Add unicode.Rune.isNumeric, isDigit, isDecimal #7545
Conversation
|
In your PR, only ASCII digits are added. It isn't correct. |
lib/pure/unicode.nim
Outdated
| proc isDigit*(c: Rune): bool {.rtl, extern: "nuc$1", procvar.} = | ||
| ## Returns true iff ``c`` is a digit character | ||
| var c = RuneImpl(c) | ||
| return c in digitRanges |
There was a problem hiding this comment.
Looks like I can't check for digit like this... I need to rewrite this line.
|
Which Unicode version you used? See also nim-unicodedb and nim-unicodeplus libraries. Just for compatibility reasons. |
|
What kind of compatibility you are mean? As described, nim-unicodeplus is not drop-in replacement, so it must be used like this: import unicode except isDigit # like README.md example
import unicodeplusI think if we have Rune.isAlpha, we need to have Rune.isDigit in std lib. :) |
|
I mean that it would be nice to have |
|
You are right, but unfortunately I'm neither Unicode guru nor English guru. As described here, there are:
Should we use same naming rules or not? If yes, I just implemented |
|
You implemented checking for a decimal digits. LGTM. Let's wait for other opinions.
Me too :). Some useful links: |
nxshock
changed the title
|
The only thing I have to add to the table is that it looks like all those new procedures could share a common template implementation. |
|
Nice job, but all of these tables need to be generated from the Unicode spec. Otherwise it's a maintenance burden. I know the current code does it this way too, but I don't want to add more of the same bad stuff. |
|
Like this library - https://github.com/nitely/nim-unicodedb ? It generates all info from UCD |
|
@Yardanico Probably. |
|
No feedback, closing. |
No description provided.