uf5x7: add Hiragana (U+3040) and Katakana (U+30A0) #6

Merged
Lephenixnoir merged 3 commits from Neiviv-ui/unicode-fonts:master into master 2021-05-04 07:37:17 +02:00
Contributor
No description provided.
Neiviv-ui added 2 commits 2021-05-01 22:44:38 +02:00
Author
Contributor

Added Katakana block

Added Katakana block
Neiviv-ui changed title from WIP: master uf5x7: add Katakana (U+30A0) to master uf5x7: add Katakana (U+30A0) 2021-05-02 08:21:31 +02:00
Neiviv-ui changed title from master uf5x7: add Katakana (U+30A0) to Wip: master uf5x7: add Hiragana (U+3040) and Katakana (U+30A0) 2021-05-02 09:49:09 +02:00
Author
Contributor

It seems that there is a problem with gint's font manager, because U+3099 and U+309A characters are combinating characters, as said on the wikipedia page, but gint doesn't merge them together, instead, it put them side by side, which is actually a really good thing because the characters's drawing can't handle it. But, when this is used, since the characters are supposed to be merged, the dakuten and handakuten are on the top-right corner, far away from the rest of the glyph. Actually, I think putting the (han)dakuten on the top-left corner would be the best solution, except there will be twice the same character, U+3099 == U+309B & U+309A == U+309C. Asking for help.

It seems that there is a problem with gint's font manager, because U+3099 and U+309A characters are combinating characters, as said on the [wikipedia page](http://en.wikipedia.org/wiki/Table_des_caract%C3%A8res_Unicode/U3040), but gint doesn't merge them together, instead, it put them side by side, which is actually a really good thing because the characters's drawing can't handle it. But, when this is used, since the characters are supposed to be merged, the dakuten and handakuten are on the top-right corner, far away from the rest of the glyph. Actually, I think putting the (han)dakuten on the top-left corner would be the best solution, except there will be twice the same character, U+3099 == U+309B & U+309A == U+309C. Asking for help.
Owner

You are correct, although this is less of a problem and more of a non-feature; I never implemented much of the Unicode logic except for decoding.

Because gint has a family of dsize() functions that return pixel-based dimensions of strings, it would be possible to simply extend the font manager to support combining characters without breaking any applications.

Not combining and simply having the character on the left is an option. Whether U+3099 being the same as U+309B is a problem depends on whether you need to be able to separate them in order to understand the text. I am not familiar with these syllabaries, so I don't know whether this is a requirement. If not, then there is no problem having the same glyph twice.

Note that in a proportional font, we could get the best results by performing actual character combining, but setting up the U+3099 and U+309A glyphs so that they are moved to the right. This would make sure that at least the character being modified remains legible.

Because uf5x7 is a fixed-size font, there are only two possibilities; either the combining diacritic fits in the original glyph, or we have to add another 5 pixels. The former option seems unlikely, so I'm OK with the latter. More spacing is not a critical legibility problem, while overlapping glyphs certainly is.

You are correct, although this is less of a problem and more of a non-feature; I never implemented much of the Unicode logic except for decoding. Because gint has a family of `dsize()` functions that return pixel-based dimensions of strings, it would be possible to simply extend the font manager to support combining characters without breaking any applications. Not combining and simply having the character on the left is an option. Whether U+3099 being the same as U+309B is a problem depends on whether you need to be able to separate them in order to understand the text. I am not familiar with these syllabaries, so I don't know whether this is a requirement. If not, then there is no problem having the same glyph twice. Note that in a proportional font, we could get the best results by performing actual character combining, but setting up the U+3099 and U+309A glyphs so that they are moved to the right. This would make sure that at least the character being modified remains legible. Because uf5x7 is a fixed-size font, there are only two possibilities; either the combining diacritic fits in the original glyph, or we have to add another 5 pixels. The former option seems unlikely, so I'm OK with the latter. More spacing is not a critical legibility problem, while overlapping glyphs certainly is.
Author
Contributor

I share your point of you, and even if I am not a master of Japanese I can tell that U+3099 and U+309A are totally useless because sometimes the character they create already exist, and else it actually doesn't even exist in japanese language ! Moreover to my mind there is no distinction between U+3099 and U+309B & U+309A and U+309C, they are never written without context (another Hiragana/Katakana to combine with), so they are supposed to be the same.

