Blame for render.py

"""Turn untrusted API JSON (via view_model) into RTL HTML. Escape first; only our tags are added.

Layout follows the 7881 debug playground: status, answer, claims table (with a verified
column), episode triage table, warnings. Styling follows the 2026-09-28 Space redesign.
"""

from __future__ import annotations

import html
import re
from urllib.parse import urlsplit

from textproc import DisplayOpts, highlight_terms, strip_fffd, trim_at_word_boundary
from view_model import EpisodeView, ViewData

_BULLET = re.compile(r"^\s*[-*\u2022]\s+(.*)$")
_NUMBERED = re.compile(r"^\s*\d+\.\s+(.*)$")
_BOLD = re.compile(r"\*\*(.+?)\*\*|__(.+?)__")
_ITALIC = re.compile(
    r"(?<!\*)\*(?!\s|\*)(.+?)(?<!\s|\*)\*(?!\*)"
    r"|(?<![\w\u0590-\u05FF])_(?!_)(.+?)(?<!_)_(?![\w\u0590-\u05FF])"
)


def _e(text: object) -> str:
    return html.escape(str(text or ""), quote=True)


def render_notice(message: str, css_class: str = "notice") -> str:
    return f'<div class="{css_class}" dir="rtl" role="status">{_e(message)}</div>'


def _inline(escaped_line: str) -> str:
    def bold(match: re.Match[str]) -> str:
        return f"<strong>{match.group(1) if match.group(1) is not None else match.group(2)}</strong>"

    def italic(match: re.Match[str]) -> str:
        return f"<em>{match.group(1) if match.group(1) is not None else match.group(2)}</em>"

    return _ITALIC.sub(italic, _BOLD.sub(bold, escaped_line))


def markdown_to_safe_html(text: str) -> str:
    """Escape HTML, then a small Markdown subset. Never emits links or raw tags."""
    lines = html.escape(strip_fffd(text).replace("\r\n", "\n").replace("\r", "\n"), quote=True).split("\n")
    blocks: list[str] = []
    i = 0
    while i < len(lines):
        if not lines[i].strip():
            i += 1
            continue
        for pattern, tag in ((_BULLET, "ul"), (_NUMBERED, "ol")):
            if pattern.match(lines[i]):
                items = []
                while i < len(lines) and pattern.match(lines[i]):
                    items.append(f"<li>{_inline(pattern.match(lines[i]).group(1))}</li>")  # type: ignore[union-attr]
                    i += 1
                blocks.append(f"<{tag}>{''.join(items)}</{tag}>")
                break
        else:
            para = []
            while i < len(lines) and lines[i].strip() and not (_BULLET.match(lines[i]) or _NUMBERED.match(lines[i])):
                para.append(_inline(lines[i]))
                i += 1
            blocks.append("<p>" + "<br>".join(para) + "</p>")
    return "".join(blocks)


def is_http_url(url: str) -> bool:
    candidate = (url or "").strip()
    if not candidate or any(ch.isspace() or ord(ch) < 32 for ch in candidate):
        return False
    try:
        parts = urlsplit(candidate)
    except ValueError:
        return False
    return parts.scheme in ("http", "https") and bool(parts.netloc)


def episode_link(ep: EpisodeView, view: dict) -> str:
    title = strip_fffd(ep.title).strip()
    if not title or is_http_url(title):
        title = str(view["missing_title"])
    inner = _e(title)
    if is_http_url(ep.url):
        return (f'<a class="episode-link" href="{_e(ep.url.strip())}" target="_blank" '
                f'rel="noopener noreferrer">{inner}</a>')
    return f'<span class="episode-name">{inner}</span>'


def _date(ep: EpisodeView) -> str:
    date = strip_fffd(ep.date).strip()
    return "" if is_http_url(date) else f'<span class="src-date" dir="ltr">{_e(date)}</span>'


def _cell(label: str, content: str, css: str = "") -> str:
    cls = f' class="{css}"' if css else ""
    return f'<td data-label="{_e(label)}"{cls}>{content}</td>'


def _table(css: str, heading: str, headers: list[str], rows: list[str]) -> str:
    head = "".join(f"<th>{_e(h)}</th>" for h in headers)
    return (f'<section class="{css}"><h2>{_e(heading)}</h2><div class="table-wrap"><table>'
            f'<thead><tr>{head}</tr></thead><tbody>{"".join(rows)}</tbody></table></div></section>')


def render_result(data: ViewData, question: str, view: dict, opts: DisplayOpts) -> str:
    status_text = view["status_labels"].get(data.status, data.status)
    parts = ['<div class="result" dir="rtl">',
             f'<div class="status status-{_e(data.status)}" role="status">'
             f'<span class="status-label">{_e(view["label_status"])}</span> {_e(status_text)}</div>']

    answer_html = markdown_to_safe_html(data.answer_markdown or "") or f"<p>{_e(view['msg_refusal'])}</p>"
    parts.append(f'<article class="answer-card"><h2>{_e(view["label_answer"])}</h2>'
                 f'<div class="answer-body">{answer_html}</div></article>')

    if data.claims:
        h = view["claims_headers"]
        rows = []
        for c in data.claims:
            quote = trim_at_word_boundary(strip_fffd(c.quote), opts.max_quote_chars, opts.ellipsis)
            quote_html = f'<q class="quote-text">{highlight_terms(quote, question, opts)}</q>'
            if c.verified:
                check = f'<span class="badge ok" title="{_e(view["verified_title"])}">✓ {_e(view["verified_badge"])}</span>'
            else:
                check = f'<span class="badge unverified" title="{_e(view["unverified_title"])}">{_e(view["unverified_badge"])}</span>'
            rows.append("<tr" + ("" if c.verified else ' class="row-unverified"') + ">"
                        + _cell(h[0], _e(c.text), "claim-text")
                        + _cell(h[1], quote_html, "claim-quote")
                        + _cell(h[2], check, "claim-verified")
                        + _cell(h[3], episode_link(c.episode, view), "col-episode")
                        + _cell(h[4], _date(c.episode), "col-date")
                        + _cell(h[5], f'<span dir="ltr">{_e(c.timestamp)}</span>', "col-time")
                        + "</tr>")
        parts.append(_table("claims", view["label_claims"], h, rows))

    if data.triage:
        h = view["triage_headers"]
        rows = []
        for t in data.triage:
            label = view["triage_labels"].get(t.label, t.label)
            rows.append("<tr>" + _cell(h[0], episode_link(t.episode, view), "col-episode") + _cell(h[1], _date(t.episode), "col-date")
                        + _cell(h[2], f'<span class="chip chip-{_e(t.label)}">{_e(label)}</span>', "col-label")
                        + _cell(h[3], _e(t.reason), "triage-reason") + "</tr>")
        parts.append(_table("triage", view["label_triage"], h, rows))

    if data.warnings:
        items = "".join(f'<li dir="auto">{_e(w)}</li>' for w in data.warnings)
        parts.append(f'<section class="warnings"><h2>{_e(view["label_warnings"])}</h2>'
                     f'<p class="warn-note">{_e(view["warnings_note"])}</p><ul>{items}</ul></section>')
    parts.append("</div>")
    return "".join(parts)