Files
BrickTracker/bricktracker/sidecar_set.py
T

295 lines
10 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import logging
import re
from datetime import datetime, timezone
from html.parser import HTMLParser
from typing import Any
from flask import current_app
from markupsafe import Markup, escape
from .sidecar import BrickSidecar
logger = logging.getLogger(__name__)
# Brickset descriptions ship as HTML (<p>, <ul>, <li>, <i>, ...). Keep only a
# small allowlist of formatting tags and drop every attribute so the markup can
# be rendered as-is without opening an XSS hole.
_ALLOWED_TAGS = frozenset({
'p', 'br', 'ul', 'ol', 'li', 'b', 'strong', 'i', 'em', 'u',
'h3', 'h4', 'h5', 'h6', 'small', 'sub', 'sup',
})
_VOID_TAGS = frozenset({'br'})
# Build a normalized, template-friendly summary of the sidecar data for a set:
# enrichment (description, dimensions, retired, ...) plus the three-way price
# comparison (paid / retail MSRP / BrickLink market value).
#
# Metadata comes from the persistent cache (a local read, effectively
# permanent). The BrickLink
# market value is read from cache only here (no network) so a page render never
# blocks; refreshing it is an explicit user action. Returns None when the
# sidecar is disabled or has no data for the set.
def summarize(
set_ref: str,
/,
*,
purchase_price: float | None = None,
fetch_price: bool = False,
) -> dict[str, Any] | None:
if not BrickSidecar.enabled():
return None
data = BrickSidecar.get_set(set_ref)
if data is None:
return None
summary: dict[str, Any] = {
'description': _description(data.get('description')),
'pieces': data.get('pieces'),
'minifigs': data.get('minifigs'),
'year': data.get('year'),
'theme': _clean_str(data.get('theme')),
'subtheme': _clean_str(data.get('subtheme')),
# Set designer(s); web-scraped, the API does not expose it.
'designer': _clean_str(data.get('designer')),
'dimensions': _dimensions(data),
'weight': _weight(data),
'instructions_count': data.get('instructionsCount'),
'additional_image_count': _to_int(data.get('additionalImageCount')),
# Carousel + extra cover sources are opt-in; off keeps the plain cover.
'additional_images_enabled': bool(
current_app.config['SIDECAR_ADDITIONAL_IMAGES']
),
'tags': _tags(data.get('tags')),
# Brickset "Notes" blurb (web-scraped; contains light HTML). Sanitised
# like the description so it renders safely.
'notes': _description(data.get('notes')),
}
# Retired status from the exit date. exit_date is the full formatted date
# (BrickData/Brickset has it even for sets retired long ago, which the
# upcoming-retirements CSV does not cover, see #37).
retired, exit_year = _retired(data.get('exitDate'))
summary['retired'] = retired
summary['exit_year'] = exit_year
summary['exit_date'] = _format_date(data.get('exitDate'))
summary['launch_year'] = _year(data.get('launchDate'))
# --- Prices: paid / retail / market --------------------------------
msrp = BrickSidecar.retail_price(data)
# When auto-fetch is on, hit the TTL-aware path (network only on a cache
# miss/expiry); otherwise stay cache-only so the render never blocks.
if fetch_price:
price_payload = BrickSidecar.get_price(set_ref)
price_fetched_at = None
else:
price_payload, price_fetched_at = BrickSidecar.cached_price(set_ref)
market_new = _to_float(price_payload.get('new_avg')) if price_payload else None
market_used = _to_float(price_payload.get('used_avg')) if price_payload else None
paid = _to_float(purchase_price)
prices: dict[str, Any] = {
'paid': paid,
'msrp': msrp,
'msrp_currency': BrickSidecar.retail_currency(),
# Inflation-adjusted RRP, web-scraped by the sidecar (single value in
# its own currency, independent of the configured retail region).
'msrp_inflated': _to_float(data.get('rrpInflated')),
'msrp_inflated_currency': _clean_str(data.get('rrpInflatedCurrency')),
'market_new': market_new,
'market_used': market_used,
'market_min': _to_float(price_payload.get('new_min')) if price_payload else None, # noqa: E501
'market_max': _to_float(price_payload.get('new_max')) if price_payload else None, # noqa: E501
'market_used_min': _to_float(price_payload.get('used_min')) if price_payload else None, # noqa: E501
'market_used_max': _to_float(price_payload.get('used_max')) if price_payload else None, # noqa: E501
'market_currency': (price_payload or {}).get('currency_code'),
'market_fetched_at': (price_payload or {}).get('fetched_at') or price_fetched_at, # noqa: E501
'has_market': price_payload is not None,
}
# Savings vs MSRP (positive = paid less than retail).
if msrp is not None and paid is not None:
prices['savings_vs_msrp'] = round(msrp - paid, 2)
