From: Mo Zhou Date: Sun, 12 Oct 2025 01:00:56 +0200 Subject: Prevent downloads during Sphinx documentation build Forward: not-needed --- examples/applications/plot_out_of_core_classification.py | 4 +++- examples/applications/plot_stock_market.py | 3 +++ sklearn/datasets/_base.py | 3 +++ sklearn/datasets/_openml.py | 3 +++ 4 files changed, 12 insertions(+), 1 deletion(-) diff --git a/examples/applications/plot_out_of_core_classification.py b/examples/applications/plot_out_of_core_classification.py index ad0ff96..3200cde 100644 --- a/examples/applications/plot_out_of_core_classification.py +++ b/examples/applications/plot_out_of_core_classification.py @@ -18,6 +18,7 @@ features (words) may appear in each batch. # SPDX-License-Identifier: BSD-3-Clause import itertools +import os import re import sys import tarfile @@ -162,7 +163,8 @@ def stream_reuters_documents(data_path=None): sys.stdout.write("\rdownloaded %s / %s" % (current_sz_mb, total_sz_mb)) archive_path = data_path / ARCHIVE_FILENAME - + if int(os.getenv('DEBIAN_POLICY_SECTION_4_9_NO_NETWORK_ACCESS', '0')) > 0: + raise TimeoutError('Debian Policy Section 4.9 prohibits network access during build') urlretrieve(DOWNLOAD_URL, filename=archive_path, reporthook=progress) if _not_in_sphinx(): sys.stdout.write("\r") diff --git a/examples/applications/plot_stock_market.py b/examples/applications/plot_stock_market.py index 40f778c..80791e2 100644 --- a/examples/applications/plot_stock_market.py +++ b/examples/applications/plot_stock_market.py @@ -23,6 +23,7 @@ that are linked tend to fluctuate in relation to each other during a day. # `data.nasdaq.com `_ and # `alphavantage.co `_. +import os import sys import numpy as np @@ -98,6 +99,8 @@ for symbol in symbols: "https://raw.githubusercontent.com/scikit-learn/examples-data/" "master/financial-data/{}.csv" ) + if int(os.getenv('DEBIAN_POLICY_SECTION_4_9_NO_NETWORK_ACCESS', '0')) > 0: + raise TimeoutError('Debian Policy Section 4.9 prohibits network access during build') quotes.append(pd.read_csv(url.format(symbol))) close_prices = np.vstack([q["close"] for q in quotes]) diff --git a/sklearn/datasets/_base.py b/sklearn/datasets/_base.py index e6e6939..66a322f 100644 --- a/sklearn/datasets/_base.py +++ b/sklearn/datasets/_base.py @@ -1488,6 +1488,9 @@ def _fetch_remote(remote, dirname=None, n_retries=3, delay=1): f"re-downloading from {remote.url} ." ) + if int(os.getenv('DEBIAN_POLICY_SECTION_4_9_NO_NETWORK_ACCESS', '0')) > 0: + raise IOError('Debian Policy Section 4.9 prohibits network access during build') + # We create a temporary file dedicated to this particular download to avoid # conflicts with parallel downloads. If the download is successful, the # temporary file is atomically renamed to the final file path (with diff --git a/sklearn/datasets/_openml.py b/sklearn/datasets/_openml.py index 537f6cd..44cbb20 100644 --- a/sklearn/datasets/_openml.py +++ b/sklearn/datasets/_openml.py @@ -994,6 +994,9 @@ def fetch_openml( dtypes: category(9), int64(6) memory usage: 2.7 MB """ + if int(os.getenv('DEBIAN_POLICY_SECTION_4_9_NO_NETWORK_ACCESS', '0')) > 0: + raise TimeoutError('Debian Policy Section 4.9 prohibits network access during build') + if cache is False: # no caching will be applied data_home = None