file_path = file_selector(folder)
+
+    if not file_path:
+        st.info("Please select a file from the sidebar.")
+        return
+
+    doc = get_doc(file_path)
+
+    with st.sidebar:
+        settings = render_sidebar_controls(doc, file_path)
+
+    col1, col2 = st.columns(2)
+    with col1:
+        with st.container(height=700):
+            st.write(doc.pages[settings["page_no"]].text)
+
+    with col2:
+        if settings["plot_layouts"]:
+            st.image(doc.pages[settings["page_no"]].get_image_with_boxes(layout=True))
+        else:
+            st.image(doc.pages[settings["page_no"]].image)
+
+
+def parse_args() -> argparse.Namespace:
+    """Parse command line arguments."""
+    parser = argparse.ArgumentParser(description="Document viewer with Streamlit")
+    parser.add_argument(
+        "folder", type=str, nargs="?", default=".", help="Root folder path"
+    )
+    return parser.parse_args()
+
+
+def main() -> None:
+    """Main entry point."""
+    folder = parse_args().folder
+
+    if runtime.exists():
+        run_streamlit(folder)
+    else:
+        try:
+            subprocess.run(
+                [sys.executable, "-m", "streamlit", "run", __file__, "--", folder],
+                check=True,
+            )
+        except KeyboardInterrupt:
+            print("\nStreamlit app terminated by user.")
+        except subprocess.CalledProcessError as e:
+            print(f"Error while running Streamlit: {e}")
+
+
+if __name__ == "__main__":
+    main()

{vlmparse-0.1.0.dist-info → vlmparse-0.1.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vlmparse
-Version: 0.1.0
+Version: 0.1.3
 Requires-Python: >=3.12.0
 Description-Content-Type: text/markdown
 License-File: LICENSE
@@ -19,6 +19,7 @@ Requires-Dist: pypdfium2>=4.30.0
 Requires-Dist: fire>=0.7.1
 Requires-Dist: lxml>=6.0.2
 Requires-Dist: tabulate>=0.9.0
+Requires-Dist: beautifulsoup4>=4.14.2
 Provides-Extra: dev
 Requires-Dist: jupyter; extra == "dev"
 Provides-Extra: docling-core
@@ -182,3 +183,13 @@ server.start()
 
 server.stop()
 ```
+
+
+Converter with automatic server deployment:
+
+```python
+from vlmparse.converter_with_server import ConverterWithServer
+
+converter_with_server = ConverterWithServer(model="mineru2.5")
+documents = converter_with_server.parse(inputs=["file1.pdf", "file2.pdf"], out_folder="./output")
+```

vlmparse-0.1.3.dist-info/RECORD

@@ -0,0 +1,50 @@
+vlmparse/base_model.py,sha256=4U4UPe8SNArliKnUf8pp8zQugWYsnhg9okylt7mrW1U,381
+vlmparse/build_doc.py,sha256=9evdU6GwVAZU15dZ1Qght6hNo_QxBQN8X3gmYdU2ltg,1965
+vlmparse/cli.py,sha256=tQma1IkOsFnqPKqqHVO1PJh18n1w82gp4ewA7oraJkE,15855
+vlmparse/converter.py,sha256=5wTA_cFyDMDSY8YgLzZV9SVBKmHjEbJCW8KPoJjmVVA,6880
