Create app,py

app,py

@@ -0,0 +1,303 @@
+import gradio as gr
+import openai
+import math
+import os
+from pypdf import PdfReader
+
+# ── CONFIG ───────────────────────────────────────────────────────────────────
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")  # ← fixed: use environment variable
+# If testing locally, you can temporarily do:
+# OPENAI_API_KEY = "sk-......................."
+
+if not OPENAI_API_KEY:
+    raise ValueError("OPENAI_API_KEY environment variable not set")
+
+client = openai.OpenAI(api_key=OPENAI_API_KEY)
+
+EMBED_MODEL = "text-embedding-3-small"
+CHAT_MODEL  = "gpt-4o-mini"
+CHUNK_SIZE  = 100
+OVERLAP     = 40
+TOP_K       = 2
+
+# ── GLOBALS ──────────────────────────────────────────────────────────────────
+stored_chunks     = []
+stored_embeddings = []
+doc_name          = ""
+
+# ── CORE LOGIC ───────────────────────────────────────────────────────────────
+def split_chunks(text, size=CHUNK_SIZE, overlap=OVERLAP):
+    words, chunks, start = text.split(), [], 0
+    while start < len(words):
+        chunks.append(" ".join(words[start:start + size]))
+        start += size - overlap
+    return chunks
+
+def embed(text):
+    return client.embeddings.create(model=EMBED_MODEL, input=text).data[0].embedding
+
+def cosine_similarity(a, b):
+    dot   = sum(x * y for x, y in zip(a, b))
+    normA = math.sqrt(sum(x * x for x in a))
+    normB = math.sqrt(sum(x * x for x in b))
+    return dot / (normA * normB) if normA * normB != 0 else 0
+
+def get_top_chunks(question, k=TOP_K):
+    q_vec  = embed(question)
+    scores = [(cosine_similarity(q_vec, cv), c)
+              for cv, c in zip(stored_embeddings, stored_chunks)]
+    scores.sort(reverse=True)
+    return [c for _, c in scores[:k]]
+
+def upload_pdf(file):
+    global stored_chunks, stored_embeddings, doc_name
+    if file is None:
+        return ("— No file selected.", gr.update(interactive=False))
+
+    reader    = PdfReader(file.name)
+    full_text = "".join(page.extract_text() or "" for page in reader.pages)
+    doc_name  = os.path.basename(file.name)
+
+    stored_chunks     = split_chunks(full_text)
+    stored_embeddings = [embed(c) for c in stored_chunks]
+
+    return (
+        f"✦  {doc_name}\n→  {len(stored_chunks)} chunks indexed and ready.",
+        gr.update(interactive=True)
+    )
+
+def chat(question, history):
+    if not stored_chunks:
+        return history + [(question, "Please index a document in the Upload tab first.")]
+
+    context = "\n\n".join([f"[{i+1}] {c}" for i, c in enumerate(get_top_chunks(question))])
+    prompt  = (
+        "Answer strictly from the document text below.\n"
+        "If not found, reply: \"This information is not in the document.\"\n\n"
+        f"TEXT:\n{context}\n\nQUESTION: {question}\n\nANSWER:"
+    )
+    response = client.chat.completions.create(
+        model=CHAT_MODEL,
+        messages=[
+            {"role": "system", "content":
+                "Strict document assistant. Only use provided text. "
+                "If answer absent, say: This information is not in the document."},
+            {"role": "user", "content": prompt}
+        ]
+    )
+    return history + [(question, response.choices[0].message.content)]
+
+# ── CSS (cleaner, higher contrast, Apple-like feel) ──────────────────────────
+css = """
+:root {
+    --bg:           #fdfdfd;
+    --surface:      #ffffff;
+    --surface-2:    #f5f5f7;
+    --border:       #d2d2d7;
+    --border-2:     #c7c7cc;
+    --accent:       #0071e3;       /* Apple blue */
+    --accent-light: #4090ff;
+    --accent-glow:  rgba(0, 113, 227, 0.12);
+    --green:        #34c759;
+    --text-1:       #1d1d1f;
+    --text-2:       #3c3c43;
+    --text-3:       #6e6e73;
+    --font-sans:    system-ui, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif;
+    --font-mono:    ui-monospace, SFMono-Regular, Menlo, Monaco, Consolas, monospace;
+    --r:            12px;
+    --r-lg:         20px;
+    --ease:         cubic-bezier(0.25, 0.1, 0.25, 1);
+    --shadow:       0 1px 3px rgba(0,0,0,0.08), 0 4px 12px rgba(0,0,0,0.05);
+}
+
+html, body, .gradio-container {
+    font-family: var(--font-sans) !important;
+    background: var(--bg) !important;
+    color: var(--text-1) !important;
+}
+
+.gradio-container {
+    max-width: 1100px !important;
+    margin: 0 auto !important;
+    padding: 32px 24px 80px !important;
+}
+
+::-webkit-scrollbar {
+    width: 6px;
+}
+::-webkit-scrollbar-track {
+    background: transparent;
+}
+::-webkit-scrollbar-thumb {
+    background: #c0c0c5;
+    border-radius: 3px;
+}
+::-webkit-scrollbar-thumb:hover {
+    background: #a0a0a8;
+}
+
+/* Masthead */
+.mast-title {
+    font-size: 2.4rem;
+    font-weight: 600;
+    letter-spacing: -0.02em;
+    color: var(--text-1);
+}
+.mast-title em {
+    color: var(--accent);
+    font-style: normal;
+    font-weight: 700;
+}
+.mast-sub {
+    font-family: var(--font-mono);
+    font-size: 0.68rem;
+    color: var(--text-3);
+    letter-spacing: 0.06em;
+    text-transform: uppercase;
