Class Query

src/licensedcode/query.py:155–652

A query represent a whole file or string being scanned for licenses. It holds known tokens, known token line positions, unknown tokens positions, and query runs. It also tracks which parts have been matched as matching progresses. For positions, we track primarily the absolute

class Query(object):
    """
    A query represent a whole file or string being scanned for licenses. It
    holds known tokens, known token line positions, unknown tokens positions,
    and query runs. It also tracks which parts have been matched as matching
    progresses.

    For positions, we track primarily the absolute position of known tokens.
    Unknown tokens are tracked only as the number of unknown in reference to a
    known token position. (using -1 for unknown tokens that precede the first
    known token.)

    Line positions (e.g. line numbers) start at 1 with a possibility
    to set the start_line to another value for special cases.

    A query is broken down in one or more "runs" that are slices of tokens used
    as a matching unit.
    """

    __slots__ = (
        'location',
        'query_string',
        'idx',
        'line_threshold',
        'tokens',
        'line_by_pos',
        'unknowns_by_pos',
        'unknowns_span',
        'stopwords_by_pos',
        'shorts_and_digits_pos',
        'query_runs',
        '_whole_query_run',
        'high_matchables',
        'low_matchables',
        'spdx_lid_token_ids',
        'spdx_lines',
        'has_long_lines',
        'is_binary',
        'start_line',
    )

    def __init__(
        self,
        location=None,
        query_string=None,
        idx=None,
        line_threshold=LINES_THRESHOLD,
        start_line=1,
        _test_mode=False,
    ):
        """
        Initialize the query from a file `location` or `query_string` string for
        an `idx` LicenseIndex.
        Break query in runs when there are at least `line_threshold` empty lines
        or junk-only lines.
        Line numbers start at ``start_line`` which is 1-based by default.
        """
        assert (location or query_string) and idx