Function build_detection_from_node

src/cluecode/copyrights.py:526–589

Return a ``cls`` Detection object from a pygmars.tree.Tree ``node`` with a space-normalized string value or None. Filter ``node`` Tokens with a type found in the ``ignored_labels`` set of ignorable token types. For copyright detection, include trailing "All rights reserved" if

(
    node,
    cls,
    ignored_labels=frozenset(),
    include_copyright_allrights=False,
    refiner=None,
)

def build_detection_from_node(
    node,
    cls,
    ignored_labels=frozenset(),
    include_copyright_allrights=False,
    refiner=None,
):
    """
    Return a ``cls`` Detection object from a pygmars.tree.Tree ``node`` with a
    space-normalized string value or None.

    Filter ``node`` Tokens with a type found in the ``ignored_labels`` set of ignorable
    token types.

    For copyright detection, include trailing "All rights reserved" if
    ``include_copyright_allrights`` is True.

    Apply the ``refiner`` callable function to the detection string.
    """
    include_copyright_allrights = (
        cls == CopyrightDetection
        and include_copyright_allrights
    )

    leaves = list(filter_tokens(node, ignored_labels=ignored_labels))

    if include_copyright_allrights:
        filtered = leaves
    else:
        filtered = []

        for token in leaves:
            # FIXME: this should operate on the tree and not on the leaves
            # ALLRIGHTRESERVED: <NNP|NN|CAPS> <RIGHT> <NNP|NN|CAPS>? <RESERVED>

            # This pops ALL RIGHT RESERVED by finding it backwards from RESERVED
            if token.label == 'RESERVED':
                if (
                    len(filtered) >= 2
                    and filtered[-1].label == 'RIGHT'
                    and filtered[-2].label in ('NN', 'CAPS', 'NNP')
                ):
                    filtered = filtered[:-2]
                elif (
                    len(filtered) >= 3
                    and filtered[-1].label in ('NN', 'CAPS', 'NNP')
                    and filtered[-2].label == 'RIGHT'
                    and filtered[-3].label in ('NN', 'CAPS', 'NNP')
                ):
                    filtered = filtered[:-3]
            else:
                filtered.append(token)

    node_string = ' '.join(t.value for t in filtered)
    node_string = ' '.join(node_string.split())

    if refiner:
        node_string = refiner(node_string)