`diamond view` sometimes produces faulty BTOP strings

[imLew]

When diamond blastx is run with --outfmt 100 and the resulting .daa file is viewed with diamond view the option --outfmt 6 btop produces a different result than what is produced by diamond blastx ... --outfmt 6 btop.

In particular BTOP strings generated by diamond view sometimes have identical consecutive letters, which should not be possible because consecutive letters indicate mismatches. The BTOP strings produced by diamond blastx always seem to have X in the one of the repeated positions.

For example

-  "btop": "18IV8LV5NNLLNNVGLFNNFFNNQQSSVIVVDH1Q-1RH1LV6IL3NG10",
+  "btop": "18IV8LV5XNXLXNXGXFXNXFXNXQXSXIXVDH1Q-1RH1LV6IL3NG10",

[bbuchfink]

This probably happens because these letters have been masked due to repeat masking (you can check using --masking 0). The view workflow does not know about the masking any more, so it will print the unmasked letters. Correctly considering the masking when producing the output is a known issue that still needs to be addressed in a future version.