Difference between revisions of "WX notation"

From Apertium
Jump to navigation Jump to search
Line 51: Line 51:
* [http://sanskrit.inria.fr/DATA/wx.html WX notation: Overview]
* [http://sanskrit.inria.fr/DATA/wx.html WX notation: Overview]
* [http://mirror.umoss.org/mozdev/indicime/wx_keyboard.html WX Keyboard Mappings]
* [http://ltrc.iiit.net/downloads/nlpbook/nlp-panini.pdf NLP: A paninian perspective] (page 191)
* [http://ltrc.iiit.net/downloads/nlpbook/nlp-panini.pdf NLP: A paninian perspective] (page 191)

Revision as of 06:38, 18 July 2009

WX notation is used to represent the Devanagari alphabet (used by Hindi, Nepali, Marathi, Bengali and many other Indian languages) in ASCII.



<anudev> there r some issues of assigning some letters of hindi with Unicode 
<anudev> still unresolved
<anudev> actually there is the issue of separate vowels and matras
<avinesh_> could u give an example
<anudev> we don't need the vowels and matras(markers) differently
<avinesh_> because for every matra there is a mapping in wx
<anudev> like a, aa, ii, u r there
<avinesh_> yeah 
<anudev> but again ी ू े 
<anudev> are not needed
<spectie> matras = ?
<avinesh_> matra is the later representation
<anudev> matras= markers
<anudev> ka
<anudev> kaa
<anudev> we will write kaa as kA in wx
<anudev> in unicode there is a separate place for both A and the marker aa
<anudev> we need a same code for both of them,
<avinesh_> sry still not getting ur point why should we use wx instead of unicode?
<avinesh_> but people only follow one convention either the A or aa 
<avinesh_> not both
<avinesh_> i mean if u see a document 
<avinesh_> it will generally be consistent
<anudev> I mean we write A for both the vowel and matra
<avinesh_> oh..
<avinesh_> ok
<avinesh_> got it
<anudev> but unicode will write differently for A as a vowel and matra
<avinesh_> k got it
<anudev> so it creates unnecessary complication
<spectie> so the problem is that in unicode
<spectie> combining characters have a separate code point
<spectie> and in WX they are unified to one code point?
<spectie> = letter
<anudev> yes
<spectie> why not use unicode normalisation ?


  • राम = र्+आ+म्+अ (rAma)
  • कृष्ण = क्+ऋ+ष्+ण्+अ (kqRNa)

External links