Upload 3 files

.env

@@ -0,0 +1,2 @@
+AZURE_ENDPOINT=https://studystone-ocr.cognitiveservices.azure.com/
+AZURE_KEY=2ivOvI1kZFPpAQ9UeWyP8I2ZeX0T8jrBR0ESGyWEk4fkjqrs6QRPJQQJ99BIACYeBjFXJ3w3AAAFACOGy1Tj

azure_ocr.py

@@ -0,0 +1,139 @@
+import time
+import os
+import requests
+import mimetypes
+from PyPDF2 import PdfReader, PdfWriter
+import tempfile
+import re
+
+AZURE_ENDPOINT = os.environ.get("AZURE_ENDPOINT")
+AZURE_KEY = os.environ.get("AZURE_KEY")
+
+if not AZURE_ENDPOINT or not AZURE_KEY:
+    raise RuntimeError("Set AZURE_ENDPOINT and AZURE_KEY in .env")
+
+AZURE_ENDPOINT = AZURE_ENDPOINT.rstrip("/")
+
+
+def read_file_bytes(path):
+    with open(path, "rb") as f:
+        return f.read()
+
+
+def detect_content_type(file_path: str):
+    mime, _ = mimetypes.guess_type(file_path)
+    return mime or "application/octet-stream"
+
+
+def submit_read_api(file_path):
+    """Submit file to Computer Vision Read API"""
+    url = f"{AZURE_ENDPOINT}/vision/v3.2/read/analyze"
+    headers = {
+        "Ocp-Apim-Subscription-Key": AZURE_KEY,
+        "Content-Type": "application/octet-stream"
+    }
+    data = read_file_bytes(file_path)
+
+    resp = requests.post(url, headers=headers, data=data)
+    print("Azure OCR request URL:", url)
+    print("Azure OCR response status:", resp.status_code)
+    print("Azure OCR response headers:", resp.headers)
+
+    resp.raise_for_status()
+    op_location = resp.headers.get("Operation-Location")
+    if not op_location:
+        raise RuntimeError(f"No Operation-Location header. Response: {resp.text}")
+    return op_location
+
+
+def poll_read_result(operation_location, timeout=180, interval=2.0):
+    """Poll until Computer Vision OCR completes"""
+    headers = {"Ocp-Apim-Subscription-Key": AZURE_KEY}
+    deadline = time.time() + timeout
+
+    while time.time() < deadline:
+        r = requests.get(operation_location, headers=headers)
+        r.raise_for_status()
+        j = r.json()
+        status = j.get("status", "").lower()
+        if status in ("succeeded", "failed"):
+            break
+        time.sleep(interval)
+
+    if status != "succeeded":
+        raise RuntimeError(f"OCR failed. Status={status}, Response={j}")
+
+    analyze_result = j.get("analyzeResult", {})
+    lines = []
+    for read_result in analyze_result.get("readResults", []):
+        for line in read_result.get("lines", []):
+            lines.append(line["text"])
+
+    print(f"✅ Extracted {len(lines)} lines of text")
+    return "\n".join(lines)
+
+
+
+def split_pdf_into_chunks(pdf_path, chunk_size=2):
+    reader = PdfReader(pdf_path)
+    total_pages = len(reader.pages)
+    chunk_files = []
+
+    for start in range(0, total_pages, chunk_size):
+        writer = PdfWriter()
+        for p in range(start, min(start + chunk_size, total_pages)):
+            writer.add_page(reader.pages[p])
+        tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf")
+        with open(tmp.name, "wb") as f:
+            writer.write(f)
+        chunk_files.append(tmp.name)
+    return chunk_files
+
+
+def clean_extracted_text(text: str) -> str:
+    # Remove page markers
+    text = re.sub(r"--- Page.*?---", "", text)
+    # Remove chunk markers
+    text = re.sub(r"\(chunk\)", "", text)
+    # Remove junk words
+    text = re.sub(r"\b(?:stone|Stegaumen|studystone\.in)\b", "", text, flags=re.IGNORECASE)
+    # Remove roll numbers and codes
+    text = re.sub(r"Z-\d+", "", text)
+    # Remove P.T.O
+    text = re.sub(r"P\.T\.O\.", "", text, flags=re.IGNORECASE)
+
+    # Normalize per-line spacing but preserve newlines
+    lines = [re.sub(r"\s+", " ", line).strip() for line in text.splitlines()]
+    return "\n".join([l for l in lines if l])
+
+
+def poll_read_result(operation_location, timeout=180, interval=2.0):
+    headers = {"Ocp-Apim-Subscription-Key": AZURE_KEY}
+    deadline = time.time() + timeout
+
+    while time.time() < deadline:
+        r = requests.get(operation_location, headers=headers)
+        r.raise_for_status()
+        j = r.json()
+        status = j.get("status", "").lower()
+        if status in ("succeeded", "failed"):
+            break
+        time.sleep(interval)
+
+    if status != "succeeded":
+        raise RuntimeError(f"OCR failed. Status={status}, Response={j}")
+
+    analyze_result = j.get("analyzeResult", {})
+    pages = analyze_result.get("pages", [])
+    content = analyze_result.get("content", "")
+
+    pages_text = []
+    for page in pages:
+        page_num = page.get("pageNumber", "?")
+        spans = page.get("spans", [])
+        text_parts = [content[s["offset"]: s["offset"] + s["length"]] for s in spans]
+        joined = "\n".join(text_parts).strip() or "(No text detected)"
+        pages_text.append(f"--- Page {page_num} ---\n{joined}")
+
+    print(f"✅ Processed {len(pages)} pages successfully")
+    return "\n\n".join(pages_text)

