select on text attribute that contains ',' gives server error

I was testing mine app that has a list of widgets of OneLineListItem, to know which one to click I choose to select by text it contains,
in my case it is a location in this format; Country, City, so for example "Argentina, Buenos Aires"

The problem with telenium is that if I try this xpath locator:
`'//OneLineListItem[@text="Argentina, Buenos Aires"]'` I got server error:
```
> select: ('//OneLineListItem[@text="Argentina, Buenos Aires"]',)
Traceback (most recent call last):
File "/home/magowiz/.virtualenvs/holiday/lib/python3.10/site-packages/telenium/client.py", line 36, in __call__
return response["result"]
KeyError: 'result'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/usr/lib/python3.10/code.py", line 90, in runcode
exec(code, self.locals)
File "", line 1, in
File "/home/magowiz/.virtualenvs/holiday/lib/python3.10/site-packages/telenium/client.py", line 60, in wait_click
if self.wait(selector, timeout=timeout):
File "/home/magowiz/.virtualenvs/holiday/lib/python3.10/site-packages/telenium/client.py", line 50, in wait
matches = self.select(selector)
File "/home/magowiz/.virtualenvs/holiday/lib/python3.10/site-packages/telenium/client.py", line 38, in __call__
raise TeleniumHttpException(response["error"]["message"])
__main__.TeleniumHttpException: Server error
```
A workaround that I find is to avoid using comma in attribute text matching rule, using this one:
`'//OneLineListItem[@text~="Buenos Aires"]'`, this is ok for me because in search field I already insert whole location and there should not be other Buenos Aires in Argentina.

Is there a way to escape comma properly before sending xpath to telenium?

Start with the select and wait flow shown in telenium/client.py, then reproduce the reported XPath using an attribute value containing a comma. Compare it with the working ~= selector and verify that the original selector returns the matching OneLineListItem without a server error.

Written by the indexing model from the issue text.