BTW for now on I'm not able to draw some characters (missing creativty), but I think of asking my japanese classe's mates (maybe my teacher) if they have a idea how to pixel art them, and also if they can read the one I have already made.

I share your point of you, and even if I am not a master of Japanese I can tell that U+3099 and U+309A are totally useless because sometimes the character they create already exist, and else it actually doesn't even exist in japanese language ! Moreover to my mind there is no distinction between U+3099 and U+309B & U+309A and U+309C, they are never written without context (another Hiragana/Katakana to combine with), so they are supposed to be the same. BTW for now on I'm not able to draw some characters (missing creativty), but I think of asking my japanese classe's mates (maybe my teacher) if they have a idea how to pixel art them, and also if they can read the one I have already made.
Neiviv-ui added 1 commit 2021-05-02 20:40:38 +02:00
Author
Contributor

I haven't done U+3091 : ゑ, and probably never will be able to, because they are too many lines, but that isn't serious, since it is just an old Hiragana that isn't used anymore, and that many students don't learn.

For me, the other are not that bad, but I will still ask other people tomorrow 😃, then, I will remove "Wip : master".

I haven't done U+3091 : ゑ, and probably never will be able to, because they are too many lines, but that isn't serious, since it is just an old Hiragana that isn't used anymore, and that many students don't learn. For me, the other are not that bad, but I will still ask other people tomorrow 😃, then, I will remove "Wip : master".
Owner

This looks really good! Thanks a lot. No problem on leaving out glyphs that just don't work at all, I've done that a couple times myself as you can see.

I've attempted to draw it and ended up with this result, though I'm not sure whether it's good enough.

This font will definitely end up being one of the most complete options for the calculator, which is really great. :D

This looks really good! Thanks a lot. No problem on leaving out glyphs that just don't work at all, I've done that a couple times myself as you can see. I've attempted to draw it and ended up with this result, though I'm not sure whether it's good enough. This font will definitely end up being one of the most complete options for the calculator, which is really great. :D
Author
Contributor

Honestly I don't know if it is good or not because I really never read it, but it may still help (even if it isn't perfectly accurate), because for example if a new web browser come up (with gint's USB features), someone wanting to read pages where this character is written may find this useful.

Talking about that, I think the fact that the font doesn't support kanji isn't a problem because there are Furigana, and when you copy and paste a text (I have tried directly by hardcoding a sentence into a program), they are next to the kanji they precise. And because gint doesn't write unknown characters, it write furigana instead of kanji which is a really good thing to help reading (all in hiragana isn't perfect, but is way better than hiragana with blank spaces instead of kanji).

I let you add the ゑ when you've merged the repositories.

PS : I have asked other people and yes, it is actually readable. ;-)

Honestly I don't know if it is good or not because I really never read it, but it may still help (even if it isn't perfectly accurate), because for example if a new web browser come up (with gint's USB features), someone wanting to read pages where this character is written may find this useful. Talking about that, I think the fact that the font doesn't support kanji isn't a problem because there are Furigana, and when you copy and paste a text (I have tried directly by hardcoding a sentence into a program), they are next to the kanji they precise. And because gint doesn't write unknown characters, it write furigana instead of kanji which is a really good thing to help reading (all in hiragana isn't perfect, but is way better than hiragana with blank spaces instead of kanji). I let you add the ゑ when you've merged the repositories. PS : I have asked other people and yes, it is actually readable. ;-)
Neiviv-ui changed title from Wip: master uf5x7: add Hiragana (U+3040) and Katakana (U+30A0) to uf5x7: add Hiragana (U+3040) and Katakana (U+30A0) 2021-05-03 22:48:56 +02:00
Owner

This is super cool, thanks a lot for your hard word. :D

This sounds reasonable to me, so I'll go with this. I didn't know about the furigana working like this, but if they show up instead of the kanji, then it's a big plus!

Do you have anything in mind for the "Finish Hiragana and Katakana" line? It seems to be that the blocks are complete.

This is super cool, thanks a lot for your hard word. :D This sounds reasonable to me, so I'll go with this. I didn't know about the furigana working like this, but if they show up instead of the kanji, then it's a big plus! Do you have anything in mind for the "Finish Hiragana and Katakana" line? It seems to be that the blocks are complete.
Lephenixnoir merged commit 4ab32f199b into master 2021-05-04 07:37:17 +02:00
Sign in to join this conversation.
No reviewers
No Label
No Milestone
No Assignees
2 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: Lephenixnoir/unicode-fonts#6
No description provided.