subject:"safer Unicode parsing"

Re: safer Unicode parsing

2016-11-19 Thread Bruno Haible

Paul Eggert wrote: > > -#if FULL_SAFETY || CONFIG_UNICODE_SAFETY > > Since this removes the only use of FULL_SAFETY, its definition in > tests/unistr/test-u32-mbtouc.c should also be removed. Good catch; thanks. Pushed with this additional change. Bruno -- In memoriam Farhád Asdaqí

Re: safer Unicode parsing

2016-11-17 Thread Paul Eggert

On 11/17/2016 10:14 AM, Bruno Haible wrote: -#if FULL_SAFETY || CONFIG_UNICODE_SAFETY Since this removes the only use of FULL_SAFETY, its definition in tests/unistr/test-u32-mbtouc.c should also be removed. Otherwise the change looks good to me; thanks.

safer Unicode parsing

2016-11-17 Thread Bruno Haible

Hi, The unistr/* modules and, with them, libunistring contain code for parsing byte streams to Unicode characters. Currently it is optimized for speed at the expense of safety (garbage in - garbage out). The code for safer parsing (garbage in - error code) is already present for years, but needs t