character decomposition

[cypherpunks2]

Are there any dictionaries or anything (preferably downloadable) with character decomposition information? I.e. what parts a hanzi is made up from. Heisig's remembering the hanzi for example, is one such a list.

For example:

昌=日+日

时=日+寸

時=日+寺

寺=土+寸

国=囗+玉

玉=王+丶

I vaguely remember seeing such a project, but now i can't find it. (Or maybe it was for Japanese?)

[chaxiu]

Wenlin does... but you have to buy it or get it some other way. But I think it's a pretty good investment:mrgreen:

www.wenlin.com

Chaxiu

Edited October 20, 2008 at 04:50 AM by chaxiu

[liuzhou]

Can we change the expression from decomposition to deconstruction, please?

I'm getting images of rotten, putrid characters, stinking and turning to messy globs of slime, attracting vultures and other carrion eaters leaving behind only oracle-less bones.

[imron]

leaving behind only oracle-less bones

All in less than 30 minutes I'd guess

Wenlin does... but you have to buy it

The cheaper option is surely just to look at the characters in any normal dictionary and figure it out yourself - for the large majority of characters it's really not that tricky.

[Hofmann]

Try 春.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

It's 艸 + 屯 + 日. 屯 is the phonetic element.

[roddy]

Yellowbridge.com might do the trick - eg.

[imron]

t's 艸 + 屯 + 日. 屯 is the phonetic element.

Here you are talking about etymology though rather than composition. I would argue that neither 艸 + 屯 are part of the composition of the character. Rather you have  and 日.

[davidj]

The cangjie input method does something like this, although I think that each key encodes for more than one possible component. I think the Unihan database (unicode.org) has the cangjie codes for every character.

[lemur]

Cangjie also cuts through components sometimes. I don't think it is the preferred way to learn character composition. I'm not saying Cangjie is useless but I think it is better used to *retain* knowledge of composition rather than *learn* it in the first place. (Disclosure: I'm learning Cangjie these days.)

[jbradfor]

I recommend http://chinese-characters.org/ , but it's on-line only.

For most characters, it breaks each character into the radical part and the non-radical part, with clickable links for each.

[claire1987lr]

i'm a chinese,i just want to know how do you type different part of chinese character,like the part of 春 without 日？

[davidj]

how do you type different part of chinese character,like the part of 春 without 日？

If they are not radicals or characters in their own right, you will probably need a custom font and to use the Unicode private use codes.

I have a copy of the Unicode 4.0 standard and if they do exist there, they are not indexed, so there is no practical way of finding them. I suspect they don't exist.

If you can find a web page that does display them, without using image replacement, I'll have a look and see how they've done it.

(For fonts which permit derivatives, e.g. GPLed fonts, like the Arphic ones, you should be able to use a font editor to cut out the sun from Spring.)

[roddy]

If you use Sogou's IME, typing shui will give you 氵, shou 扌, cao 艹, etc.

[imron]

For the large majority of radicals, no special fonts have been necessary from Windows XP onwards. As for how I physically typed it in, I use Wubi, so it's quite easy, just type DWI

D for 三

W for 人

and I to resolve the conflict.

Other radicals are similar, I just type them in by their component shapes.

If you want to learn Wubi, I recommend 五笔快打 for learning, which you can download from here, and 极点五笔 for normal typing, which can be downloaded here. As a native speaker, if you spend 10-15 mins a day with 五笔快打 you should be typing at reasonable speeds in about a month.

If you use Sogou's IME, typing shui will give you 氵, shou 扌, cao 艹, etc.

So what do you need to type to make  appear?

[claire1987lr]

thanks a lot!i think i'll try wubi....

[m.ellison]

I've just found http://glyph.iso10646hk.net/doc/normal_char.txt but I haven't checked it yet; maybe it's what you want but it's only for traditional characters.

[Hofmann]

MDBG does it for some characters.

[david808]

Arch Chinese does it for all the simplified Chinese characters. Traditional Chinese character support is coming soon.

[tooironic]

I second the previous poster's suggestion of getting Wenlin. It gives you the background, etymology, deconstruction, or whatever you want to call it for every character you could ever want (and then some). On top of that, it is a fantastic resource for a Chinese learner, with a whole myriad of tools that come in handy (pronunciation, document/notepad tool, handwriting input, stroke order animations, list generator, etc and of course a pretty sturdy Chi-Eng dictionary).

In terms of an actual book, 《中文字普》 (Chinese Characters: A Genealogy and Dictionary) is not bad, although certainly not exhaustive, nor entirely modern either.

[m.ellison]

I've analysed the file that I referred to earlier in this thread and I have pulled out decompositions similar to the following. It covers about 25,000 characters, so it might be useful for anyone that wants to go on after they have finished Heisig's 3000 characters (traditional characters only though). Also at the moment it has some problems (eg primitives that do not have Unicode values). Also, I have not checked the output, only generated it from the data, so I cannot vouch for its correctness.

㐖 吉⺃

㐜 求九

㐡 九耎

㐤 九真

㐨 予予

㐩 井井

㐫 ⼇凶

㐬 ⼇⼛川

㐭 ⼇回

㐮 ⾐八꯴

㐰 ⺅⼝

㐱 ⼈⼺

㐲 ⺅⼤

㐳 ⺅⺎

㐴 ⺅⼣

㐶 ⺅⺒

㐸 ⺅欠

㐹 ⺅⽓

㐺 ⺅⺅⼈