Version bumped to 1.1.6 — a major update: new no‑compile HTML parsing core, optional success‑request logging, and several dependency/packaging fixes. See CHANGELOG.md for details.
jwebs is a complete, high‑performance library for web scraping, crawling automation, and content analysis. It supports both HTTP/1.1 and HTTP/2 (user selectable) and includes built‑in caching, rate limiting, robots.txt handling, dynamic proxy rotation, distributed crawling (via Redis), data extraction, content differencing, uptime monitoring, Sitemap/RSS generation, and optional AI‑powered extraction. HTML parsing runs on a fast, no‑compile core (a direct ctypes binding to libxml2's parser and XPath engine, ~1.7x lxml's speed) that installs cleanly on constrained platforms like Android/Termux, with automatic BeautifulSoup fallback. Tested on Python 3.9 – 3.13.
from jwebs import JWebs
j = JWebs()
resp = j.GET("https://example.com")
print(f"Status: {resp.status}")
print(f"Content length: {len(resp.text)}")· HTTP – HTTP/1.1 and HTTP/2 (user selectable), Keep‑Alive, automatic redirects, batch concurrent requests.
· Fast HTML Parsing – A no‑compile ctypes binding straight to libxml2's parser and XPath engine (~1.7x lxml's speed), with automatic BeautifulSoup fallback. lxml itself is still available as an optional accelerator (jwebs[lxml]) for platforms that can compile it.
· Request Management – Two‑layer cache (memory + SQLite), rate limiting (Token Bucket), robots.txt respect, session management.
· Security & Flexibility – User‑Agent rotation, dynamic proxy rotation, client certificates (mTLS), SSL and security headers checking, HTTP Basic Auth (auth=).
· Crawling & Automation – Simple crawler and distributed crawler (Redis) that can run across multiple machines.
· Data Extraction – Extract text, links, emails, phone numbers, prices, JSON‑LD, meta tags, images, social media links. · Content Analysis – Sentiment analysis, automatic translation, content differencing (diff).
· Monitoring – Uptime monitoring, performance testing (TTFB, page size), SEO and security audits.
· Logging – Structured, rotating file/console logging; optionally log successful requests too, not just errors (off by default, toggle anytime with SET_LOG_SUCCESS()).
· Utilities – Sitemap.xml generator, RSS feed generator, GraphQL client, async client.
· AI (optional) – Intelligent data extraction via natural language instructions (DeepSeek/OpenAI) and text summarization.
# Basic installation (core dependencies only, Redis client included)
pip install jwebs
# With HTTP/2 support
pip install jwebs[http2]
# With the lxml accelerator (fastest HTML parsing; needs a C compiler,
# so it may not install on Android/Termux -- the fast no-compile core
# above is used automatically wherever lxml isn't available)
pip install jwebs[lxml]
# All optional features
pip install jwebs[all]redis (the Python client) is installed automatically as a core dependency. You still need the Redis server itself running for DistributedCrawler -- install it with your package manager:
· Ubuntu/Debian: sudo apt install redis · Termux (Android): pkg install redis · macOS: brew install redis
Or download from redis.io
If pip install jwebs[lxml] fails to build on Android/Termux (no prebuilt wheel for lxml there), you don't need it -- just pip install jwebs and the library automatically uses its built‑in no‑compile HTML core instead, at close to the same speed.
from jwebs import JWebs
j = JWebs(http_version='2', use_cache=True)
title = j.GET_TITLE("https://http2.golang.org/")
print(f"Title: {title}")By default, caching is disabled for every HTTP method, even when
use_cache=True. You must explicitly opt in per method via
cache_methods:
from jwebs import JWebs
# Cache GET responses only
j = JWebs(use_cache=True, cache_methods=['GET'])
# Cache GET and POST responses
j = JWebs(use_cache=True, cache_methods=['GET', 'POST'])
# Cache every supported method
j = JWebs(use_cache=True, cache_methods='all')
# Change it later on an existing instance
j.set_cache(True, methods=['GET', 'PUT'])Caching non-GET methods (POST, PUT, PATCH) can be unsafe if those requests have side effects (e.g. creating a resource) -- only enable it for endpoints you know are idempotent. Cache keys for these methods include a hash of the request body, so different payloads to the same URL are cached separately.
from jwebs import JWebs
j = JWebs()
emails = j.EXTRACT_EMAILS("https://example.com")
links = j.GET_LINKS("https://example.com", internal=True)
print(f"Emails: {emails}\nInternal Links: {len(links)}")from jwebs import JWebs
j = JWebs()
crawler = j.create_distributed_crawler(redis_url="redis://localhost:6379/0")
crawler.add_seed("https://example.com", depth=0)
crawler.crawl_worker(max_pages=10, max_depth=2, strict_page_limit=True)
results = crawler.get_all_results()
for url, info in results.items():
print(f"{url} → {info.get('title', 'no title')}")from jwebs import JWebs
j = JWebs()
report = j.SECURITY_AUDIT("https://example.com")
print(f"SSL valid: {report.ssl_valid}")
print(f"Security grade: {report.grade}")from jwebs import JWebs
j = JWebs()
snap1 = j.TAKE_SNAPSHOT("version1", "Hello world")
snap2 = j.TAKE_SNAPSHOT("version2", "Hello jwebs")
diff = j.COMPARE_SNAPSHOTS(snap1, snap2)
print(f"Similarity: {j.SIMILARITY('Hello world', 'Hello jwebs')}")from jwebs import JWebs
import time
j = JWebs()
j.MONITOR_URL("https://example.com", expected_status=200)
j.START_MONITORING()
time.sleep(5)
j.STOP_MONITORING()from jwebs import JWebs
j = JWebs(enable_logging=True, log_success=True) # also logs successful requests, not just errors
resp = j.GET("https://example.com")
j.SET_LOG_SUCCESS(False) # keep logging on, stop logging successes onlyYou can report bugs via GitHub Issues or submit fixes via pull requests.
· GitHub repository: https://github.com/JCode-JCode/jwebs
· PyPI page: https://pypi.org/project/jwebs/
This project is licensed under the Apache License 2.0 – see the LICENSE file for details.
Designed and built with love by J Code❤️