leeper / rio

…e package. They all seem to be 1.0 (#259)

* isfiletext() checks whether a file is text or binary in a cross platform matter, completing ticket #236. This can be useful when a file extension is missing or ambiguous

* Minor tweak-- removed redundant 'fwf' entry from 'txtformats' in the test for isfiletext()

* Incorporating feedback from #239#pullrequestreview-335144814 and merging in changes to master

* Incorporating review feedback

… check

…e8.3 setting, letting the current default (and newer) version of xcode be used instead. Also took the before_install script back out because now the data.table binary is again available. Fixes #246. Fixes #242

…visCI (see #246, #242, #247)

… headers for XML where '&' and ' ' both get replaced by '.'. Fixes #234

…e failing to install, from upstream: Rdatatable/data.table#3938

* Added dynamic argument checking for 'readODS:read_ods()' as proposed in #223. Also more detailed tests.

* '.import.rio_xls()' no longer passing '...' to 'read_xls()' because that functions cannot accept '...' arguments. Instead, the length of '...' is checked and if > 0 a warning is issued. With corresponding tests. Also went back and changed my earlier 'read_xlsx()' test to be 'expect_warning()' instead of 'expect_true()' since by that point that's what needs testing.

* Add support for pzfx files

* Completion of .pzfx addition by documenting in NEWS.md

* Give message if pzfx test is skipped due to missing library

* Remove pzfx as import requirement

…to 'load()' nor to 'readRDS()' because those underlying functions cannot accept '...' arguments. Instead, the length of '...' is checked and if > 0 a warning is issued. With corresponding tests. As proposed in #223 (#229)

…opefully I guessed the project settings correctly from the existing files (#227)

…t as per #222 (review) . Please note my response in that thread to the last question about the 'rows' argument.

* Removed duplicated example line in import.R

* Run devtools::document() to propagate change to import.Rd.