LWT Diary: Legal Workflow Automation Suite

TESSERACT_CONFIG = "--psm 10 --oem 3 -c tessedit_char_whitelist=0123456789"

def solve_captcha(page: Page) -> tuple[str, str]:
    page.wait_for_selector("canvas[data-testid='captcha-canvas']", timeout=DEFAULT_TIMEOUT)
    
    # Wait until canvas is actually drawn (data URL length > 1000 bytes)
    page.wait_for_function(
        "() => { const c = document.querySelector('canvas[data-testid="captcha-canvas"]'); return c && c.toDataURL().length > 1000; }",
        timeout=DEFAULT_TIMEOUT
    )
    
    img_bytes = page.locator("canvas[data-testid='captcha-canvas']").screenshot()
    img = Image.open(io.BytesIO(img_bytes)).convert('L')
    raw_text = pytesseract.image_to_string(img, config=TESSERACT_CONFIG).strip()
    cleaned_text = ''.join(filter(str.isdigit, raw_text))
    return raw_text, cleaned_text

def print_page_to_pdf(page: Page, pdf_path: str) -> bool:
    try:
        # Expand all accordion/collapse elements so all sections are visible
        page.evaluate("""() => {
            document.querySelectorAll(".collapse").forEach(el => {
                el.classList.add("show");
                el.style.display = "block";
            });
        }""")

        # Print via Chrome CDP session (works in both headed and headless modes)
        cdp = page.context.new_cdp_session(page)
        res = cdp.send("Page.printToPDF", {
            "printBackground": True,
            "preferCSSPageSize": True
        })
        pdf_bytes = base64.b64decode(res["data"])
        with open(pdf_path, "wb") as f:
            f.write(pdf_bytes)
        return True
    except Exception as e:
        print(f"Failed to print page to PDF: {e}")
        return False

case_status_data = {}

# Intercept initial case data load directly from backend API
def intercept_case_status(response):
    nonlocal case_status_data
    try:
        if "getCaseStatus" in response.url and response.status == 200:
            data = response.json()
            if data and data.get("status") == 200:
                case_status_data = data.get("data", {})
    except Exception:
        pass

page.on("response", intercept_case_status)
submit_meta = submit_form(page)

def ocr_captcha(png: bytes) -> str:
    gray = ImageOps.autocontrast(Image.open(io.BytesIO(png)).convert("L"))
    w, h = gray.size
    big = gray.resize((max(w * 4, 400), max(h * 4, 140)))
    
    variants = [
        big,
        ImageEnhance.Contrast(ImageEnhance.Sharpness(big).enhance(2.0)).enhance(1.8),
        big.point(lambda x: 0 if x < 150 else 255),
        ImageOps.invert(big),
        big.filter(ImageFilter.MedianFilter(3)).point(lambda x: 0 if x < 155 else 255),
    ]
    
    hits: List[str] = []
    for img in variants:
        for psm in (8, 7, 13):
            raw = pytesseract.image_to_string(img, config=f"--psm {psm} -c tessedit_char_whitelist={WHITELIST}") or ""
            cleaned = re.sub(r"[^0-9a-zA-Z]", "", raw)
            if len(cleaned) == 6:
                return cleaned
            if cleaned:
                hits.append(cleaned)
                
    six = [c for c in hits if len(c) == 6]
    return Counter(six).most_common(1)[0][0] if six else ""