Hi,

On Sun, Jul 26, 2015 at 4:53 AM, Drew Neil <[email protected]> wrote:
> I agree with h_east that if you’re planning to run the :substitute command
> across multiple files, it makes sense to use:
>
>     :cfdo %s/pattern/replacement/g
>
> and not:
>
>     :cdo s/pattern/replacement/g
>

Depending on the task, you can use either the ":cdo" or the ":cfdo" commands.

If you want to perform text search/replace across all the files in the quickfix
list, then the ":cfdo" command with ":%s/.../g" is the correct option (as it is
more optimal).

This is similar to using the "g" argument to the ":substitute" command to
replace all the matching text in a single line. And using the "%" range
to replace text in all the lines. If you want to replace multiple matching
text across all the lines in a file, then you have to pass both "%" and "g".
If you don't, then the text will not be correctly replaced. This is not a
problem with the ":substitute" command.

>
> I can also see myself wanting to use the :cdo command in combination with
> :normal for certain types of task. But I’ve come across another problem.
> Suppose that we have a text file containing these four lines of text:
>
>     http://example.com
>     http://example.org
>     http://example.net
>     http://example.com http://example.org http://example.net
>
> Now let’s say that we want to turn each occurrence of ‘http’ to ‘https’. (We
> could use the :substitute command here, but let me use this to demonstrate a
> problem with using :normal). We’ll use :vimgrep to populate the quickfix
> list with 6 matches:
>
>     :vimgrep /http\zs:/g %
>
> Then we’ll insert the ’s’ character in front of the colon with this command:
>
>     :cdo normal is
>
> The resulting text looks like this:
>
>     https://example.com
>     https://example.org
>     https://example.net
>     https://example.com httsp://example.org htstp://example.net
>
> In the last line, we get ‘https’, then ‘httsp’, then ‘htstp’. Not ideal!
>
> The problem here is that the quickfix list records line and column numbers.
> If characters are added or removed near the start of the line, the column
> numbers for later matches on that line will no longer line up with the match
> that created the original quickfix list entry.
>
> I’m not sure if this is a problem with the quickfix list, with :cdo, or with
> :normal.
>

This is a problem with the quickfix list functionality. Currently when a line
is added or removed, then the line numbers in the quickfix list entries
are updated. But when a line is modified, the column numbers in the
quickfix list entries are not updated. Refer to the qf_mark_adjust()
function.

In the above example, you should use ":s/../g" instead of the ":normal"
command.

