about read.get_aligned_pairs(matches_only=True, with_seq=True)

[bioinformatica]

I want to count the number of reads for the mutation T>C, I used the following code：

        query_seq = read.query_sequence
        ref_seq = reference.fetch(read.reference_name, read.reference_start, read.reference_end)

        alignments = read.get_aligned_pairs(matches_only=True, with_seq=True)

        for query_pos, ref_pos, ref_base in alignments:
            if ref_base.upper() == 'T' and query_seq[query_pos].upper() == 'C':
                tc_count += 1
                print(query_seq)
                print(ref_seq)
                break

Output：
CATTTCTTAGTCCTACACACTAAAAACAGTAATGTGCTGCTTTGAGTGTGTAGGACTAAGAAATGGGATTCAGAGTAGTAAAGAGAAAAGTGGAATTTCCAAGCACTATGAATTACTGTTCTTTAAAAAACAGCAAAAATC

ACTTTAAAAACAGTAATGTGCTGCTTTGAGTGTGTAGGACTAAGAAATGGGATTCAGAGTAGTAAAGAGAAAAGTGGAATTTCCAAGCACTATGAATTACTGTTCTTTAAAAAACAGCAAAAATC

I set matches_only=True. Normally, tccount is always equal to 0. Why are their lengths different? causing tccount to increase

[bioinformatica]

If matches_only=True is set, will only the exact same base pairs be returned? How should I change the code to count the number of reads with t>C?