Model/scripts/hyde/elmhurst_lib.py

"""Reusable Elmhurst RdSAP-10-Online automation helpers (campaign edition).

Why this exists: the per-cert build in the `expand-sap-accuracy-corpus` loop is
the slow, flaky part. `elmhurst_fill.py` hard-codes one cert's values and relies
on a *persisted* login session — but Elmhurst's auth is a **session cookie** that
Playwright's persistent context does NOT reliably restore across process launches.
This module sidesteps both problems:

  * **Log in fresh each run** (creds from env `ELM_ACCESS`/`ELM_PWD`, else a
    gitignored `scripts/hyde/.elmhurst-creds.json` {"access","pwd"}). Assessment
    DATA persists server-side via *Save & Close*, so re-login per run is fine and
    you can build/verify one page at a time across separate processes.
  * **Headed on the noVNC display** (`DISPLAY=:99`; run `start_viewer.sh` first).

Hard-won Elmhurst-DOM gotchas baked into the helpers below (see SKILL.md too):
  * Most controls AutoPostBack — set one field, wait for the postback, re-locate.
  * Sub-tabs (Doors, Air Pressure Test, Lighting, Water-Heating FGHRS…) are
    AjaxControlToolkit tabs: their fields are in the DOM but NOT visible until the
    tab span `#__tab_<panelId>` is clicked. Use `click_tab()`.
  * Delete is a Yes/No MODAL (not a native dialog): click the row delete, then
    `#..._LinkButtonYes`. Its ModalBackground intercepts clicks — use JS .click().
  * The window grid will NOT delete the LAST row, and **each Add wipes the
    previously-added row's editable cells** — so multi-row window entry via the
    add-form does not stick. Enter ONE combined window row (total glazed area,
    one orientation); per worked-ref guidance that's an accepted approximation.
  * Window glazing: "...known data" enables per-window U/g but then DEMANDS
    Frame Factor + Data-Source + Location + Building Part per row (Recommendations
    validation fails otherwise). Prefer a standard band, e.g.
    "Double post or during 2022" (auto-derives U≈1.4/g≈0.72), and still set
    Orientation + Location="External wall" + Building Part="Main".
  * The "U-value known" override checkboxes (walls/roof/floor) don't reliably
    toggle via automation — accept the natural RdSAP age-band U-value.
  * The **Recommendations** page is the validation gate; it MUST show zero errors
    before the Energy Report Summary / worksheet PDFs are downloadable.
    Common required-but-blank fields: insulated-door U-value, air-pressure-test
    certificate number (when method=Blower Door).
  * Truth only after Save & Close + reload: the live (unsaved) grid mis-displays
    older rows as 0.00 — never delete a row by its *live* width; reload first.

Usage:
    import elmhurst_lib as E
    with E.session() as (ctx, page):
        E.goto(page, "Dimensions", "WebFormDimensions.aspx")
        E.set_text(page, E.FP + "...TextBoxFloorAreaLowestFloor", "44.93")
        E.save_close(page)
"""

from __future__ import annotations

import json
import os
import re
from contextlib import contextmanager
from pathlib import Path
from typing import Callable, Generator, Optional

from playwright.sync_api import (
    BrowserContext,
    Page,
    TimeoutError as PlaywrightTimeoutError,
    sync_playwright,
)

HERE = Path(__file__).parent
SESSION_DIR = HERE / ".elmhurst-session"
CREDS_FILE = HERE / ".elmhurst-creds.json"  # gitignored; {"access":..,"pwd":..}

# The single reusable campaign assessment ("Khalim-test"). Overwrite it per UPRN.
ASSESSMENT_GUID = "B44A0DB4-4C08-4241-B818-86F060172105"
ENTRY_URL = (
    "https://rdsap10online.elmhurstenergy.co.uk/Processing/WebFormAddress.aspx"
    f"?Guid={ASSESSMENT_GUID}"
)
FP = "ContentBody_ContentPlaceHolder1_"  # control-id prefix
NAV = "ContentBody_buttonAction"  # left-rail section nav id prefix
SETTLE_MS = 700


def _creds() -> tuple[str, str]:
    if CREDS_FILE.exists():
        d = json.loads(CREDS_FILE.read_text())
        return d["access"], d["pwd"]
    access = os.environ.get("ELM_ACCESS")
    pwd = os.environ.get("ELM_PWD")
    if not access or not pwd:
        raise SystemExit(
            "No Elmhurst creds: set ELM_ACCESS/ELM_PWD or write "
            f"{CREDS_FILE} = {{'access':..,'pwd':..}}"
        )
    return access, pwd


