slugify() function of PyTrakt seems to not rename exactly as Trakt does

[typhoe]

Hi,

this issue is a follow up from an issue I have on PlexTraktSync available here: Taxel/PlexTraktSync#615

Two examples of shows that are not slugified correctly:
https://trakt.tv/shows/re-zero-starting-life-in-another-world
https://trakt.tv/shows/so-i-m-a-spider-so-what

I will quote the collaborator from PlexTraktSync:

On RE-Zero:

That's a very specific issue because of this show's name using - symbol : Re:ZERO -Starting Life in Another World-
The trailing dash has been removed in Trakt slug version : re-zero-starting-life-in-another-world whereas the pytrakt module does not remove it with its own slugify() function : re-zero-starting-life-in-another-world- so it doesn't match and trakt API returns 404 error.

On So-I'm-a-spider:

it's the i'm slugified into im by PyTrakt module and into i-m in trakt database.
It's clearly a problem in slugify() function of PyTrakt module. It does not slugify exactly as official trakt.tv does.

Thank you

[glensc]

Are you willing to provide a fix?

The current slugify is defined here:

• PyTrakt/trakt/utils.py

  Lines 10 to 20 in 8a6d4f1

  	def slugify(value):
  	"""Converts to lowercase, removes non-word characters (alphanumerics and
  	underscores) and converts spaces to hyphens. Also strips leading and
  	trailing whitespace.
  	Adapted from django.utils.text.slugify
  	"""
  	nfkd_form = unicodedata.normalize('NFKD', value)
  	decoded = nfkd_form.encode('ascii', 'ignore').decode('utf-8')
  	value = re.sub(r'[^\w\s-]', ' ', decoded).strip().lower()
  	return re.sub(r'[-\s]+', '-', value)

[glensc]

ps: edit your issue and use backticks for title, so it's better visible when rendered:

`some text inside backticks`

[typhoe]

Hi, I added the backticks for readability but I'm sorry but I won't be able to provide any fix.
I only used a workaround to bypass the issue on PlexTraktSync.

[glensc]

@typhoe are you able to find trakt.tv official docs or description how they slugify? perhaps ask their support?

[typhoe]

No sorry.

I only posted my issue here because one of the dev from PlexTraktSync told me the issue seemed to be from the slugify() function :
Taxel/PlexTraktSync#615 (comment)
and asked me to post here :
Taxel/PlexTraktSync#615 (comment)

I asked on the Trakt ticket support but that could take a while ;-)
https://support.trakt.tv/support/tickets/10789

[typhoe]

Hi,

Justin of Trakt support answered me, I'll copy/paste their answer below:

The rules would be defined in https://github.com/norman/friendly_id/ but in general I'd recommend using IDs to match and not relying on a slug. If you know the Trakt ID, ​that can be used in the URL or you use some of the lookup functionality we have like this:

https://trakt.docs.apiary.io/#reference/search/id-lookup/get-id-lookup-results
https://trakt.docs.apiary.io/#introduction/website-media-links

I'd highly recommend using IDs since Trakt could have different titles (and thus different slugs) compared to TMDB or TVDB.