Find every third value and insert cr or newline in VIM_问答_开发者

So I have several large datasets that I need to make more readable and currently I'm having to go in and move to each 3rd value and insert a newline. I've tried several things in VIM to get this to work, but none seem to return the value that I'm looking for. Here's some of my data:

(0.96260310749184663, 4.3830008206495812, 0.84922658632317849),
(0.96260310749184663, 5.0000002088986637, 1.049701855818201),
(0.96260310749184697, 5.6169993576359696, 0.8492264385213405),
(0.96260310749184719, 5.9983257940402384, 0.32437568665165911),
(0.96260310749184719, 5.9983258053918069, -0.32437572732698844),
(0.96260310749184719, 5.6169994358349786, -0.84922691097323821),
(0.96260310749184697, 5.0000000093711492, -1.0497019267383632)

What I need to do is make it look like this instead:

(0.96260310749184663, 4.3830008206495812, 
 0.84922658632317849),
(0.96260310749184663, 5.0000002088986637, 
 1.049701855818201),
(0.96260310749184697, 5.6169993576359696, 
 0.8492264385213405),
(0.96260310749184719, 5.9983257940402384, 
 0.32437568665165911),
(0.96260310749184719, 5.9983258053918069, 
 -0.32437572732698844),
(0.96260310749184719, 5.6169994358349786, 
 -0.84922691097323821),
(0.96260310749184697, 5.0000000093711492, 
 -1.0497019267383632)

I used this on a selection to try and get it to move that 3rd value down, but it duplicated the entire thing and then moved the entire line down:

:'<,'>s/\([-[0-9.]\+,\s[-[0-9.]\+,\)/\1\r&/g

I also tried removing the 1 to make it work, but that didn't work either. So is there any way I can capture that 3rd value and insert a carriage return or newline to make this work? Thanks.

Edit---

I apologize for mis-communicating part of my problem: The data at the top is all on one line, not several lines. It looks like this:

(0.31852533878680489, 0.10352149350126813, -0.0046069731261429991), (0.31852526554320226, -0.103521543762028, -0.0046069731261429991), (0.19682845687859715,开发者_运维百科 -0.27102285045297636, -0.004606973126142444), (-8.1184530326649769e-05, -0.33500267513407755, -0.0046069731261416669), (-0.19699089317458821, -0.27102292369657782, -0.0046069731261408897), (-0.31868770183919259, -0.10352150714022559, -0.0046069731261403346), (-0.31868770183919221, 0.10352156674487166, -0.0046069731261403346),

:g/./norm! 2Wi^M

Explanation:

:g/./{cmd} will run the {cmd} on every line
norm! will execute the following string as normal mode commands
2Wi^M Move 2 WORDS then insert a return
^M is accomplished by pressing <c-v><cr> or <c-q><cr>.

It is very tempting to do %norm! 2Wi^M, but this will fail as it messes up the lines that are being worked on.

Two approaches

1. ex commands and formatting

My first thought was textwidth:

:se tw=50
:g/./norm! gqq
:%s/^[^(]/ &/g

This

sets text width to 50 chars
reformats each (nonempty) line
inserts a space at the start of any lines which don't start with (

2. use a macro

Alternatively: make a macro

gg
qq2f,a<CR><Esc>j0q
100000@q

Rationale:

go to begin of buffer
record macro q (qq starts recording, q ends recording)
- 2f, - forward to second comma
- a<CR><Esc> - append newline
- j0 - next line, move caret to first character
rinse and repeat (100000 is the number of times to repeat; processing will stop when the macro fails (e.g. at end of file)

Since the entire text to break is in one line initially, one can use a single short substitution command,

:s/,.\{-},\|), /&\r/g

The way I would do this would be something like this:

'<,'>s/^[^,]\+,[^,]\+,\zs/\r

Translated:

'<,'>     " Over the visual range
s/X/Y     " Substitute X with Y
^         " Start of line
[^,]\+    " Anything that isn't a comma, one or more, as many as possible
,         " A comma (end of first field)
[^,]\+    " Anything that isn't a comma, one or more again
,         " A comma (send of second field)
\zs       " Mark this point as the start of the match so we don't have to bother including all of the above in the result
\r        " We're replacing nothing at the end of the above match with a new-line (\r)

Another alternative:

'<,'>s/^\(.\{-},\)\{2}/&\r

Translated:

'<,'>s/X/Y     " As before
^              " Start of line
\(...\)        " A group containing:
.\{-},         " Everything up to and including the first comma
\{2}           " Match the preceding group twice (so up to and including the second comma)
&\r            " Replace with what was already there followed by a new-line

And another... given that you have equal data widths, navigate to the space before the third column of the first row. Hit Ctrl-V and go down so that you've selected the whole column of spaces (6j in your example). Press either s or Shift-i (depending on whether you want to keep the space) and then ENTER and then ESC.

Use the macros to format a line and go to the begining of the next line. Then apply that macro as may times as you need.

So, starting to record a macro is q_ where _ is a register, then edit your line, press q again to save the macro. And finally, apply it using @_ which you can prefix with a number of times you want to repeat it.

Make a recursive macro (as an alternative to @sehe's macro-for-100000-times answer).

ggqa2f,a<CR><Esc>j@aq

Then @a to run the command on the second line (you shouldn't need to move your cursor).

Visualization:

gg

Find every third value and insert cr or newline in VIM