Recently, I subscribed the news feed of ArchWiki (Arch Linux Wiki) New Pages with Feedly. This allows me to know what is new about Arch Linux in some sense. Then, I found this, Rime IME. In Windows, I used Sogou Pinyin Input which is very nice, because it intelligently solve the sentence problem.
In Linux, previously I used IBus input. Due to the buggy upgrade of IBus in Arch Linux previously, then I switched to Fcitx. Fcitx has cleaner interface and configuration. The default Pinyin input method works well. However, the Pinyin using Fcitx or IBus does not work intelligently like Sogou Pinyin. Other than default Pinyin module, I have tried Google Pinyin, Sun Pinyin, libpinyin, and even Sogou Pinyin modules of Fcitx. I can’t really differentiate the advantages or disadvantages between these Pinyin input method, except the configuration interface. Yet, Sogou Pinyin does not work well in my computer.
Then, due to the character found in Taixuanjing (太玄经) 𤕠 (if you can read, then your computer support this font), I found that Pinyin cannot produce this character. Some of the Chinese characters are too weird and I don’t know how to pronounce. Thus, Wubi input method is more feasible. That is to input the Chinese character based on the look instead of the pronunciation. That is why I learn Wubi86 input method. There is another version, Wubi98. However, Wubi86 is more common.
The character such as 𤕠 in Taixuanjing is considered very weird that normal Wubi86 cannot input. Thus, to input this kind of characters, in Fcitx I have to use wubi-large module.
Rime IME is a little weird. It is installed as a module of Fcitx or Ibus. And it is an IME that can use multiple types of input methods (not simultaneously). It has Wubi86, Luna Pinyin, Terra Pinyin, Combo Pinyin, etc. The configuration is simple but without any GUI. We have to edit the configuration files manually, where configuration files are text file. It allows to input the characters like using wubi-large module, yet extra configuration required. For instance, create “wubi86.custom.yaml” in the user directory (refers to this page), then add in
patch: translator/enable_charset_filter: false
Interestingly, most of the input methods are China locale, yet Rime IME is Taiwan locale. However, Rime IME Luna Pinyin allows to produce Simplified Chinese characters.
Furthermore, comparing Rime IME Wubi86 and Fcitx Wubi86 module, Fcitx Wubi86 module uses OpenCC (Open Chinese Converter) to convert the characters between Simplified Chinese or Traditional Chinese. However, Rime IME does not use OpenCC. Thus, Rime IME Wubi86 does not allow to input Traditional Chinese by using Simplified Chinese key strokes. Other than that, Rime IME Wubi86 will adjust the candidate characters intelligently.
Quite a promising solution. Like!
(PS: And I personally feel that the author is a very interesting guy.)