@contextmanager
def session() -> Generator[tuple[BrowserContext, Page], None, None]:
    """Launch headed on :99, log in fresh, land on the assessment. Closes on exit."""
    with sync_playwright() as p:
        ctx = p.chromium.launch_persistent_context(
            user_data_dir=str(SESSION_DIR),
            headless=False,
            accept_downloads=True,
            viewport={"width": 1400, "height": 1000},
        )
        page = ctx.pages[0] if ctx.pages else ctx.new_page()
        page.on("dialog", lambda d: d.accept())
        page.goto(ENTRY_URL, wait_until="networkidle", timeout=60_000)
        if "WebFormLogin" in page.url:
            access, pwd = _creds()
            page.fill("#ctl00_ctl00_ContentBody_ContentBody_TextBoxAccessCode", access)
            page.fill("#ctl00_ctl00_ContentBody_ContentBody_TextBoxPassword", pwd)
            with page.expect_navigation(wait_until="networkidle", timeout=60_000):
                page.click("#ctl00_ctl00_ContentBody_ContentBody_ImageButtonEnter")
            page.wait_for_timeout(1000)
        try:
            yield ctx, page
        finally:
            ctx.close()


def commit(page: Page, action: Callable[[], object]) -> None:
    """Run a control change that triggers an ASP.NET postback; wait for it."""
    try:
        with page.expect_navigation(wait_until="load", timeout=8_000):
            action()
    except PlaywrightTimeoutError:
        page.wait_for_load_state("networkidle", timeout=8_000)
    page.wait_for_timeout(SETTLE_MS)


def goto(page: Page, nav: str, url_fragment: str) -> None:
    """Navigate to a left-rail section (no-op if already on it — re-clicking
    double-postbacks). Navigating saves the current page into the server session."""
    if url_fragment in page.url:
        return
    with page.expect_navigation(wait_until="networkidle", timeout=60_000):
        page.click(f"#{NAV}{nav}_Link", timeout=15_000)
    page.wait_for_timeout(SETTLE_MS)


def click_tab(page: Page, panel_id: str) -> None:
    """Activate an AjaxControlToolkit sub-tab by its panel id, e.g.
    'TabContainer_TabPanelDoorsPanel'. Client-side; no postback."""
    page.evaluate(
        "(id)=>{const e=document.getElementById('__tab_'+id); if(e)e.click();}",
        f"{FP}{panel_id}",
    )
    page.wait_for_timeout(500)


def set_text(page: Page, suffix: str, value: str, autopostback: bool = True) -> None:
    loc = page.locator(f"#{FP}{suffix}")
    loc.fill(value)
    if autopostback:
        commit(page, lambda: loc.press("Tab"))
    else:
        page.wait_for_timeout(150)


def set_select(page: Page, suffix: str, value: str, autopostback: bool = True) -> None:
    loc = page.locator(f"#{FP}{suffix}")
    if autopostback:
        commit(page, lambda: loc.select_option(value))
    else:
        loc.select_option(value)


def select_options(page: Page, suffix: str) -> list[tuple[str, str]]:
    """Return [(value, visible_text), …] for the <select> at FP+suffix — lets a
    sweep enumerate the *real* Elmhurst options instead of hard-coding label
    strings that drift between RdSAP releases. Skips the blank placeholder."""
    return page.evaluate(
        """(id)=>{const s=document.getElementById(id); if(!s)return [];
        return [...s.options].map(o=>[o.value,(o.text||'').trim()])
          .filter(([v,t])=>v!=='' && t!=='');}""",
        f"{FP}{suffix}",
    )


