version 0.12.3:
^^^^^^^^^^^^^^^

* parameter `buffering` for :func:`parsing.open`

version 0.12.2 and before :
^^^^^^^^^^^^^^^^^^^^^^^^^^^

* ...I get carried away and I forget to keep track...

version 0.10.1:
^^^^^^^^^^^^^^^

* Fixed quality for CSFASTQ file (the quality line was the same as CSFASTA)


version 0.10.0:
^^^^^^^^^^^^^^^

* renamed module 'string' to 'ngsp_string' (dark/ugly corner in Python's
  import system caused issues on some systems)

* added parameter `--buffer-size` to script xsq-convert
 
* fixed issue with FASTQ parsing when the quality line starts with '@'

* fixed ZeroDivisionError on when converting XSQ files for which there
  is close to no data for a given barcode

* added switch `--unassigned` to xsq-convert to be able to recover
  reads with unassigned barcodes

version 0.9.1:
^^^^^^^^^^^^^^

* fixed unwanted trimming of the last kmer when using the functions
  :func:`kmers.kmerbytes_frombit2_iter`, :func:`kmers.kmerbit2_frombit2_iter`,
  :func:`poskmers.kmerbit2_frombit2_iter`

version 0.9.0:
^^^^^^^^^^^^^^

* refactored the options for converting XSQ files. Two families of data formats
  can be chosen (FASTA-like or FASTQ-like), and for each either colour-space
  or base-space data can be obtained.

* add new module colorspace, and new script to perform colorspace conversions


version 0.8.2:
^^^^^^^^^^^^^^

* option to convert XSQ files into CSFASTQ files


version 0.8.1:
^^^^^^^^^^^^^^

* Report OS errors when trying to create missing directories for the XSQ reports

* Handle F5 reads in XSQ files when having paired-end RNASeq data

version 0.8.0:
^^^^^^^^^^^^^^

* :func:`_libdna.slicebit2_frombit2bytes` to slice DNA while remaining bitpacked.

* New :func:`kmers.poskmerbit2_frombit2_iter`.

version 0.7.6:
^^^^^^^^^^^^^^

* Fixed HTML QC report when subdirectories were not existing
  (failed to create them)

* QC plots now resized with the window

version 0.7.5:
^^^^^^^^^^^^^^

* Fixed HTML QC report for paired-end runs

* Added parameters maxhead and maxseq to limit the size of the header
  and of the sequence when reading FASTAB files (since there are
  otherwise no checks while reading binary files).

* Fixed HTML QC report when the target directory is already existing
 
version 0.7.4:
^^^^^^^^^^^^^^

* ASCII representation of quality values was incorrect for QUAL files.

* When converting XSQ files, libraries are stored inside directories
  corresponding to the respective XSQ file names.

* fixed anchor in the HTML quality reports

* fixed relative path for the QC values when the HTML report is not generated
  in the current directory ('.')

version 0.7.3:
^^^^^^^^^^^^^^

* ASCII representation of quality values was incorrect for XSQ conversion.
  SOLiD is using the Sanger scoring, not the illumina one.

version 0.7.2:
^^^^^^^^^^^^^^

* XSQ files can be inconsistent in may ways. Attempt at getting more robust
  against that

version 0.7.1:
^^^^^^^^^^^^^^

* new exception fastq.FastqError when problems while parsing FASTQ files

* fixed handling of flag -i for xsq-convert


version 0.7.0:
^^^^^^^^^^^^^^

* functions in _libdna to make the reverse, complement, and reverse complement
  while keeping the DNA bit-packed.

* new methods reversed, complemented, and reversecomplemented for the class dna.PackedDNABytes.

