Incorrect broken links reported #70
Description
Describe the bug
I noticed this with the blog plugin for mkdocs material. Any link to a blog post referenced as "/blog/posts/blah.md" will be identified as not having a source file.
This turns out to be based on how the files dictionary is used. At the time that the key for the files is set up, it uses the url which isn't going to be what is looked up in the blog case. The current setup assumes that the dest_uri will the the url at the time of on_files. That isn't the case with the blog plugin and might not be the case with other files (at this time, I have only verified with the blog plugin).
At the time of "on_files", the information about what the dest_uri's will be is not yet available. It is the same value as "url". However, by the time we are checking to see if a path exists, each File object has been updated and the dest_uri is correct.
The bug can be fixed by making self.files a list and then in find_source_file, iterate over the list of all files and compare search_path to the dest_uri of each File. If the two match, we've found the correct file object to return.
To Reproduce
Turn off use_directory_urls, run html-proofer on a site that is using the mkdocs-material blog plugin and linking to posts like /blog/posts/foo.md.
Environment (please complete the following information):
The occurs in GitHub CI using the following config: https://github.com/ponylang/ponylang-website/blob/main/.github/workflows/pr.yml#L33
It occurs for me locally with Ubuntu 22.
Local python version is Python 3.10.12
pip list:
Package Version
-------------------------- ----------------------
Babel 2.14.0
beautifulsoup4 4.12.3
blinker 1.4
certifi 2024.2.2
charset-normalizer 3.3.2
click 8.1.7
colorama 0.4.6
cryptography 3.4.8
dbus-python 1.2.18
distro 1.7.0
distro-info 1.1build1
ghp-import 2.1.0
gitdb 4.0.11
GitPython 3.1.41
httplib2 0.20.2
idna 3.6
importlib-metadata 4.6.4
jeepney 0.7.1
Jinja2 3.1.3
keyring 23.5.0
launchpadlib 1.10.16
lazr.restfulclient 0.14.4
lazr.uri 1.0.6
Markdown 3.5.2
MarkupSafe 2.1.5
mergedeep 1.3.4
mkdocs 1.5.3
mkdocs-ezlinks-plugin 0.1.14
mkdocs-htmlproofer-plugin 1.1.0.dev0
mkdocs-material 9.5.10+insiders.4.52.2
mkdocs-material-extensions 1.3.1
mkdocs-rss-plugin 1.12.1
more-itertools 8.10.0
oauthlib 3.2.0
packaging 23.2
paginate 0.5.6
pathspec 0.12.1
pip 22.0.2
platformdirs 4.2.0
Pygments 2.17.2
PyGObject 3.42.1
pygtrie 2.5.0
PyJWT 2.3.0
pymdown-extensions 10.7
pyparsing 2.4.7
python-apt 2.3.0+ubuntu2.1
python-dateutil 2.8.2
PyYAML 5.4.1
pyyaml_env_tag 0.1
regex 2023.12.25
requests 2.31.0
SecretStorage 3.3.1
setuptools 59.6.0
six 1.16.0
smmap 5.0.1
soupsieve 2.5
tzdata 2024.1
unattended-upgrades 0.1
urllib3 2.2.0
wadllib 1.3.6
watchdog 4.0.0
wheel 0.37.1
zipp 1.0.0
Contents of mkdocs.yml can be seen at:
https://github.com/ponylang/ponylang-website/blob/main/mkdocs.yml
I'm happy to open a PR with the fix detailed earlier. If you would like me to do so, please let me know. I'm going to prep that soon and switch us to using our fork with the fix. Opening a PR would be very easy after that.