def select_by_contains(page: Page, suffix: str, needle: str, autopostback: bool = True) -> Optional[str]:
    """Select the option whose visible text contains `needle` (case-insensitive),
    by its value. Returns the chosen text, or None if no option matched — robust
    to label drift where the exact string ("CA Cavity" vs "Cavity (CA)") is
    uncertain until seen live."""
    for value, text in select_options(page, suffix):
        if needle.lower() in text.lower():
            set_select(page, suffix, value, autopostback)
            return text
    return None


def save_close(page: Page) -> None:
    """Save & Close — commits the whole server session to the DB (so the data
    survives the next fresh-login run). The body button id-suffix avoids the
    top-bar LinkButton copy. NEVER touches Submit."""
    with page.expect_navigation(wait_until="networkidle", timeout=60_000):
        page.click("#ContentBody_ButtonSaveClose", timeout=15_000)
    page.wait_for_timeout(SETTLE_MS)


# --- Window grid helpers (Openings page) ---------------------------------
_WP = "TabContainer_TabPanelWindowsPanel_"
_GRID = "GridViewExtendedWidows"


def window_row_count(page: Page) -> int:
    return page.evaluate(
        f"""()=>document.querySelectorAll("[id*='{_GRID}_DeleteButton_']").length"""
    )


def _jsclick(page: Page, element_id: str) -> None:
    page.evaluate("(id)=>{const e=document.getElementById(id); if(e)e.click();}", element_id)


def add_combined_window(
    page: Page, total_area_m2: float, orientation: str,
    glazing: str = "Double post or during 2022",
) -> None:
    """Add ONE window row of `total_area_m2` (width=area, height=1.0). Sets the
    required per-row fields (orientation/location/building-part) so the
    Recommendations validation passes. Caller then deletes the prior rows."""
    set_select(page, f"{_WP}DropDownListExtGlazing", glazing)
    page.locator(f"#{FP}{_WP}TextBoxExtWidth").fill(str(total_area_m2))
    h = page.locator(f"#{FP}{_WP}TextBoxExtHeight")
    h.fill("1.00")
    commit(page, lambda: h.press("Tab"))
    page.locator(f"#{FP}{_WP}DropDownListExtOrientation").select_option(orientation)
    page.locator(f"#{FP}{_WP}DropDownListExtBuildingPartId").select_option("Main")
    page.locator(f"#{FP}{_WP}DropDownListExtLocation").select_option("External wall")
    page.wait_for_timeout(200)
    before = window_row_count(page)
    _jsclick(page, f"{FP}{_WP}ButtonAddWindow")
    for _ in range(20):
        page.wait_for_timeout(200)
        if window_row_count(page) > before:
            break


def delete_first_window(page: Page) -> None:
    """Delete the first window row via the Yes/No modal. Will refuse to drop the
    LAST row (Elmhurst blocks it) — so add the replacement BEFORE deleting."""
    before = window_row_count(page)
    first = page.evaluate(
        f"""()=>{{const e=document.querySelector("[id*='{_GRID}_DeleteButton_']"); return e?e.id:null;}}"""
    )
    if not first:
        return
    _jsclick(page, first)
    page.wait_for_selector(f"#{FP}DeleteWindowDialog_LinkButtonYes", state="visible", timeout=5_000)
    _jsclick(page, f"{FP}DeleteWindowDialog_LinkButtonYes")
    for _ in range(20):
        page.wait_for_timeout(200)
        if window_row_count(page) < before:
            break


# --- Heating dialogs (boiler search + cascade) ---------------------------
# Hard-won (campaign certs 100021943298 / 10096028301): every Elmhurst modal
# dialog (PCDF boiler search, the SelectHeatingDialog cascade for water-heating
# method and main-heating controls) sits ABOVE a `modalBackground` overlay that
# makes Playwright's actionability hit-test see the background "intercepting
# pointer events" — so element .click()/.fill()/.select_option() all fail even
# with force. The reliable pattern is:
#   * open the dialog with a JS .click() on its button,
#   * set <select> values by JS (set .value + dispatch 'change') — this drives
#     the AutoPostBack cascade without a real click,
#   * commit by a raw `page.mouse.click(x, y)` at the OK/Select control's centre
#     (coordinate clicks land on the topmost element = the dialog, not the
#     background).
# Typing into the boiler search box needs force-click-then-keyboard (the box is
# overlapped); result cells are clicked by coordinate (the model column is
# display:none at >=1024px, so match the row by its Ref-No cell and click the
# first VISIBLE cell). Direct-writing the PCDF reference number is COSMETIC — it
# does NOT re-resolve the boiler type/efficiency; only the search dialog does.
MH1 = (
    "ContentBody_ContentPlaceHolder1_TabContainer_TabPanelMainHeating1_"
    "WebUserControlMainHeating1_"
)
HEATING_DIALOG = "ContentBody_OutsideUpdatePanel_SelectHeatingDialog_"


