"omit_na=TRUE" keeps the string as is if the replacement is NA

[yutannihilation]

Fixes #267.

[gagolews]

Cool. Can you update documentation in R/sub.R too?

@param omit_na single logical value; if \code{TRUE}, missing values in \code{from},
#' \code{to}, or \code{length} will result in an unchanged input; replacement function only

should rather now be "....missing values in any of the arguments provided will result..."

[yutannihilation]

Thanks! I did update.

Really thanks for your suggestion of the paragraph, as I'm not so good at English 👍

[yutannihilation]

By the way, why is the CI falling with STRINGI_DISABLE_ICU_BUNDLE="1"? The error says below. Is this up to my PR?

checking with pkg-config for the system ICU4C... no
*** pkg-config did not detect ICU4C-devel libraries installed
*** Trying with "standard" fallback flags
checking whether we may build an ICU4C-based project... yes
checking programmatically for sufficient U_ICU_VERSION_MAJOR_NUM... no
*** The available ICU4C cannot be used
*** Failed to configure package build with forced system ICU usage
*** Consider installing, e.g., libicu-devel (RHEL, Fedora) or libicu-dev (Debian, Ubuntu)
*** together with pkg-config
ERROR: configuration failed for package ‘stringi’
* removing ‘/home/travis/build/gagolews/stringi/stringi.Rcheck/stringi’

[gagolews]

Thanks! Don't worry about the failing travis CI checks, it seems like I have to upgrade the build settings.

[yutannihilation]

Oh, I see. Thanks for merging!