Upload folder using huggingface_hub

README.md

@@ -1,12 +1,12 @@
+
 ---
-title: Highlight Pdf
-emoji: 🌖
-colorFrom: blue
-colorTo: red
+title: highlight_pdf 
+emoji: 🔥
+colorFrom: indigo
+colorTo: indigo
 sdk: gradio
 sdk_version: 5.20.1
-app_file: app.py
+app_file: run.py
 pinned: false
+hf_oauth: true
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

requirements.txt

@@ -0,0 +1,2 @@
+gradio_pdf>=0.0.22
+pymupdf>=1.25.3

run.ipynb

@@ -0,0 +1 @@
+{"cells": [{"cell_type": "markdown", "id": "302934307671667531413257853548643485645", "metadata": {}, "source": ["# Gradio Demo: highlight_pdf"]}, {"cell_type": "code", "execution_count": null, "id": "272996653310673477252411125948039410165", "metadata": {}, "outputs": [], "source": ["!pip install -q gradio gradio_pdf>=0.0.22 pymupdf>=1.25.3"]}, {"cell_type": "code", "execution_count": null, "id": "288918539441861185822528903084949547379", "metadata": {}, "outputs": [], "source": ["# Downloading files from the demo repo\n", "import os\n", "!wget -q https://github.com/gradio-app/gradio/raw/main/demo/highlight_pdf/Lorem_ipsum.pdf"]}, {"cell_type": "code", "execution_count": null, "id": "44380577570523278879349135829904343037", "metadata": {}, "outputs": [], "source": ["import gradio as gr\n", "from gradio_pdf import PDF\n", "import pymupdf\n", "import os\n", "from pathlib import Path\n", "\n", "current_dir = Path(os.path.abspath(''))\n", "\n", "def highlight_text_in_pdf(pdf_file: Path, highlight_text: str):\n", "    page_number = 0\n", "    doc = pymupdf.open(pdf_file)\n", "    for page in doc:\n", "        text_instances = page.search_for(highlight_text)\n", "        if len(text_instances) > 0:\n", "            page_number = page.number\n", "        for inst in text_instances:\n", "            page.add_highlight_annot(inst)\n", "\n", "    new_pdf_file = str(pdf_file.parents[0]) + \"/new_\" + pdf_file.name\n", "    doc.save(new_pdf_file)\n", "\n", "    if page_number is None:\n", "        page_number = 0\n", "    \n", "    return new_pdf_file, page_number + 1\n", "\n", "def ask(query): \n", "    result = f\"Something about : {query}\"\n", "    sources = \"Document 1\"\n", "    pdf_path = current_dir / \"Lorem_ipsum.pdf\"\n", "    pdf_name = \"Document 1\"\n", "    context_to_highlight = \"Ut velit mauris\"\n", "\n", "    pdf, page_number = highlight_text_in_pdf(pdf_path, context_to_highlight)\n", "    return result, sources + f\" - Page {page_number}\", PDF(pdf, label=pdf_name, starting_page=page_number, interactive=True)\n", "\n", "\n", "if __name__ == \"__main__\":\n", "    with gr.Blocks() as demo:\n", "        title = gr.HTML(f\"<center><h1>Bot</h1></center>\")\n", "        with gr.Row():\n", "            with gr.Column(scale=2):\n", "                input = gr.Textbox(label=\"Question\", autofocus=True, interactive=True)\n", "                btn = gr.Button(\"Ask\", variant=\"primary\")\n", "                output = gr.Markdown(label=\"Anwser\")\n", "            with gr.Column(scale=2):\n", "                srcs = gr.Textbox(label=\"Sources\", interactive=False)\n", "                pdf = PDF(label=\"Document\")\n", "            \n", "        btn.click(fn=ask, inputs=input, outputs=[output, srcs, pdf])\n", "\n", "    demo.launch()\n"]}], "metadata": {}, "nbformat": 4, "nbformat_minor": 5}

run.py

@@ -0,0 +1,52 @@
+import gradio as gr
+from gradio_pdf import PDF
+import pymupdf
+import os
+from pathlib import Path
+
+current_dir = Path(os.path.abspath(''))
+
+def highlight_text_in_pdf(pdf_file: Path, highlight_text: str):
+    page_number = 0
+    doc = pymupdf.open(pdf_file)
+    for page in doc:
+        text_instances = page.search_for(highlight_text)
+        if len(text_instances) > 0:
+            page_number = page.number
+        for inst in text_instances:
+            page.add_highlight_annot(inst)
+
+    new_pdf_file = str(pdf_file.parents[0]) + "/new_" + pdf_file.name
+    doc.save(new_pdf_file)
+
+    if page_number is None:
+        page_number = 0
+    
+    return new_pdf_file, page_number + 1
+
+def ask(query): 
+    result = f"Something about : {query}"
+    sources = "Document 1"
+    pdf_path = current_dir / "Lorem_ipsum.pdf"
+    pdf_name = "Document 1"
+    context_to_highlight = "Ut velit mauris"
+
+    pdf, page_number = highlight_text_in_pdf(pdf_path, context_to_highlight)
+    return result, sources + f" - Page {page_number}", PDF(pdf, label=pdf_name, starting_page=page_number, interactive=True)
+
+
+if __name__ == "__main__":
+    with gr.Blocks() as demo:
+        title = gr.HTML(f"<center><h1>Bot</h1></center>")
+        with gr.Row():
+            with gr.Column(scale=2):
+                input = gr.Textbox(label="Question", autofocus=True, interactive=True)
+                btn = gr.Button("Ask", variant="primary")
+                output = gr.Markdown(label="Anwser")
+            with gr.Column(scale=2):
+                srcs = gr.Textbox(label="Sources", interactive=False)
+                pdf = PDF(label="Document")
+            
+        btn.click(fn=ask, inputs=input, outputs=[output, srcs, pdf])
+
+    demo.launch()