# Watched words: in Persian, content is affected without containing the word

**URL:** https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851
**Category:** Support
**Created:** [5월 4, 2019, 6:26오후 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851 "2019-05-04T18:26:40Z")
**Posts on this page:** 8
**Page:** 1

<div class="post-metadata">

### Author: ![Pad\_Pors](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/pad_pors/32/52016_2.png) [@Pad\_Pors](https://meta.discourse.org/u/Pad_Pors)
#### Post date: [5월 4, 2019, 6:26오후 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/1 "2019-05-04T18:26:40Z")

</div>

is it possible that watched words work inaccurately in **other languages** like Arabic or Persian?

currently we’ve added some words to watched, but topics without such words get affected.

---

<div class="post-metadata">

### Author: ![HAWK](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/hawk/32/86627_2.png) [@HAWK](https://meta.discourse.org/u/HAWK)
#### Post date: [5월 6, 2019, 1:53오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/2 "2019-05-06T01:53:22Z")

</div>

Can you give some more details with examples?

---

<div class="post-metadata">

### Author: ![Pad\_Pors](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/pad_pors/32/52016_2.png) [@Pad\_Pors](https://meta.discourse.org/u/Pad_Pors)
#### Post date: [5월 6, 2019, 4:17오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/3 "2019-05-06T04:17:49Z")

</div>

I’d, I’m just not sure what kind of details or examples may be useful.

by “example”, do you mean a list of Persian words which we have added to watched word, and a Persian text which got affected without containing the words? (would it be useful if the words and the text are not in English?)

---

<div class="post-metadata">

### Author: ![schleifer](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/schleifer/32/86416_2.png) [@schleifer](https://meta.discourse.org/u/schleifer)
#### Post date: [5월 6, 2019, 4:22오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/4 "2019-05-06T04:22:02Z")

</div>

> [@Pad\_Pors](#):
>
> Do you mean I should provide a list of Persian words that we have added to the watched words, followed by a Persian text that was affected despite not containing those words?

Yes. A specific example in Persian showing “this is a text that triggered it, but it does not contain any of the watched words from this list” would be necessary.

---

<div class="post-metadata">

### Author: ![Pad\_Pors](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/pad_pors/32/52016_2.png) [@Pad\_Pors](https://meta.discourse.org/u/Pad_Pors)
#### Post date: [5월 6, 2019, 4:38오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/5 "2019-05-06T04:38:09Z")

</div>

ok, I got it! we have this word in our watched word list: **دین**

and then the last posts that needed approval contained words like this:

اسکاندیناوی  
بدین  
ندین  
دادین  
…

these 4 words are all different words with different meanings, but they contain those three letters in the same sequence.

* * *

should I add spaces to the watched word? something like this: " دین " instead of this “دین”

---

<div class="post-metadata">

### Author: ![codinghorror](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/codinghorror/32/110067_2.png) [@codinghorror](https://meta.discourse.org/u/codinghorror)
#### Post date: [5월 9, 2019, 11:12오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/6 "2019-05-09T11:12:22Z")

</div>

The way you should trigger is including the regex for matching spaces or non-word characters on each side. For example let’s say you want to hide the following (kinda) naughty word in English

> ass

so you replace it with

> butt

Well, that’s called a cl **butt** ic mistake … see [Obscenity Filters: Bad Idea, or Incredibly Intercoursing Bad Idea?](https://blog.codinghorror.com/obscenity-filters-bad-idea-or-incredibly-intercoursing-bad-idea/)

---

<div class="post-metadata">

### Author: ![david](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/david/32/157490_2.png) [@david](https://meta.discourse.org/u/david)
#### Post date: [5월 9, 2019, 11:21오전 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/7 "2019-05-09T11:21:52Z")

</div>

The way we define word boundaries doesn’t work well for non-latin characters at the moment. There is some discussion on the issue here:

> [@Censored words do not respect word boundaries in non-latin alphabet](https://meta.discourse.org/t/censored-words-do-not-respect-word-boundaries-in-non-latin-alphabet/102916):
>
> A censored word: ебля (f…ck) A word in post text: употреблять (To Consume)

---

<div class="post-metadata">

### Author: ![david](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/david/32/157490_2.png) [@david](https://meta.discourse.org/u/david)
#### Post date: [5월 10, 2019, 5:11오후 UTC](https://meta.discourse.org/t/watched-words-in-persian-content-is-affected-without-containing-the-word/116851/8 "2019-05-10T17:11:59Z")

</div>

This topic was automatically closed after 29 hours. New replies are no longer allowed.
