The never-ending encoding story got another twist for me today: on preparing an example, I noted that garbled Umlauts are exactly mirror-errored when using ui/server and single file ( runApp(list()) shiny examples.
For the ui/server app, try:
runGist("https://gist.github.com/f734287e9ebf4c59806b")
For the single-file app, see
https://gist.github.com/8a99754f2ac4c6917cb8.gitsessionInfo() is shown there; this is under Windows 7, German, and latest shiny. Garbled text in one app is good
in the other app, and vice versa. I have marked good and bad (=Umlaut-garbled) text in the app.
Hi Dieter, sorry for this problem. Can you try seeing if setting options(encoding="UTF-8") before calling runGist helps?
> options(encoding="UTF-8") > library(shiny) > runGist("https://gist.github.com/f734287e9ebf4c59806b") Downloading https://gist.github.com/f734287e9ebf4c59806b/download Listening on http://127.0.0.1:4904 Warning in readLines(conn) : invalid input found on input connection '' Warning in readLines(conn) : incomplete final line found on '' Warning message: In download.file(url, ...) : downloaded length 622 != reported length 200
Okay, we are trying to put an end to this never-ending encoding
problem under Windows now, and we will really appreciate it if you can
help us test it:
umlautOk = nlevels(na.omit(as.factor(iconv(c("Pre","Prä")))))==2
failed, but generating a 1-page report with knitr trashed 5 Minutes and terminated with an index out of range (2 seconds on runApp(), or with forced Language settings)Thanks to Jeff Allen and Joe Cheng for their heroic efforts. And thanks to you: For me the output of
runGist("https://gist.github.com/f734287e9ebf4c59806b")
with your modified Shiny version looks good.
Dieter