How can i contribute?

[E]

Hi,

I had a very similar idea as KanjiVG (representation of Han characters stroke order in an open, re-usable manner) and then my friend pointed me here. This looks really close to what I have in mind, so I'm really eager to contribute and make this better. Question is, I don't know where to start.

My interest lies in internationalization. Seems like this area does need work but it is unclear what needs to be done. Pointers? Also, it seems data entry is done manually by volunteers editing XML files. Is that so?

Things I know -- Python, Cantonese, English, Korean

Things I don't know -- Perl, Japanese

Hope to be useful.

Cheers,

E

[Alexandre Courbot]

Hi,

Sorry for the late reply and thanks for your interest in KanjiVG.
Where to start really depends on what your skills are and what you
want to contribute. The SVG files for KanjiVG can be edited with any
good SVG editor, but the additional information about components and
layout is indeed added by hand. There is really a need for making a
real web-based KanjiVG editor that would allow to easily edit the
information.

Regarding internationalization, what do you have in mind? Do you mean
you want to also cover other languages using these characters, like
Chinese?

Cheers,
Alex.

> --
> --
> You received this message because you are subscribed to the "KanjiVG" group.
> For options and unsubscribing, visit this group at
> http://groups.google.com/group/kanjivg
> ---
> You received this message because you are subscribed to the Google Groups
> "KanjiVG" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to kanjivg+u...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Ahmed Fasih]

On Saturday, March 29, 2014 12:38:09 PM UTC-4, Alexandre Courbot wrote:

 the additional information about components and
layout is indeed added by hand. There is really a need for making a
real web-based KanjiVG editor that would allow to easily edit the
information.

Alexandre, I have (some) exposure in *using* the KanjiVG database but very little in terms of *authoring* it, so I don't have a good sense of the constellation of features that one would like to see in such a webapp. Could you sketch the different KanjiVG workflows, both so we can get a sense of what all goes into the production side and to guide the creation of any webapp? Drawing strokes? Rearranging nested groups? Changing stroke orders and adding variants?

[Alexandre Courbot]

Basically there are two main dimensions to a character described using KanjiVG:

1) The layout, which is the path information that forms the kanji.
There should be as many paths as there are strokes ; ideally to keep
things coherent a path should be a slightly rearranged copy of another
path of similar type (the number of control points *might* become of
importance for some applications, so paths of the same type should
have the same number of control points).

2) The semantics of the character, e.g. its components, sub-components
and how they relate to each-other.

There are quite a few good tools for 1) (e.g. inkscape), but very
little for 2), which is basically done by copy-pasting as far as I am
concerned. All the same this data is hard to visualize. So a web tool
that would address these issues would be most welcome.

Note that I come to believe less and less that such a tool should be a
direct reflection of the information represented in the .svg. SVG is a
purely hierarchical format, and that constraint does not really
reflect the reality of how kanji information is organized. For
instance, as you noticed some components are split in two, and these
two parts might not even share the same parent. A tool that provides
an easy view and editing capabilities of this kind of fact, and that
is also capable of outputting the corresponding svg, would be a great,
great help.

Actually, and if that's what it takes, I am not at all against using
an alternate source format that would be better suited for this kind
of work, with SVGs being generated from it.

In any case this will require some thorough thinking, but if you are
willing to take on that task, this would be a great contribution to
this project.

[msk...@ansuz.sooke.bc.ca]

On Sun, 17 Aug 2014, Alexandre Courbot wrote:
> 2) The semantics of the character, e.g. its components, sub-components
> and how they relate to each-other.
>
> There are quite a few good tools for 1) (e.g. inkscape), but very
> little for 2), which is basically done by copy-pasting as far as I am

Let me put in a plug for my IDSgrep tool, which is not Web-based nor
"visual" but deals with semantic structure in a way that may be of
interest. It can use KanjiVG as an input source. Software here:
http://tsukurimashou.sourceforge.jp/idsgrep.php.en
and an academic preprint about it here:
http://arxiv.org/abs/1404.5585

--
Matthew Skala
msk...@ansuz.sooke.bc.ca People before principles.
http://ansuz.sooke.bc.ca/

[Ahmed Fasih]

Matthew, thanks for the pointer! Aside from my question on the Sourceforge JP bugtracker about OS X, I got IDSgrep built in a fresh Ubuntu virtual machine. In an earlier thread I asked about KanjiVG's representation of 蔵 and IDSgrep is certainly a valuable tool for studying decompositions like this. According to CJKVI, 【蔵】⿱<艹>⿻十丨⿵戊臣, which makes much more sense---for this specific decomposition task---than KanjiVG's nested groups, though some on the list have raised the possibility that KanjiVG has an error there. Indeed, the KanjiVG-derived .eids file has "【蔵】⿱艹?" :(

Hope you have/had a good journey to Europe and best of luck there!

Ahmed