NEWS

rtracklayer 1.30

Add readGFF(), a fast and flexible GFF/GTF parser implemented in C. See ?readGFF for more information.

import.gff() now uses readGFF() internally which makes it at least 5x faster in most use cases and dramatically reduces its memory footprint.

ucscGenomes() retrieves organism information
New function exportToTabix() exports a GRanges to a tabix-indexed tab separated file that contains all of the metadata columns. Use import,TabixFile to load specific ranges of data back into a GRanges.
BigWig import/export to/from Integer/Numeric/RleList is now much more efficient, and uses a more efficient storage format within the BigWig file, when possible.

import,BigWigFile gains an asRle parameter that returns the data as an RleList (assuming it tiles the sequence); much faster than importing a GRanges and calling coverage() on it.
add export,RleList,BigWigFile method for direct (and much more efficient) output of RleLists (like coverage) to BigWig files.

handle different Set-Cookie header field casing; often due to proxies (thanks to Altuna Akalin)
attempt to more gracefully handle UCSC truncation of large data downloads
handle re-directs due to UCSC mirroring (thanks Martin Morgan)

All methods with asRangedData argument now have it default to FALSE, instead of TRUE. A warning is issued if the argument is missing. Eventually, we will drop all support for RangedData import (export will still work via automatic coercion to GRanges).

Chromosome list for a genome is now downloaded from the table browser, instead of the Genome Browser page. This supports genomes with more than 1000 contigs.
BEDX+Y formats now work when a track line is present, and the extraCols argument is used for the column names.
path.expand() is now called for paths passed off to the Kent library.
Order of metadata columns upon GFF import no longer depends on LC_COLLATE.

Export of GappedAlignments objects as BAM files. There is also a corresponding import method.
Support for BEDX+Y import using the extraCols parameter to import.
Direct export of RleList objects to BigWig files.

Integrated with tabix via Rsamtools. BED and GFF export methods now generate a tabix index, if index=TRUE. Most import() methods gain a 'which' argument that takes advantage of tabix, when available.
Added wigToBigWig() function for efficient conversion of WIG to BigWig.
Added SeqinfoForBSGenome() and SeqinfoForUCSCGenome() for conveniently retrieving Seqinfo objects for a given genome.
Added support for FASTA import/export via Biostrings.
GTF and GVF files are now parsed as GFF.

The import/export API is now based on RTLFile objects, which wrap a file path, URL or connection. There is an RTLFile subclass for every file format. This makes it easier to extend rtracklayer (export, import) with new file types. The existing API is still supported (and even encouraged for most uses).
Handle CSV attributes in GFF3 using CharacterList columns.
BED columns thickStart/thickEnd translate to an IRanges column named "thick". The blockStarts/Sizes/Count columns now map to a single RangesList "blocks" column.

Numerous fixes in the import/export methods, as a result of implementing a full unit test suite. If something was not working for you in the past, please try again.
Compression and connections should now work fairly uniformly across file types. (start date: 29 March, 2012)