You want to use a regular expression to delete the previous character of a particular character.

Wilwilco has answered himself, but r'^[^XY]*' does not seem to meet the requirements.
Because you have deleted any string that is not X or Y, for example:

>>re.sub(r'^[^XY]*',', '●N◆YYXY clock')
"YYXY watch".
>>re.sub(r'^[^XY]*',', '●N◆YYYX Clock')
"YYYX watch".
>>>re.sub(r'^[^XY]*', '', '●N◆X clock')
"X clock".

>>re.sub(r'^.*(XY.*)',r'\1', '●N◆XY clock')
"XY watch".
>>re.sub(r'^.*(XY.*)',r'\1', '●N◆YYXY Clock')
"XY watch".
>>re.sub(r'^.*(XY.*)',r'\1', '●N◆X clock')
'●N◆X clock'

Any string that starts with XY is captured and used as a replacement string in r'\1'.

>>re.sub(r'^.*(?=XY)',r', '●N◆YYXY clock')
"XY watch".
>>>re.sub(r'^.*(?=XY)',r', '●N◆XY clock')
"XY watch".
>>>re.sub(r'^.*(?=XY)',r', '●N◆X clock')
'●N◆X clock'

Use (?=...) to replace it with an empty string until just before XY.

Watch takes a variety of kanji and hiragana

This condition is not included in my reply.
Any characters other than kanji and hiragana are allowed.
Willwilco's answer is the same, so I'm going through it this time.
(Regular expressions that match only kanji and hiragana are a bit troublesome.)

After much consideration, I might have been able to use the following regular expressions as expected.

import re

a=re.sub(r'^[^XY]*',', '●N◆XY clock')

print(a)