Hi,
New upstream.
Sat May 16 2026 Caolán McNamara (caolan.mcnamara@collabora.com)
* bump release to 3.4.8
+ respect bufsize in prepbuffer
+ widen fp_Create bufsize from uint4 to size_t (ABI break;
SONAME bumped libexttextcat-2.0.so.0 -> libexttextcat-2.0.so.1
via -version-info 1:0:0)
+ fix leak in textcat_Init bailout path
+ avoid forming &dest[-1] in wg_trim (undefined behavior)
+ update README for current UDHR source location (efele.net,
after unicode.org/udhr retired in January 2024)
+ add Telugu, Kabyle, Nenets, Karakalpak, Morisyen, Arpitan,
Sidamo, Quechua (North Bolivian), Inari Saami and Skolt Saami
+ rename ha-NG to ha-Latn-NG to match BCP-47/LibreOffice
Ben file:
title = "libexttextcat";
is_affected = .depends ~ "libexttextcat-2.0-0" | .depends ~ "libexttextcat-2.0-1";
is_good = .depends ~ "libexttextcat-2.0-1";
is_bad = .depends ~ "libexttextcat-2.0-0";
(direct) r-deps rebuilt, they all build (frog after the new ucto), so
bin-NMUs will work
Regards,
Rene