[vim/vim] [Doc] A strange criterion for fuzzy search scoring (#8577)

13 views
Skip to first unread message

naohiro ono

unread,
Jul 15, 2021, 10:50:33 PM7/15/21
to vim/vim, Subscribed

This is one of the criteria for fuzzy search scoring written in :h fuzzy-match:

    - Matches after a camel case character or a path separator or a hyphen.

I think it makes sense that matches after

  • a path separator or
  • a hyphen

will affect the score, but I wonder if it is true that matches AFTER

  • a camel case character

will also affect. (I'm assuming here that "a camel case character" stands for an uppercase letter in a lower-camel-cased word, e.g. "C" in "camalCase".)

So I think this may be some kind of mistake. Perhaps this patch may fix it:

diff --git a/runtime/doc/pattern.txt b/runtime/doc/pattern.txt
index 962882765..8e2d8a45e 100644
--- a/runtime/doc/pattern.txt
+++ b/runtime/doc/pattern.txt
@@ -1472,7 +1472,8 @@ criteria:
     - The number of characters (distance) between two consecutive matching
       characters.
     - Matches at the beginning of a word
-    - Matches after a camel case character or a path separator or a hyphen.
+    - Matches at a camel case character.
+    - Matches after a path separator or a hyphen.
     - The number of unmatched characters in a string.
 The matching string with the highest score is returned first.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or unsubscribe.

naohiro ono

unread,
Jul 15, 2021, 11:00:47 PM7/15/21
to vim/vim, Subscribed

cc @yegappan

Yegappan Lakshmanan

unread,
Jul 16, 2021, 10:52:16 AM7/16/21
to vim_dev, reply+ACY5DGCJBPYOKBKD2D...@reply.github.com, vim/vim, Subscribed
Hi,

On Thu, Jul 15, 2021 at 7:50 PM naohiro ono <vim-dev...@256bit.org> wrote:

This is one of the criteria for fuzzy search scoring written in :h fuzzy-match:

    - Matches after a camel case character or a path separator or a hyphen.

I think it makes sense that matches after

  • a path separator or
  • a hyphen

will affect the score, but I wonder if it is true that matches AFTER

  • a camel case character

will also affect. (I'm assuming here that "a camel case character" stands for an uppercase letter in a lower-camel-cased word, e.g. "C" in "camalCase".)

So I think this may be some kind of mistake. Perhaps this patch may fix it:

diff --git a/runtime/doc/pattern.txt b/runtime/doc/pattern.txt
index 962882765..8e2d8a45e 100644
--- a/runtime/doc/pattern.txt
+++ b/runtime/doc/pattern.txt
@@ -1472,7 +1472,8 @@ criteria:
     - The number of characters (distance) between two consecutive matching
       characters.
     - Matches at the beginning of a word
-    - Matches after a camel case character or a path separator or a hyphen.
+    - Matches at a camel case character.
+    - Matches after a path separator or a hyphen.
     - The number of unmatched characters in a string.
 The matching string with the highest score is returned first.

Yes. The description looks good to me. The matches starting with a camel
case character are given an additional score.

Regards,
Yegappan 

vim-dev ML

unread,
Jul 16, 2021, 10:52:34 AM7/16/21
to vim/vim, vim-dev ML, Your activity

Hi,

On Thu, Jul 15, 2021 at 7:50 PM naohiro ono ***@***.***>
wrote:


> This is one of the criteria for fuzzy search scoring written in :h
> fuzzy-match:
>
> - Matches after a camel case character or a path separator or a hyphen.
>
> I think it makes sense that matches after
>
> - a path separator or
> - a hyphen

>
> will affect the score, but I wonder if it is true that matches AFTER
>
> - a camel case character

>
> will also affect. (I'm assuming here that "a camel case character" stands
> for an uppercase letter in a lower-camel-cased word, e.g. "C" in
> "camalCase".)
>
> So I think this may be some kind of mistake. Perhaps this patch may fix it:
>
> diff --git a/runtime/doc/pattern.txt b/runtime/doc/pattern.txt
> index 962882765..8e2d8a45e 100644--- a/runtime/doc/pattern.txt+++ b/runtime/doc/pattern.txt@@ -1472,7 +1472,8 @@ criteria:

> - The number of characters (distance) between two consecutive matching
> characters.
> - Matches at the beginning of a word- - Matches after a camel case character or a path separator or a hyphen.+ - Matches at a camel case character.+ - Matches after a path separator or a hyphen.

> - The number of unmatched characters in a string.
> The matching string with the highest score is returned first.
>
>
>
Yes. The description looks good to me. The matches starting with a camel
case character are given an additional score.

Regards,
Yegappan

naohiro ono

unread,
Jul 16, 2021, 11:17:56 AM7/16/21
to vim/vim, vim-dev ML, Comment

@yegappan Thank you so much for the check. Then we can include this in the next update of runtime files. :)

—
You are receiving this because you commented.

Bram Moolenaar

unread,
Jul 16, 2021, 4:10:44 PM7/16/21
to vim/vim, vim-dev ML, Comment

Closed #8577.

—
You are receiving this because you commented.

Bram Moolenaar

unread,
Jul 16, 2021, 4:10:44 PM7/16/21
to vim/vim, vim-dev ML, Comment

It is more precise, I'll include that change.

—
You are receiving this because you commented.

Reply all
Reply to author
Forward
0 new messages