Allow multiple characters surround

[tonicebrian]

Hi,

I'd like to use your plugin for editing Trac wiki pages. In this language sections have more than one = level, for instance:

= Section1 =

== Section2 ==

=== Section3 ===

would it be possible to specify a number before the character to wrap? For instance if I want to convert a text into Section3 header it will be something like cs<text_selection>3=

Thank you

[tangledhelix]

This would also be useful for Mediawiki: ''italic'', '''bold'''

And of course Markdown: bold

[tpope]

I guess it'd be possible, though I'm not too keen on implementing it. In the past, I've just created one-offs for those, e.g.:

let g:surround_{char2nr('o')} = "**\r**"

[fncll]

I'm resurrecting this because I have a question about your philosophy on this and a specific question regarding multiple characters. I am wondering why
ysw<em>
works but
ysw**
does not (to surround the word with two * marks)

[tpope]

All surroundings are invoked as a single character. < is a special surrounding that prompts for the tag.

[isovector]

I think the majority of use cases for this would be covered by allowing a prefix for the surround plugin; 2ysw* really should run the surround twice, which would give the desired behavior.

[tommcdo]

That seems like a reasonable solution. Today, the count for the operator is unused.

You can still supply a count for the motion (e.g. ys2w*); I'm pretty sure that comes for free when implementing custom operators.

[haakenlid]

How about using < for this as well? Since XML/HTML tags must start with a letter, and the multi-character tags in markdown and friends usually are non-letter characters, this could make sense.

Currently you can use cs*<**> to turn *bold* into <**>bold</**>

That isn't very useful, since <**> is not a valid xml tag name. Instead this could result in **bold**, which is what you want if you write ReStructured Text.

[isovector]

That strikes me as being a hack; then you wouldn't be able to do multiple
surrounds of alphanumeric characters.

[haakenlid]

What do you mean "multiple surrounds of alphanumeric characters"?

[isovector]

This proposal fails to allow surrounding with the characters "tt", since
that is valid html and would result in "..." rather than
"tt...tt". Furthermore it makes surrounding with "***********" a lot of
work.

On Thu, Nov 26, 2015 at 6:07 PM, Håken Lid notifications@github.com wrote:

What do you mean "multiple surrounds of alphanumeric characters"?

—
Reply to this email directly or view it on GitHub
#15 (comment).

[derBingle]

I know it's a bit of a hack, but for anyone else trying to figure this out, you can use gv to solve this problem. For instance, I'm using a custom leader mapping to bold sentences, and it looks like this:

nmap <silent> <leader>sb ysis*gvS*

Gets the job done.

[RubenVerborgh]

Since the solutions in this thread involved a lot of keypresses, I started the Vim Markup Assistant plugin, which toggles single words in Markdown to italic or bold. Thereby, I lose the flexibility of vim-surround, but I gain simplicity. Examples here.

[feistiny]

with vim-repeat, an alternative is:
e.g. now has text 123, and the cursor is at 1
type ysf3" result in "123", now the cursor is at " not 1
again, type ysf3/ result in /"123/", now the cursor is at /
once more, type ysf3? result in ?/"123?/", now the cursor is at ?
so, if " / ? is the same char like a ", the result will be """123""", and just type once ysf3", also the later two can be repeat with ., unfortunately vim-repeat is neither not support for count operation, otherwise repeat may be more easy

[jacktose]

Just adding a possible use-case: Jinja2 has this kind of stuff:

• {{ ... }} (doable with counts or repeats)
• {% ... %}
• {# ... #}

and it's common to put a space between those and the contents. So I'd love to do something like this:

1. test comment please ignore
2. a vim-surround miracle occurs
3. {# test comment please ignore #}