diff --git a/pyproject.toml b/pyproject.toml
index 82303986..082982da 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -66,6 +66,7 @@ test = [
"pytest-env==1.1.5",
"vcrpy==7.0.0; python_version >='3.10'",
"mock==5.2.0",
+ "responses==0.25.8",
]
dev = [
{include-group = "test"},
diff --git a/ricecooker/utils/downloader.py b/ricecooker/utils/downloader.py
index 859ff26e..df553d8a 100644
--- a/ricecooker/utils/downloader.py
+++ b/ricecooker/utils/downloader.py
@@ -390,8 +390,12 @@ def download_assets(
# if we're really stuck, just default to HTML as that is most likely if this is a redirect.
if not ext:
ext = ".html"
- subpath = os.path.dirname(filename)
- filename = "index{}".format(ext)
+
+ subpath = filename
+ # Add the existing filename in front of index.xxx, this can contain slashes and those will result
+ # in subdirectories created in the downloaded version. This ensures multiple instances of extensionless
+ # resources referenced from a page won't clobber each other.
+ filename = filename + "/index{}".format(ext)
os.makedirs(os.path.join(destination, subpath), exist_ok=True)
@@ -429,7 +433,7 @@ def js_content_middleware(content, url, **kwargs):
return content
def css_node_filter(node):
- if "rel" in node:
+ if "rel" in node.attrs:
return "stylesheet" in node["rel"]
return node["href"].split("?")[0].strip().endswith(".css")
diff --git a/tests/test_downloader.py b/tests/test_downloader.py
index 911e08a6..a306a544 100644
--- a/tests/test_downloader.py
+++ b/tests/test_downloader.py
@@ -1,8 +1,49 @@
+import mimetypes
import os
+import re
+import tempfile
import unittest
+from pathlib import Path
+from urllib.parse import unquote
+from urllib.parse import urlparse
+
+import responses
from ricecooker.utils import downloader
+TESTCONTENT = Path(__file__).resolve().parent / "testcontent"
+
+# Portless host used to serve the checked-in sample tree. A real host (no
+# ":port") keeps the archived paths free of colons, which are illegal in
+# Windows filenames - the reason the previous embedded-server version of this
+# test failed on Windows.
+SAMPLE_HOST = "https://example.org"
+
+
+def _serve_sample_tree(request):
+ """responses callback: serve files from tests/testcontent as a static site.
+
+ Maps the request URL path onto the on-disk sample tree so the fixtures
+ checked in for this test are served verbatim, no local web server needed.
+ """
+ rel_path = unquote(urlparse(request.url).path).lstrip("/")
+ file_path = TESTCONTENT / rel_path
+ if not file_path.is_file():
+ return (404, {}, b"")
+ ext = file_path.suffix
+ if ext == ".html":
+ content_type = "text/html; charset=utf-8"
+ elif ext == "":
+ # Extensionless resources in this fixture are stylesheets; Google Fonts
+ # serves its CSS from an extensionless URL, which is what this test
+ # exercises.
+ content_type = "text/css"
+ else:
+ content_type = (
+ mimetypes.guess_type(str(file_path))[0] or "application/octet-stream"
+ )
+ return (200, {"Content-Type": content_type}, file_path.read_bytes())
+
class TestArchiver(unittest.TestCase):
def test_get_archive_filename_absolute(self):
@@ -70,3 +111,63 @@ def test_archive_path_as_relative_url(self):
link_filename, page_filename
)
assert rel_path == "../kolibri_1.2.3.png"
+
+
+# The sample tree lives under a subject folder mirroring a real PreTeXt book:
+# activecalculus.org (the page) references a stylesheet from fonts.googleapis.com
+# via an extensionless URL, whose CSS in turn references a font on
+# fonts.gstatic.com.
+SAMPLE_ROOT = "samples/PreTeXt_book_test_manually_cleaned_urls"
+# The Google Fonts stylesheet URL is extensionless; the original had a colon,
+# replaced with an underscore so the fixture file checks out on Windows.
