Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 2 additions & 24 deletions pyproject.toml
Original file line number Diff line number Diff line change
Expand Up @@ -729,30 +729,8 @@ verify = [

# Web Module - Web utilities
# Use: pip install scitex[web]
web = [
"aiohttp",
"beautifulsoup4",
"readability-lxml",
"requests",
"Pillow",
"matplotlib",
"tqdm",
"joblib",
"scikit-learn",
"pytest-asyncio",
"ruamel.yaml",
"xarray",
"seaborn",
"scipy",
"markdown2",
"anthropic",
"openai",
"google-genai",
"groq",
# # Heavy dependencies handled by _AVAILABLE flags
# "torch",
# "umap-learn",
]
# Real implementation lives in the standalone scitex-web package.
web = ["scitex-web[readability]>=0.1.0"]

# Clew Module - Hash-based verification for reproducible science (Ariadne's thread)
# Use: pip install scitex[clew]
Expand Down
51 changes: 18 additions & 33 deletions src/scitex/web/__init__.py
Original file line number Diff line number Diff line change
@@ -1,35 +1,20 @@
#!/usr/bin/env python3
"""Web-related utilities module for scitex."""
"""SciTeX web — thin compatibility shim for scitex-web.

from ._scraping import get_image_urls, get_urls
from ._search_pubmed import (
_fetch_details,
_get_citation,
_parse_abstract_xml,
_search_pubmed,
)
from ._search_pubmed import batch__fetch_details as _batch__fetch_details
from ._search_pubmed import fetch_async as _fetch_async
from ._search_pubmed import format_bibtex as _format_bibtex
from ._search_pubmed import get_crossref_metrics
from ._search_pubmed import parse_args as _parse_args
from ._search_pubmed import run_main as _run_main
from ._search_pubmed import save_bibtex as _save_bibtex
from ._search_pubmed import search_pubmed
from ._summarize_url import crawl_to_json, crawl_url
from ._summarize_url import extract_main_content as _extract_main_content
from ._summarize_url import summarize_all as _summarize_all
from ._summarize_url import summarize_url
from .download_images import download_images
Aliases ``scitex.web`` to the standalone ``scitex_web`` package via ``sys.modules``.
``scitex.web is scitex_web``.

__all__ = [
# Public API
"search_pubmed",
"get_crossref_metrics",
"summarize_url",
"crawl_url",
"crawl_to_json",
"get_urls",
"download_images",
"get_image_urls",
]
Install: ``pip install scitex[web]`` (or ``pip install scitex-web``).
See: https://github.com/ywatanabe1989/scitex-web
"""

import sys as _sys

try:
import scitex_web as _real
except ImportError as _e: # pragma: no cover
raise ImportError(
"scitex.web requires the 'scitex-web' package. "
"Install with: pip install scitex[web] (or: pip install scitex-web)"
) from _e

_sys.modules[__name__] = _real
162 changes: 0 additions & 162 deletions src/scitex/web/_scraping.py

This file was deleted.

Loading
Loading