Infinite loop in inline_line_width

[p-vitt]

I'm constantly running into an infinte loop in inline_line_width when trying to dump a particular website. With others, it runs fine.

Some system details:

$ pip list --format=columns | grep WeasyPrint
WeasyPrint        0.42
$ python --version
Python 3.6.4
$ uname -mrs
Darwin 15.6.0 x86_64

new_resume_at stays to be 0 and resume_at stays to be 55 in every iteration, thus this loop won't end. At the beginning of the loop, some values get added to lines, but later on it keeps adding 0.9431584 on and on again.

How to reproduce:

link = 'http://www.zeit.de/politik/deutschland/2018-01/gruene-robert-habeck-palastrevolte-fluegelstreit-amt-und-mandat/komplettansicht'
weasyprint.HTML(url=link).write_pdf("test.pdf")

Do I misuse weasyprint? Do I need to add some template or configuration in order to dump a website?

[liZe]

Do I misuse weasyprint?

No, you don't. There's a bug!

[liZe]

I've tried hard to build a small test case to reproduce the problem and add it to WP's tests, but the real case is so strange (::after pseudo-tag in a display: table a tag with content: attr(href) and overflow-wrap: break-word) that I failed.

Never mind, bug is fixed in 995a9a3, the cause was easy to spot and the fix is clean.