jobcore 0.2.0

A lightweight Python client for job listings. No API key, no headless browser, no scraping service.

jobcore

A lightweight Python client for job listings. No API key, no browser automation.

Currently supports Apple career listings (jobs.apple.com). More sources planned.

Install

Available on PyPI:

pip install jobcore

Or install from source:

git clone https://github.com/Mikedan37/jobcore.git
cd jobcore
pip install -e .

Requires Python 3.10+ and httpx.

Quick example

from jobcore import JobCoreClient

client = JobCoreClient()

for job in client.search(query="iOS engineer", limit=5):
    print(f"{job.title} — {job.team.name}")
    print(f"  {job.location} · {job.posting_date}")
    print(f"  {job.url}")
    print()

client.close()

CLI

~6,000+ live positions across teams and locations.

jobcore count                          # total open positions
jobcore search "Swift"                 # search by keyword
jobcore search --team "Software"       # filter by team
jobcore search "iOS" --location "US"   # combine filters
jobcore page 5                         # fetch a specific page
jobcore dump > jobs.jsonl              # all jobs as JSON lines

Or via module: python -m jobcore count

Why this exists

Major employers don't always provide public job APIs.

This package gives you a clean, stable interface over public job listings so you can:

• search roles programmatically
• build job matching pipelines
• automate job discovery workflows

One dependency (httpx). No Selenium. No API key.

What this does

• Fetches job listings from public career pages
• Parses structured job data from server-rendered HTML (no headless browser)
• Returns typed Python models (Job, JobCorePage)
• Handles pagination, retries, and rate limiting
• Provides a small CLI for quick lookups and bulk export

What this does NOT do

• Does not apply to jobs on your behalf
• Does not bypass authentication or access controls
• Does not scrape aggressively or circumvent rate limits
• Does not enrich, score, or rank jobs

Responsible use

This library parses data already served to standard browser requests. Built-in safeguards:

• 0.5s default delay between page requests. Do not set it to zero.
• 3 retries max on transient failures. No infinite loops.
• Honest User-Agent with a project URL. Do not spoof it.

If a source changes their site structure or blocks automated access, this library will stop working. That is their right. File an issue and we will adapt.

API reference

JobCoreClient

client = JobCoreClient(delay=0.5, retries=3, timeout=30.0)

Method	Returns	Description
search(query=, location=, team=, limit=)	list[Job]	Search with client-side filtering
search_page(page=1)	JobCorePage	Fetch a single page (20 jobs)
iter_jobs(max_pages=300)	Iterator[Job]	Auto-paginate all jobs
get_total_count()	int	Total open positions
close()	—	Close HTTP client

Supports context manager: with JobCoreClient() as client: ...

Job

Field	Type	Description
position_id	str	Unique position ID
title	str	Job title
summary	str	2-3 paragraph description
url	str	Direct link to posting
posted_at	datetime?	UTC timestamp
posting_date	str	Human-readable date
team	JobTeam	Department (name, code)
locations	list[JobLocation]	One or more locations
home_office	bool	Remote-eligible
weekly_hours	int	40 = full-time

Properties: location (primary as string), is_remote, is_part_time

JobCorePage

Field	Type	Description
jobs	list[Job]	Up to 20 jobs
total_records	int	Total matching jobs
page	int	Current page (1-indexed)

Properties: total_pages, has_next

Development

git clone https://github.com/Mikedan37/jobcore.git
cd jobcore
pip install -e ".[dev]"
pytest tests/ -v              # 32 tests, ~0.5s, fully offline

Tests parse saved HTML fixtures. No network calls. If a source changes their site and tests break, re-capture the fixture and update the parser.

Limitations

• Client-side filtering. Source pages may not support server-side text queries. search() fetches pages and filters locally.
• Summaries only. The summary field is from the search listing. Full detail pages are not fetched.
• HTML parsing is inherently fragile. If a source redesigns their site, the parser will break. File an issue.

Maintained by

Maintained by Danylchuk Studios, LLC (Seeker project).

Unofficial job data tooling. Not affiliated with any company whose listings are accessed.

Intended for personal use, research, and tooling. Not for high-volume scraping or commercial redistribution of job listings.

License

MIT