+CSS_URL_PART = "css2_family_Material+Symbols+Outlined_opsz,wght,FILL,GRAD@24,400,0,0"
+
+
+@responses.activate
+def test_pretextbook_css_fetch():
+ """A stylesheet linked via an extensionless URL is archived and rewritten.
+
+ Regression test for two bugs when archiving a PreTeXt book: the CSS
+ ```` (matched by ``rel="stylesheet"``, not a ``.css`` extension) was
+ skipped, and the generated ``index.css`` for the extensionless resource was
+ dumped at the archive root instead of nested under the resource's own path,
+ which made its relative font references escape the archive directory.
+ """
+ responses.add_callback(
+ responses.GET,
+ re.compile(re.escape(SAMPLE_HOST) + r"/.*"),
+ callback=_serve_sample_tree,
+ )
+
+ page_url = "{}/{}/activecalculus.org/single2e/sec-5-2-FTC2.html".format(
+ SAMPLE_HOST, SAMPLE_ROOT
+ )
+
+ with tempfile.TemporaryDirectory() as download_root:
+ archive = downloader.ArchiveDownloader(download_root)
+ archive.get_page(page_url)
+
+ book_dest_dir = Path(download_root) / "example.org" / SAMPLE_ROOT
+
+ # The page's stylesheet link is rewritten to the nested index.css.
+ page_html = (
+ book_dest_dir / "activecalculus.org" / "single2e" / "sec-5-2-FTC2.html"
+ ).read_text()
+ assert "../../fonts.googleapis.com/" + CSS_URL_PART + "/index.css" in page_html
+
+ # The extensionless stylesheet is nested under its own path as index.css,
+ # not clobbered at the archive root.
+ index_css = book_dest_dir / "fonts.googleapis.com" / CSS_URL_PART / "index.css"
+ css_contents = index_css.read_text()
+ assert "fonts.gstatic.com/s/materialsymbolsoutlined" in css_contents
+
+ # The font referenced from the CSS is downloaded to its own nested path.
+ font_path = (
+ book_dest_dir
+ / "fonts.gstatic.com"
+ / "s"
+ / "materialsymbolsoutlined"
+ / "v290"
+ / "material_symbols.woff"
+ )
+ assert font_path.stat().st_size > 0
diff --git a/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/activecalculus.org/single2e/sec-5-2-FTC2.html b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/activecalculus.org/single2e/sec-5-2-FTC2.html
new file mode 100644
index 00000000..4d2a6fac
--- /dev/null
+++ b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/activecalculus.org/single2e/sec-5-2-FTC2.html
@@ -0,0 +1,22 @@
+
+
+
+
+
+
+
+
+
+
+ AC The Second Fundamental Theorem of Calculus
+
+
+
+
+
+
+
+
+
+
+
diff --git a/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.googleapis.com/css2_family_Material+Symbols+Outlined_opsz,wght,FILL,GRAD@24,400,0,0 b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.googleapis.com/css2_family_Material+Symbols+Outlined_opsz,wght,FILL,GRAD@24,400,0,0
new file mode 100644
index 00000000..00d15183
--- /dev/null
+++ b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.googleapis.com/css2_family_Material+Symbols+Outlined_opsz,wght,FILL,GRAD@24,400,0,0
@@ -0,0 +1,21 @@
+@font-face {
+ font-family: 'Material Symbols Outlined';
+ font-style: normal;
+ font-weight: 400;
+ src: url("../fonts.gstatic.com/s/materialsymbolsoutlined/v290/material_symbols.woff") format('woff');
+}
+
+.material-symbols-outlined {
+ font-family: 'Material Symbols Outlined';
+ font-weight: normal;
+ font-style: normal;
+ font-size: 24px;
+ line-height: 1;
+ letter-spacing: normal;
+ text-transform: none;
+ display: inline-block;
+ white-space: nowrap;
+ word-wrap: normal;
+ direction: ltr;
+ -moz-font-feature-settings: 'liga';
+}
diff --git a/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.gstatic.com/s/materialsymbolsoutlined/v290/material_symbols.woff b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.gstatic.com/s/materialsymbolsoutlined/v290/material_symbols.woff
new file mode 100644
index 00000000..025a3daa
--- /dev/null
+++ b/tests/testcontent/samples/PreTeXt_book_test_manually_cleaned_urls/fonts.gstatic.com/s/materialsymbolsoutlined/v290/material_symbols.woff
@@ -0,0 +1 @@
+Not really a font, just a placeholder file.
