Checkpoints could be accessed as rules when using rules

[iromeo]

E.g here foo is accessed via checkpoints and rules

checkpoint foo:
    output:
        key1="{sample}.tsv",
        key2="{sample}.txt"


def boo_input(wildcards):
    path = checkpoints.foo.get(**wildcards).output.key1
    df = pd.read_csv(path, sep="\t")
    return tuple(df[df["mask"]]["File"])


rule boo:
    input: boo_input
    output: "too/{sample}.txt"
    log: "too/{sample}.txt.log"
    shell:
        "echo  {input:q} 1>{output} 2>{log}"

rule all:
    input:
        expand(
            [
                rules.foo.output.key1,
                rules.foo.output.key2,
                rules.boo.output
            ],
            sample=["A", "B"]
        ),