static/index.html

@@ -0,0 +1,103 @@
+<!doctype html>
+<html lang="en">
+<head>
+  <meta charset="utf-8" />
+  <title>Studystone OCR Tool</title>
+  <meta name="viewport" content="width=device-width,initial-scale=1" />
+  <style>
+    body {
+      font-family: "Segoe UI", Arial, sans-serif;
+      background:#f4f6f9;
+      margin:0; padding:20px;
+      color:#222;
+    }
+    .container {
+      max-width:800px;
+      margin:20px auto;
+      background:#fff;
+      padding:20px;
+      border-radius:10px;
+      box-shadow:0 4px 10px rgba(0,0,0,0.08);
+    }
+    h1 { color:#1a73e8; margin:0 0 10px }
+    input[type="file"] { margin-top:10px }
+    button {
+      background:#1a73e8;
+      color:#fff; border:none;
+      padding:10px 14px; border-radius:6px;
+      margin-top:10px; cursor:pointer;
+    }
+    button:hover { opacity:.9 }
+    textarea {
+      width:100%; min-height:250px;
+      margin-top:20px; padding:12px;
+      border:1px solid #ccc;
+      border-radius:8px;
+      font-family:inherit; font-size:14px;
+      line-height:1.5;
+      resize:vertical;
+    }
+    #buttons { display:flex; gap:10px; margin-top:10px }
+  </style>
+</head>
+<body>
+  <div class="container">
+    <h1>Azure OCR Extractor</h1>
+    <input id="fileInput" type="file" accept=".png,.jpg,.jpeg,.pdf" />
+    <div id="buttons">
+      <button id="uploadBtn">Upload & Extract</button>
+      <button id="copyBtn" style="background:#28a745">Copy Text</button>
+      <button id="downloadBtn" style="background:#6c63ff">Download TXT</button>
+    </div>
+    <div id="status"></div>
+    <textarea id="ocrTextarea" placeholder="Extracted OCR text will appear here..."></textarea>
+  </div>
+
+<script>
+  const fileInput = document.getElementById("fileInput");
+  const uploadBtn = document.getElementById("uploadBtn");
+  const statusEl = document.getElementById("status");
+  const ocrTextarea = document.getElementById("ocrTextarea");
+  const copyBtn = document.getElementById("copyBtn");
+  const downloadBtn = document.getElementById("downloadBtn");
+
+  uploadBtn.addEventListener("click", async () => {
+    if (!fileInput.files.length) {
+      alert("Please choose a file.");
+      return;
+    }
+    statusEl.textContent = "Processing OCR...";
+    const fd = new FormData();
+    fd.append("file", fileInput.files[0]);
+    try {
+      const r = await fetch("/upload", { method:"POST", body: fd });
+      const j = await r.json();
+      if (!r.ok) {
+        statusEl.textContent = "OCR failed: " + (j.error || j.details);
+        return;
+      }
+      ocrTextarea.value = j.text || "";
+      statusEl.textContent = "Done.";
+    } catch (err) {
+      statusEl.textContent = "Error: " + err.message;
+    }
+  });
+
+  copyBtn.addEventListener("click", () => {
+    ocrTextarea.select();
+    document.execCommand("copy");
+    statusEl.textContent = "Text copied!";
+  });
+
+  downloadBtn.addEventListener("click", () => {
+    const text = ocrTextarea.value;
+    const blob = new Blob([text], { type: "text/plain" });
+    const link = document.createElement("a");
+    link.href = URL.createObjectURL(blob);
+    link.download = "ocr_text.txt";
+    link.click();
+    statusEl.textContent = "Text file downloaded.";
+  });
+</script>
+</body>
+</html>