def dialog_commit(page: Page, label: str = "OK") -> bool:
    """Coordinate-click a dialog's commit control (a styled <span>, e.g. 'OK' or
    'Select') — bypasses the modalBackground that defeats element clicks."""
    pt = page.evaluate(
        """(lbl)=>{const e=[...document.querySelectorAll('[id*=SelectHeatingDialog] span,[id*=SelectHeatingDialog] a,[id*=SelectHeatingDialog] input,[id*=SelectBoilerDialog] span')]
        .find(x=>new RegExp('^'+lbl+'$','i').test((x.value||x.innerText||'').trim()) && x.offsetParent);
        if(!e)return null;const c=e.getBoundingClientRect();return {x:c.x+c.width/2,y:c.y+c.height/2};}""",
        label,
    )
    if not pt:
        return False
    page.mouse.click(pt["x"], pt["y"])
    page.wait_for_timeout(SETTLE_MS * 3)
    return True


def set_heating_dialog(page: Page, open_button_suffix: str, *level_regexes: str) -> None:
    """Open a SelectHeatingDialog cascade (water-heating method or main-heating
    controls) via the button at FP+`open_button_suffix`, set DropDownList1..N to
    the options matching each regex (JS, fires the cascade), then coordinate-OK.
    e.g. water (combi): set_heating_dialog(page, MH..ButtonWaterHeatingCode,
    'From Space Heating', 'From the primary heating system');
    control 2106:       set_heating_dialog(page, ..ButtonMainHeatingControls,
    '^Boilers', '^Standard', 'CBE Programmer, room thermostat and TRVs')."""
    page.evaluate("(id)=>{const e=document.getElementById(id); if(e)e.click();}", f"{FP}{open_button_suffix}")
    page.wait_for_timeout(SETTLE_MS * 3)
    for i, rx in enumerate(level_regexes, start=1):
        page.evaluate(
            """(a)=>{const s=document.getElementById(a[0]);if(!s)return;
            const o=[...s.options].find(o=>new RegExp(a[1],'i').test(o.text));
            if(o){s.value=o.value;s.dispatchEvent(new Event('change',{bubbles:true}));}}""",
            [f"{HEATING_DIALOG}DropDownList{i}", rx],
        )
        page.wait_for_timeout(SETTLE_MS * 4)  # AutoPostBack re-renders the next level
    dialog_commit(page, "OK")


def select_boiler(page: Page, query: str, ref: str) -> str:
    """Set the main-heating boiler to PCDB `ref` via the search dialog. `query`
    is a full-text search (brand/model) that surfaces `ref` on page 1. TWO-STEP
    commit: click the result ROW (highlights it) then the top 'Select' span."""
    page.evaluate("(id)=>{const e=document.getElementById(id); if(e)e.click();}", f"{MH1}ButtonBoilerReference")
    page.wait_for_selector("[id*='SelectBoilerDialog_TextBoxFullSearch']", state="visible", timeout=10_000)
    page.locator("[id*='SelectBoilerDialog_TextBoxFullSearch']").click(force=True)
    page.keyboard.type(query)
    page.locator("[id*='SelectBoilerDialog_HyperLinkActionBoilerSearch']").click(force=True)
    page.wait_for_timeout(SETTLE_MS * 4)
    pt = page.evaluate(
        """(ref)=>{const t=document.querySelector("[id*=SelectBoilerDialog] [id*=GridView]");if(!t)return null;
        for(const r of t.querySelectorAll('tr')){const tds=[...r.querySelectorAll('td')];
        if(tds.length>2 && tds[2].innerText.replace(/\\s+/g,' ').trim()===ref){
        const v=tds.find(td=>td.offsetParent&&td.getClientRects().length);
        if(v){const c=v.getBoundingClientRect();return {x:c.x+c.width/2,y:c.y+c.height/2};}}}return null;}""",
        ref,
    )
    if not pt:
        return "row-not-found"
    page.mouse.click(pt["x"], pt["y"])  # highlight the row
    page.wait_for_timeout(SETTLE_MS)
    page.locator("[id*=SelectBoilerDialog] span", has_text="Select").filter(
        visible=True
    ).first.click(force=True, timeout=8_000)
    page.wait_for_timeout(SETTLE_MS * 3)
    return "ok"


