I am finding a common issue I have with vscode multi-cursor editing mode is how to deal with "fields" of different lengths. To give an example:-
I have an sql file which contains multiple lines of the form
INSERT INTO settings (name,value) VALUES('key_for_value', 'value_to_set'); -- a comment about this setting
Which I want to turn into
UPDATE settings SET name = 'value_to_set' WHERE name = 'key_for_value'; -- a comment about this setting
Obviously each 'key_for_value' and 'value_to_set' are different lengths in each line.
Its fairly easy to grab a bit of the first line of INSERT INTO... and with a control-D get multiple cursors from there, and edit the lines to to get to UPDATE settings SET name = but then I am stuck - I can't find a key binding to (for instance) jump to the next comma - then start a selection and jump the next closing round bracket. I could then do a multiline cut, move the cursor back to start of line, more forward to after the = sign and do a multiline paste.
Is this sort of thing possible?
you can do it with a regex search replace
Search for
INSERT INTO (\w+) \((\w+),(\w+)\) VALUES\('([^']+)', '([^']+)'\);(.*)
replace with
UPDATE $1 SET $2 = '$5' WHERE $2 = '$4';$6
Related
In a snake_cased language, I would want to navigate variablewise and not word_wise and also exclude sigils like #, % or / from these stops.
Example:
|$here_she_goes_again; #the pipe marks my cursor position
With one Ctrl+Right, I want to land on the space before the semicolon,
$here_she_goes_again|; #the pipe marks my cursor position
then, with a Ctrl+Left, I want to return to the beginning of the line.
|$here_she_goes_again; #the pipe marks my cursor position
Somebody got this to work?
Put this into your settings.json:
"[javascript]": {
"editor.wordSeparators": "`~!##%^&*()-=+[{]}\|;:'",.<>/?"
}
Use whatever your language identifier is. I deleted the $ from the default separators to get your example to work for javascript. You can remove the other characters you indicated. The underscore was already not in the default for me. Just make sure those characters are not in the language-specific setting shown above.
You can use the extension Select By and the command moveby.regex
You are able to define a regex to search and bind this to Ctrl+Left and another to Ctrl+Right
In the key binding you can limit this to a particular languageID.
What's an easy way to convert a multi-line comment (e.g. JSDoc with each line separated by line breaks) into a paragraph without any line breaks that I can copy into an email or another document?
I know I can use search & replace with regular expressions, but is there a more ergonomic way to do it?
You probably knew that you can use multiple cursors to change multiple lines at once, but did you know you can also use them to remove line breaks? Assume you start with this comment:
/**
* Returns a new `Temporal.LocalDateTime` instance representing the first
* valid time during the current calendar day and time zone of `this`.
*
* The local time of the result is almost always `00:00`, but in rare cases it
* could be a later time e.g. if DST starts at midnight in a time zone. For
* example:
* ```
* const ldt = Temporal.LocalDateTime.from('2015-10-18T12:00-02:00[America/Sao_Paulo]');
* ldt.startOfDay; // => 2015-10-18T01:00-02:00[America/Sao_Paulo]
* ```
*/
First part: use multiple cursors to remove the prefix characters on each line.
Click on the upper-left corner of the comment (the /**).
Now hold down Cmd+Shift (Alt+Shift on PC) and click after the */ on the last line of the comment section.
This will create a columnar, multi-line selection that includes the non-text prefix characters on each line. If the selection doesn't include all the prefix characters, you can hold down the Shift key and use the left or right arrow keys to adjust the width of the selection.
Press the Delete key to remove prefix characters on all lines.
Second part: it's time to delete the line breaks and replace them with spaces. I discovered today that you can use multiple cursors for this part too!
After you've deleted the prefix text above, but before you've pressed any other keys, press the backspace key. It will delete the line breaks but leave each cursor in the same place!
Type the spacebar once to insert one space to replace each line break.
Press ESC to clear multiple selections, and delete the extra space at the start of the line. You may have an extra space(s) at the end of the line too that may need trimming.
Copy the resulting one-line text.
Use Cmd+Z (Ctrl+Z on Windows) to undo the last few changes so your code comment will be back to normal.
Now you can paste the copied text into an email!
The same solution works to replace line breaks with spaces in any multi-line text, not only code comments.
I'm sure that many of you already knew how to do this trick, but I found it so easy and so cool that I thought it was worth sharing as a Q&A here so others can learn about this trick too.
Here's what the steps look like in the VSCode IDE:
Before deleting, you should see something like this:
After deleting prefix characters:
After deleting line breaks (note the multiple cursors are still there):
After inserting spaces in place of the deleted line breaks:
I usually select the first line break, then hit/hold command+D repeatedly to add cursors at all line endings I want to edit. Then, just hit space once.
Say you have the following text:
abc
123
abc
456
789
abc
abc
I want to remove all "abc" lines and just keep one. I don't mind sorting. The result should be like this:
abc
123
456
789
If the order of lines is not important
Sort lines alphabetically, if they aren't already, and perform these steps:
(based on this related question: How do I find and remove duplicate lines from a file using Regular Expressions?)
