2017-02-04 | Paul Boddie | raw annotate files changeset graph | Removed recoding to UTF-8 since this failed for ISO-8859-15, causing UTF-8 recodings of byte sequences to occur, not producing such undesirable data for ISO-8859-1 only because of it being special-cased. This change may break other ASCII-incompatible encodings because UTF-8 is likely to be the safe form of such data, permitting the parser to understand it, and without such recoding the parser will no longer recognise the grammar's tokens. |