+    margin-top: 6px;
+}
+
+/* Chatbot fixes – most important for readability */
+.chatbot-wrap > div {
+    background: var(--bg) !important;
+    border: 1px solid var(--border) !important;
+    border-radius: var(--r-lg) !important;
+}
+.message-wrap .user-row .message {
+    background: var(--accent) !important;
+    color: white !important;
+    border-radius: 20px 20px 4px 20px !important;
+    padding: 12px 16px !important;
+    font-size: 0.94rem !important;
+    line-height: 1.5 !important;
+}
+.message-wrap .bot-row .message {
+    background: var(--surface) !important;
+    color: var(--text-1) !important;
+    border: 1px solid var(--border-2) !important;
+    border-radius: 20px 20px 20px 4px !important;
+    padding: 12px 16px !important;
+    font-size: 0.94rem !important;
+    line-height: 1.5 !important;
+}
+
+/* Inputs */
+textarea, input[type="text"] {
+    background: white !important;
+    border: 1px solid var(--border-2) !important;
+    border-radius: var(--r) !important;
+    color: var(--text-1) !important;
+    font-size: 0.96rem !important;
+    padding: 12px 16px !important;
+}
+textarea:focus, input[type="text"]:focus {
+    border-color: var(--accent) !important;
+    box-shadow: 0 0 0 3px var(--accent-glow) !important;
+}
+textarea::placeholder {
+    color: var(--text-3) !important;
+}
+
+/* Buttons */
+button.primary {
+    background: var(--accent) !important;
+    color: white !important;
+    border: none !important;
+    border-radius: 10px !important;
+    font-weight: 500 !important;
+}
+button.primary:hover {
+    background: var(--accent-light) !important;
+}
+
+/* Tabs & labels */
+label > span, .gr-form label span {
+    font-family: var(--font-mono) !important;
+    font-size: 0.64rem !important;
+    color: var(--text-3) !important;
+    letter-spacing: 0.04em;
+    text-transform: uppercase;
+}
+"""
+
+# ── UI ────────────────────────────────────────────────────────────────────────
+with gr.Blocks(css=css, title="Document Intelligence") as demo:
+
+    gr.HTML("""
+    <div style="display:flex; justify-content:space-between; align-items:flex-start; margin-bottom:32px; padding-bottom:20px; border-bottom:1px solid #d2d2d7;">
+        <div>
+            <div style="font-size:2.4rem; font-weight:600; letter-spacing:-0.02em;">Document <span style="color:#0071e3;">Intelligence</span></div>
+            <div style="font-family:ui-monospace,monospace; font-size:0.68rem; color:#6e6e73; letter-spacing:0.06em; text-transform:uppercase; margin-top:6px;">
+                Retrieval-Augmented Generation · Semantic Search · GPT-4o mini
+            </div>
+        </div>
+        <div style="text-align:right;">
+            <div style="font-size:0.9rem; font-weight:500; color:#1d1d1f;">Your Name / Portfolio</div>
+            <div style="font-size:0.7rem; color:#6e6e73;">Gradio + OpenAI</div>
+        </div>
+    </div>
+    """)
+
+    with gr.Tabs():
+        with gr.TabItem("Upload & Index"):
+            with gr.Row():
+                with gr.Column(scale=1):
+                    pdf_input = gr.File(label="Upload PDF document", file_types=[".pdf"])
+                    upload_btn = gr.Button("Index Document", variant="primary")
+                    upload_status = gr.Textbox(
+                        label="Status",
+                        lines=3,
+                        interactive=False,
+                        value="— Waiting for PDF…",
+                    )
+
+                with gr.Column(scale=1):
+                    gr.Markdown("### Settings & Pipeline")
+                    gr.Markdown("""
+                    - Embedding model: **text-embedding-3-small**  
+                    - Chat model: **gpt-4o-mini**  
+                    - Chunk size: **100** words  
+                    - Overlap: **40** words  
+                    - Retrieval: **top-2** chunks
+                    """)
+
+        with gr.TabItem("Chat"):
+            chatbot = gr.Chatbot(
+                height=520,
+                show_label=False,
+                bubble_full_width=False,
+                elem_classes=["chatbot-wrap"],
+            )
+            with gr.Row():
+                msg_box = gr.Textbox(
+                    placeholder="Ask anything about the document…",
+                    container=False,
+                    scale=5,
+                )
+                send_btn = gr.Button("Send", variant="primary", scale=1, min_width=80)
+
+            gr.Markdown(
+                "*Only answers grounded in the uploaded document. Clear chat to start over.*",
+                elem_classes=["small-footer"]
+            )
+
+    # Events
+    upload_btn.click(
+        fn=upload_pdf,
+        inputs=pdf_input,
+        outputs=[upload_status, send_btn]
+    )
+
+    send_btn.click(
+        fn=chat,
+        inputs=[msg_box, chatbot],
+        outputs=chatbot
+    ).then(lambda: "", None, msg_box)
+
+    msg_box.submit(
+        fn=chat,
+        inputs=[msg_box, chatbot],
+        outputs=chatbot
+    ).then(lambda: "", None, msg_box)
+
+demo.launch()