Control+F
Toggle "Replace mode"
Toggle "Use Regular Expression" (the icon with the .* symbol)
In the search field, type ^(.*)(\n\1)+$
In the "replace with" field, type $1
Click ("Replace All").
If the order of lines is important so you can't sort
In this case, either resort to a solution outside VS Code (see here), or - if your document is not very large and you don't mind spamming the Replace All button - follow the previous steps, but in steps 4 and 5, enter these:
(based on Remove specific duplicate lines without sorting)
Caution: Blocks for files with too many lines (1000+); may cause VS Code to crash; may introduce blank lines in some cases.
search: ((^[^\S$]*?(?=\S)(?:.*)+$)[\S\s]*?)^\2$(?:\n)?
replace with: $1
and then click the "Replace All" button as many times as there are duplicate occurrences.
You'll know it's enough when the line count stops decreasing when you click the button. Navigate to the last line of the document to keep an eye on that.
Coming in vscode v1.62 is a command to eliminate duplicate lines from a selection:
Delete Duplicate Lines in the Command Palette
or
editor.action.removeDuplicateLines as a command in a keybinding
(there is no default keybinding for this command)
Here is a very interesting extension: Transformer
Features:
Unique Lines As New Document
Unique Lines
Align CSV
Align To Cursor
Compact CSV
Copy To New Document
Count Duplicate Lines As New Document
Encode / Decode
Filter Lines As New Document
Filter Lines
Join Lines
JSON String As Text
Lines As JSON String Array
Normalize Diacritical Marks
Randomize Lines
Randomize Selections
Reverse Lines
Reverse Selections
Rotate Backward Selections
Rotate Forward Selections
Select Highlights
Select Lines
Selection As JSON String
Sort Lines By Length
Sort Lines
Sort Selections
Split Lines After
Split Lines Before
Split Lines
Trim Lines
Trim Selections
Unique Lines
Removes duplicate lines from the document Operates on selection or
current block if no selection
Unique Lines As New Document
Unique lines are opened in a new document Operates on selection or
current block if no selection
I haven't played with it much besides the "Unique Lines" command but it seems quite nicely done (including attempting a macro recorder!).
To add to #Marc.2377 's reply.
If the order is important and you don't care that you just keep the last of the duplicate lines, simply search for the following regexp if you want to only remove duplicte non-empty lines
^(.+)\n(?=(?:.*\n)*?\1$)
If you also want to remove duplicate empty lines, use * instead of +
^(.*)\n(?=(?:.*\n)*?\1$)
and replace with nothing.
This will take a line and try to find ahead some more (maybe 0) lines followed by the exact same line taken. It will remove the taken line.
This is just a one-shot regex. No need to spam the replace button.
This now also takes the comment of #awk into account, in where the last line has to have a linefeed in order to be identified as a duplicate. This is no longer the case now by excluding the \n from the line to search and adding a $ to the line found.
I just had the same issue and found the Visual Studio Code package "Sort lines". See the Visual Studio Code market place for details (e.g. Sort lines).
This package has the option "Sorting lines (unique)", which did it for me. Take care of any white spaces at the beginning/end of lines. They influence whether lines are considered unique or not.
Install the DupChecker extension, hit F1, and type "Check Duplicates".
It will check for duplicates and ask if you want to remove them.
Try find and replace with a regular expression.
Find:
^(.+)((?:\r?\n.*)*)(?:\r?\n\1)$
Replace:
$1$2
It is possible to introduce some variance in the first group.
If you don't mind some Vim in your VS Code. You can install Vim emulation plugin.
Then you can use vim commands
:sort u
It will sort lines and it will remove duplicates
Sublime Text 3
It has blisteringly fast native permutation functions.
Edit > Permute Lines > Unique or ⇧⌘U, and
Edit > Permute Selections > Unique
Visual Studio Code is my daily driver. But, I keep Sublime Text on standby for these situations.
Not actually in Visual Studio Code, but if it works, it works.
Open a new Excel spreadsheet
Paste the data into a column
Go to the Data tab
Select the column of data (if you haven't already)
Click Remove Duplicates (somewhat in the middle of the bar)
Click OK to remove duplicates.
It is not the best answer, as you specified Visual Studio Code, but as I said: If it works, it works :)
I have a Word document with fields of the reference variety, which occur in the form "[field].[field]"--in other words, there's a period between the two fields. I want to globally replace this with a space.
Word offers the ^d special character to search for fields, but for some reason the query "^d.^d" does not find anything. However, ".^d" does. Now comes the problem, however--what do I specify as the replacement text in order to retain the field code? If using regular expressions, I could use a "Find What Expression" such as \1, but with regexp ("wild card") mode the ^d is not permitted.
I guess I could write a macro...
I would like to add to Bibadia's solution.
An example of an index entry field; we want to change a name we misspelled.
Make sure hidden formatting is displayed (toggle with SHIFT+CTRL+F8).
Make sure wildcards option is not selected. To search for fields, use the opening and closing field braces code (optionally use ^w for spaces, as Bibadia suggested):^19 XE "Deo, John" ^21
Replace won't recognize field braces character, but will allow to insert the clipboard's content. ;). To do that, insert in text the correct entry. CTRL+F9 to insert field and type:XE "Doe, John"
Select the field above and copy
Use ^c in the replace box
Hit Replace All
Ta-da!
