Encoding sleuthing function

Sometimes you know what a character should be, and you know how it's stored in a file, and you want to figure out what encoding it might have been (e.g. https://github.com/hadley/readr/issues/252). It would be helpful to have something to help automate the search, e.g.

``` R
x <- "Ü"
encs <- names(stri_enc_list())
conv <- lapply(encs, function(to) stri_conv(x, "UTF-8", to, to_raw = TRUE)[[1]])
names(conv) <- encs
Filter(function(x) identical(x[[1]], charToRaw("\x86")), conv)
```

Start by reproducing the R example with stri_enc_list() and stri_conv(), including the target raw byte \x86. Define an interface that automates this encoding search and verify that it identifies matching encodings without requiring the caller to write the lapply and Filter steps.

Written by the indexing model from the issue text.