I have a text file with multiple lines such as:
amanda: foo
robert: bla
amanda: bar
peter: da
I'd like to remove all lines with amanda. I use ctrl-s and kill each line individually. Is possible to remove all lines at once?
M-x delete-matching-lines. It's possible to use regular expression.
One way is to use query-replace-regexp with a regular expression of ^.*amanda.*$ to an empty string.
Although #Marcos's answer is idiomatic one (you should use it), let's explore other variants. Let's say you have a text buffer and you want to delete a line containing li in it:
Vladimir
Ilich
Ulyanov
Remember, that ^ matches beginning of line and $ end of line in regex. $ doesn't touch the newline character after the line, so replace with regex ^.*li.*$ will produce an empty line, as per #ataylor's answer:
Vladimir
Ulyanov
For some reason it's impossible to match before ^ and after $ in regex, therefore \s-^.*li.*$ nor ^.*li.*$\s- won't work. Note, \s- matches any whitespace character, (i.e. space, tab, newline and carriage return), so intuitively the regexes should've deleted the newline too, as newline is the only possible whitespace character before ^ or after $. To match exactly newline, you should enter it verbatim, C-q C-j by default. Emacs frequently denotes the newline in separate font color as ^J, it's not a sequence of ^ and J, but a single character, please pay attention.
Therefore to delete a line containing li, you could run command query-replace-regexp on string ^.*li.*^J, where ^J is newline:
Vladimir
Ulyanov
Related
How can I wrap text in Notepad++, but leave paragraph lines intact?
Example:
The original text looks like this:
This
is
paragraph
one.
This
is
paragraph
two.
I would like the text to look like this:
This is paragraph one.
This is paragraph two.
But currently, when I go to Edit > Blank Operations > Remove Unnecessary Blank and EOL, the text ends up with no paragraph line in between, like this:
This is paragraph one. This is paragraph two.
How can I fix this?
Thank you.
You may try the following find and replace, in regex mode:
Find: ([^.\s])\r?\n
Replace: $1[ ]
Demo
The regex pattern ([^.\s])\r?\n will match and capture any letter which appears before a CR?LF character, so long as that character is not a full stop. It then replaces with that same character, followed by just a single space, thereby removing/replacing the original CR?LF.
This will replace every single linebreak with a space.
Ctrl+H
Find what: [^\r\n]+\K\R(?!\R)
Replace with: # a space
CHECK Wrap around
CHECK Regular expression
Replace all
Explanation:
[^\r\n]+ # 1 or more non linebreak
\K # forget them
\R # any kind of linebreak (i.e. \r, \n, \r\n)
(?!\R) # negative lookahead, make sure we haven't another linebreak after
Screenshot (before):
Screenshot (after):
In Emacs, I would like to search certain characters and replace them. They can be separated by their unicode number. For example, below 3 characters has different unicode number.
á(#xe1), ⓐ(#x24d0), 𝒶(#x1d4b6)
What if I want to search characters between the range #x1d000 to #x1dfff, and then I will use regxp replace to add a double quote("") for each of these characters?
First of all, you can enter Unicode characters by their hexadecimal codes using the key binding C-x 8 C-m (the command is called insert-char). So type C-x 8 C-m, enter 1d000, and then hit RET to insert the character with Unicode code point 1d000.
Then we can use this to search and replace.
Type C-M-% to run the command query-replace-regexp
For the search expression, enter [, then C-x 8 C-m 1d000 RET , then -, and then C-x 8 C-m 1dfff RET, and finally ]. That is, search for any character in the range between 1d000 and 1dfff. (This is similar to the "normal" regexp [a-z], which matches all characters between a and z, i.e. all lowercase characters.)
For the replace expression, enter "\&". \& is a special sequence for inserting the text matched by the search expression, so we're going to wrap every matching character in double quotes.
Then, hit y to replace matches one by one, or ! to replace all remaining matches.
Suppose you have this file:
x
a
b
c
x
x
a
b
c
x
x
and you want to find the sequence abc (and select the whole 3 lines) with Notepad++ . How to express the newline in regex, please?
Notepad++ can do that comfortably, you don't even need regexes
In the find dialogue box look in the bottom left and switch your search mode to Extended which allows \n etc.
As odds on you're working on a file in windows format you'll be looking for \r\n (carriage return, newline)
a\r\nb\r\nc
Will find the pattern over three lines
Update 18th June 2012
With the new Notepad++ v6, you can indeed search for newlines with regexes. So you can just use
a\r\nb\r\nc
even with regular expressions to accomplish what you want. Note \r\n is Windows encoding of line-breaks. In Unix files, its just \n.
Unfortunately, you can't do that in Notepad++ when using regex search. Notepad++ is based on the Scintilla editor component, which doesn't handle newlines in regex.
You can use extended search for newline searching, but I don't think that will help you search for 3 lines.
More info here.
Update: Robb and StartClass0830 were right about extended search. It does work, but not when using regular expressions search.
^a\x0D\x0Ab\x0D\x0Ac
This will work \x0D is newline and \x0A is carriage return. Assumption is that each line in your file ends with ascii 10 and 13.
I found a workaround for this.
Simply, in Extended mode replace all \r\n to a string that didn't exist in the rest of the document eg. ,,,newline,,, (watch out for special regexp chars like $, &, and *).
Then switch to Regexp mode, do some replacements (now newline is ,,,newline,,,).
Next, switch to Extended mode again and replace all ,,,newline,,, to \r\n.
