RcppCWB v0.3.2 ‘Il Postino’ solidifies corpus download.

A new, previously unknown bug users reported when trying to use the GermaParl package to download and install the GermaParl corpus (function germaparl_install_corpus()) generated some urgency to release of a new version of the cwbtools R package: We suddenly saw errors (false positives!) when checking whether GermaParl can be downloaded from Zenodo. A switch from RCurl::url.exists() to httr::http_error() solves this small yet nasty problem.

A second important bug fix is that it is now possible to encode more than one positional attribute at a time – without hacks. We are relieved that this limitation of the usefulness of cwbtools is overcome.

But cwbtools v0.3.2 also brings new features:

• When a corpus is downloaded from Zenodo, the md5 checksum is checked to ensure the integrity of the data.
• It is now possible to download corpora from Amazon S3. This may be very useful for projects that cannot yet publish data publicly.

So data delivery is more robust and more comprehensive. This is why the release is labelled ‘Il Postino’. To learn more about cwbtools v0.3.2 ‘Il Postino’, please consult the package changelog.

Many thanks to the users who reported the issue that downloading GermaParl did not work, temporarily! Good news: Downloading GermaParl from Zenodo works again the way it should.