VI editor: Search and Replace text

IMPORTANT! Before start changing your text file, please remember the commands to UNDO and REDO:

:u    or  :undo[n]
:red  or  :redo[n]

where "n" is the number of changes. The default value is "n=1".


With the command (in ESC mode):

:%s/foo/bar/gc

the active file will be searched for the pattern "foo" and replaced with "bar". For every replacement you'll be asked for a confirmation.

With

:%s/foo/bar/g

you won't be asked for a confirmation.

:%s/foo/bar/gi is case-insensitive
:%s/foo/bar/gI is case-sensitive

With ESC or "q" you terminate the searching.

Note that above we used the % symbol. This is a shortcut that means "from the first line to the end of file".

The "substitute" (:s) command accepts, before of itself, a range of lines, e.g.:

:28s/text1/text2/        only on line 28
:.s/text1/text2/         only on the current line
:.,20s/text1/text2/      from current to line 20
:1,$s/text1/text2/       from line 1 to the end of file
:%s/text1/text2/         over the whole file. Like 1,$

Morevore please notice that if you want to use special characters like [ you have to escape them, like:


:.,$s/lego/\[lego/gc


this command substitutes all the occurrences of "lego" with "[lego", from the current line to the end of file, asking for confirmation before each substitution.

IMPORTANT. Please notice that you have to escape symbols ONLY within the substituting text, not in the text you look for. E.g.:

:.,$s/(ciao)/\[\(ciao)\)\]/gc

substitutes "(ciao)" with "[(ciao)]".

History Brush on Photoshop CS3: Make a Black&White Image with coloured areas

Open an image in Photoshop CS3.



Edit the image as necessary. Open the Image->Adjustements->Black&White in order to get a nice b&w image.



Click on "History" and click on the little square on the left of a status of the image just before the black&white conversion. This status will be taken as "Source" for the "History Brush".



Now come back to the image clicking on the last status. Select the "History Brush" and "paint" an area of your image.


You'll see the area coming back to its original colour!

Rename files

Thanks to Stephan [my room-mate :-) ] I discovered the command "rename"!

N.B. The following works on Debian:

Let's say I want to rename a file callled "test11.txt" to "test12.txt"; the command which does the job is:

rename 's/11/12/' test11.txt

where 's/11/12/' is a perl-style Regular Expression (RE), where 's' stands for "string".

If you are not sure about your RE you can test it with the "-n" option, as:

rename -n 's/11/12/' test11.txt

it just shows to you the result of the operation without actually doing it.

If you have many files in a folder and you want to rename all of them you can use the "*" wildcard:

rename 's/11/12/' *

the command above renames all files containing a string "11" in the name, changing it to "12".

N.B. And the following works on SLC4 (and I guess on Fedora Core then)

the command:

rename .htm .html *.html

changes all the files ending with .htm into .html

P.S.
And if you want, for example, to rename or to add a suffix to a list of files in a directory you can use these commands:

for FILE in * ; do mv $FILE $FILE.txt ; done

where we add the suffix .txt to all files in the directory.
(Note for physicists: Useful to rename AOD or DPD data files to .root files ;-) )

VI editor: comment a whole line or N lines

If you want to comment N lines of a Bash script (for example) with the VI (or VIM) editor you have to type in ESC- mode (type ESC before typing the command):


:.,25s/^/# /

where ".,25s" stands for "from the current line to line number 25" and the sequence added to comment a line is "# " (hash plus a white space).
For a single line:

:11s/^/# /

for the 11th line or

:.s/^/# /

for the current line pointed by the cursor.



To uncomment:

:1,$s/^# //

where "$" stands for "last line", i.e. "from line 1 to the end of file".

Print the Copiright Sign in Python

To print the Copiright Sign with Python we need to use the Unicode string literal format.

A quoted string with a "u" or a "U" just before the leading quote is an Unicode string.
The escape sequence "\N{name}" prints the character specified by "name", where "name" is a standard Unicode name, a s listed in www.unicode.org/charts/

Example:

print u'\N{Copyright Sign}'

prints the Copyright sign.