+vlmparse/converter_with_server.py,sha256=9yoqfv8akB0xZZ7Snjq3aHW5NPNam2AgbK7_rfFqNkk,3909
+vlmparse/registries.py,sha256=TdSR1fx1Tz3roGk4Tk5ckIK6Iz-e4UD4erWUk96fFpQ,5846
+vlmparse/utils.py,sha256=jZWbNMwpZSZL--ZzvL8wPG_7mwpw9Pi36qTO9TjvHZU,1239
+vlmparse/benchpdf2md/create_dataset.py,sha256=0o4I0O3pHm1W7NYOTnW1JvPmgxJM8KLElKFvAbPAIic,1855
+vlmparse/benchpdf2md/run_benchmark.py,sha256=LMHElWyWIgB4ppBL0s-qjfMz5FZQnZOEm5mXxd0p0C8,9800
+vlmparse/benchpdf2md/utils.py,sha256=Q62vtvLIzxOEzSi-w210d7qnaRz-q_5ykmLNTkmbs-8,1732
+vlmparse/benchpdf2md/bench_tests/benchmark_tsts.py,sha256=XISZKlN1CiPBGDHgiDkjFkhGWxPxMFRu1GfStiOprmo,69527
+vlmparse/benchpdf2md/bench_tests/utils.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+vlmparse/benchpdf2md/olmocrbench/repeatdetect.py,sha256=SWCfYgAVmtyNypCsEZnuj6HoNGjFjTOqPBn2fCnEsiQ,5509
+vlmparse/benchpdf2md/olmocrbench/run_olmocr_bench.py,sha256=XnVlN5Sy7mXkr7nqWbcCgROsoavFBXqn4haSoxVsyLM,8495
+vlmparse/benchpdf2md/olmocrbench/tests.py,sha256=bszavrGQNCB3l60gPrsMgxtcMJvDjwZBBL_0B_K1oFA,53625
+vlmparse/benchpdf2md/olmocrbench/katex/__init__.py,sha256=DD9Knd52Ur9WiatymQGI9B1-yJ7OkL2w5E0quITDkPY,72
+vlmparse/benchpdf2md/olmocrbench/katex/render.py,sha256=l8mYRnz2I-10RE3JoxYrMfZ2UhxECTiRFazuf-dHnqU,18219
+vlmparse/benchpdf2md/st_visu_benchmark/app.py,sha256=ng8o_q20rY88BFQSEi62dBylklV6yyGyrrMiTrl7xtM,8921
+vlmparse/benchpdf2md/st_visu_benchmark/highligh_text.py,sha256=vsGtd9nPrtL5uDqALMumVlsllLvVZr2E6_d9TOLQoN4,3915
+vlmparse/benchpdf2md/st_visu_benchmark/test_form.py,sha256=qNmFZoSdbWcw1EJKesgO7rYV_LpslYPzij_gyJEuDP8,3701
+vlmparse/benchpdf2md/st_visu_benchmark/ui_elements.py,sha256=WkKncexShO3SU-DO7dPT4DOe-8UNjsCaHlj9L1B2mkI,572
+vlmparse/benchpdf2md/st_visu_benchmark/utils.py,sha256=JSmOJQY1DDETtWmjWv07SlQlORE6yBewiMcE5qRZI_Q,1109
+vlmparse/clients/chandra.py,sha256=zfu-A6Slh-fIAyrtrlVoCb6QHLBimnimefap_K9YwYw,9775
+vlmparse/clients/deepseekocr.py,sha256=iCG5wI5yPv98hIPgVJX4gkkkH1OekblZjFhh5ORVWAk,1813
+vlmparse/clients/docling.py,sha256=K-Grl_nZiSdooEdEaflevprE56l3Keby9xSMBtFwdis,5355
+vlmparse/clients/dotsocr.py,sha256=9ygvIVVOi9UhTUJwmrI-h6AjMV9vL9J2vMaBfUyTorY,9895
+vlmparse/clients/granite_docling.py,sha256=EQpsv5qSJG0HtMSacmJStER2sq4TGf1EMU5_NmJsl4g,4634
+vlmparse/clients/hunyuanocr.py,sha256=Xw0Q1l-3pQzaEgFngnfM8vrSWpnT3I99QvDaGZ8XooM,1712
+vlmparse/clients/lightonocr.py,sha256=wx1Im8Z3wlRWwYbPqnSd3LqTtdAU8CnX5mzu1BuCUY8,1314
+vlmparse/clients/mineru.py,sha256=OL56O6cMGksoVMVDyMYCQvwIvZEjKiAZXjEMEshqXaY,3549
+vlmparse/clients/nanonetocr.py,sha256=BT5vaeerCsK5agvOaHK3NvLUqWd1FfDmrMmDYbp646I,1543
+vlmparse/clients/olmocr.py,sha256=mQEDpfyLY8a80Zlps5mG0QaWytIgnNQZVEVWKWjPIjk,1849
