Seqan read compressed stream

Question

[pmarijon]

Hi,

I want read a sequence file (fasta fastq bam, etc), so I read Seqan tutorial. But If I want know my position in file I need use std::ifstream (for generate a progress bar) , it's not a problem, I write this test code:

#include <iostream>
#include <fstream>

#include <seqan/seq_io.h>


int main (int argc, char ** argv) {
    std::streampos begin,end;
    std::ifstream myfile (argv[1], std::ios::in | std::ios::binary);

    begin = myfile.tellg();

    seqan::SeqFileIn seq_file(myfile);
    seqan::CharString id;
    seqan::Dna5String seq;
    seqan::CharString qual;

    while(!seqan::atEnd(seq_file))
    {
    seqan::readRecord(id, seq, qual, seq_file);
    std::cout<<"pos: "<<myfile.tellg()<<" id "<<id<<std::endl;
    }

    end = myfile.tellg();

    myfile.close();

    std::cout << "begin: "<< begin << " end: "<< end << std::endl;
    std::cout << "size is: " << (end-begin) << " bytes.\n"<<std::endl;
    return 0;
}

But when I try this code on compressed fastq read, Seqan throw an exception terminate called after throwing an instance of 'seqan::ParseError'

My question :

• Use std::ifstream is the only solution to get the current position in file ?
• How I can say to Seqan this stream are a compressed stream ?
• Can I generate an uncompressed stream from my compressed stream (with SeqAn or zlib)

Thanks.

[Pierre Lindenbaum]

why would you want to know the position of a fastq record in a compressed file ? unless you're using bgzf, there is no way to 'fseek ' a bgzip file...

[pmarijon]

I want generate a progress bar, the post required an edit. For compress file we can have a good approximation with size of compressed file and the position in compressed file.

[Pierre Lindenbaum]

then I would create a custom std::streambuf to count the number of bytes... e.g: https://artofcode.wordpress.com/2010/12/12/deriving-from-stdstreambuf/

[pmarijon]

I use a std::ifstream to get current position in file during seqan parsing, it's easy. But when I try my code on compressed file, seqan parsing failed. So seqan didn't detect my stream contain compressed data or seqan can't work on compressed stream, but isn't documented.

[kloetzl]

So seqan didn't detect my stream contain compressed data or seqan can't work on compressed stream, but isn't documented.

Usually it is the other way round: Things don't work on compressed data, unless documented.

[pmarijon]

Is documented

These classes provide an API for accessing sequence files in different file formats, either compressed or uncompressed.

Source : https://seqan.readthedocs.io/en/master/Tutorial/InputOutput/SequenceIO.html

[kloetzl]

Well, there is compressed .bam and compressed .gz.

[pmarijon]

And bz2 according to http://docs.seqan.de/seqan/master/group_FileCompressionTags.html#FileCompressionTags%23BgzFile