Modifying the Convert Timezone macro in DBT to accept the timezone parameter from a column - jinja2

Question

I’m trying to modify the convert_timezone macro from the dbt_date package so that I can supply the target timezone from a column in my dataset, but I can’t seem to figure out how to get it working. I’m relatively new to DBT and Jinja so hoping one of you brilliant people can help me figure it out!
The macro in its current form can be seen below. Thanks!
{%- macro convert_timezone(column, target_tz=None, source_tz=None) -%}
{%- set source_tz = "UTC" if not source_tz else source_tz -%}
{%- set target_tz = var("dbt_date:time_zone") if not target_tz else target_tz -%}
{{ adapter.dispatch('convert_timezone', 'dbt_date') (column, target_tz, source_tz) }}
{%- endmacro -%}
{% macro default__convert_timezone(column, target_tz, source_tz) -%}
{%- if not source_tz -%}
cast(convert_timezone('{{ target_tz }}', {{ column }}) as {{ dbt_utils.type_timestamp() }})
{%- else -%}
cast(convert_timezone('{{ source_tz }}', '{{ target_tz }}', {{ column }}) as {{ dbt_utils.type_timestamp() }})
{%- endif -%}
{%- endmacro -%}
{%- macro bigquery__convert_timezone(column, target_tz, source_tz=None) -%}
timestamp(datetime({{ column }}, '{{ target_tz}}'))
{%- endmacro -%}
{%- macro spark__convert_timezone(column, target_tz, source_tz) -%}
from_utc_timestamp(
to_utc_timestamp({{ column }}, '{{ source_tz }}'),
'{{ target_tz }}'
)
{%- endmacro -%}
{% macro postgres__convert_timezone(column, target_tz, source_tz) -%}
{%- if source_tz -%}
cast({{ column }} at time zone '{{ source_tz }}' at time zone '{{ target_tz }}' as {{ dbt_utils.type_timestamp() }})
{%- else -%}
cast({{ column }} at time zone '{{ target_tz }}' as {{ dbt_utils.type_timestamp() }})
{%- endif -%}
{%- endmacro -%}
{%- macro redshift__convert_timezone(column, target_tz, source_tz) -%}
{{ return(dbt_date.default__convert_timezone(column, target_tz, source_tz)) }}
{%- endmacro -%}

Answer 1

convert_timezone as a macro is mostly useful for other package developers, so you can support timezone operations across a number of different databases, which all have their own slight syntax variations.
You didn't tag your RDBMS, but assuming you're on Snowflake, you could just use the built-in convert_timezone function, passing a field name as the <target_tz> argument.
If you really do want to modify the macro, it should be as simple as removing the single quotes around the tz arguments, so that you can pass in a field reference instead of a string literal:
{%- macro convert_timezone(column, target_tz=None, source_tz=None) -%}
{%- set source_tz = "'UTC'" if not source_tz else source_tz -%}
{%- set target_tz = var("dbt_date:time_zone") if not target_tz else target_tz -%}
{{ adapter.dispatch('convert_timezone', 'dbt_date') (column, target_tz, source_tz) }}
{%- endmacro -%}
{% macro default__convert_timezone(column, target_tz, source_tz) -%}
{%- if not source_tz -%}
cast(convert_timezone({{ target_tz }}, {{ column }}) as {{ dbt_utils.type_timestamp() }})
{%- else -%}
cast(convert_timezone({{ source_tz }}, {{ target_tz }}, {{ column }}) as {{ dbt_utils.type_timestamp() }})
{%- endif -%}
{%- endmacro -%}
{%- macro bigquery__convert_timezone(column, target_tz, source_tz=None) -%}
timestamp(datetime({{ column }}, {{ target_tz}}))
{%- endmacro -%}
{%- macro spark__convert_timezone(column, target_tz, source_tz) -%}
from_utc_timestamp(
to_utc_timestamp({{ column }}, {{ source_tz }}),
{{ target_tz }}
)
{%- endmacro -%}
{% macro postgres__convert_timezone(column, target_tz, source_tz) -%}
{%- if source_tz -%}
cast({{ column }} at time zone {{ source_tz }} at time zone {{ target_tz }} as {{ dbt_utils.type_timestamp() }})
{%- else -%}
cast({{ column }} at time zone {{ target_tz }} as {{ dbt_utils.type_timestamp() }})
{%- endif -%}
{%- endmacro -%}
{%- macro redshift__convert_timezone(column, target_tz, source_tz) -%}
{{ return(dbt_date.default__convert_timezone(column, target_tz, source_tz)) }}
{%- endmacro -%}
Assuming you have data:
created_at
tz
2022-06-08 T12:00:00Z
America/New_York
2022-06-07 T12:00:00Z
America/Denver
You can then call this like so:
select
{{ convert_timezone("created_at", "tz") }} as use_tz_field,
{{ convert_timezone("created_at", "'America/Los_Angeles'" }} as use_tz_literal