Replacing first occurence except comments using sed - sed

I have a text file of the form:
a = 1
#b = [2,3]
c = 4
d = [5,6]
e = [7,8]
I want to replace the pattern inside the brackets (and the brackets) with a number, but ignore matches in the comments, preferably using sed.
For files with exactly one matching line, I've used
sed -i "/^#/!s/\[.*\]/9/" myfile
How can this be modified to replace only the first match if there are more?

This is the correct, because changing just the first occurence.
awk '!end && /^[^#]+ = \[/ {$3="9"; end=1}1' myfile
if there is not end flag and line is not beggining from # and match to = [, then change the third column and set the flag to prevent changing in next occurences.
a = 1
#b = [2,3]
c = 4
d = 9
e = [7,8] <--- this is not changed as you want

This one-liner should do the job:
sed '/^\s*#/!{s/\[[^]]*\]/9/}' file
add the -i option if you like to do the change in place.

Related

How can I extract first and last line from multiple text blocks separated with new line?

I have a file containing multiple tests with detailed action written one beneath another. All test blocks are separated one from another by new line. I want to extract only first and last line from the all blocks and put it on one line for each block into a new file. Here is an example:
input.txt:
[test1]
duration
summary
code=
Results= PASS
[test2]
duration
summary=x
code=
Results=FAIL
.....
[testX]
duration
summary=x
code=
Results= PASS
output.txt should be sometime like this:
test1 PASS
test2 FAIL
...
testX PASS
eg2:
[Linux_MP3Enc_xffv.2_Con_37_003]
type = testcase
summary = MP3 encoder test
ActionGroup[Linux_Enc] = PASS
ActionGroup[Linux_Playb] = PASS
ActionGroup[Linux_Pause_Resume] = PASS
ActionGroup[Linux_Fast_Seek] = PASS
Duration = 230.607398987 s
Total_Result = PASS
[Composer__vtx_007]
type = testcase
summary = composer
Background[0xff000000] = PASS
Background[0xffFFFFFF] = PASS
Background[0xffFF0000] = PASS
Background[0xff00FF00] = PASS
Background[0xff0000FF] = PASS
Background[0xff00FFFF] = PASS
Background[0xffFFFF00] = PASS
Background[0xffFF00FF] = PASS
Duration = 28.3567230701 s
Total_Result = PASS
[Videox_Rotate_008]
type = testcase
summary = rotation
Rotation[0] = PASS
Rotation[1] = PASS
Rotation[2] = PASS
Rotation[3] = PASS
Duration = 14.0116529465 s
Total_Result = PASS
Thank you!
Short and simple gnu awk:
awk -F= -v RS='' '{print $1 $NF}' file
[Linux_MP3Enc_xffv.2_Con_37_003] PASS
[Composer__vtx_007] PASS
[Videox_Rotate_008] PASS
If you do not like the brackets:
awk -F'[]=[]' -v RS='' '{print $2 $NF}' file
Linux_MP3Enc_xffv.2_Con_37_003 PASS
Composer__vtx_007 PASS
Videox_Rotate_008 PASS
Using sed as tagged (although other tools would probably be more natural to use) :
sed -nE '/^\[.*\]$/h;s/^Results= ?//;t r;b;:r;H;x;s/\n/ /;p'
Explanation :
/^\[.*\]$/h # matches the [...] lines, put them in the hold buffer
s/^Results= ?// # matches the Results= lines, discards the useless part
t r;b # on lines which matched, jump to label r;
# otherwise jump to the end (and start processing the next line)
:r;H;x;s/\n/ /;p # label r; append the pattern space (which contains the end of the Results= line)
# to the hold buffer. Switch Hold buffer and pattern space,
# replace the linefeed in the pattern space by a space and print it
You can try it here.
One way to solve this is using a regular expression such as:
(?<testId>test\d+)(?:.*\n){4}.*(?<outcome>PASS|FAIL)
The regex matches your sample output and stores the test id (e.g. "test1") in the capture group named "testId" and the outcome (e.g. "PASS") in the capture group "outcome".
(Test it in regexr)
The regex can be used in any language with regex support. The below code shows how to do it in Python.
(Test it in repl.it)
import re
# Read from input.txt
with open('input.txt', 'r') as f:
indata = f.read()
# Modify the regex slightly to fit Python regex syntax
pattern = '(?:.*)(?P<testId>test\d+)(?:.*\n){4}.*(?P<outcome>PASS|FAIL)'
# Get a generator which yeilds all matches
matches = re.finditer(pattern, indata)
# Combine the matches to a list of strings
outputs = ['{} {}'.format(m.group('testId'), m.group('outcome')) for m in matches]
# Join all rows to one string
output = '\n'.join(outputs)
# Write to output.txt
with open('output.txt', 'w') as f:
f.write(output)
Running the above script on input.txt containing:
[test1]
duration
summary
code=
Results= PASS
[test2]
duration
summary=x
code=
Results=FAIL
[test444]
duration
summary=x
code=
Results= PASS
yields a file output.txt containing:
test1 PASS
test2 FAIL
test444 PASS
In order to print the first and last line from the block, how about:
awk -v RS="" '{
n = split($0, a, /\n/)
print a[1]
print a[n]
}' input.txt
Result for the 1st example:
[Linux_MP3Enc_xffv.2_Con_37_003]
Total_Result = PASS
[Composer__vtx_007]
Total_Result = PASS
[Videox_Rotate_008]
Total_Result = PASS
The man page of awk tells:
If RS is set to the null string, then records are separated by blank lines.
You can easily split the block with blank lines with this feature.
Hope this helps.

How to update a specific key in a specific section using sed?

I have an ini file similar to the following one:
[seciton1]
key1 = 0
key2 = 0
[section2]
key1 = 0
key2 = 0
I want to update key1 of section "section2" from 0 to 1. what will be the accurate sed command for the same.
also is sed command supported in SSH.NET c# library? if not then what are the ways using ssh.net we can update ini file.
This will update key2 once when its after section2:
awk '/section2/ {f=1} /key2/ && f {$3=1;f=0} 1' file
[seciton1]
key1 = 0
key2 = 0
[section2]
key1 = 0
key2 = 1
To write it back to the original file, do:
awk '/section2/ {f=1} /key2/ && f {$3=1;f=0} 1' file > tmp && mv tmp file
This might work for you (GNU sed):
sed -i '/^\[/h;G;/section2/s/\(key1 = \).*/\11/m;P;d' file
Make a copy of the section header and append it to each line. If the current section header is section2 substitute 1 for the value of key1.
N.B. The m flag in the substitute command restricts the .* to the remainder of the first line (of the now two lines in the pattern space i.e. the current line and the appended section header) only.

Sed: How to insert a pattern which includes single ticks?

I'm trying to use sed to replace a specific line within a configuration file:
The pattern for the line I want to replace is:
ALLOWED_HOSTS.*
The text I want to insert is:
'$PublicIP' (Including the single ticks)
But when I run the command:
sed 's/ALLOWED_HOSTS.*/ALLOWED_HOSTS = ['$PublicIP']/g' /root/project/django/mysite/mysite/settings.py
The line is changed to:
ALLOWED_HOSTS = [1.1.1.1]
instead of:
ALLOWED_HOSTS = ['1.1.1.1']
How shall I edit the command to include the single ticks as well?
You could try to escape the single ticks , or better you can reassign the variable including the simple ticks:
PublicIP="'$PublicIP'".
By the way even this sed without redifining var, works ok in my case:
$ a="3.3.3.3"
$ echo "ALLOWED_HOSTS = [2.2.2.2]" |sed 's/2.2.2.2/'"'$a'"'/g'
ALLOWED_HOSTS = ['3.3.3.3']
Even this works ok:
$ echo "ALLOWED_HOSTS = [2.2.2.2]" |sed "s/2.2.2.2/'$a'/g"
ALLOWED_HOSTS = ['3.3.3.3']

use sed to change a text report to csv

I have a report looks like this:
par_a
.xx
.yy
par_b
.zz
.tt
I wish to convert this format into csv format as below using sed 1 liner:
par_a,.xx
par_a,.yy
par_b,.zz
par_b,.tt
please help.
With awk:
awk '/^par_/{v=$0;next}/^ /{$0=v","$1;print}' File
Or to make it more generic:
awk '/^[^[:blank:]]/{v=$0;next} /^[[:blank:]]/{$0=v","$1;print}' File
When a line starts with par_, save the content to variable v. Now, when a line starts with space, change the line to content of v followed by , followed by the first field.
Output:
AMD$ awk '/^par_/{v=$0}/^ /{$0=v","$1;print}' File
par_a,.xx
par_a,.yy
par_b,.zz
par_b,.tt
With sed:
sed '/^par_/ { h; d; }; G; s/^[[:space:]]*//; s/\(.*\)\n\(.*\)/\2,\1/' filename
This works as follows:
/^par_/ { # if a new paragraph begins
h # remember it
d # but don't print anything yet
}
# otherwise:
G # fetch the remembered paragraph line to the pattern space
s/^[[:space:]]*// # remove leading whitespace
s/\(.*\)\n\(.*\)/\2,\1/ # rearrange to desired CSV format
Depending on your actual input data, you may want to replace the /^par_/ with, say, /^[^[:space:]]/. It just has to be a pattern that recognizes the beginning line of a paragraph.
Addendum: Shorter version that avoids regex repetition when using the space pattern to recognize paragraphs:
sed -r '/^\s+/! { h; d; }; s///; G; s/(.*)\n(.*)/\2,\1/' filename
Or, if you have to use BSD sed (as comes with Mac OS X):
sed '/^[[:space:]]\{1,\}/! { h; d; }; s///; G; s/\(.*\)\n\(.*\)/\2,\1/' filename
The latter should be portable to all seds, but as you can see, writing portable sed involves some pain.

Flip array index with sed

I have some java code declaring a 2d array that I want to flip.
Content is like:
zData[0][0] = 198;
zData[0][1] = 198;
zData[0][2] = 198;
...
And I want to flip indices to have
zData[0][0] = 198;
zData[1][0] = 198;
zData[2][0] = 198;
So I tried doing it with sed:
sed -r 's#zData[([0-9]*)][([0-9]*)]#zData[\2][\1]#g' DataSample1.java
But unfortunately sed says:
sed: -e expression #1, char 43: Unmatched ) or \)
Might the string "zData" hold kind of flag or option?
I tried not using the -r option but I have the same kind of message for:
sed 's#zData[\(\[\0\-\9\]\*\)][\(\[\0\-\9\]\*\)]#zData[\2][\1]#g' DataSample1.java
Thanks for your help
Simples:
$ sed -r 's/(zData)(\[[^]]+])(\[[^]]+])/\1\3\2/' file
zData[0][0] = 198;
zData[1][0] = 198;
zData[2][0] = 198;
Regexplanation:
# Match
(zData) # Capture the variable name we want to transpose
( # Start capture group for first index
\[ # Opening bracket escaped to mean literal [
[^]]+ # One or more none ] characters i.e the digits
] # The closing literal ] doesn't need escaping here.
) # Close the capture
(\[[^]]+]) # Same regexp as before for the second index
# Replace
\1\3\2 # Switch the indexes but rearranging the 2nd and 3rd capture groups
Note: Switch \[[^]]+] to if it is clearer \[[0-9]+] for you, so instead of saying match an opening square bracket followed by one or more none-closing brackets followed by a closing bracket you are saying match an opening square bracket followed by one or more digit followed by a closing bracket.
Try that one:
sed 's#\([a-zA-Z0-9_-]\+\)\(\[[^]]*\]\)\(\[[^]*]\]\)\(.*$\)#\1\3\2\4#'
It adds four captures for the variable name, the first index, the second index and the rest and then switches order.
Edit: #Sudo_O's solution with extended regular expressions is much more readable. Thx for that! Nevertheless, on some systems sed -r may not be available, since it is not part of basic POSIX.