\ No newline at end of file
diff --git a/uv.lock b/uv.lock
index a19a1a99..3f4dbb54 100644
--- a/uv.lock
+++ b/uv.lock
@@ -1459,6 +1459,20 @@ wheels = [
{ url = "https://files.pythonhosted.org/packages/e1/d5/de8f089119205a09da657ed4784c584ede8381a0ce6821212a6d4ca47054/requests_file-3.0.1-py2.py3-none-any.whl", hash = "sha256:d0f5eb94353986d998f80ac63c7f146a307728be051d4d1cd390dbdb59c10fa2", size = 4514, upload-time = "2025-10-20T18:56:41.184Z" },
]
+[[package]]
+name = "responses"
+version = "0.25.8"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+ { name = "pyyaml", marker = "platform_python_implementation == 'CPython'" },
+ { name = "requests", marker = "platform_python_implementation == 'CPython'" },
+ { name = "urllib3", marker = "platform_python_implementation == 'CPython'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/0e/95/89c054ad70bfef6da605338b009b2e283485835351a9935c7bfbfaca7ffc/responses-0.25.8.tar.gz", hash = "sha256:9374d047a575c8f781b94454db5cab590b6029505f488d12899ddb10a4af1cf4", size = 79320, upload-time = "2025-08-08T19:01:46.709Z" }
+wheels = [
+ { url = "https://files.pythonhosted.org/packages/1c/4c/cc276ce57e572c102d9542d383b2cfd551276581dc60004cb94fe8774c11/responses-0.25.8-py3-none-any.whl", hash = "sha256:0c710af92def29c8352ceadff0c3fe340ace27cf5af1bbe46fb71275bcd2831c", size = 34769, upload-time = "2025-08-08T19:01:45.018Z" },
+]
+
[[package]]
name = "ricecooker"
source = { editable = "." }
@@ -1503,6 +1517,7 @@ dev = [
{ name = "pytest", marker = "platform_python_implementation == 'CPython'" },
{ name = "pytest-env", marker = "platform_python_implementation == 'CPython'" },
{ name = "requests-cache", marker = "platform_python_implementation == 'CPython'" },
+ { name = "responses", marker = "platform_python_implementation == 'CPython'" },
{ name = "ruff", marker = "platform_python_implementation == 'CPython'" },
{ name = "vcrpy", marker = "python_full_version >= '3.10' and platform_python_implementation == 'CPython'" },
]
@@ -1511,6 +1526,7 @@ test = [
{ name = "pytest", marker = "platform_python_implementation == 'CPython'" },
{ name = "pytest-env", marker = "platform_python_implementation == 'CPython'" },
{ name = "requests-cache", marker = "platform_python_implementation == 'CPython'" },
+ { name = "responses", marker = "platform_python_implementation == 'CPython'" },
{ name = "vcrpy", marker = "python_full_version >= '3.10' and platform_python_implementation == 'CPython'" },
]
@@ -1550,6 +1566,7 @@ dev = [
{ name = "pytest", specifier = "==8.4.2" },
{ name = "pytest-env", specifier = "==1.1.5" },
{ name = "requests-cache", specifier = "==1.2.1" },
+ { name = "responses", specifier = "==0.25.8" },
{ name = "ruff", specifier = ">=0.11" },
{ name = "vcrpy", marker = "python_full_version >= '3.10'", specifier = "==7.0.0" },
]
@@ -1558,6 +1575,7 @@ test = [
{ name = "pytest", specifier = "==8.4.2" },
{ name = "pytest-env", specifier = "==1.1.5" },
{ name = "requests-cache", specifier = "==1.2.1" },
+ { name = "responses", specifier = "==0.25.8" },
{ name = "vcrpy", marker = "python_full_version >= '3.10'", specifier = "==7.0.0" },
]