+vlmparse/clients/openai_converter.py,sha256=QBHomw3_K_BpMVboaMd4rlhrntsqEFT46kxpg9KqBIs,5636
+vlmparse/clients/paddleocrvl.py,sha256=tmaqg3boV4edywiiiNiNiI3dBHi111wz4dFb52OISXw,1376
+vlmparse/clients/prompts.py,sha256=-J60lqxgRzlkQ9VsQLxmWsIMaDt-gNqWqWoqHIw9CLc,4228
+vlmparse/clients/pipe_utils/cleaner.py,sha256=oxBkBTOkluN1lmeNbzajRIe0_D__ZGwUOBaI_Ph0uxE,2396
+vlmparse/clients/pipe_utils/html_to_md_conversion.py,sha256=cFFqzD2jCNw_968_eu3Wt--Ox7iJj2Rn5UoP_DZWosU,4112
+vlmparse/clients/pipe_utils/utils.py,sha256=935ecIO446I0pstszE_1nrIPHn1Ffrxunq7fVd0dsd8,315
+vlmparse/data_model/box.py,sha256=lJsh4qhjgYXZF5vTSJ1qMXD5GVlBi2_SBedBMlfJikU,16868
+vlmparse/data_model/document.py,sha256=pdCZvWzRFkez53ZJpNaB4ezUW-OVUlbR3_SBmmgVzGQ,4217
+vlmparse/servers/docker_server.py,sha256=nI7K8CEzJwSZxLY7Jg9IuYHHLR5YQpOSgY8Ln71POJA,6608
+vlmparse/servers/utils.py,sha256=gMk5Y8FA1nlSxi7JzKxZu7XyljkYUZ5AnsTb3YFqu28,8821
+vlmparse/st_viewer/fs_nav.py,sha256=7GNH68h2Loh5pQ64Pe72-D2cs2BLhqRXevEmKdFmPX0,1616
+vlmparse/st_viewer/st_viewer.py,sha256=m2rQTtk5rlwErNmivNAg-4rkHkvNkvLhoJZxFQi7Dwk,2105
+vlmparse-0.1.3.dist-info/licenses/LICENSE,sha256=3TKJHk8hPBR5dbLWZ3IpfCftl-_m-iyBwpYQGZYxj14,1080
+vlmparse-0.1.3.dist-info/METADATA,sha256=JkSI4uFnnF59WReyhfRFZZVoe6KLk0ZJrjG0FQkUIPI,5112
+vlmparse-0.1.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+vlmparse-0.1.3.dist-info/entry_points.txt,sha256=gD5berP6HwE2wNIkls-Lw5goiceA8uMgPEd7ifnFJXs,47
+vlmparse-0.1.3.dist-info/top_level.txt,sha256=k4ni-GNH_iAX7liQEsk_KY_c3xgZgt8k9fsSs9IXLXs,9
+vlmparse-0.1.3.dist-info/RECORD,,

vlmparse-0.1.0.dist-info/RECORD

@@ -1,13 +0,0 @@
-vlmparse/base_model.py,sha256=4U4UPe8SNArliKnUf8pp8zQugWYsnhg9okylt7mrW1U,381
-vlmparse/build_doc.py,sha256=9evdU6GwVAZU15dZ1Qght6hNo_QxBQN8X3gmYdU2ltg,1965
-vlmparse/cli.py,sha256=tQma1IkOsFnqPKqqHVO1PJh18n1w82gp4ewA7oraJkE,15855
-vlmparse/converter.py,sha256=5wTA_cFyDMDSY8YgLzZV9SVBKmHjEbJCW8KPoJjmVVA,6880
-vlmparse/converter_with_server.py,sha256=9yoqfv8akB0xZZ7Snjq3aHW5NPNam2AgbK7_rfFqNkk,3909
-vlmparse/registries.py,sha256=TdSR1fx1Tz3roGk4Tk5ckIK6Iz-e4UD4erWUk96fFpQ,5846
-vlmparse/utils.py,sha256=jZWbNMwpZSZL--ZzvL8wPG_7mwpw9Pi36qTO9TjvHZU,1239
-vlmparse-0.1.0.dist-info/licenses/LICENSE,sha256=3TKJHk8hPBR5dbLWZ3IpfCftl-_m-iyBwpYQGZYxj14,1080
-vlmparse-0.1.0.dist-info/METADATA,sha256=4xxtT6rE3pJyqfqbVjl8Llj7C5Az99TeusxXmMHzMMQ,4788
-vlmparse-0.1.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-vlmparse-0.1.0.dist-info/entry_points.txt,sha256=gD5berP6HwE2wNIkls-Lw5goiceA8uMgPEd7ifnFJXs,47
-vlmparse-0.1.0.dist-info/top_level.txt,sha256=k4ni-GNH_iAX7liQEsk_KY_c3xgZgt8k9fsSs9IXLXs,9
-vlmparse-0.1.0.dist-info/RECORD,,