def read_default_uvalue(
    page: Page, suffix: Optional[str] = None, label_regex: str = r"Default U-value"
) -> Optional[float]:
    """Read the read-only "Default U-value [W/m²K]" the tool computes for the
    active fabric element (wall/roof/floor), after the type/insulation/thickness
    postbacks have settled.

    The override checkbox ("U-value Known") is documented as unreliable to toggle,
    but the *default* field is a plain display (a SPAN whose text IS the value,
    e.g. `…WebUserControlWallMain_LabelUValueDefault`) that RdSAP recomputes on
    every postback — that recomputed number is Elmhurst's
    age-band×insulation×thickness answer, which is exactly what the probe wants.

    Preferred: pass the exact `suffix` (after FP) of that label span. Without it,
    falls back to LABEL PROXIMITY (match `label_regex`, read the nearest numeric),
    which is brittle near the thickness box. Returns None if not found (caller
    should fall back to the screenshot)."""
    if suffix is not None:
        raw_text = page.evaluate(
            "(id)=>{const e=document.getElementById(id); return e?(e.textContent||'').trim():null;}",
            f"{FP}{suffix}",
        )
        if raw_text is not None:
            m = re.search(r"-?\d+(?:\.\d+)?", str(raw_text))
            if m:
                return float(m.group(0))
        return None
    raw = page.evaluate(
        """(rx)=>{
        const re=new RegExp(rx,'i');
        // Candidate label nodes: any visible element whose own text matches.
        const labels=[...document.querySelectorAll('label,span,td,div,th')]
          .filter(e=>e.offsetParent && re.test((e.textContent||'').trim()));
        const numFrom=(s)=>{const m=(s||'').match(/-?\\d+(?:\\.\\d+)?/); return m?m[0]:null;};
        for(const lab of labels){
          // 1) a sibling/descendant input carrying the value
          let scope=lab.closest('tr,div,td')||lab.parentElement||lab;
          const inp=scope.querySelector("input[type='text'],input:not([type])");
          if(inp && numFrom(inp.value)!==null) return inp.value;
          // 2) the next cell's text (label | value table layout)
          let sib=lab.nextElementSibling;
          for(let i=0;i<3 && sib;i++,sib=sib.nextElementSibling){
            const n=numFrom(sib.textContent);
            if(n!==null && !re.test(sib.textContent)) return n;
          }
          // 3) trailing number in the label's own container, after the label text
          const tail=numFrom(scope.textContent.replace(lab.textContent,''));
          if(tail!==null) return tail;
        }
        return null;}""",
        label_regex,
    )
    if raw is None:
        return None
    try:
        return float(str(raw).strip())
    except ValueError:
        return None


def clear_hot_water_cylinder(page: Page, panel: str = "TabContainer_TabPanelWaterHeating_") -> None:
    """Uncheck the HW cylinder (combi). The checkbox won't uncheck via
    .uncheck()/commit (a regular-boiler default re-asserts it); JS-set
    checked=false + fire click + change drives the AutoPostBack."""
    cid = f"{FP}{panel}CheckBoxHotWaterCylinder"
    if not page.locator(f"#{cid}").is_checked():
        return
    try:
        with page.expect_navigation(wait_until="load", timeout=8_000):
            page.evaluate(
                """(id)=>{const c=document.getElementById(id);if(c){c.checked=false;
                c.dispatchEvent(new Event('click',{bubbles:true}));
                c.dispatchEvent(new Event('change',{bubbles:true}));}}""",
                cid,
            )
    except PlaywrightTimeoutError:
        page.wait_for_timeout(2_000)