DokuWiki can convert various character combinations to their typographical equivalent.

e.g.

<<  to « 
>>  to » 
=>  to ⇒

see also typography.

Details of the conversions are stored in the conf/entities.conf file. You can add your own entity conversions by creating a conf/entities.local.conf file and placing your conversions in that file. It is possible to modify Dokuwiki's conf/entities.conf file, however this could result in your changes being overwritten when upgrading Dokuwiki.

It seems that it is possible to use this table to adapt DokuWiki to some French typography rules that necessitate non-breaking spaces before or after some punctuation marks. For example, in the conf/entities.conf file, I write rules like this:

<< &laquo;&nbsp;
? &nbsp;?
: &nbsp;:
! &nbsp;!
; &nbsp;;

To use typography for esperanto x-writing:

cx	─Й
Cx	─И
CX	─И
gx	─Э
Gx	─Ь
GX	─Ь
hx	─е
Hx	─д
HX	─д
jx	─╡
Jx	─┤
JX	─┤	
sx	┼Э
Sx	┼Ь
SX	┼Ь
ux	┼н
Ux	┼м
UX	┼м

The typography for esperanto writing did not show right in the code section above; for those interested these are the characters as they should be displayed: (— 'Ryan Chappelle' 2008/05/28 06:04)

CX  Ĉ     cx   ĉ
GX  Ĝ     gx   ĝ
HX  Ĥ     hx   ĥ
JX  Ĵ     jx   ĵ
SX  Ŝ     sx   ŝ
UX  Ŭ     ux   ŭ

As said in entities.conf, you can use HTML entities to configure your entities, but it is not recommended, because it may break non-HTML rendering. So, UTF-8 chars are used directly instead. To achieve this, please either use an UTF-8 compliant text editor, like notepad++; or an hexadecimal editor, like hexplorer and an UTF-8 character map. However regular text editors should NOT be used for this, since some characters used by UTF-8 encoding are not reachable through a regular keyboard.

As instance, the equivalent for ≠ (HTML: ≠) char in UTF-8 is 0xE2; 0×89; 0xA0; and 0xA0 shows as a regular space even if it's not one. Also, do NOT try to calculate UTF-8 yourself without any clue about what UTF-8 is, or your entities might not work.