Sati
Sati
Reverting to an older version of `sp` solved the problem.
> 类似这种？ > > https://github.com/hankcs/HanLP/blob/doc-zh/plugins/hanlp_demo/hanlp_demo/zh/demo_custom_dict_stl.py#L16 以上連結所示這個例句：`"正向匹配商品和服务、任何和服务必按上述切分"`的後半子句，其實最自然的句讀方式應該是：「`任何 / 和服务 / 必 / 按 / 上述 / 切分。`」因為這裡指的是「和服務」短語未被切分的狀態。這個短語本身不成詞，屬於特殊語境之下的特殊用語，這種情況在書面上一般會用開關引號標示出來的。 且「務必」這個詞在自然語當中很常見，「和服」也是自然語中的專有名詞，如果沒有語境及上下文，就算讓人來做句讀，都會有爭議。例如，在這個「強制校正」條例之下，像「`和服務必妥善收藏`」這樣的例句就會無法正確句讀。 猜想像這類情況務必要由語境上來解決。電腦按語境為詞組做區分應該不會太困難。
對，這不失為一種可行的方案。我認為，如果能夠按照不同語境來製作這類的規則模組，效果應該會很不錯。
@quackduck can this feature be implemented?
+1 to fix selection lag issue. One workaround, before this issue is fixed, could be [supporting markup types](#179) that require just *two clicks* to select a region bounded by two...
Regarding @darcamo's `nov` epub-reader support request above, I find: ```elisp (defun org-remark-source-find-file-name () "Assumes that we are currently in the source buffer. Returns the filename for the source buffer. We...
Since this is separate from the main topic of this issue, I have opened a new issue for it.
> if a) when you create a remark the notes buffer also opens and is active, and b) you cd hit `C-c C-c` or similar when done to save and...