Fix: Tier 2 numbered list, sub-bullet checkmarks, Relocation Cost page, ISP FINANCING, green NB

app/services/normalizer.py

@@ -129,6 +129,12 @@ def normalize_section(
     # ── doc_v1 ──
     if layout_norm == "doc_v1" and isinstance(section_json.get("blocks"), list):
         blocks.extend(_normalize_doc_v1(section_json["blocks"], skip_title=title))
+        # Post-process breakdown section for Relocation Cost layout
+        if key_norm == "program_features_breakdown":
+            blocks = _postprocess_breakdown(blocks, section_json["blocks"])
+        # Post-process Tier 2 section for sub-bullet styling
+        if key_norm == "summary_of_universities_cosigner":
+            blocks = _postprocess_tier2(blocks)
         return blocks
 
     # ── Fallback ──
@@ -513,6 +519,230 @@ def _normalize_table_v2(json_data: dict) -> RenderBlock:
     )
 
 
+# ───────────────────────────────────────────────────────────────
+# Breakdown section post-processor
+# ───────────────────────────────────────────────────────────────
+
+def _postprocess_breakdown(
+    blocks: list[RenderBlock],
+    raw_blocks: list,
+) -> list[RenderBlock]:
+    """Rewrite the breakdown section to match the reference layout.
+
+    - "Relocation Cost" becomes a banner heading with page-break-before
+    - The relocation table gets a merged right cell (rowspan) with the
+      cost-coverage note moved inside it
+    - "ISP FINANCING" becomes an inline note with mixed bold/italic
+    - "NB: CREDIT FACILITY" is styled green
+    - Dollar amounts in parentheticals keep their original $ format
+    """
+    from markupsafe import Markup
+
+    # Find raw blocks for the relocation cost table (pre-normalised, $ intact)
+    raw_reloc_table = None
+    raw_note_after_table = None
+    found_reloc = False
+    for i, rb in enumerate(raw_blocks):
+        if not isinstance(rb, dict):
+            continue
+        if rb.get("type") == "subheading" and "relocation" in str(rb.get("text", "")).lower():
+            found_reloc = True
+            continue
+        if found_reloc and rb.get("type") == "table_v1" and raw_reloc_table is None:
+            raw_reloc_table = rb
+            continue
+        if found_reloc and raw_reloc_table and rb.get("type") == "paragraph" and raw_note_after_table is None:
+            raw_note_after_table = rb
+            break
+
+    result: list[RenderBlock] = []
+    i = 0
+    while i < len(blocks):
+        blk = blocks[i]
+
+        # ── Detect "Relocation Cost" heading ──
+        if (blk.block_type == "heading_2"
+                and "relocation" in blk.data.get("text", "").lower()):
+
+            # Banner heading with page break
+            result.append(RenderBlock(
+                block_type="heading_2",
+                css_class="hb-heading-2 hb-banner-heading page-break",
+                data={"text": blk.data["text"]},
+            ))
+            i += 1
+
+            # Replace the next table with spanning variant that has merged cell
+            if i < len(blocks) and blocks[i].block_type == "table" and raw_reloc_table:
+                raw_rows = raw_reloc_table.get("rows", [])
+                # Build the note text for the merged right cell
+                note_text = ""
+                if raw_note_after_table:
+                    note_text = str(raw_note_after_table.get("text", ""))
+
+                spanning_rows = _build_relocation_spanning_rows(raw_rows, note_text)
+                result.append(RenderBlock(
+                    block_type="table",
+                    css_class="hb-table hb-relocation-table",
+                    data={"rows": spanning_rows, "variant": "spanning"},
+                ))
+                i += 1  # skip the original table
+
+                # Skip the paragraph that was moved into the merged cell
+                if (i < len(blocks)
+                        and blocks[i].block_type == "paragraph"
+                        and note_text):
+                    i += 1
+                continue
+
+        # ── "ISP FINANCING" heading → inline note with mixed formatting ──
+        if (blk.block_type == "heading_2"
+                and "isp financing" in blk.data.get("text", "").lower()):
+            # Next block should be the interest rate paragraph
+            rate_text = ""
+            if i + 1 < len(blocks) and blocks[i + 1].block_type == "paragraph":
+                rate_text = blocks[i + 1].data.get("text", "")
+            result.append(RenderBlock(
+                block_type="note",
+                css_class="hb-note hb-isp-financing",
+                data={
+                    "parts": [
+                        {"text": "ISP FINANCING", "style": "bold"},
+                        {"text": " (" + _extract_rate_italic(rate_text) + "): " if rate_text else "", "style": "italic"},
+                        {"text": _extract_rate_amount(rate_text), "style": "bold"},
