added gene multicopies handling. new function detect_copies, and now the target_genome_path of child features and the transfer depend on the parent feature

changed how the features are transfered. now after the transfer of a gene, all its child features are transfered with it. this automatically sorts the output gff, and makes easier the handling of the cov and id filters (if a gene didnt pass, dont transfer its chids, etc)

now always use the function get_seg_occ to get the right segment occurence. so now the option -ann outputs the target genome gff with all the copies of the genes if -cov and -id are at 0

added a function to get the right segment occurence, given the segment_id,walk_name,feature_id,copy_id

modified how the feature paths in the target genomes are found; now several can be found, on several walks (chr/ctg); adapted how the features are transfered (only for the genome annotation transfer, not for the graph_gaf/gff, variations and alignment)

the dict segments_on_target_genome now stores the segment twice if it is present twice in a given walk. adapted the rest of the code to use the last occurence (like before, will change that later)

inverted the start and stop positions in the output gff when the stop was before the start (in inverted regions)

the function that generates the target genomes gff now doesnt transfer if the coverage or id is too low

adapted to keep the orientation of the segments from the walks. added functions to better handle inversions

adapted the code to use the options for the output files. changed every instance of 'clustal' to 'aln' in the code

the functions that read the gfa file now read a file with only the segments or only the walk (generated with getSegmentsCoordinates), to handle smaller files