On 1/14/2021 1:46 PM, Andreas Falkenhahn via curl-library wrote:
Is there any specification concerning the encoding used by the string copied to
CURLOPT_ERRORBUFFER in case of an error?
On my German Windows system I often seem to be getting ISO 8859-1. For example,
the error message for CURLE_SSL_CONNECT_ERROR is this:
"schannel: next InitializeSecurityContext failed: Unknown error (0x80092013) -
Die Sperrfunktion konnte die Sperrung nicht überprüfen, da der Sperrserver offline
war."
Note the German umlauts "ü" in the error message.
Is there any way to make curl return something more i18n-proof, e.g. UTF-8? If
not, consider this a feature request;)
We left extended error messages in the received encoding intentionally.
curl errors are in English but may append a system error like whatever
strerror gives us or for some Windows errors we convert to multibyte to
be consistent and keep local encoding. [1]
[1]: https://github.com/curl/curl/blob/curl-7_74_0/lib/strerror.c#L666-L678
-------------------------------------------------------------------
Unsubscribe: https://cool.haxx.se/list/listinfo/curl-library
Etiquette: https://curl.se/mail/etiquette.html
