3 ms·
In Egypt, it also helps to be able to read the arabic numerals — eastern arabic [0] that is. [0] https://en.wikipedia.org/wiki/Arabic_numerals#Comparison_of_di
by car 5y ago
In Egypt, it also helps to be able to read the arabic numerals — eastern arabic [0] that is.
[0] https://en.wikipedia.org/wiki/Arabic_numerals#Comparison_of_different_numerals https://en.wikipedia.org/wiki/Arabic_numerals#Comparison_of_...
Eastern Arabic ٠ ١ ٢ ٣ ٤ ٥ ٦ ٧ ٨ ٩ ١٠
- pjmlp 5y agoYeah, those street markets where there is a "small" deviation between western and eastern numerals if prices happen to be shown.
- contravariant 5y agoFor some reason I didn't expect you to list them from right to left.
- pajko 5y agoIt's "automatic", see https://www.w3.org/International/articles/inline-bidi-markup/uba-basics https://www.w3.org/International/articles/inline-bidi-markup... and https://unicode.org/reports/tr9/ https://unicode.org/reports/tr9/ However that algorithm has a couple of flaws in mixed left and right ordered text, and a couple of control characters have been introduced to solve the issue: https://www.w3.org/International/questions/qa-bidi-unicode-controls https://www.w3.org/International/questions/qa-bidi-unicode-c... Could play a bit in https://www.babelstone.co.uk/Software/BabelPad.html https://www.babelstone.co.uk/Software/BabelPad.html to see how things work.
- contravariant 5y agoAh is that what it was. Mixing ltr and rtl text always trips me up. To make matters even weirder it seems that sequences of nummerals, such as ١٠ are ltr again. That has some very unintuitive consequences when you try to edit that sentence.