Friday 29 August 2003
Arabica Utils contains a small (but slowly increasing) number of code conversion facets. One of these coverted from ISO8859:1 (Latin1) encoded text to UTF8, and another converted from UTF16 to UTF8. I realised however, that they were inconsistently named, and have now changed this.
From now on, the facets will be named
. The UTF16 to UTF8 is therefore called
, and reading a stream imbued with this facet converts from
UTF8. Writing to such a stream converts from
Facets with a name like [something
]codecvt.h are byte stream encoders/decoders where [something] is an external encoding. Reading a stream imbued with
will un-Base64 the stream, while writing to it will Base64 everything.
The upshot of this is that some of the names of existing facets have changed in CVS -
. I now this is a bit of a pain in the arse, but I do believe it will make thing clearer going forward.