It's usually better to go the macro route when finding fields because, as you say, the find algorithm that Word uses doesn't work the way you might hope with fields.
But if you know exactly what the fields contain, you can specify a search pattern that will probably work (however not in wildcard mode).
For example, if you want to look for figure number field pairs such as
{ STYLEREF 1 \s }.{ SEQ Figure \* ARABIC \s 1 }
(which would typically be the same set of fields everywhere in the document)
If you only really need to look for the following:
{ STYLEREF 1 \s }.<any field>
you could ensure that field codes are displayed and search for
^d STYLEREF 1 \s ^21.^d
or
^19 STYLEREF 1 \s ^21.^19
If you need to be more precise, you can spell out the second field as well.
"^d" only works for finding the field beginning, not the field end.
It's a shame that ^w wants to find at least 1 whitespace character because otherwise it would be more robust to look for
^19^wSTYLEREF^w1^w\s^w^21.^19
Perhaps someone else knows how to work around that without using wildcards?
Torzaburo,
I suggest that you do this using a macro. You can start by recording the macro, and later refining your processing steps within the macro.
First turn on the hidden characters by navigating to Home > Paragraph > toggle the show/hide Paragraph symbol. Also, select all and toggle the field codes on (right-click and select "Toggle Field Codes".
Open a new blank Word doc in addition to the one you have open. You will use this later. Start the macro recording and find the field using the "^d" (field code) as you said.
When the field is found, copy only the field text within the brackets, and not the full field reference. While the macro is still recording, ALT + TAB to the new blank document and paste the field code in as plain text.
At this point, do the necessary find & replace processing to the field codes. Highlight the processed field codes, copy, ALT + TAB back to the original document, and paste back between the { } brackets.
Stop the macro recording. Add any further custom processing to the macro VBA.
Select-All and re-toggle the field codes. Update the field codes.
You don't need a macro. Just toggle all field codes on by using Alt+F9. Then do a find and replace for what you want to change. Once the replacement is complete, use Alt+F9 again to toggle the field codes back off.
Disclaimer: I didn't originate this solution, but it's clean and elegant and I thought it should be included here:
(Adapted from Search & Replace Field Codes in Word):
Create or find a single instance of the field you want to convert text to
Toggle Field Codes visible (AltF9)
Copy the code for the field you want to use to the Clipboard (highlight and CtrlC)
Open the Replace dialog box (CtrlH), insert the text you want to replace in the Find What box and then enter ^c in the Replace With box.
This will replace your text with the contents of the Clipboard, turning it into the field code you copied in step 3. It also copies formatting information (font, color, etc.), to control how the field will appear when hidden. (Caveat: I've tested this with Word 2003 under Windows 7 only.)
Coming in late on this, probably way too late for Beth (sorry Beth). And this may not be quite what Beth was looking for. But for anyone interested ...
It sounds like Beth may have created captions throughout the document using INSERT CAPTION (hence the presence of field codes). This means these captions will have been (automatically) created in CAPTION style.
To globally replace the separator "." with " " (space) in such captions, take two steps:
[1] Go to REFERENCES | INSERT CAPTION, then click on NUMBERING and replace the SEPARATOR "." with "EM-DASH". This will replace all separators in captions for the selected label in the CAPTION Window. If you have other labels in use in the document (e.g. FIGURE), select the other labels one by one and repeat this process.
[2] Do a find/replace searching for special character "em-dash" (^+) in style CAPTION, replacing with " ". Click REPLACE ALL.
Voila!
NOTE: This presumes that em-dash does not appear in the caption text anywhere. If it does, then you'll need to do a pre- and post- "fiddle" to ensure these em-dashes are not touched by the global replace above.
The "pre-fiddle" is to do a global find/replace across captions, replacing the em-dash ("^+") with some other string (e.g. "EM-DASH") that doesn't ever occur in any caption's text. Then you do the separator change as described above. Finally, the "post-fiddle" is to restore the em-dashes that were in the captions, by doing a global replace of the string "EM-DASH" with the actual em-dash character "^+".
I'm new to n++, but I have been most impressed with this tool so far. I've been trying to record a macro that do a search/replace, but the 'search' part seems to have the initial search text from the recording 'hard-coded' in the macro.
What I want is:
Manually locate the cursor at the beginning of the first line of a fixed format code segment, then Macro actions:
move cursor two lines down
move cursor right x characters
mark charters from pos x to x+n
search and replace all occurrences of the selected text with "{p_'selected text'}"
In an more advanced version, I'd like to add some logic to step 4: only execute the replace part if the # of occurrences are > 1 (e.g. by first adding a count statement, but I'm not sure how to obtain the returned count # from the dialog box)
Is this possible?
While I'm a big fan of Notepad++, this sounds like something I would accomplish with AutoHotKey. You would select the text and copy it to the clipboard. AutoHotKey would read the clipboard, replace the text as you desire, and either replace the clipboard contents, or send it back to your document. Let me know if you would like to go that route.