On 06/26/2018 08:20 AM, Arthur Zakirov wrote:
Hello hackers, I'd like to propose the patch which syncs PostgreSQL snowball stemmers. As Tom pointed [1] stemmers haven't synced for a very long time. I copied all source files without changes, except replacing '#include "../runtime/header.h"' with '#include "header.h"' and removing includes of standard headers from utilities.c. Hungarian language uses ISO-8859-1 and UTF-8 charsets in Postgres HEAD. But in Snowball HEAD it is ISO-8859-2 per commit [2]. This patch changes hungarian's charset from ISO-8859-1 to ISO-8859-2 too. Additionally updated files in the patch are: - utilities.c - header.h Will add to the next commitfest. Any comments? 1 - https://www.postgresql.org/message-id/5689.1519054983%40sss.pgh.pa.us 2 - https://github.com/snowballstem/snowball/commit/4bcae97db044253ea2edae1dd3ca59f3cddd4b9d
I agree with Tom that we should sync with the upstream before we do anything else. This is a very large patch but with fairly limited impact. I think now at the start of a dev cycle is the right time to apply it.
I don't know if we have a buildfarm animal testing Hungarian. Maybe we need a buildfarm animal or two testing a large number of locales.
cheers andrew -- Andrew Dunstan https://www.2ndQuadrant.com PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
