# problem breaking URL when using tex4ht only

This problem only shows up when compiling to HTML.

I have some long URL's and a URL with underscore do not break inside table when compiled to HTML as it does when compiled to PDF.

I tried all the tricks I could find, and htlatex still does not want to break the URL inside the table. MWE

\documentclass[12pt]{article}%
\usepackage{longtable}
\usepackage{hyperref}
\makeatletter %from http://tex.stackexchange.com/questions/3033/forcing-linebreaks-in-url
\makeatother

%\usepackage[hyphenbreaks]{breakurl}
%\usepackage[hyphens]{url}

\begin{document}
\begin {longtable}{|p{.2in}|p{2.2in}|p{1in}|p{.8in}|}\hline
1&
\url{http://www.foo.com/questions/aaaaaa/aaaaaaa-aa-aaa-aaaaaaaaa-aaaaaaaaaa-aaaaa-aa-aaaaaaa-aaaaaa}
&
more&1/5/2016
\\\hline
\end{longtable}

Another table

\begin {longtable}{|p{.2in}|p{2.2in}|p{1in}|p{.8in}|}\hline
2&
\url{https://www.foo.com/post/aaa_aaa_aaaa_aaaaaaaa_aaaaa_aa_aaaaaaaaa_aaaaaa_aaa_aaa_aaaaaaaaa_aa_aaaaaaa_aaa_aaaaaaaa_aaaaaaa}
&another good question&1/5/2016\\\hline
\end{longtable}
another one

\begin {longtable}{|p{.2in}|p{2.2in}|p{1in}|p{.8in}|}\hline
1&
aaaaaa aaaaaaa aa aaa aaaaaaaaa aaaaaaaaaa aaaaa aa aaaaaaa
aaaaaa aaaaa aaaaaaaaaa aaaaaaaa aaaaa aaaaaa
aaaaaa aaaaaaa aa aaa aaaaaaaaa aaaaaaaaaa aaaaa aa aaaaaaa
&more&1/5/2016
\\\hline
\end{longtable}
\end{document}


When compiled to pdf using pdflatex, this is the output, as expected:

When compiled to HTML using

 htlatex foo.tex "htm,p-width"


this is the output

The problem shows up only with url with underscore. With hyphens in URL, it seems to do it fine.

Is there a way to make tex4ht break URL with underscore also?

TL 2015

tex4ht doesn't do anything with line breaks in URLs. It works for hyphens because browsers do that, but they don't break words with underscores. We need to insert zero-width space character after underscores to enable the linebreaking. Actions on certain characters may be defined using \g@addto@macro\UrlSpecials command, possibly in the .cfg file:

\Preamble{xhtml,p-width}
\makeatletter
\def\novspace{\HCode{&\#x200B;}}
\makeatother
\begin{document}
\EndPreamble


we defined the \novspace command which inserts XML entity for zero width space unicode character and this command is inserted after the underscore using

\g@addto@macro\UrlSpecials{\do\_{\_\novspace}}


the result:

• Thanks Michal. Nice solution. But for some reason it only works on Firefox and IE 11. On Chrome it does not wrap. This might be a browser thing. Earlier I found a hack using CSS in the .cfg file, but that worked only in Chrome and not in Firefox! Chrome version 50.0.2661.94 m. FF version 46.0.1 – Nasser May 9 '16 at 17:36
• @Nasser that's definitely a browser thing. There are also some related CSS properties - word-break and word-wrap, they don't seem to do anything in Firefox, maybe it would work in Chrome – michal.h21 May 9 '16 at 21:17
• yes, those what I tried and they work in Chrome only. But there is supposed to be a way to do it using CSS that will work in all browsers, based on what I read, but all the attempts I made failed so far. Not worth it. So I gave up. Your solution is simple and works. Who cares about Chrome users :) – Nasser May 9 '16 at 22:29
• @Nasser I've also tried the CS way first, but it didn't work in Firefox, only after that I've tried the actual solution. I would care about Chrome users and add also the CSS, it will work for everybody then (maybe except Links users :) – michal.h21 May 10 '16 at 6:16