+                    ],
+                    "inline": True,
+                },
+            ))
+            i += 1  # skip the heading
+            if rate_text:
+                i += 1  # skip the paragraph
+            continue
+
+        # ── "NB: CREDIT FACILITY" note → green styling ──
+        if (blk.block_type == "note"
+                and "credit facility" in blk.data.get("text", "").lower()):
+            result.append(RenderBlock(
+                block_type="note",
+                css_class="hb-note hb-credit-note",
+                data=blk.data,
+            ))
+            i += 1
+            continue
+
+        result.append(blk)
+        i += 1
+
+    return result
+
+
+def _build_relocation_spanning_rows(
+    raw_rows: list, note_text: str,
+) -> list[list[dict]]:
+    """Build spanning rows for the relocation cost table.
+
+    Row 0: normal 2-column (consultation fees | Covered in the contribution)
+    Rows 1-7: left cell per row, right cell merged (rowspan) with italic note
+    Rows 8+: left cell only, empty right
+    """
+    from markupsafe import Markup
+
+    if not raw_rows:
+        return []
+
+    rows: list[list[dict]] = []
+
+    # Row 0 — has "Covered in the contribution"
+    first = raw_rows[0] if raw_rows else ["", ""]
+    rows.append([
+        {"text": Markup(emphasize_keywords(str(first[0] if len(first) > 0 else ""))), "colspan": 1, "rowspan": 1},
+        {"text": Markup("<em>" + h(str(first[1] if len(first) > 1 else "")) + "</em>"), "colspan": 1, "rowspan": 1},
+    ])
+
+    # Rows 1-7: items with dollar amounts that get the merged right cell
+    # These are the visa/fee/rent/ticket rows (have parenthetical dollar amounts)
+    merged_start = 1
+    merged_end = min(8, len(raw_rows))  # Visa Integrity through Air ticket
+
+    for idx in range(merged_start, len(raw_rows)):
+        cell_text = str(raw_rows[idx][0] if len(raw_rows[idx]) > 0 else "")
+        left = {"text": Markup(emphasize_keywords(cell_text)), "colspan": 1, "rowspan": 1}
+
+        if idx == merged_start and note_text:
+            # First merged row gets the rowspan cell
+            span_count = merged_end - merged_start
+            note_html = note_text.replace("\n\n", "<br/><br/>")
+            right = {
+                "text": Markup('<em class="hb-merged-note">' + h(note_html).replace("&lt;br/&gt;&lt;br/&gt;", "<br/><br/>") + "</em>"),
+                "colspan": 1,
+                "rowspan": span_count,
+            }
+            rows.append([left, right])
+        elif idx < merged_end:
+            # Subsequent merged rows — no right cell (covered by rowspan)
+            rows.append([left])
+        else:
+            # Remaining rows — empty right cell
+            rows.append([
+                left,
+                {"text": "", "colspan": 1, "rowspan": 1},
+            ])
+
+    return rows
+
+
+def _extract_rate_italic(text: str) -> str:
+    """Extract the italic portion: 'Interest rate of 12% – 15% Market Rate PA'."""
+    # Text is like: "Interest rate of 12% – 15% Market Rate: UP TO USD 10,000"
+    m = re.match(r"(Interest rate.*?(?:Market Rate|PA))", text, re.IGNORECASE)
+    if m:
+        return m.group(1).rstrip(": ")
+    # Fallback: everything before the colon
+    if ":" in text:
+        return text.split(":")[0].strip()
+    return text
+
+
+def _extract_rate_amount(text: str) -> str:
+    """Extract the amount portion: 'UP TO USD 10,000'."""
+    m = re.search(r"(UP TO.*)", text, re.IGNORECASE)
+    if m:
+        return m.group(1).strip()
+    if ":" in text:
+        return text.split(":", 1)[1].strip()
+    return ""
+
+
+# ───────────────────────────────────────────────────────────────
+# Tier 2 (cosigner) section post-processor
+# ───────────────────────────────────────────────────────────────
+
+def _postprocess_tier2(blocks: list[RenderBlock]) -> list[RenderBlock]:
+    """Style the Tier 2 section to match the reference layout.
+
+    - Second consecutive bullet_list (sub-bullets under Sources of Funds)
+      gets checkmark styling instead of arrows.
+    """
+    result: list[RenderBlock] = []
+    prev_was_bullet = False
+    for blk in blocks:
+        if blk.block_type == "bullet_list":
+            if prev_was_bullet:
+                # This is the sub-bullet list → use checkmark class
+                result.append(RenderBlock(
+                    block_type="bullet_list",
+                    css_class="hb-bullet-list hb-sub-bullets",
+                    data=blk.data,
+                ))
+            else:
+                result.append(blk)
+            prev_was_bullet = True
+        else:
+            prev_was_bullet = False
+            result.append(blk)
+    return result
+
+
 def _normalize_doc_v1(blocks: list, *, skip_title: str = "") -> list[RenderBlock]:
     """Normalise doc_v1 blocks into typed RenderBlocks.
 

