Pyfaidx

- `--regex` option for `faidx` script allows matching only certain records (55)
- `filt_function` argument for `Faidx` and `Fasta` allows filtering certain records based on a matching function (jsilter)
- fixed a bug (56) when calling `FastaRecord.long_name` when using `filt_function` or `split_char`.
- fai index is not created if the modification time == fasta file (66)

- Zero-length slices return an empty string (53 thanks vejnar)
- Added FastaRecord.long_name property to extract the actual defline for a sequence from the file. (54)
- fai index is not created if the modification time == fasta file (66)

- Files with modification times newer than their indices are re-indexed (50 - thanks deannachurch)
- Test data is now downloaded from NCBI using biopython, and unit tests are more independent (51)
- fai index is not created if the modification time == fasta file (66)

- Fixes bug number 47, which did not allow sequences containing a lowercase "n" to be complemented (thanks pschaughency)
- Implements better error handling in 48 for sequences with invalid characters.
- Added `--version` flag to faidx script (42)
- Now only allow integer values for `read_ahead` (41)

- --delimiter option for cli script and split_char argument for Fasta
and Faidx

- --split-files option writes each returned sequence to an individual
file. Names are generated based on the sequence name and region
coordinates.
- --stats option prints the name and sequence length for each entry,
suitable for use as a UCSC-style
[chrom.sizes](http://genome.ucsc.edu/goldenpath/help/hg19.chrom.sizes)
file.
- Sequence longname attribute allows access to "chr:start-end
(complement)" formatted names