For Notepad 6 and beyond, do this as a regular expression:
Select Search Mode > Regular expression (w/o . matches newline)
And in the Find what Textbox : a[\r\n]b[\r\n]+c[\r\n]
or if you are looking at the (Windows or Unix) file to see its line breaks as \r\n or \n then you may find it easier to use Extended Mode:
Select Search Mode > Extended (\n, \r, \t, \0, \x...)
And in the Find what Textbox for Windows: a\r\nb\r\nc\r\n
Or in the Find what Textbox for Unix: a\nb\nc\n
Wasn't clear if the OP intent is to select the trailing line return (after the 'c') as well, as would be necessary to remove the lines.
To not select the trailing line return, as appropriate for replacing with a non-empty string, simply remove the final line return from the matching statement.
Note that if there should be a match on the last line of the string, without a matching trailing line return, the match fails.
a\r\nb\r\nc works for me, but not ^a\x0D\x0Ab\x0D\x0Ac
Hmm, too bad that newline is not working with regular expressions. Now I have to go back to Textpad again. :(
Select Search Mode Which is
Extended (\n, \r, \t, \0, \x...)
\n is new line and such
This is Manuel
Find: "(^a.$)\r\n(b.)\r\n^(c.*)$" - pickup 3 whole lines, only storing data
Replace with: "\1\2\3" - Put down (replay) data
Works fine in Regex with Notepad++ v7.9.5
Place holders: ^ Start and $ End of line can be inside or out of ()store as shown, though clearly not necessary in given example. Note "[^x]" is different - here "^" is "NOT".
Advantage of storing and replay allows much more complicated pattern match without having to type in again what you want to end up with, and even change of replay: "\2\3\1" for "bca"
I have run accross this little issue when the document is windows CR/LF
If you click the box for . to match newlines you need .. to match CR/LF so if you have
<blah><blah>",
"<more><blah>
you need to use ",.." to match some string comma cr/lf another string
In Notepad++ you can also try highlighting the desired part of the text and then pressing CTRL+J.
That would justify the text and thus removing all line endings.
Right now I'm using:
(setq show-trailing-whitespace t)
In my .emacs to show trailing whitespace for my CC mode. I can't seem to figure out how to have it not show the whitespace font for whitespace only lines.
Empty lines separating indenting code are sometimes indented at the code level and sometimes not indented at all, and I don't want to draw my attention to a line I don't care to change.
I'd like to stick to built-in emacs modules, but I'm willing to use whitespace.el but it's not clear how to configure it to do this.
Since you want to use built-in modules, I'd advise using the whitespace.el link you specified - as it is shipped with Emacs 23. This answer works when using that whitespace.
As long as you have 'trailing in your 'whitespace-style variable (which it is by default), the following change to the regular expression for what indicates "trailing" whitespace should give you what you want:
(setq whitespace-trailing-regexp
"\\b\\(\\(\t\\| \\|\xA0\\|\x8A0\\|\x920\\|\xE20\\|\xF20\\)+\\)$")
Note: It is just the default value, with the \b prepended, indicating that the whitespace should follow a word.
With
"\\b.*\\(\\(\t\\| \\|\xA0\\|\x8A0\\|\x920\\|\xE20\\|\xF20\\)+\\)$"
the word does not need to be directly in front of the trailing whitespaces but there can be e.g. punctuation characters between them (i.e. this also highlights trailing whitespaces behind non-word characters).
Edit:
Using
"\\b.*?\\(\\(\t\\| \\|\xA0\\|\x8A0\\|\x920\\|\xE20\\|\xF20\\)+\\)$")
highlights all the trailing whitespaces and thus eliminates the drawback mentioned in comment #1.
If I have the following text in my Eclipse editor:
Text Line 1
Text Line 2
I would like to concatenate the text into:
Text Line 1Text Line 2
My first idea was to search for carriage return character '\n' and replace it with '' to concatenate it.
I tried using the search function of Eclipse, but it does not recognize carriage return character.
Are there any other editor that can do this?
Eclipse does this if you:
turn on regular expression mode in search/replace
enter \R for the newline
Just use Edit -> Find/Replace, switch on the Regular Expressions checkbox, search for \n and replace it by space.
I tried it in Eclipse 3.4 and it worked well.
Short answer:
I decided to use \s++ as separator in multi-line search expressions (with regular expressions enabled) and \Qfoo\E to escape special characters if required.
Long answer:
As soru already answered, for any "Unicode linebreak sequence" a regular expression search with \R can be used.
A pure carriage return is represented by \r. Upper and lower case make a difference. \R represents any unicode linebreak sequence (for example \r\n).
I found this SO question because I wanted to search for a multi-line expression in Eclipse, including line breaks and tabs:
#Override
#Transient
In order to include the white spaces in my regular search expression I used (on Windows platform)
#Override\r\n\t*#Transient
Following expressions also work:
#Override\R\t*#Transient
#Override\s++#Transient
Please note that the second expression also matches #Override #Transient
without a line break, which is fine for me.
Following expressions did not! work for me:
#Override\r\t*#Transient
#Override\n\t*#Transient
Explanation of some regular expressions:
\R represents any unicode linebreak sequence (for example \r\n)
\s represents any white space
\t represents a tab
* matches zero or more occurrences
++ matches one ore more occurrences
\Q and \E escape wrapped content. Use them if your original multi line expression includes special regex characters, for example
\Q/**\E\s++\Q*\E
matches
/**
*
Also see:
Difference between \n and \r?
https://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html
Most find and replace tasks in editors (at least, TextPad) have the ability to replace via a regex. If you can find this option in eclipse, then just use that.
\r is the correct regular expression for carriage return. But Eclipse standard editor does not find it.
So use external editor, for example notepad++