Filtering

Overview

All models support mutation using JMESPath expressions.

Select

The select method allows selecting subsets of data from within a model:

from beancount import loader
from bdantic import parse_loader

entries, errors, options = loader.load_file("ledger.beancount")
result = parse_loader(entries, errors, options)

print(result.entries[0].select("postings[].units"))

The result is dependent on the selection expression and will be in the form of one or more nested dictionaries/lists. In other words, the original models will be in dict form, however, they can be converted back like so:

from bdantic.models import Amount

amounts = result.entries[0].select("postings[].units")
amounts = [Amount.parse_obj(a) for a in amounts]

Filter

For models which wraps lists, the filter method can be used for filtering the list down:

from beancount import loader
from bdantic import parse_loader

entries, errors, options = loader.load_file("ledger.beancount")
result = parse_loader(entries, errors, options)

txns = result.entries.filter("[?ty == 'Transaction']")

Unlike the select method, filtering will attempt to preserve the formats of models. This is important to note because if the expression mutates the models in such a way that validation fails then an exception will be raised.