I removed scikit-bio's dependency by copying here the code that was used in scikit-bio 0.2.3. I've also adapted the needed code to work with Python 3.
We can potentially improve the code and the speed of the code here if we make the code less general. At this moment, this code is expecting a split libraries output from Qiime 1, so it is not a general FASTQ format. Given this, we may be able to speed up the parsers if we assume that the FASTQ are 4 lines per sequence. However, the amount of code copied over is significant, and I was uncomfortable doing all this optimizations at this moment.
I removed scikit-bio's dependency by copying here the code that was used in scikit-bio 0.2.3. I've also adapted the needed code to work with Python 3.
We can potentially improve the code and the speed of the code here if we make the code less general. At this moment, this code is expecting a split libraries output from Qiime 1, so it is not a general FASTQ format. Given this, we may be able to speed up the parsers if we assume that the FASTQ are 4 lines per sequence. However, the amount of code copied over is significant, and I was uncomfortable doing all this optimizations at this moment.