From 3cec7dcb201568a09ee9cebf0ea1e43cba9d334a Mon Sep 17 00:00:00 2001 From: Jonas Bernoulli Date: Tue, 17 Jan 2023 17:33:37 +0100 Subject: [PATCH] Use utf-8 coding-system instead of utf-8-auto MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This fixes an issue that was merely triggered by a change in Emacs [1]. [2: 127bb98] started binding `coding-system-for-write' and `coding-system-for-read' to `utf-8-auto'. As Eli points out at [3] `utf-8-auto' was probably a mistake and `utf-8' should have been used instead. 1: https://github.com/emacsmirror/emacs/commit/cfd2b3504ab5de6eb5 2: 2014-02-02 127bb98df20d84117d34822f50ea584af81b19c3 Set coding system to utf-8-auto for sqlite. 3: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=60872#11. > Setting coding-system-for-write to utf-8-auto makes no sense; > that coding-system's _only_ raison d'être is for using in > coding-system-for-read, i.e. when decoding stuff that may or > may not start with a BOM. Since EmacSQL is not a general-purpose database but instead stores data encoded by Emacs for later retrieval by Emacs only, we do not have to with the unlikely appearance of a BOM, so we can use `utf-8' instead of `utf-8-auto' for writes *and* reads. Re #104, #106, magit/forge#533, magit/forge#535, bug#60872. --- emacsql-sqlite.el | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/emacsql-sqlite.el b/emacsql-sqlite.el index 1193147..7b8876f 100644 --- a/emacsql-sqlite.el +++ b/emacsql-sqlite.el @@ -73,8 +73,9 @@ used.") ((connection emacsql-sqlite-connection) &rest _rest) (emacsql-sqlite-ensure-binary) (let* ((process-connection-type nil) ; use a pipe - (coding-system-for-write 'utf-8-auto) - (coding-system-for-read 'utf-8-auto) + ;; See https://debbugs.gnu.org/cgi/bugreport.cgi?bug=60872#11. + (coding-system-for-write 'utf-8) + (coding-system-for-read 'utf-8) (file (slot-value connection 'file)) (buffer (generate-new-buffer " *emacsql-sqlite*")) (fullfile (if file (expand-file-name file) ":memory:"))