After converting to Arabic-Indic digits, the numbers inside my store link written without https changed; how do I avoid that?
Write the link in full, starting with https://, before you convert. Link protection in the digit converter recognises a link by how it starts: a scheme such as https:// or the prefix www., plus emails and IP addresses. A bare domain such as mystore.com/p/125 looks like ordinary text to it, so its digits get converted.
What exactly happened
This line holds the same link twice, once bare and once in full:
اطلب من mystore.com/p/125 أو https://mystore.com/p/125
After converting to Arabic-Indic digits with the default settings:
اطلب من mystore.com/p/١٢٥ أو https://mystore.com/p/125
The tool converted 3 of the 6 digits, protected 3, and listed one protected token: https://mystore.com/p/125. The first link never appeared in that list, which is why it now ends in ١٢٥. A link with Arabic-Indic digits is no longer the same link: it may open a different page or nothing at all.
How protection recognises a link
Protection works on simple text patterns; it does not understand what the text means. It protects these shapes only:
- Anything starting with a scheme followed by ://, such as https:// or http://.
- Anything starting with www., even with no scheme in front.
- An email address in the form name@domain.
- An IPv4 address such as 192.168.1.10, and a dotted version number such as 1.2.3.
A bare domain matches none of them. We tried a line containing instagram.com/store2026, www.mystore.com/p/125 and the address sales2@mystore.com: the www. link and the email were protected, and the Instagram link came out as store٢٠٢٦.
A safe routine before converting
- Look through your text for links written without a beginning, and add https:// to the front of each.
- Open the Arabic/Persian/Western digit converter, paste the text, and make sure 'Protect URLs, emails, IP addresses and version numbers' is on.
- After converting, read the 'Protected tokens, left untouched' line under the result and compare its count with the number of links in your text. Any link missing from it has had its digits converted.
- For a long text with many links, search the result for your domain, such as mystore.com, and look at what follows it.
If a link has already been converted
Do not try to repair it inside the Arabic text. Copy the original link from its source, or convert that part alone back to Western digits. Note that switching protection off will not help here: it makes the tool convert the digits of full links too, so https://mystore.com/p/125 becomes https://mystore.com/p/١٢٥.
If the link itself contains Arabic letters or spaces and you want to see how it is actually sent, try it in the URL encoder/decoder, which shows the encoded form the browser uses.
Common questions
Does the tool protect links that already contain Arabic-Indic digits?
Protection recognises a link by its beginning, so a link starting with https:// is copied as is whatever its digits. It does not convert those digits to Western either; it leaves the link as it came.
Why doesn't the tool detect bare domains?
The tool sticks to the unambiguous shapes. A pattern that caught every word containing a dot, such as a file name like list.pdf, would also protect ordinary text whose digits you want converted.
Do I lose my links when converting to Western digits?
No. Links already use Western digits, so converting to Western changes nothing in them.
