{% extends "brickwork/shell/app.html" %}
{% comment %}
A long-form, printable performance report: the document someone actually
circulates at month end, not a dashboard.

COPY THIS FILE into your project and edit it. It is not on the template loader
path, so you cannot extend it (ADR-056).

What your view must supply:
    nav_items / nav_active     as in list.html
    top_accounts_rows          bw_ranked_list rows: label/amount/value pairs,
                                a list of dicts a Django template cannot build
                                inline (see tests/test_examples.py for the
                                exact shape)
Everything else, the headline figures, the narrative, the breakdown tables and
the methodology notes, is typed straight into the template. That is what makes
a report genuinely copy-paste: replace the words and the numbers, keep the
structure.

ON PRINT. This page ships NO print stylesheet: there is not one @media print
rule anywhere in brickwork's shipped CSS (ICO-023 leaves print to the
consumer, deliberately). Printing this page today prints the app shell too:
sidebar, topbar, skip link and all. What the page DOES give you structurally,
so that adding print styles later is your only job rather than a rewrite:
every section is a real <section> under one heading, the data lives in real
<table> elements rather than styled divs, and no figure's meaning depends on
colour (every delta in this report pairs its glyph with visually hidden
direction text). A `@media print` block in your own stylesheet that hides
[data-bw-shell-chrome] siblings of #bw-main and widens bw-prose's measure is
normally enough; brickwork does not ship that block for you.

States: none of its own: a static document render. The ranked list composed
  below carries its own loading/empty states (see _ranked_list.html's header)
  if your view ever renders it that way; a circulated report is normally
  rendered once, fully populated.
Accessibility: inherits shell/app.html's skip link, sidebar/drawer nav and
  page-header region. The report body is real semantic markup throughout
  (h2/h3, p, table with scope="col"/scope="row" headers, an ordered ranked
  list), which is what bw-prose styles rather than replaces (see
  sections/content/prose-block.html's own header). Every wide table sits in
  its own scrollable, keyboard-reachable region (role="region" plus
  aria-label, tabindex="0") so a narrow viewport never scrolls the whole
  page sideways. Every trend figure pairs a decorative glyph with visually
  hidden increased/decreased/unchanged text, so no reading here rides on
  colour alone.
Responsive: no breakpoint switch of its own; inherits shell/app.html's
  sidebar-to-drawer collapse at --bw-breakpoint-md (48rem). bw-prose holds a
  fixed 65ch measure at every width; each data table scrolls horizontally
  within its own region rather than reflowing.
{% endcomment %}
{% load brickwork_components brickwork_nav %}

{% block page_title %}Q2 performance report - Northwind{% endblock %}

{% block sidebar %}{% bw_nav nav_items nav_active %}{% endblock %}
{% block sidebar_mobile %}{% bw_nav nav_items nav_active %}{% endblock %}
{% block brand_wordmark %}Northwind{% endblock %}

{% block page_header %}
  {% include "brickwork/components/_page_header.html" with title="Q2 2026 performance report" description="1 April to 30 June 2026, prepared for the leadership team." %}
{% endblock %}

{% block page_actions %}
  {% bw_button "Download PDF" variant="secondary" icon="download" href="/reports/q2-2026.pdf" %}
  {% bw_button "Share" variant="ghost" icon="external-link" href="/reports/q2-2026/share/" %}
{% endblock %}

{% block content %}
<div class="bw-section-stack">

  {% comment %}
  Executive summary. This is the paragraph a director reads if they read
  nothing else, so it carries the whole story in prose before a single table
  appears: what happened, why, and what changes next quarter.
  {% endcomment %}
  <section class="bw-content-section">
    <div class="bw-prose">
      <h2>Executive summary</h2>
      <p class="bw-prose__lede">
        Collections improved for the third straight quarter while the invoice
        book kept growing, which is the combination we set out to prove was
        possible rather than a trade-off.
      </p>
      <p>
        Revenue billed reached £612,400 across the quarter, up 9.4% on Q1. Average
        days to pay fell to 18.2 days, the lowest since we began tracking it, and
        overdue balances fell in absolute terms even as the number of invoices
        raised grew by 14%. The reminder-schedule change rolled out in March
        (see Methodology) accounts for most of the improvement: accounts on the
        new schedule pay 3.1 days faster on average than accounts still on the
        old one.
      </p>
      <p>
        The one area to watch is the Northern region, where overdue balances rose
        against the trend. That is broken out in its own section below, with the
        two accounts responsible for most of it.
      </p>
    </div>
  </section>

  {% comment %}
  Headline figures. bw-stat-grid reflows on its own; every trend here renders
  through _stat.html's composed _trend_indicator.html, which ALWAYS pairs the
  glyph with visually hidden increased/decreased/unchanged text, so the
  direction reaches a screen reader in words and never rides on green or red.

  Note what that hidden word actually says: trend names the DIRECTION THE
  NUMBER MOVED, not whether the news is good. Days to pay and overdue balance
  both FELL this quarter, so both take trend="down" even though falling is
  the outcome we wanted; the judgement goes in trend_label, in words. Passing
  trend="up" for "this improved" would announce "increased" over a figure the
  reader can see has gone down.
  {% endcomment %}
  <section class="bw-content-section">
    <div class="bw-stat-grid">
      {% include "brickwork/components/_stat.html" with label="Revenue billed" value="£612,400" trend="up" trend_label="9.4% on Q1" %}
      {% include "brickwork/components/_stat.html" with label="Invoices raised" value="1,486" trend="up" trend_label="14% on Q1" %}
      {% include "brickwork/components/_stat.html" with label="Average days to pay" value="18.2" trend="down" trend_label="2.6 days lower than Q1, which is better" %}
      {% include "brickwork/components/_stat.html" with label="Overdue balance" value="£41,900" trend="down" trend_label="6.7% lower than Q1, which is better" %}
    </div>
  </section>