app/static/css/print.css

@@ -215,7 +215,7 @@ ul.hb-bullet-list li::before {
 .hb-numbered-list,
 .ol,
 ol.hb-numbered-list {
-    list-style: decimal;
+    list-style: decimal !important;
     margin: 0 0 8pt 18pt;
     padding: 0;
     font-size: 9pt;
@@ -234,7 +234,18 @@ ol.hb-numbered-list li {
 .hb-numbered-list li::before,
 .ol li::before,
 ol.hb-numbered-list li::before {
-    content: none;
+    content: none !important;
+}
+
+/* Sub-bullets — checkmarks, indented under parent bullet */
+.hb-sub-bullets {
+    margin-left: 32pt !important;
+}
+
+.hb-sub-bullets li::before {
+    content: "\2713" !important;
+    color: #000000;
+    font-size: 9pt;
 }
 
 /* ------------------------------
@@ -1088,4 +1099,75 @@ table.programs td:nth-child(5) {
 .page-content img {
     max-width: 100%;
     height: auto;
+}
+
+/* ------------------------------
+   BREAKDOWN — RELOCATION COST
+   ------------------------------ */
+
+/* Banner heading (teal background bar) */
+.hb-banner-heading {
+    background-color: #199970;
+    color: #FFFFFF !important;
+    padding: 6pt 10pt;
+    margin: 10pt 0 6pt;
+    font-size: 11pt;
+    font-weight: 700;
+}
+
+/* Page break before Relocation Cost */
+.sec-breakdown .page-break {
+    page-break-before: always;
+    break-before: page;
+}
+
+/* Relocation cost table — merged note cell */
+.hb-relocation-table {
+    margin: 0 0 10pt;
+}
+
+.hb-relocation-table td {
+    vertical-align: top;
+    padding: 3pt 6pt;
+    border: 1pt solid #CCCCCC;
+    font-size: 9.5pt;
+}
+
+.hb-merged-note {
+    font-style: italic;
+    font-weight: 400;
+    font-size: 9.5pt;
+    line-height: 1.4;
+    display: block;
+    padding: 4pt 2pt;
+}
+
+/* ISP FINANCING line */
+.hb-isp-financing {
+    margin: 14pt 0 6pt;
+    padding: 6pt 0;
+    border-top: 1.5pt solid #000000;
+    border-bottom: 1.5pt solid #000000;
+    text-align: center;
+    font-size: 10pt;
+}
+
+.hb-isp-financing strong {
+    font-weight: 700;
+    color: #000000;
+}
+
+.hb-isp-financing em {
+    font-style: italic;
+    color: #199970;
+    font-weight: 700;
+}
+
+/* NB: CREDIT FACILITY — green */
+.hb-credit-note {
+    text-align: center;
+    color: #199970;
+    font-size: 10pt;
+    font-weight: 700;
+    margin: 6pt 0;
 }

app/templates/partials/blocks/heading.html

@@ -1,6 +1,6 @@
 {# Block partial: heading_1 / heading_2 #}
 {% if block.block_type == 'heading_1' %}
-<h2 class="hb-heading-1">{{ block.data.text | e }}</h2>
+<h2 class="{{ block.css_class | default('hb-heading-1') }}">{{ block.data.text | e }}</h2>
 {% elif block.block_type == 'heading_2' %}
-<h3 class="hb-heading-2">{{ block.data.text | e }}</h3>
+<h3 class="{{ block.css_class | default('hb-heading-2') }}">{{ block.data.text | e }}</h3>
 {% endif %}

app/templates/partials/blocks/note.html

@@ -1,16 +1,20 @@
 {# Block partial: note (standalone or inline-parts) #}
 {% if block.data.inline and block.data.parts %}
-<div class="hb-note">
+<div class="{{ block.css_class | default('hb-note') }}">
     {% for part in block.data.parts %}
     {% if part.style == 'red_bold' %}
     <strong class="hb-note-keyword">{{ part.text | e }}</strong>
+    {% elif part.style == 'bold' %}
+    <strong>{{ part.text | e }}</strong>
+    {% elif part.style == 'italic' %}
+    <em>{{ part.text | e }}</em>
     {% else %}
     {{ part.text | e }}
     {% endif %}
     {% endfor %}
 </div>
 {% else %}
-<div class="hb-note">
+<div class="{{ block.css_class | default('hb-note') }}">
     {% set text = block.data.text | default('') %}
     {# Highlight NOTE / ONLY IF keywords in bold + red; rest stays bold via CSS #}
     {% if text.upper().startswith('NOTE:') %}