Workflow-App/Workflow/scripts/_capture_periods.py

"""
Direct capture script for JS MINING AND ENGINEERING EMP201 periods.
Uses .click() on the EMP201 link (form loads on same page via JS).
"""
import base64
import json
import os
import re
import subprocess
import tempfile
import time


AB = "C:\\Users\\SJ\\AppData\\Roaming\\npm\\agent-browser.cmd"

PERIOD_URLS = {
    "202603": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=168198289&TaxPayerID=31142387&ReadOnly=1&Period=202603&TPReturnITID=0&TPReturnPITID=79672519&HistoryInd=1",
    "202604": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=168885394&TaxPayerID=31142387&ReadOnly=1&Period=202604&TPReturnITID=0&TPReturnPITID=80251983&HistoryInd=1",
    "202605": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=169701876&TaxPayerID=31142387&ReadOnly=1&Period=202605&TPReturnITID=0&TPReturnPITID=80776676&HistoryInd=1",
    "202606": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=170268751&TaxPayerID=31142387&ReadOnly=1&Period=202606&TPReturnITID=0&TPReturnPITID=81270492&HistoryInd=1",
    "202607": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=170816319&TaxPayerID=31142387&ReadOnly=1&Period=202607&TPReturnITID=0&TPReturnPITID=81856371&HistoryInd=1",
    "202608": "https://secure.sarsefiling.co.za/EFDotNet/LoadPage.aspx?PageCode=EMP2013LANDING&ReturnTypeID=36&ReturnGroupID=171605060&TaxPayerID=31142387&ReadOnly=1&Period=202608&TPReturnITID=0&TPReturnPITID=82341489&HistoryInd=1",
}


def run_ab(*args, timeout=30):
    tmp = tempfile.NamedTemporaryFile(mode="w", delete=False, encoding="utf-8", errors="replace", prefix="ab_", suffix=".out")
    tmp_path = tmp.name
    tmp.close()
    try:
        result = subprocess.run([AB, *args], stdout=open(tmp_path, "w", encoding="utf-8", errors="replace"), stderr=subprocess.DEVNULL, timeout=timeout)
        with open(tmp_path, "r", encoding="utf-8", errors="replace") as f:
            out = f.read().strip()
    finally:
        try: os.unlink(tmp_path)
        except OSError: pass
    if result.returncode != 0:
        raise RuntimeError(f"agent-browser {args!r} failed: {out}")
    return out


def eval_js(js):
    b64 = base64.b64encode(js.encode("utf-8")).decode("ascii")
    out = run_ab("eval", "-b", b64)
    if out == "null" or not out:
        return None
    if out.startswith('"') and out.endswith('"'):
        out = json.loads(out)
    return out


def navigate_iframe_to(url):
    js = f"document.querySelectorAll('iframe')[0].src = {json.dumps(url)}; 'ok'"
    eval_js(js)


def click_emp201_link():
    """Click the EMP201 link inside the iframe; form loads on same page via JS."""
    js = """(() => {
        const ifr = document.querySelectorAll('iframe')[0];
        const doc = ifr.contentDocument;
        const links = Array.from(doc.querySelectorAll('a'));
        // Find link with text 'EMP201' or href with BPEMP201Form
        let target = links.find(a => a.href && a.href.includes('BPEMP201Form'));
        if (!target) target = links.find(a => a.textContent.trim() === 'EMP201');
        if (!target) return JSON.stringify({ clicked: false, reason: 'no link' });
        target.click();
        return JSON.stringify({ clicked: true, href: target.href });
    })()"""
    return eval_js(js)


def read_input_xml():
    js = """(() => {
        const ifr = document.querySelectorAll('iframe')[0];
        const doc = ifr.contentDocument;
        const all = Array.from(doc.querySelectorAll('input'));
        const x = all.find(i => i.name === 'InputXML');
        return x ? x.value : '';
    })()"""
    res = eval_js(js)
    return res or ""


def parse_input_xml(xml_str):
    xml_str = (xml_str.replace("&lt;", "<").replace("&gt;", ">")
               .replace("&quot;", '"').replace("&amp;", "&"))
    def get(tag):
        m = re.search(rf"<{tag}>([^<]*)</{tag}>", xml_str)
        return m.group(1) if m else ""
    return {
        "company_name":   get("TRADING_NAME"),
        "paye_no":        get("PAYE_REF_NO"),
        "payment_period": get("PAYMENT_PERIOD"),
        "paye_payable":   get("PAYE_AMT"),
        "sdl_payable":    get("SDL_AMT"),
        "uif_payable":    get("UIF_AMT"),
        "total_payable":  get("TOTAL_PAYMENT"),
        "payment_ref_no": get("PAYMENT_REF_NO"),
    }


def main():
    periods_wanted = ["202603", "202604", "202605", "202606", "202607", "202608"]
    results = []
    for period in periods_wanted:
        if period not in PERIOD_URLS:
            continue
        href = PERIOD_URLS[period]
        print(f"\n{period}: navigating to work page...", flush=True)
        navigate_iframe_to(href)
        time.sleep(3.5)
        print(f"  clicking EMP201 link...", flush=True)
        click_emp201_link()
        # Wait longer for form to load
        for i in range(6):
            time.sleep(2)
            xml = read_input_xml()
            if xml:
                break
        if not xml:
            print(f"  {period}: no InputXML after 12s", flush=True)
            continue
        fields = parse_input_xml(xml)
        fields["period"] = period
        results.append(fields)
        print(f"  {period}: PAYE={fields['paye_payable']} SDL={fields['sdl_payable']} UIF={fields['uif_payable']} Total={fields['total_payable']} Ref={fields['payment_ref_no']}", flush=True)
        # Navigate back to work page (back to landing) so next iteration's navigate works
        # The navigate_iframe_to sets a new URL, so we don't need to click Back

    print("\n\n=== FINAL RESULTS ===", flush=True)
    print(json.dumps(results, indent=2), flush=True)
    out_path = r"C:\Users\SJ\Downloads\js_mining_emp201_results.json"
    with open(out_path, "w", encoding="utf-8") as f:
        json.dump(results, f, indent=2)
    print(f"\nSaved to {out_path}", flush=True)


if __name__ == "__main__":
    main()