PDF link with a hash causes PDF-parsing by accident #663
Comments
|
The But the main issue is that PDFs shouldn't be parsed by Nokogiri. (Edit: Deleted a sequence of errors on my part.) |
|
There is a workaround. HTMLProofer.check_directory(
"./path",
{ url_ignore: [/.*\.pdf(#.*)?/] }
).run |
|
@stevecheckoway |
|
@scivola, @stevecheckoway, Simple minimal reproducible example Bash: touch exists.pdf
wget https://file-examples-com.github.io/uploads/2017/10/file-example_PDF_1MB.pdf -O large.pdf
echo '<!doctype html>
<html>
<head><title>Example</title></head>
<body>
<a href="exists.pdf#page=2">Exists</a>
<a href="large.pdf#page=2">Large</a>
<a href="missing.pdf#page=2">Missing</a>
</body>
</html>' > example.htmlRuby: require 'html-proofer'
HTMLProofer.check_file(
"example.html",
{ :url_swap => {/(?<=\.pdf)#.*$/ => ""} }
).run
# * internally linking to missing.pdf#page=2, which does not exist (line 6)
# <a href="missing.pdf#page=2">Missing</a>Note that the error reports the original full link w/o stripped hash, but the link is in fact checked post- |
|
@riccardoporreca |
|
Thanks @riccardoporreca! Hopefully, html-proofer itself will be fixed, but in the mean time, that's a good workaround. |
If we want to make a link to a specified page of a PDF file, we can write as follows:
But it causes HTMLProofer to parse of the PDF file as an HTML file.
There are three problems.
[1] This useless parsing wastes CPU time and memory.
[2] htmlproofer will report a failure as follows:
[3] Sometimes HTMLProofer raises an ArgumentError such as:
The exception tends to occur for relatively complicated PDF file.
But I have yet to find the exact conditions for reproducing.
The text was updated successfully, but these errors were encountered: