There is a command to replace bbb to ccc, if the line contains abc.
echo "abc yyy bbb xzy" | sed -e "/abc/ s/bbb/ccc/"
Does anyone know what the command would be, if I want to do the replacement, only if the line contains both abc and xyz?
Because it doesn't matter which one is matched first, you can look for abc first, then make the substitution if also xyz matches1:
sed '/abc/{/xyz/s/bbb/ccc/}'
or, considerably less elegant:
sed '/abc.*xyz\|xyz.*abc/s/bbb/ccc/'
but no nesting.
1BSD sed requires a semicolon before the closing brace.
Just use awk and you can code it as you'd write it, with && between the conditions:
awk '/abc/ && /xyz/ { sub(/bbb/,"ccc") } 1'
Try writing:
awk '(/abc/ && /xyz/) || (/def/ && (/ghi/ || /klm/)) { sub(/bbb/,"ccc") } 1'
or any other more interesting compound condition with sed. Awk is available everywhere sed is and the above is fully portable and will work as-is in every awk in every UNIX installation.
Related
I have one command which is used to extract lines between two string patterns 'string1' and 'string2'. This is stored in variable called 'var1'.
var1=$(awk '/string1/{flag=1; next} /string2/{flag=0} flag' text.txt)
This command works well and the output is a set of lines.
Do you hear the people sing?
Singing a song of angry men?
It is the music of a people
Who will not be slaves again
I want the output of the above command to be inserted after a string pattern 'string3' in another file called stat.txt. I used sed as follows
sed '/string3/a'$var1'' stat.txt
I am having trouble getting the new output. Here, the $var1 seems to be working partially i.e. only one line -
string3
Do you hear the people sing?
Any other suggestions to solve this?
I would be tempted to use sed to extract the lines, and awk to insert them into the other text:
lines=$(sed -n '/string1/,/string2/ p' text.txt)
awk -v new="$lines" '{print} /string3/ {print new}' stat.txt
or perhaps both tasks in a single awk call
awk '
NR == FNR && /string1/ {flag = 1}
NR == FNR && /string2/ {flag = 0}
NR == FNR && flag {lines = lines $0 ORS}
NR == FNR {next}
{print}
/string3/ {printf "%s", lines} # it already ends with a newline
' text.txt stat.txt
It's a data format problem...
Appending a multi-line block of text with the sed append command requires that every line in the block to be appended ends with a \ -- except for the last line of that block. So if we take the two lines of code that didn't work in the question, and reformat the text as required by the append command, the original code should work as expected:
var1=$(awk '/string1/{flag=1; next} /string2/{flag=0} flag' text.txt)
var1="$(sed '$!s/$/\\/' <<< "$var1")"
sed '/string3/a'$var1'' stat.txt
Note that the 2nd line above contains a bashism. A more portable version would be:
var1="$(echo "$var1" | sed '$!s/$/\\/')"
Either variant would convert $var1 to:
Do you hear the people sing?\
Singing a song of angry men?\
It is the music of a people\
Who will not be slaves again
To delete/comment 3 lines befor a pattern (including the line with the pattern):
how can i achive it through sed command
Ref:
sed or awk: delete n lines following a pattern
the above ref blog help to achive the this with after a pattern match but i need to know before match
define host{
use xxx;
host_name pattern;
alias yyy;
address zzz;
}
the below sed command will comment the '#' after the pattern match for example
sed -e '/pattern/,+3 s/^/#/' file.cfg
define host{
use xxx;
#host_name pattern;
#alias yyy;
#address zzz;
#}
like this how can i do this for the before pattern?
can any one help me to resolve this
If tac is allowed :
tac|sed -e '/pattern/,+3 s/^/#/'|tac
If tac isn't allowed :
sed -e '1!G;h;$!d'|sed -e '/pattern/,+3 s/^/#/'|sed -e '1!G;h;$!d'
(source : http://sed.sourceforge.net/sed1line.txt)
Reverse the file, comment the 3 lines after, then re-reverse the file.
tac file | sed '/pattern/ {s/^/#/; N; N; s/\n/&#/g;}' | tac
#define host{
#use xxx;
#host_name pattern;
alias yyy;
address zzz;
}
Although I think awk is a little easier to read:
tac file | awk '/pattern/ {c=3} c-- > 0 {$0 = "#" $0} 1' | tac
This might work for you (GNU sed):
sed ':a;N;s/\n/&/3;Ta;/pattern[^\n]*$/s/^/#/mg;P;D' file
Gather up 4 lines in the pattern space and if the last line contains pattern insert # at the beginning of each line in the pattern space.
To delete those 4 lines, use:
sed ':a;N;s/\n/&/3;Ta;/pattern[^\n]*$/d;P;D' file
To delete the 3 lines before pattern but not the line containing pattern use:
sed ':a;N;s/\n/&/3;Ta;/pattern[^\n]*$/s/.*\n//;P;D'
What is this sed command doing? and is there any online utility that kind of explains sed a little bit, like regex?
sed -i '1s/$/|,a Type,b Type,c Type/;/./!b;1!s/$/|,,,/' textflile.txt
I think in the beginning it is adding csv a type, b type, c type at the end of the line but what does the rest of the command too
I don't know of any such utility, but let me explain using a text editor:
sed -i '1s/$/|,a Type,b Type,c Type/;/./!b;1!s/$/|,,,/' textflile.txt
^ ^ ^ ^ ^^ ^^ ^
| | | | || || |
modify | End Non-empty || || input
the | of lines || |Negation, file
file | line only || |i.e. lines 2,3,...
in | || |
place | || First
First line Negation, i.e.| line
empty lines only|
Branch to
script end,
i.e. skip the rest
In other words, it adds |,a type, b Type,c Type to the first line, doesn't change empty lines, and adds |,,, to all the remaining lines.
sed -i '1s/$/|,a Type,b Type,c Type/;/./!b;1!s/$/|,,,/' textflile.txt
can be written as
sed -i '
1 s/$/|,a Type,b Type,c Type/
/./! b
1! s/$/|,,,/
' textflile.txt
on line 1 only, add some text to the end of the line
if the line is empty ("matches 1 character, not"), goto next "cycle" (i.e., print current line and go to next line)
on every line except line 1, add "|,,," to the end of the line
So, it looks like you're adding some blank fields to a CSV file.
info sed contains the complete sed manual.
This doesn't answer your question but it's important for people to know and requires more space and formatting than a comment so: FYI to do what #choroba says that sed script does, i.e.
it adds |,a type, b Type,c Type to the first line,
doesn't change empty lines,
and adds |,,, to all the remaining lines.
is just this in awk:
awk '
NR==1 { print $0 "|,a type, b Type,c Type"; next }
!NF { print }
NF { print $0 "|,,," }
'
or if you're familiar with ternary expressions and want to remove the redundant code:
awk '{
sfx = "|," (NR==1 ? "a type, b Type,c Type" : ",,")
print $0 (NF ? sfx : "")
}'
I want to execute some sed command for any line that matches either the and or or of multiple commands: e.g., sed '50,70/abc/d' would delete all lines in range 50,70 that match /abc/, or a way to do sed -e '10,20s/complicated/regex/' -e '30,40s/complicated/regex/ without having to retype s/compicated/regex/
Logical-and
The and part can be done with braces:
sed '50,70{/abc/d;}'
Further, braces can be nested for multiple and conditions.
(The above was tested under GNU sed. BSD sed may differ in small but frustrating details.)
Logical-or
The or part can be handled with branching:
sed -e '10,20{b cr;}' -e '30,40{b cr;}' -e b -e :cr -e 's/complicated/regex/' file
10,20{b cr;}
For all lines from 10 through 20, we branch to label cr
30,40{b cr;}
For all lines from 30 through 40, we branch to label cr
b
For all other lines, we skip the rest of the commands.
:cr
This marks the label cr
s/complicated/regex/
This performs the substitution on lines which branched to cr.
With GNU sed, the syntax for the above can be shortened a bit to:
sed '10,20{b cr}; 30,40{b cr}; b; :cr; s/complicated/regex/' file
To delete lines from 10 to 20 and 30 to 40 matching your complicated regex with GNU sed:
sed -e '10,20bA;30,40bA;b;:A;s/complicated/regex/;d' file
or:
sed -e '10,20bA' -e '30,40bA' -e 'b;:A;s/complicated/regex/;d' file
bA: jump to label :A
b: a jump without label -> jump to end of script
d: delete line
I don't think sed has the facility for multiple selection criteria, my advice would be to step up to awk, where you can do something like:
awk 'NR >= 50 && NR <= 70 && /abc/ {next} {print}' inputFile
awk '(NR >= 10 and NR <= 20) || (NR >= 30 && NR <= 40) {
sub("from-regex", "to-string", $0); print }'
sed is excellent for simple substitutions on individual lines but for anything else just use awk for clarity, robustness, portability, maintainability, etc...
awk '
(NR>=50 && NR<=70) && /abc/ { next }
(NR>=10 && NR<=20) || (NR>=30 && NR<=40) { sub(/complicated/,"regex") }
{ print }
' file
I need to replace if ($_SESSION['POST']['*']==1){ with if (isset($_SESSION['POST']['*']) && $_SESSION['POST']['*']==1){
(I'm using * as a wild card)
I've tried sed -i "s/if ($_SESSION['POST']['.*']/if (isset($_SESSION['POST']['.*']) && $_SESSION['POST']['.*']/g" *.php and a few other variations without success.
Here goes...
sed "s/\(if (\)\(\$_SESSION\['POST']\['\([^']*\)']\)==1/\1isset(\2) \&\& \$_SESSION['POST']['\3']==1/" file
Using double quotes means that the $ symbols must be escaped, otherwise they will be interpreted as shell variables. The square brackets need to be escaped, otherwise they will be interpreted as the beginning of a range. It's OK to leave the closing square brackets as they are.
In order to capture the key, I have used a character class [^']*. This means zero or more characters that are not a single quote.
In the replacement, the captured groups (the parts between parentheses in the match) are referred to using \1, \2, etc.
Testing it out:
$ cat file
if ($_SESSION['POST']['foo']==1){
// do something
}
if ($_SESSION['POST']['bar']==1){
// do something else
}
$ sed "s/\(if (\)\(\$_SESSION\['POST']\['\([^']*\)']\)==1/\1isset(\2) \&\& \$_SESSION['POST']['\3']==1/" file
if (isset($_SESSION['POST']['foo']) && $_SESSION['POST']['foo']==1){
// do something
}
if (isset($_SESSION['POST']['bar']) && $_SESSION['POST']['bar']==1){
// do something else
}
By the way it makes the command a few characters shorter if you use extended regexp mode (-r or -E). In extended mode, the parentheses enclosing capture groups don't have to be escaped but literal ones do, so your command would then be:
sed -r "s/(if \()(\$_SESSION\['POST']\['([^']*)'])==1/\1isset(\2) \&\& \$_SESSION['POST']['\3']==1/" file
This sed should work:
s="if (\$_SESSION['POST']['name']==1){"
sed -r 's/(if +)\((([^=]+)[^\)]+)/\1(isset(\3) \&\& \2/' <<< "$s"
if (isset($_SESSION['POST']['name']) && $_SESSION['POST']['name']==1){
PS: Use sed -E instead of sed -r on OSX.
Here's another.
This is what we need to produce:
Pattern: if (\$_SESSION\['POST'\]\['\([^']*\)'\]
Replacement: if (isset($_SESSION['POST']['\1']) \&\& $_SESSION['POST']['\1']
When quoted in shell level:
Pattern: "if (\\\$_SESSION\['POST'\]\['\([^']*\)'\]"
Replacement: "if (isset(\$_SESSION['POST']['\1']) \\&\\& \$_SESSION['POST']['\1']"
Putting it together:
sed -i "s|if (\\\$_SESSION\['POST'\]\['\([^']*\)'\]|if (isset(\$_SESSION['POST']['\1']) \\&\\& \$_SESSION['POST']['\1']|g" file
Test:
# sed "s|if (\\\$_SESSION\['POST'\]\['\([^']*\)'\]|if (isset(\$_SESSION['POST']['\1']) \\&\\& \$_SESSION['POST']['\1']|g" <<'EOF'
> if ($_SESSION['POST']['ABC']==1){
> EOF
if (isset($_SESSION['POST']['ABC']) && $_SESSION['POST']['ABC']==1){