Chinese Input methods

[RichardVV]

I'm living Hong Kong at the moment and I'm trying to learn to type chinese. Any recommendation to a foreigner living in HK who is learning the language? Also are there any good supporting softwares that help me?

[skylee]

I use the pinyin input that comes with Windows. You should be able to find it in the language setting of your PC.

[foodtarget]

You should definitely download and install the Google Pinyin Input Method. It's amazing! And whatever you do, DON'T try to learn the Cangjie input method (倉頡輸入法). I have a friend from Hong Kong who says that the Cangjie IM is popular there and helps you remember how the character is written. But I found it to be way too complicated for a foreigner and only loosely based on the way the character is written... but yea, download the Google thing

[RichardVV]

Is Pinyin the most popular far us learners? I've been trying to learn 倉頡. I think it is very difficult, anyone else having problems with 倉頡 too? If yes, what are they and please tell me the advantages and disadvantages of using Pinyin over 倉頡.

[foodtarget]

whatever you do, DON'T try to learn the Cangjie input method (倉頡輸入法)

I repeat, DON'T waste your time with Cangjie. It is really a bitch to learn. It's supposedly based on radicals, but it's really more like the vague shape of the character. There are so many rules you have to memorize, and nearly all the explanations I could find for Cangjie are in Chinese. If you're a beginner learning Chinese, it will be a real pain. The only potential benefits of Cangjie is that you can type characters that you don't know the pronunciation for and you can type some characters that are only used in Cantonese. So basically I only installed Cangjie on my computer for when I didn't know how to type something in pinyin. But it turned out that most of the time I couldn't freakin figure out how to type it in Cangjie either! So since you will probably become quite familiar with pinyin as you learn Chinese, it will be a lot easier to just type in pinyin. If you are dead set on learning an input method based on character composition, their is some method that is probably the most popular alternative to pinyin on the mainland.... I wanna say Wubihua (五笔画) but I'm not sure. But anyway, none of my Chinese friends had even heard of Cangjie. I'll ask them about the other method and get back to you.

[imron]

Wubizixing 五笔字型 is what you are talking about. Wubihua 五笔画 is the stroke input method you find on most mobile phones. I highly recommend learning Wubizixing. It is based on character shapes (sometimes these are radicals, sometimes not), and it's a good way to learn how to break a character into its component parts, and it's also good for practicing active recall when typing - unlike pinyin input methods which use passive recall. There are several threads on the forums already discussing wubi, here's a good place to start.

[renzhe]

Is there a Wubi input method for SCIM?

[skylee]

I agree with foodtarget.

Basically I would only use input methods that don't require me to learn a new set of rules/skills. Thus I use pinyin. And I think the stroke input method on mobile phones is good too.

There is a stroke-based input method called 九方, which is very easy IMO. But I don't use it as I don't want to install any software.

[imron]

I'd be surprised if there wasn't

[foodtarget]

Wubizixing 五笔字型 is what you are talking about. Wubihua 五笔画 is the stroke input method you find on most mobile phones.

Oh wow, that clears up a lot of confusion and makes a lot more sense. When I asked a Chinese friend "What's that other way to type?" he just said 五笔, so I thought that meant 五笔画. But when I looked up info on 五笔画, it said you only need five keys. I thought that must be a really odd way to type on the computer...

[RichardVV]

Thanks everyone, It seems that倉頡 really is hard, Are there any books that teach chinese input methods (english text of course), like books on pingyin and五笔画 or even 倉頡. I'd like to try to learn each method and compare them so please post books you know of about Chinese input methods.

Thanks in advance!

[imron]

The definitive explanation of 五笔字型 for English speakers can be found here. The same author has also done a writeup of 五笔画

[RichardVV]

I'm thinking of giving up on Cangjie because it really is too hard compared to Pingyin. I haven't tried Wubi yet but It should be similar to Cellphones's way for typing according to previous posts... Can't find any paperbacks that teach chinese input in english yet though.

[foodtarget]

I'm thinking of giving up on Cangjie because it really is too hard compared to Pingyin.

I guess the difficulty of Cangjie is just something one has to experience for oneself. It seems like a good idea - typing based on radicals and character shapes and whatnot. But it's actually damn hard, especially for a foreigner. And fyi, there's no "g" in Pinyin.

I haven't tried Wubi yet but It should be similar to Cellphones's way for typing according to previous posts...

I was confused about this too until recently. The way for typing on a cellphone is Wubihua (五笔画) and the computer input method is Wubizixing (五笔字型)。 五笔画 only uses 5 keys to represent the 5 basic strokes, and you type by inputting the strokes in the order that you would write the character. 五笔字型 uses the whole QWERTY keyboard and is more complicated. I haven't learned it (yet) but it seems similar (at least in the basic idea) to Cangjie, in that typing is based on radicals and character shapes.

Can't find any paperbacks that teach chinese input in english yet though.

Yea can't help you there, as I don't know of any. But you shouldn't really need one for pinyin as it is pretty straight-forward. As for the other two methods, I doubt there is a big demand for English-language Chinese typing tutorials, so maybe such books don't exist. I have heard that in HK you can buy books that have tables of characters with their Cangjie codes or something like that, which might be of some use, but they are still written in Chinese. For English language explanations, I think the previously mentioned websites might be your best bet. Wikipedia also has articles in English on Cangjie and Wubizixing.

[lemur]

Is there a Wubi input method for SCIM?

There is. It appears under "Chinese (Simplified) -> Wubi". I've tried it recently but I've decided to put Wubi aside for now. I need to type in traditional characters but there is very little explanation on how to do that with Wubi. It is definitely possible to do it with SCIM's version of Wubi but it is just not evident. I met my Waterloo when trying to figure out how to type 學. The best I was able to achieve was to type the simplified version and select the traditional version from the list of choices presented by SCIM.

