Issue with -ido ending word

[ssoto]

Hello guys,
I found a extrange behaviour trying to translate a word:

  from urllib.parse import urlencode
  import requests
  
  def translate_to_and(sentence, vaf):
       encoded_url_params = urlencode({"spanish": sentence, "vaf": vaf})
       response = requests.get(f"https://api.andaluh.es/epa?{encoded_url_params}")
       return response.json()
  
  es_sentence = "he sido yo"
  for vaf in ["ç", "z", "s", "h"]:
      print(translate_to_and(es_sentence, vaf))

The output of this piece of code is:

{'spanish': 'que ha sido eso', 'andaluh': 'que a çido eço', 'rules': {'vaf': 'ç', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'que ha sido eso', 'andaluh': 'que a zío ezo', 'rules': {'vaf': 'z', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'que ha sido eso', 'andaluh': 'que a sío eso', 'rules': {'vaf': 's', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'que ha sido eso', 'andaluh': 'que a hido eho', 'rules': {'vaf': 'h', 'vvf': 'h', 'escapeLinks': False}}

As you can see, depending on vaf parameter the api return a result. Applying EPA rules, the response is correct for z and s vaf parameters and looks wrong for ç and h ones.

[ssoto]

I tried to replicate this behaviour with other sentences

With me gusta mucho el cocido it works nice:

In [20]: sentence = "me gusta mucho el cocido"

In [21]: for vaf in ["ç", "z", "s", "h"]:
    ...:     print(translate_to_and(sentence, vaf))
    ...: 
{'spanish': 'me gusta mucho el cocido', 'andaluh': 'me gûtta muxo er coçío', 'rules': {'vaf': 'ç', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'me gusta mucho el cocido', 'andaluh': 'me gûtta muxo er cozío', 'rules': {'vaf': 'z', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'me gusta mucho el cocido', 'andaluh': 'me gûtta muxo er cosío', 'rules': {'vaf': 's', 'vvf': 'h', 'escapeLinks': False}}
{'spanish': 'me gusta mucho el cocido', 'andaluh': 'me gûtta muxo er cohío', 'rules': {'vaf': 'h', 'vvf': 'h', 'escapeLinks': False}}