hyphenation issues

[juh2]

If I simply put a \hyphenation{...} somewhere in the style file, it does not work.

Is there a special location where to put hyphenation?

[ousia]

Not that I know. But I would recommend right after \sethyphenationfeatures.

The commands are new:

\registerhyphenationexception[MacOS]
\registerhyphenationexception[de][MacOS]

[juh2]

Thanks! These commands seems to be undocumented. Why "MacOS"?

[ousia]

The hyphenation engine is a (somehow) new one.

The hyphenator was announced at the mailing list.

And I guess I found the command reading the source.

Please, feel free to improve the ConTeXt wiki :smirk:.

[juh2]

Interesting link. In the thread you complain that the new hyphenator does not work with \hyphenation.

https://mailman.ntg.nl/pipermail/ntg-context/2014/080082.html

I tried the registerhyphenationexception with your MWE from the mailinglist but Nietz-sche is still incorrectly hyphenated.

[ousia]

It works with \registerhyphenationexception.

[juh2]

Finally I understood. The example with MacOS disturbed me, because I thought it refers to ConTeXt on Mac OS X. I added a very incomplete new page in the wiki.

[juh2]

I am afraid that it does not work in fptc:

Maybe you can try to compile a small text with the word "Wirtschaftsoligarchie"

It is hyphenated after "Wirtschaft".

The exception \registerhyphenationexception[Wirt-schafts-o-li-gar-chie]

does not work here.

[ousia]

I tried the following (after updating pandoc-xhtml):

<div lang="de">
<span class="hyphenatedword">Wirtschaftsoligarchie</span>
</div>

And it is hyphenated Wirt-schaft-so-lig-ar-chie.

From what you are describing, the word isn’t tagged as German. See the difference after compiling this sample with ConTeXt:

\starttext
\hyphenatedword{Wirtschaftsoligarchie}
\de\hyphenatedword{Wirtschaftsoligarchie}
\stoptext

Now I can only guess: did you set the proper lang value in the metadata?

If i’m missing something, please explain with a minimal sample.

BTW, didn’t you have another issues?

[juh2]

I have

---
locale: de-DE
lang: de-DE
---

in my YAML-Codeblock. I hope this is enough to set mainlanguage in the process.

[ousia]

I guess lang: de-DE should be enough.

But anyway, what do you get from this as a single document?

---
title: Deutsche Silbentrennung
lang: de-DE
...

<span class="hyphenatedword">Wirtschaftsoligarchie</span>

Did you get the right hyphenation when adding the Markdown snippet to your document?

Please, update pandoc-xhtml.tex. Otherwise, the class would be useless.

[juh2]

No, it does not work.

I put

\registerhyphenationexception[Wirt-schafts-o-li-gar-chie]

after the setuphyphenation-commands in my style files.

The result is: Wirt-schaft-so-lig-ar-chie

It does not work with lang: de-DE and/or locale: de-DE

[ousia]

It works fine here. I’m afraid the word is hyphenated with English hyphenation rules in your system.

BTW, do you have the following line in your copy of pandoc-xhtml.tex?

\installlanguage [de-DE] [de-de]

Add it, if not. And replace in the same file:

\startxmlsetups xml:hyphenatedword
    \hyphenatedword{\xmlflush{#1}}
\stopxmlsetups

with

\startxmlsetups xml:hyphenatedword
    \hyphenatedword{\xmlflush{#1}} \currentdate
\stopxmlsetups

and tell me what happens. You should get a German date, not an English one.

[juh2]

Wirt-schaft-so-lig-ar-chie 11. Februar 2016

[ousia]

What happens when you add in the book you’re composing (I mean, one of your production documents) the following?

<span class="hyphenatedword">Wirtschaftsoligarchie</span>

But comment the hyphenation exception before.

[juh2]

I added some more samples:

Wirt-schaft-so-lig-ar-chie 12. Februar 2016 Haus-macht 12. Februar 2016 Gesell-schafts-wis-sen-schaf-ten 12. Februar 2016

With \registerhyphenationexception I could not put a hyphen at Ge-sell-schafts...

Somehow the whole mechanism is not working here:

context --version

resolvers       | trees | analyzing 'home:texmf'
mtx-context     | ConTeXt Process Management 0.63
mtx-context     |
mtx-context     | main context file: /home/juh/context/tex/texmf-context/tex/context/base/mkiv/context.mkiv
mtx-context     | current version: 2016.02.06 14:06

luajittex --version
This is LuajitTeX, Version beta-0.89.0 (TeX Live 2016/dev)

[ousia]

Leaving the issue with Ge-schell-schaft aside (it isn’t a hyphenation exception), does hyphenation work fine now in German?

TeX defines a minimum number of letters on the left and on the right to hyphenate words (lefthyphenmin and righthyphenmin). In Spanish, the default values are 2 and 2. In German, default values are 3 and 3.

Because of that, you can get co-mo in Spanish (which is wrong) and Methode in German.

You could define a minimum length to apply hyphenation (to prevent co-mo in Spanish or ei-ne in German). The problem is more complex in German (see this thread.)

The way to set the minimum characters is:

\setuplanguage[de-de][lefthyphenmin=2, righthyphenmin=2]

And I’m now in a hurry. I will give the other parameter later this evening (I don’t have time now to search it).

[ousia]

The way to define a minimum word length to apply hyphenation in ConTeXt is:

\definehyphenationfeatures
       [givemefive]
       [hyphenmin=4]

\sethyphenationfeatures
       [givemefive]

Some questions that you should take into consideration:

• hyphenmin is applied to the whole document and it isn’t related to a particular language. Well, it is related to all languages. Of course, lefthyphenmin and righthyphenmin should be specified for each language.
• The length of the word (number of letters) to be hyphenated is the value of hyphenmin plus one.
• Of course, givemefive is only an identifier. Use whatever you want as identifier for \definehyphenationfeatures. It should match the value in \sethyphenationfeatures.
• The styling environment (from-pandoc-to-context.tex) already contains \sethyphenationfeatures[strict]. Add your hyphenation feature there.
• This doesn’t solve all issues in German hyphenation (see sample quoted in this message).

I hope this is clear now. Let me know, if it isn’t or if hyphenation doesn’t work as expected in your documents.

[ousia]

@juh2, I don’t know whether the issue was solved. (I guess it is, but I really don’t know.)

I suspect that the whole problem with wrong hyphenation is that the \installlanguage command was missing from your copy of pandoc-xhtml.tex.

If all problems with hyphenation commented in this issue are solved, please close it.

I’m not closing it myself, so you may reopen it when needed.