[imron]

As long as you have a version of Wubi that supports traditional character roots, the logic behind Wubi is exactly the same whether you are using traditional or simplified characters. Any literature or learning materials you see for Wubi and simplified characters is directly applicable to typing traditional characters. All you are doing is combining shapes in the order that they are written, so the only difference will be that some traditional characters will use different key combinations than their simplified counterparts because the characters are written differently, e.g. 医 ATDI vs 醫 ATDG

Some traditional and simplified characters will even use the same key combinations because they use the same roots, e.g. 説 and 说 both have the code YUKQ.

See this thread for a bit more detailed discussion (especially the second page).

So, as long as you are learning Wubi, the same principles can be applied to either traditional or simplified characters.

As for 學, you probably picked a slightly more complicated one to start with. It might be a better idea to start with some of the easier ones when you are just beginning, until you have a feel for how Wubi works.

Anyway, here's an explanation of this character. For long characters, the rule in wubi is to type the key for the first 3 roots, and the key for the final root. Let's first break 學 down into its component parts:

First root: 臼 - found on the V key

Second root: 乂 - found on the Q key

Third root: 乂 - found on the Q key

Fourth root: 冖 found on the P key

Fifth root: 子 found on the B key.

So, by choosing the first 3 roots, followed by the final root, we can see that the full key combination for 學 would be VQQB. Compare this to the simplified version 学 - IPBF.

[lemur]

As long as you have a version of Wubi that supports traditional character roots, the logic behind Wubi is exactly the same whether you are using traditional or simplified characters.

Yes, but I can't help have the feeling that Wubi is biased towards simplified characters. Some of that feeling is due to the fact that every online tool I've found which would give Wubi codes for specific characters only worked with simplified characters. If you ask for the Wubi code for 学, you get an answer but if you ask for 學, you're out of luck.

Anyway, here's an explanation of this character. For long characters, the rule in wubi is to type the key for the first 3 roots, and the key for the final root. Let's first break 學 down into its component parts:

First root: 臼 - found on the V key

Second root: 乂 - found on the Q key

Third root: 乂 - found on the Q key

Fourth root: 冖 found on the P key

Fifth root: 子 found on the B key.

So, by choosing the first 3 roots, followed by the final root, we can see that the full key combination for 學 would be VQQB. Compare this to the simplified version 学 - IPBF.

Thanks for the help. You are confirming what I thought. I had figured out the decomposition for the traditional and simplified characters but here's the deal: VQQB simply does not work in SCIM's Wubi mode. If I type VQQB, I get nothing at all.

To take the other examples you gave:

- 医 ATDI , 醫 ATDG: both work as expected.

- When I type YUKQ, SCIM gives me a choice between 说 and a square which indicates a missing character. It turns out the square is 説. That must be a bug in SCIM because when I use pinyin input, I get all traditional characters without any problem.

[imron]

Yes, it is more biased towards simplified, because that was what it was originally designed for, however I don't really see it as a large problem. Case in point, I've only ever learnt simplified, but I can still type traditional characters with Wubi just by looking at them and knowing how Wubi combines shapes.

Anyway, I presume you're running Linux? If you have access to a windows box at all, then you might want to check out 五笔快打, which is an excellent tool for learning 五笔 and also includes tools for looking up the codes for traditional characters. That said though, the key to learning 五笔 isn't memorising character codes. The key is to internalise the character roots through practice, and learn how to combine the roots together. You can do this by practicing with either traditional or simplified characters as there are only a handful of specialised roots for traditional characters. Most of the roots are identical, and once you've internalised which roots map to which key, then it just becomes a matter of how you combine them.

but here's the deal: VQQB simply does not work in SCIM's Wubi mode.

Unfortunately, that would be a problem with SCIM's mapping tables then. It might be worth having a hunt to see if there's a database with mappings for all the traditional characters. VQQB works fine for me using the freewb IME under windows, although I should add it doesn't work when using FIT on my Mac.

Depending on how technically inclined you are, you could just try downloading freewb and look to see if you can configure SCIM to use freewb's mapping tables. Actually, just having a quick hunt through those mapping tables now, it appears there is also another mapping for 學 - WFQB, which would use the following roots:

W - for 亻

F - for 二

Q - for 乂

B - for 子

So you might want to see if that works also.

[ipsi()]

Interesting. I get "" for VQQG on SCIM under Ubuntu 8.04, but WFQB does indeed produce 學，or 発. Do you currently have a Linux box Imron? Might be worth trying some of the other Freeweb codes to see if maybe SCIM uses those.

Still, WFQB seems a little strange for 學...

Anyway, ATDG produces 醫, YGKG produces 语 and 語, and LGHH gives 車. It defaults to simplified characters, meaning you have to select 語 by hitting '2'. Maybe this can be changed, not sure. Be very nice if it could.

[imron]

I get "" for VQQG

Is this a typo?

Anyway, under wubi, the 亻root is sometimes appropriated for questionable puposes. Normally where there there is a left-falling stroke followed by a vertical stroke, even where the vertical stroke is located to the left of the left-falling stroke e.g. in the character 段, which is WDMC

W for 亻

D for 三

M for 几

C for 又

or even more bizarrely in the character 追 which is WNNP

Still, WFQB seems a little strange for 學...

Not really, and not when you understand how 亻is sometimes used under Wubi.

So for 學, I imagine the full breakdown would be something like:

W 亻

F 二

Q 乂

Q 乂

N 乛

F 二

P 冖

B 子

I do have access to a linux box, but it involves rebooting my machine which is a pain to do when you've got a whole bunch of applications open.