Filter certain persons to add to keywords template on export

[oPromessa]

Quick question...

As I export my Photos library to Synology Photos which does not recognise the RegionName (faces) I'm adding them to keywords with "person:" prefix.

I could use --person-keyword and filter --person query tag but I'd like to include all pics and add person:Person1 keywords only for the list of persons I select, and when they exist.

My .toml looks like this but:

• if Person1 or Person2 exists it the exports all persons, say "person:Person1 person:PersonNotToInclude" get included in the keywords list.

exiftool_merge_keywords = true
exiftool_merge_persons = true
keyword_template = [
    "{var:empty,}{var:name1,Person1}{var:name2,Person2}{person|sort matches {%name1}|{%name2}?person:{person},{%empty}}",
]

I guess I must go thru the list of persons and processed them one by one... maybe with the use of an external python function?

[oPromessa]

Testing out with the filter example function and it works!!! Hopefully will not delay much the export ;)

TOML file

exiftool = true
exiftool_merge_keywords = true
exiftool_merge_persons = true
keyword_template = [
    "{person|sort|function:persons.py::myfilter}",
]

Python file:

""" Example of using a custom python function as an osxphotos template filter 

    Use in formath:
    "{template_field|template_filter.py::myfilter}"

    Your filter function will receive a list of strings even if the template renders to a single value.  
    You should expect a list and return a list and be able to handle multi-value templates like {keyword}
    as well as single-value templates like {original_name}
"""
from typing import List

def myfilter(values: List[str]) -> List[str]:
    """Custom filter to add keywords for filtered persons
    filtered_persons = ["Person1", "Person2", "Person3", "Person4"]  # Persons to include in the list

    values = ["person:" + person for person in values if person in filtered_persons]
    
    return values

# Example usage #1:
# input_strings = ["apple", "Person1", "orange", "Person3", "kiwi"]
# result = myfilter(input_strings)
# print(result)

# Example usage #2:
# input_strings = ["apple", "orange", "kiwi"]
# result = myfilter(input_strings)
# print(result)

[RhetTbull]

Your function looks like it will do the trick. The function is loaded once then run for each photo and shouldn't slow down the export more than any other template. It's probably possible to do this in the template language using complex chaining for filters but your function is the best way in this case.