  {% comment %}
  Regional breakdown. Narrative first, then the table it is describing: a
  report earns its tables rather than opening with one.
  {% endcomment %}
  <section class="bw-content-section">
    <div class="bw-prose">
      <h2>Revenue and collections by region</h2>
      <p>
        Every region grew revenue this quarter. Collections speed improved
        everywhere except the North, where two accounts (below) are carrying
        most of an increase that would otherwise read as a regional problem
        rather than a concentrated one.
      </p>

      {% comment %}
      TWO tables, not one five-column table, and this is a real constraint
      rather than a stylistic choice. A table cannot compress below the widest
      unbreakable content it holds, so five columns of currency and day counts
      have an intrinsic minimum around 425px: wider than a 375px phone. The
      wrapper below scrolls, but a table that wide still drags the DOCUMENT
      wider than the viewport, which is the WCAG 1.4.10 failure the archetype
      harness measures at 375px. Shortening the words does not rescue it
      (measured: 425px to 422px, three pixels); the column COUNT is what
      binds. Two three-column tables each measure about 327px and fit.

      Keep this in mind when you edit: adding a column here is not free, and
      the page will fail its own gate at 375px before it looks wrong to you
      on a laptop.
      {% endcomment %}
      <div class="bw-prose__table-wrap" tabindex="0" role="region" aria-label="Revenue by region, Q2 2026">
        <table>
          <caption>Revenue billed by region, Q2 2026 versus Q1 2026.</caption>
          <thead>
            <tr>
              <th scope="col">Region</th>
              <th scope="col">Revenue billed</th>
              <th scope="col">Change on Q1</th>
            </tr>
          </thead>
          <tbody>
            <tr>
              <th scope="row">South</th>
              <td>£248,100</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="up" trend_label="11.2%" %}
              </td>
            </tr>
            <tr>
              <th scope="row">Midlands</th>
              <td>£189,700</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="up" trend_label="8.6%" %}
              </td>
            </tr>
            <tr>
              <th scope="row">North</th>
              <td>£132,400</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="up" trend_label="6.9%" %}
              </td>
            </tr>
            <tr>
              <th scope="row">Wales and South West</th>
              <td>£42,200</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="up" trend_label="10.5%" %}
              </td>
            </tr>
          </tbody>
        </table>
      </div>

      <div class="bw-prose__table-wrap" tabindex="0" role="region" aria-label="Collections speed by region, Q2 2026">
        <table>
          <caption>Average days to pay by region, Q2 2026 versus Q1 2026.</caption>
          <thead>
            <tr>
              <th scope="col">Region</th>
              <th scope="col">Avg. days to pay</th>
              <th scope="col">Change on Q1</th>
            </tr>
          </thead>
          <tbody>
            <tr>
              <th scope="row">South</th>
              <td>16.4 days</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="down" trend_label="3.0 days lower, better" %}
              </td>
            </tr>
            <tr>
              <th scope="row">Midlands</th>
              <td>17.9 days</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="down" trend_label="2.1 days lower, better" %}
              </td>
            </tr>
            <tr>
              <th scope="row">North</th>
              <td>22.6 days</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="up" trend_label="1.4 days higher, worse" %}
              </td>
            </tr>
            <tr>
              <th scope="row">Wales and South West</th>
              <td>15.1 days</td>
              <td>
                {% include "brickwork/components/_trend_indicator.html" with trend="down" trend_label="1.9 days lower, better" %}
              </td>
            </tr>
          </tbody>
        </table>
      </div>

      <p>
        Read the North row alongside the account breakdown below before drawing
        a regional conclusion: this is a concentration story, not a systemic
        slowdown in that region's chasing.
      </p>
    </div>
  </section>

  {% comment %}
  Top overdue accounts. bw_ranked_list is the right shape for "which N things
  contribute most", here basis="max" so the bar reads as a leaderboard against
  the largest single balance. The package never formats numbers, so "value" is
  supplied pre-formatted by the view; see top_accounts_rows in this file's own
  header.
  {% endcomment %}
  <section class="bw-content-section">
    <div class="bw-prose">
      <h2>Where the North's overdue balance sits</h2>
      <p>
        Two accounts hold most of the North region's overdue total. Both are
        past the final reminder stage and are now with the account owner
        rather than the automated schedule.
      </p>
    </div>
    {% bw_ranked_list top_accounts_rows basis="max" label="Top overdue accounts by balance, North region, Q2 2026" %}
  </section>

  {% comment %}
  Methodology and footnotes tail. Every report worth circulating says how the
  numbers were produced, so a reader can trust or challenge them.
  {% endcomment %}
  <section class="bw-content-section">
    <div class="bw-prose">
      <h2>Methodology</h2>
      <p>
        Revenue billed is invoice value raised within the quarter, exclusive of
        VAT. Average days to pay is the mean of (payment date minus due date)
        across invoices settled within the quarter, so a late invoice still
        paid this quarter counts, and one paid next quarter does not yet. Overdue
        balance is the outstanding value of invoices past their due date as at
        30 June 2026.
      </p>
      <p>
        The revised reminder schedule (first reminder seven days before the due
        date, rather than three) rolled out to 60% of accounts on 3 March 2026
        and to the remainder on 14 April 2026. The 3.1-day improvement figure in
        the summary compares accounts on the new schedule for the full quarter
        against accounts still on the old one.
      </p>
      <hr>
      <p>
        Prepared by the finance operations team. Figures are unaudited management
        figures, not statutory accounts. Questions to
        <a href="mailto:financeops@northwind.example">financeops@northwind.example</a>.
      </p>
    </div>
  </section>

</div>
{% endblock %}