Committed by Greg Sabino Mullane <[email protected]> Minor doc changes for utf8 stuff.
--- Changes | 7 ++++++- Pg.pm | 10 +++++----- 2 files changed, 11 insertions(+), 6 deletions(-) diff --git a/Changes b/Changes index 7278ffa..cfdc24c 100644 --- a/Changes +++ b/Changes @@ -1,7 +1,11 @@ 'GSM' is Greg Sabino Mullane, [email protected] -Version ??? +Version 3.0.0? + + - Major change in UTF-8 handlind. If client_encoding is set to UTF-8, + always mark returned Perl strings as utf8. See the docs for + pg_enable_utf8 for more information. - Map SQL_CHAR back to bpchar, not char. [GSM] @@ -25,6 +29,7 @@ Version ??? - Change NOTICE to DEBUG1 in t/02attribs.t test for handle attribute "PrintWarn": implicit index creation is now quieter in Postgres [Erik Rijkers] + Version 2.19.3 Released August 21, 2012 (git commit be018f10fdaf4163f98affcb7244046e8f47420d) - Fix bug in pg_st_split_statement causing segfaults diff --git a/Pg.pm b/Pg.pm index c80a613..a8676c2 100644 --- a/Pg.pm +++ b/Pg.pm @@ -3135,14 +3135,14 @@ marks, such as geometric operators. =head3 B<pg_enable_utf8> (integer) DBD::Pg specific attribute. The behavior of DBD::Pg with regards to this flag has -changed as of version xxx. The default value for this attribute, -1, indicates -that the internal C<utf8> flag will be turned on for all strings coming back +changed as of version 3.0.0. The default value for this attribute, -1, indicates +that the internal Perl C<utf8> flag will be turned on for all strings coming back from the database if the client_encoding is set to 'UTF8'. Use of this default -is highly encouraged, and you should not need to use this attribute except -for the following two conditions: +is highly encouraged. If your code was previously using pg_enable_utf8, you can +probably remove mention of it entirely. If this attribute is set to 0, then the internal C<utf8> flag will *never* be -turned on for returned data, regardless of the current client_encoding. +turned on for returned data, regardless of the current client_encoding. If this attribute is set to 1, then the internal C<utf8> flag will *always* be turned on for returned data, regardless of the current client_encoding -- 1.7.1
