Messenger_1Million_Stats/scripts/find_holes_v2.py

from million.analyze.word_finder import retain_counts
import million.analyze.message_evaluation as msg_val
from million.model.sequence import Sequence
import million.parse.fb_exports as fb


def _find_value_around_index(messages, value, idx, amplitude) -> int:
    check_value = lambda x: msg_val.get(messages[x]) == value

    if check_value(idx):
        return idx

    for offset in range(1, amplitude):
        o_idx = idx + offset * +1
        if check_value(o_idx):
            return o_idx

        o_idx = idx + offset * -1
        if check_value(o_idx):
            return o_idx

    return -1


def _open_sequence(sequences, msg):
    sequence = Sequence(
        start=msg_val.get(msg), start_message=msg, end=-1, end_message=msg
    )

    sequences.append(sequence)


def _close_sequence(sequences):
    if len(sequences) == 0:
        return

    sequences[-1].end = msg_val.get(sequences[-1].end_message)


def _opened_sequence(sequences):
    return len(sequences) > 0 and sequences[-1].end == -1


export = fb.parse_dirfiles("./data/")
messages = retain_counts(export.messages)

current = 1
base_idx = 0
amplitude = 200

sequences = []

while base_idx < len(messages):
    curr_idx = _find_value_around_index(messages, current, base_idx, amplitude)
    print(
        f"searching {current} from [{messages[base_idx]}]\t-> {'Not found' if curr_idx == -1 else 'Itself' if curr_idx == base_idx else messages[curr_idx]}"
    )

    if curr_idx != -1:  # trouvé

        if not _opened_sequence(sequences):
            _open_sequence(sequences, messages[curr_idx])
        else:
            sequences[-1].end_message = messages[curr_idx]

        base_idx = curr_idx + 1
        current += 1
    else:  # pas trouvé

        # fermer la sequence si ouverte
        if _opened_sequence(sequences):
            _close_sequence(sequences)

        if msg_val.get(messages[base_idx]) < current:
            base_idx += 1
        else:
            current += 1