This is a specialized function for use in connection
with text reordering. It is designed to facilitate manual splitting
of block boundary boxes and typically takes a one-row block dataframe
generated by from_labelme().
reassign_tokens2(token_df, block, page = 1)a data frame generated by dair::build_token_df
a one-row data frame of the same format as token_df
the number of the page on which the block belongs
a token data frame
if (FALSE) {
new_token_df <- reassign_tokens2(token_df, new_block_df)
new_token_df <- reassign_tokens2(token_df, new_block_df, 5)
}