Open behnam opened 7 years ago
Najib to merge the content he drafted on the wiki and do a PR. We'll be using this issue for review purposes.
Review of https://w3c.github.io/alreq/#h_numbers
Let's not call the variants Arabic vs. Persian. (Shervin)
We may want to stay away from francophone and anglophone terms. (Shervin)
Instead of Middle East / Western styles, I think we better use the terms more common in this topic: Eastern style and Western style.
We may want to expand on Thousand Separator and Decimal Separator. (Shervin)
Root of Bidi_Class differences between western and easter digits.
Maybe update the example for Bidi_Class difference to include Arabic-script text, instead of English.
Would number ranges (like %12-%15) be always LTR, or there are cases that 12 can go on the right-hand-side of 15?
* Maybe update the example for Bidi_Class difference to include Arabic-script text, instead of English.
A table like this could be more suggestive
* Would number ranges (like %12-%15) be always LTR, or there are cases that 12 can go on the right-hand-side of 15?
In 12-15, -
is European Separator changed to European Number (rule W4, http://unicode.org/reports/tr9/#W4). The whole string is a "number"
In 12٪-15٪ , -
comes after ٪ European Terminator, and changed to Neutral (Rule W6). 12 comes at the right of 15 in RTL context.
BTW (and WRT previous item), this is another example where Arabic-indic digits differs from Eastern Arabic-Indic digits.
I’d expect the number ranges to always be RTL. I don’t think fractions using solidus are common or even used in Egypt, for me ١/٢ reads immediately as “2 January” not as a fraction. I have seen vulgar fractions, on the other hand, in some old school books and they were LTR.
Agree. Number ranges are to be read RTL.
At the risk of reviving an old discussion, here's a question: in the current draft is a statement about the degree sign to be placed on the right of the number without any qualification. I wondered if this is correct, as I have seen the opposite custom, placing the degree sign to the left (in logical reading order). Incidentally Najib's table above seems to suggest that both can happen.
Thanks for the reminder on this overlooked nuance. We need to add some text to align (at least) with CLDR on this:
To populate 3.8 Arabic numbering
Draft is here: https://github.com/w3c/alreq/wiki/Arabic-numerals-(Draft)
Other issues related: https://github.com/w3c/alreq/issues/59 https://github.com/w3c/alreq/issues/62 https://github.com/w3c/alreq/issues/85