SciTE Japanese Resource and wrap mode 2 proposal

These products are my non-officially work on SciTE and Scintilla. Includes Japanese translation and wrap-mode 2 support.

Note: On CVS, character-wrap is accepted on Scintilla. But because of the problem of menu-structure and some reason, SciTE on CVS still cannot select it from menu.

Document written in Japanese is available.

Fixed Note of My Local Work

Scintilla 1.62+

SciTE 1.62+

Note: For many points, I fixed only Windows version.

"Why You Requests wrap.mode=2?"


Noted below is written to explain what I want to implement in Scintilla and SciTE. But you can apply any product which treat the language described below.

Then, why you requests "wrap.mode=2"?

Just consider followings.

  1. Some Languages(like Japanese) have no word-breaking characters.
  2. But even in these languages, space-separated-words(like English or others) can be inserted.

Then what happens?

In the languages described in #1 above, wordwrap is meaningless and it displays so bad in situation #2. When I want use "line-wrap" in situation #2, what can I do? -- The only answer is "Adding new wrap mode implementation." Implemetation of new wrap mode is very easy way. Simply avoid searching white-spaces, and allow any words to be separated into two visual lines.


How to distinguish words in Japanese documents?

Based on the knowledges of it, but can'tyoureadmeinenglish? Japanese uses each 100 of Hiragana, Katakana and thousands of Kanji. So we can distinguish more easily than english case just before you see.

What's the problem of wrap.mode=1 in Japanese?

  1. Waste of space.

    As mentioned above, there is no whitespaces between words written in Japanese, but now, we use many alphabetical acronyms. For example, USB, UTF, HTML, etc :D. We can place them without whitespaces both side, like Japanese-words, but their looks are somewhat cramped. Yes, because of complex structures especially Kanji-characters have.

    So many of us use whitespaces only both side of words consist from Latin1 characters, but it causes word-wraps after/before almost EVERY Latin1-words in Scintilla with wrap.mode=1, even if the words places at the beginning/end of the line. This is the problem.

  2. Waste of time.

    Any characters allowed to turn back to leftside, so searching whitespaces before/after the word to wrap is meaningless. In addition, because there are few whitespaces, searching once itself takes much time than in other languages.

    The case of Scintilla, character-wrap needs more time when using multi-byte encodings.

Jan 29, 2005 Suzumizaki-Kimitaka, (At-mark should be placed on first dot.)
Feb 15, 2005: some information is added and fixed.