Regards,
Yegappan

>
> On Sat, Jul 25, 2015 at 6:22 PM, Yegappan Lakshmanan <[email protected]>
> wrote:
>>
>> Hi,
>>
>> On Sat, Jul 25, 2015 at 9:55 AM, h_east <[email protected]> wrote:
>> > Hi Yegappan and Bram
>> >
>> > 2015-7-25(Sat) 12:27:56 UTC+9 [email protected]:
>>
>> >> Hi Hirohito,
>> >>
>> >> On Fri, Jul 24, 2015 at 3:42 PM, h_east wrote:
>> >> > Hi Yegappan, Bram and List
>> >> >
>> >> >> >> >
>> >> >> >> > Thanks for testing the patch. I will send out an updated patch
>> >> >> >> > in a few days.
>> >> >> >> > Hopefully this time it will get included. This has been
>> >> >> >> > outstanding for more
>> >> >> >> > than two years.
>> >> >> >> >
>> >> >> >>
>> >> >> >> The updated patch (against vim 7.4.796) is attached.
>> >> >> >
>> >> >> > Thanks.  So now it's ready to include, right?
>> >> >> >
>> >> >>
>> >> >> Yes. Of course :-)
>> >> >
>> >> > I confirmed this patch.
>> >> >
>> >> > I found unexpected behaviors.
>> >> >
>> >>
>> >> Thanks for testing the patch and sending the bug report. I am attaching
>> >> an updated patch that fixes the two problems. Let me know if you see
>> >> any
>> >> issues with this attached patch.
>> >
>> > I confirmed that reported problem have been fixed.
>> > Thank you for quickly fixes.
>> >
>> > I think it is better to discuss.
>> >> > This is my opnion.
>> >> > When the search pattern exists more in a row, I think :cdo/:ldo
>> >> > confuse to use.
>> >> > and the processing time tends to be long.
>> >
>> > Do you understand that the results of the following two commands are
>> > different,
>> > When the search pattern exists more in a row?
>> >
>>
>> The ":cdo" command executes the supplied command for every valid entry
>> in the quickfix list. It is upto the supplied command to perform the
>> appropriate
>> action for every entry.
>>
>> >
>> > (1) :cdo s/\<cmdidx\>/ex_&/g | update
>> >
>>
>> In this case, the supplied substitute command replaces all the occurrences
>> of
>> cmdidx in the current line.
>>
>> >
>> > (2) :exec "cdo norm!iex_\<Esc>:w\<CR>"
>> >
>>
>> In this case, the supplied replaces only the first occurrence of cmdidx.
>> This is
>> not a problem with the ":cdo" command. This is a problem with the user
>> supplied
>> command.
>>
>> >
>> > The (1) is processed all search pattern.
>> > But, The (2) is processed first search pattern in a row.
>> >
>>
>> This is the expected behavior as this is a problem with the user
>> supplied command.
>>
>> >
>> > ':cdo' is not necessary, When use only :substitute.
>> >
>> > When we use the ':cfdo' command such as ':cdo', Speed is also faster.
>> >
>> >   :cfdo %s/\<cmdidx\>/ex_&/g | update
>> >
>> > So I propose to including patch only ':cfdo' and ':lfdo'.
>> >
>> > How do you think?
>> >
>>
>> No. In some cases the ":cdo/:ldo" commands are useful and in some
>> other cases ":cfdo/:lfdo" commands are useful.
>>
>> You are assuming that the ":cdo/:cfdo" commands will only be used
>> to perform substitutions and the results in the quickfix/location lists
>> are
>> from a search command (e.g. vimgrep). This is not always the case.
>> You can populate the quickfix list with output from various tools
>> (e.g. cscope, tags, lid, global, build output, static analysis output,
>> etc.).
>>
>> Regards,
>> Yegappan
>>
>> >> >
>> >> > Case#1
>> >> > How to reproduce:
>> >> > 1. cd to vim src dir.
>> >> >     $ cd (Vim clone dir)/vim/src
>> >> > 2. Start Vim. (including this patch version Vim)
>> >> >     $ vim -N -u NONE
>> >> > 3. Grep word "cmdidx" from source and header using vimgrep.
>> >> >     :vimgrep "\<cmdidx\>" **/*.[ch]
>> >> > 4. Open quickfix window.
>> >> >     :copen
>> >> > 5. Do :cdo command. (Intentionally forget the '| update')
>> >> >     :cdo s/\<cmdidx\>/ex_&/g
>> >> >
>> >> > Expect behavior:
>> >> > - E37 occurs once.
>> >> >
>> >> > Actual behavior:
>> >> > - E37 occurs continuously.
>> >> >
>> >> >
>> >> > --------
>> >> > Case#2
>> >> > How to reproduce:
>> >> > 1~4. (Same abobe.)
>> >> > 5. Do :cdo command. (Intentionally forget the ":w\<CR>")
>> >> >     :exec "cdo norm!iex_\<Esc>"
>> >> >
>> >> > Expect behavior:
>> >> > - E37 occurs once.
>> >> >
>> >> > Actual behavior:
>> >> > - E37 occurs continuously.
>> >> >
>> >> >   And, When press Ctrl-C after the '-- More --' display, buffer.c was
>> >> > modified unexpectedly.
>> >> >
>> >> >     [original buffer.c:4901]
>> >> >     if (eap->cmdidx == CMD_unhide || eap->cmdidx == CMD_sunhide)
>> >> >
>> >> >     [modified buffer.c:4901]
>> >> >     if
>> >> > (eap->exexexexexexexexexexexexexexexexexexexexexexexex___________________
>> >> > _____cmdidx == CMD_unhide || eap->cmdidx == CMD_sunhide)
>> >> >
>> >> >
>> >> > --------
>> >> > This is my opnion.
>> >> > When the search pattern exists more in a row, I think :cdo/:ldo
>> >> > confuse to use.
>> >> > and the processing time tends to be long.
>>

-- 
-- 
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php

--- 
You received this message because you are subscribed to the Google Groups 
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Raspunde prin e-mail lui