# Value movement vs what was paid. SIDECAR_PRICE_BASIS picks which market
# average to compare against ('used' or 'new'); fall back to the other when
# the preferred one is missing.
basis = str(current_app.config.get('SIDECAR_PRICE_BASIS', 'used')).lower()
if basis == 'new':
market_ref = market_new if market_new is not None else market_used
else:
market_ref = market_used if market_used is not None else market_new
prices['market_basis'] = basis
if market_ref is not None and paid is not None:
prices['gain_vs_paid'] = round(market_ref - paid, 2)
summary['prices'] = prices
return summary
# --- Helpers ------------------------------------------------------------
def _clean_str(value: Any) -> str | None:
if value is None:
return None
text = str(value).strip()
return text or None
# Rebuild an HTML fragment keeping only the allowlisted formatting tags, with
# all attributes stripped and all text escaped. convert_charrefs=True turns
# entities (&reg;, &#39;, ...) into real characters before they reach the data
# handler, so they render correctly.
class _DescriptionSanitizer(HTMLParser):
def __init__(self) -> None:
super().__init__(convert_charrefs=True)
self.parts: list[str] = []
def handle_starttag(self, tag: str, attrs: Any) -> None:
if tag in _ALLOWED_TAGS:
self.parts.append(f'<{tag}>')
def handle_startendtag(self, tag: str, attrs: Any) -> None:
if tag in _ALLOWED_TAGS:
self.parts.append(f'<{tag}>')
def handle_endtag(self, tag: str) -> None:
if tag in _ALLOWED_TAGS and tag not in _VOID_TAGS:
self.parts.append(f'</{tag}>')
def handle_data(self, data: str) -> None:
self.parts.append(str(escape(data)))
def _description(value: Any) -> Markup | None:
text = _clean_str(value)
if text is None:
return None
parser = _DescriptionSanitizer()
parser.feed(text)
parser.close()
cleaned = ''.join(parser.parts)
# Brickset pads descriptions with empty spacer paragraphs (<p>&nbsp;</p>);
# drop those so the blocks sit flush instead of leaving big gaps. \s also
# matches the &nbsp; (\xa0) the parser leaves behind.
cleaned = re.sub(r'<p>\s*</p>', '', cleaned).strip()
return Markup(cleaned) if cleaned else None
def _to_float(value: Any) -> float | None:
if value is None or value == '':
return None
try:
return float(value)
except (TypeError, ValueError):
return None
def _to_int(value: Any) -> int | None:
if value is None or value == '':
return None
try:
return int(value)
except (TypeError, ValueError):
return None
def _dimensions(data: dict[str, Any], /) -> str | None:
height = _to_float(data.get('height'))
width = _to_float(data.get('width'))
depth = _to_float(data.get('depth'))
if height is None and width is None and depth is None:
return None
parts = [
'{0:g}'.format(round(value, 1)) if value is not None else '?'
for value in (height, width, depth)
]
return '{0} cm'.format(' × '.join(parts))
def _weight(data: dict[str, Any], /) -> str | None:
weight = _to_float(data.get('weight'))
if weight is None:
return None
return '{0:g} kg'.format(round(weight, 2))
def _tags(value: Any) -> list[str]:
if not value:
return []
return [tag.strip() for tag in str(value).split(',') if tag.strip()]
def _parse_date(value: Any) -> datetime | None:
if not value:
return None
text = str(value).replace('Z', '+00:00')
try:
parsed = datetime.fromisoformat(text)
except ValueError:
return None
if parsed.tzinfo is None:
parsed = parsed.replace(tzinfo=timezone.utc)
return parsed
def _year(value: Any) -> int | None:
parsed = _parse_date(value)
return parsed.year if parsed is not None else None
def _format_date(value: Any) -> str | None:
parsed = _parse_date(value)
if parsed is None:
return None
return parsed.strftime(current_app.config['PURCHASE_DATE_FORMAT'])
def _retired(exit_date: Any) -> tuple[bool, int | None]:
parsed = _parse_date(exit_date)
if parsed is None:
return False, None
return parsed < datetime.now(timezone.utc), parsed.year
# Bulk retirement lookup for a list of set refs (e.g. the whole wishlist) in a
# single cached BrickData call. Returns {ref: {'date': str|None,
# 'retired': bool}} keyed by both "number-variant" and the bare number so the
# caller can match whichever form it stores. Empty dict when the sidecar is
# disabled or unreachable, so callers degrade to the CSV (see #37).
def retirement_dates(
set_refs: Any,
/,
) -> dict[str, dict[str, Any]]:
if not BrickSidecar.enabled():
return {}
refs = list({str(ref) for ref in set_refs if ref})
if not refs:
return {}
try:
bulk = BrickSidecar.get_sets_bulk(refs, price=False)
except Exception as exception:
logger.debug('sidecar retirement bulk fetch failed: %s', exception)
return {}
out: dict[str, dict[str, Any]] = {}
for ref, data in bulk.items():
retired, _ = _retired(data.get('exitDate'))
info = {
'date': _format_date(data.get('exitDate')),
'retired': retired,
}
out[ref] = info
number, _, _ = ref.partition('-')
out.setdefault(number, info)
return out