344780 - Make better breaking spec

Status: RESOLVED FIXED

(Core :: Layout: Text and Fonts, defect)

Description

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

I think that we should not break the line before '+', '=' and '-'.
Because IE doesn't break the line in this point. And we allow to break after too.
So, some table layout cases, we can see only '+' or '=' or '-' line.

E.g.,

2000-01
-
01T00:00

And '-' are used for date format in some locales. I think that we should not break both before and after if around of '-' are numeric.

Comment 1

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Attachment: Patch rv1.0

Comment 2

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Attachment: testcase

Comment 3

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Or there is another idea. That is always not breaking the line between first and second character. But I worry that way, that may have some regression.

Comment 4

[Bill Gianopoulos [:WG9s]]

(In reply to comment #3)
> Or there is another idea. That is always not breaking the line between first
> and second character. But I worry that way, that may have some regression.
> 

That would not fix the problem for the case of the '-' character.  Currently hyphenated words break before the hyphen which is incorrect at least in English.  hyphenated words are supposed to break following the hyphen.

Comment 5

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #4)
> (In reply to comment #3)
> > Or there is another idea. That is always not breaking the line between first
> > and second character. But I worry that way, that may have some regression.
> > 
> 
> That would not fix the problem for the case of the '-' character.  Currently
> hyphenated words break before the hyphen which is incorrect at least in
> English.  hyphenated words are supposed to break following the hyphen.

Right. Thank you.

Comment 6

[Dão Gottwald [:dao]]

(In reply to comment #0)
> I think that we should not break the line before '+', '=' and '-'.

And %.

Comment 7

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #6)
> (In reply to comment #0)
> > I think that we should not break the line before '+', '=' and '-'.
> 
> And %.
> 

What case is a problem?
If the previous character is numeric, not broken before '%'.

Comment 8

[Dão Gottwald [:dao]]

(In reply to comment #7)
> (In reply to comment #6)
> > And %.
> 
> What case is a problem?
> If the previous character is numeric, not broken before '%'.

It shouldn't break after % if the next character is numeric (e.g. %xx in URLs).
Also note that % is the modulo operator, so one could argue that it should be treated like *, +, / and so forth.

Comment 9

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #8)
> (In reply to comment #7)
> > (In reply to comment #6)
> > > And %.
> > 
> > What case is a problem?
> > If the previous character is numeric, not broken before '%'.
> 
> It shouldn't break after % if the next character is numeric (e.g. %xx in URLs).

Right and good.

Comment 10

[Dão Gottwald [:dao]]

Breaking "$foo" is also inappropriate (see languages like Perl or PHP) and inconsistent with both IE and Opera.

Comment 11

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Attachment: Patch rv2.0

better fix.

1. Don't break after '$' for some script. (IE only breaks if previous is numeric.)
2. Don't break after '%'. But keep to break before it for get param of URL.
3. Don't break around '+' for IE compatible.
4. Don't break around '-'. But keep to break if after char is an alphabet for long word or URL.
5. Don't break around '='. But keep to break if after char is an alphabet for long URL.
6. Don't break after '\' even if after char is alphabet for escape sequence.
7. Don't break around copyright mark(U+A9) for IE compatible.
8. Don't break around reserved mark(U+AE) for IE compatible.
9. Don't break before degree sign(U+B0) even if the before character is alphabet for IE compatible.

And don't break after '.' that is bug of bug 255990.(I don't change this table.)

We have other non-compatible cases. But I think that this patch is enough for daily use. (I think that we need refactoring for more compatible. Because current spec is difficult to keep the balance between Latin text and Japanese text. Note that current spec is based on Japanese Line Breaking spec(JIS X 4051).)

Comment 12

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Attachment: table of each UA's breaking point

Comment 13

[Dão Gottwald [:dao]]

German quotation marks („x“) missing.

Comment 14

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #13)
> German quotation marks („x“) missing.

These are not below U+0100... I think that it is not a regression of bug 255990.

Comment 15

[Dão Gottwald [:dao]]

(In reply to comment #14)
> I think that it is not a regression of bug 255990.

Do you know where similar work has been done? It's trunk-only and probably quite new.

Comment 16

[Martijn Wargers (dead)]

(In reply to comment #13)
> German quotation marks („x“) missing.

Dao, could you file a new bug on this, and mention the bug number here? Thanks.

Comment 17

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #15)
> (In reply to comment #14)
> > I think that it is not a regression of bug 255990.
> 
> Do you know where similar work has been done? It's trunk-only and probably
> quite new.
> 

Wow, I see. The patch of bug 255990 isn't affecting to only below U+100.

Comment 18

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #17)
> (In reply to comment #15)
> > (In reply to comment #14)
> > > I think that it is not a regression of bug 255990.
> > 
> > Do you know where similar work has been done? It's trunk-only and probably
> > quite new.
> > 
> 
> Wow, I see. The patch of bug 255990 isn't affecting to only below U+100.
> 

Ah, sorry. My previous comment may be wrong, I misread the code. I'll test it.

Comment 19

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to comment #16)
> (In reply to comment #13)
> > German quotation marks („x“) missing.

I understand why this happen. Because the patch of bug 255990 removed the CJK checking in nsJISx4051LineBreaker. So, the German quotations are affected the patch. I'll check the regression range.

Comment 20

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Attachment: Patch rv3.0

The General Punctuations are override as open/close class. But they are not match to Germany. We should remove the override the specifications.

Comment 21

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

backed-out the original patch. I'll retry to fix it with UAX#14. If you have some ideas and suggestions, please tell me.

-> FIXED(by backed-out)