Thanks Tomas and 'Brodie' for your expert explanation; it provides great help in understanding and solving my immediate problem.
Thomas' observation to 'do something like e.g. "only keep ASCII digits, ASCII space, ASCII underscore, but remove all other characters"' points to a basic weakness in the code I'm looking at. E.g., removing non-breaking space is probably not appropriate ('foo\ua0bar' is probably cleaned to 'foo bar' and not 'foobar'). And more generally other non-ASCII characters ('fancy' quotes, em-dashes, ...) would require special treatment. It seems like the right thing to do is to handle the raw data in its original encoding, rather than to try to clean it to ASCII. Martin On 1/5/22, 4:17 AM, "Tomas Kalibera" <tomas.kalib...@gmail.com> wrote: Hi Martin, I'd add few comments to the excellent analysis of Brodie. - \xhh is allowed and defined in Perl regular expressions, see ?regex (would need perl=TRUE), but to enter that in an R string, you need to escape the backslash. - \xhh is not defined by POSIX for extended regular expressions, neither it is documented in ?regex for those; TRE supports it, but still portable programs should not rely on that - literal \xhh in an R string is turned to the byte by R, but I would say this should not be used at all by users, because the result is encoding specific - use of \u and \U in an R string is fine, it has well defined semantics and the corresponding string will then be flagged UTF-8 in R (so e.g. \ua0 is fine to represent the Unicode no-break space) - see caveats of using character ranges with POSIX extended regular expressions in ?regex re encodings, using Perl regular expressions in UTF-8 mode is more reliable for those So, a variant of your example might be: > gsub("[\\x7f-\\xff]", "", "fo\ua0o", perl=TRUE) [1] "foo" (note that the \ua0 ensures that the text is UTF-8, and hence the UTF-8 mode for regular expressions is used, ?regex has more) However, I think it is better to formulate regular expressions to cover all of Unicode, so do something like e.g. "only keep ASCII digits, ASCII space, ASCII underscore, but remove all other characters". Best Tomas On 1/4/22 8:35 PM, Martin Morgan wrote: > I'm not very good at character encoding / etc so this might be user error. The following code is meant to replace extended ASCII characters, in particular a non-breaking space, with "", and it works in R-4-1-branch > >> R.version.string > [1] "R version 4.1.2 Patched (2022-01-04 r81445)" >> gsub("[\x7f-\xff]", "", "fo\xa0o") > [1] "foo" > > but fails in R-devel > >> R.version.string > [1] "R Under development (unstable) (2022-01-04 r81445)" >> gsub("[\x7f-\xff]", "", "fo\xa0o") > Error in gsub("[\177-\xff]", "", "fo\xa0o") : invalid regular expression '[-�]', reason 'Invalid character range' > In addition: Warning message: > In gsub("[\177-\xff]", "", "fo\xa0o") : > TRE pattern compilation error 'Invalid character range' > > There are other oddities, too, like > >> gsub("[[:alnum:]]", "", "fo\xa0o") # R-4-1-branch > [1] "\xfc\xbe\x8c\x86\x84\xbc" > >> gsub("[[:alnum:]]", "", "fo\xa0o") # R-devel > [1] "<>" > > The R-devel sessionInfo is > >> sessionInfo() > R Under development (unstable) (2022-01-04 r81445) > Platform: x86_64-apple-darwin19.6.0 (64-bit) > Running under: macOS Catalina 10.15.7 > > Matrix products: default > BLAS: /Users/ma38727/bin/R-devel/lib/libRblas.dylib > LAPACK: /Users/ma38727/bin/R-devel/lib/libRlapack.dylib > > locale: > [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8 > > attached base packages: > [1] stats graphics grDevices utils datasets methods base > > loaded via a namespace (and not attached): > [1] compiler_4.2.0 > > (I have built my own R on macOS; similar behavior is observed on a Linux machine) > > Any hints welcome, > > Martin Morgan > ______________________________________________ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel