You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
60 lines
2.2 KiB
Python
60 lines
2.2 KiB
Python
"""AAO Career Center adapter — click-based AJAX pagination."""
|
|
from __future__ import annotations
|
|
|
|
from loguru import logger
|
|
|
|
from gimme_job.adapters.base import ManifestDrivenAdapter
|
|
from gimme_job.adapters.registry import register
|
|
from gimme_job.models.dto import JobPostingCandidate, RawJobCard
|
|
|
|
|
|
@register("aaoinfo")
|
|
class AAOInfoAdapter(ManifestDrivenAdapter):
|
|
_BASE_URL = "https://careers.aaoinfo.org"
|
|
|
|
def _wait_for_results(self, page) -> None:
|
|
try:
|
|
page.wait_for_selector("div.job-tile", timeout=15000, state="attached")
|
|
except Exception:
|
|
logger.debug("[aaoinfo] Results wait timed out — may be zero results")
|
|
|
|
def collect_cards(self, page, manifest):
|
|
self._wait_for_results(page)
|
|
return super().collect_cards(page, manifest)
|
|
|
|
def paginate(self, page, page_index: int, config) -> bool:
|
|
"""Click the next page number button in the AJAX pagination."""
|
|
if page_index >= config.pagination.max_pages - 1:
|
|
return False
|
|
|
|
next_page = page_index + 2
|
|
|
|
# Try clicking the specific page number button
|
|
next_btn = page.query_selector(f"#page-item-{next_page} a")
|
|
if not next_btn:
|
|
# Fall back to the last non-active/non-disabled pagination link
|
|
next_btn = page.query_selector(
|
|
"ul.pagination li.page-item:not(.active):not(.disabled):last-child a"
|
|
)
|
|
|
|
if not next_btn:
|
|
logger.debug(f"[aaoinfo] No next page button found — stopping at page {page_index + 1}")
|
|
return False
|
|
|
|
logger.debug(f"[aaoinfo] Paginating to page {next_page}")
|
|
next_btn.click()
|
|
|
|
# Wait for the new page's active indicator to confirm AJAX completed
|
|
try:
|
|
page.wait_for_selector(f"#page-item-{next_page}.active", timeout=12000)
|
|
except Exception:
|
|
page.wait_for_timeout(3000)
|
|
|
|
return True
|
|
|
|
def normalize(self, raw: RawJobCard) -> JobPostingCandidate:
|
|
candidate = super().normalize(raw)
|
|
if candidate.job_url and candidate.job_url.startswith("/"):
|
|
candidate.job_url = self._BASE_URL + candidate.job_url
|
|
return candidate
|