add segmented listing sync
This commit is contained in:
@@ -133,9 +133,9 @@ class ListingCollector:
|
||||
|
||||
return not old_first_href
|
||||
|
||||
def open_cars_listing(self, page: Page) -> None:
|
||||
logger.info("Opening cars listing page: %s", self.settings.listing.cars_url)
|
||||
url = self.settings.listing.cars_url
|
||||
def open_cars_listing(self, page: Page, *, url_override: str | None = None) -> None:
|
||||
url = url_override or self.settings.listing.cars_url
|
||||
logger.info("Opening cars listing page: %s", url)
|
||||
last_err = None
|
||||
for attempt in range(3):
|
||||
try:
|
||||
@@ -202,16 +202,82 @@ class ListingCollector:
|
||||
# Короткая пауза вместо длинного sleep.
|
||||
time.sleep(1.0)
|
||||
|
||||
def apply_filters(self, page: Page, make: str | None = None, model: str | None = None) -> dict[str, str | None]:
|
||||
applied = {"make": None, "model": None}
|
||||
def apply_filters(
|
||||
self,
|
||||
page: Page,
|
||||
make: str | None = None,
|
||||
model: str | None = None,
|
||||
year_min: int | None = None,
|
||||
year_max: int | None = None,
|
||||
) -> dict[str, str | int | None]:
|
||||
applied: dict[str, str | int | None] = {"make": None, "model": None, "year_min": None, "year_max": None}
|
||||
if make and self._try_fill_filter_input(page, ["input[placeholder*='Make']", "input[aria-label*='Make']"], make):
|
||||
applied["make"] = make
|
||||
self.pacer.after_filter_action()
|
||||
if model and self._try_fill_filter_input(page, ["input[placeholder*='Model']", "input[aria-label*='Model']"], model):
|
||||
applied["model"] = model
|
||||
self.pacer.after_filter_action()
|
||||
if year_min is not None or year_max is not None:
|
||||
if self._apply_year_range(page, year_min, year_max):
|
||||
applied["year_min"] = year_min
|
||||
applied["year_max"] = year_max
|
||||
self.pacer.after_filter_action()
|
||||
return applied
|
||||
|
||||
def _apply_year_range(self, page: Page, year_min: int | None, year_max: int | None) -> bool:
|
||||
"""Заполняет поля фильтра Year и нажимает Apply Year."""
|
||||
if year_min is None and year_max is None:
|
||||
return False
|
||||
try:
|
||||
success = page.evaluate(
|
||||
"""([yearMin, yearMax]) => {
|
||||
const inputs = Array.from(document.querySelectorAll('input'));
|
||||
const yearInputs = inputs.filter(inp => {
|
||||
const v = parseInt(inp.value, 10);
|
||||
return !isNaN(v) && v >= 1900 && v <= 2100;
|
||||
});
|
||||
if (yearInputs.length < 2) return false;
|
||||
yearInputs.sort((a, b) => parseInt(a.value) - parseInt(b.value));
|
||||
const setVal = (el, val) => {
|
||||
const setter = Object.getOwnPropertyDescriptor(
|
||||
HTMLInputElement.prototype, 'value'
|
||||
).set;
|
||||
setter.call(el, String(val));
|
||||
el.dispatchEvent(new Event('input', {bubbles: true}));
|
||||
el.dispatchEvent(new Event('change', {bubbles: true}));
|
||||
};
|
||||
if (yearMin !== null) setVal(yearInputs[0], yearMin);
|
||||
if (yearMax !== null) setVal(yearInputs[yearInputs.length - 1], yearMax);
|
||||
const container = yearInputs[0].closest(
|
||||
'[class*="filter"], [class*="year"], section, fieldset'
|
||||
) || yearInputs[0].parentElement.parentElement;
|
||||
if (container) {
|
||||
const btn = Array.from(container.querySelectorAll(
|
||||
'button, a, [role="button"], span[class*="apply"]'
|
||||
)).find(el => /apply|\u043f\u0440\u0438\u043c\u0435\u043d/i.test(el.textContent));
|
||||
if (btn) { btn.click(); return true; }
|
||||
}
|
||||
yearInputs[yearInputs.length - 1].dispatchEvent(
|
||||
new KeyboardEvent('keydown', {
|
||||
key: 'Enter', code: 'Enter', keyCode: 13, bubbles: true
|
||||
})
|
||||
);
|
||||
return true;
|
||||
}""",
|
||||
[year_min, year_max],
|
||||
)
|
||||
if success:
|
||||
try:
|
||||
page.wait_for_load_state("domcontentloaded", timeout=15_000)
|
||||
except Exception:
|
||||
pass
|
||||
self._wait_for_listing_content(page)
|
||||
logger.info("Applied year range filter: %s — %s", year_min, year_max)
|
||||
return True
|
||||
except Exception as exc:
|
||||
logger.warning("Failed to apply year range filter: %s", exc)
|
||||
return False
|
||||
|
||||
def collect_current_page(self, page: Page, page_number: int = 1) -> ListingPageResult:
|
||||
# Считываем ссылки одним проходом по DOM.
|
||||
self._accept_cookie_banner(page)
|
||||
|
||||
Reference in New Issue
Block a user