How to use sed to delete patterns including backslash on MacOS? - sed

I'm trying to use sed to delete patterns like "\subsection{sometext}",
so from:
Important astrophysical issues could be solved...
\subsection{ Introduction\good line}
\subsection{? apple $ fast burry }white
abc
I want:
Important astrophysical issues could be solved...
white
abc
I have tried:
sed -E -i '' 's/\\(section|subsection|subsubsection)\{([^\}]+)\}//g' test.tex
but the result is:
Important astrophysical issues could be solved...
\subsection{ Introduction\good line}
white
abc
It seems that the text including backslash fails to match. I use MacOS.

} inside [] does not need escaping, so:
's/\\(section|subsection|subsubsection)\{([^}]+)\}//g'
simply works.
you could see the rules in re_format(7):
... With the exception of these and some combinations using `[' (see next paragraphs), all other special characters, including `\', lose their special significance within a bracket expression.

Related

sed is ignoring some matches

I'm trying to replace all matches of http to https using backreference:
example test3.txt file:
http://stronka.wpblog.internal http://stronka.wpblog.internal
abc
jdfgijdf dfijog http://stronka.wpblog.internal dfgtdgrtg http://stronka.wpblog.internal/ sfdgth http://stronka.wpblog.internal/dupa drgfthj
ghj gjerioghj fhjdf http://stronka.wpblog.internal/
and when I run sed against the test3.txt file:
~# sed -r 's#http(://.*.wpblog.internal)#https\1#g' test3.txt
https://stronka.wpblog.internal http://stronka.wpblog.internal
abc
jdfgijdf dfijog https://stronka.wpblog.internal dfgtdgrtg https://stronka.wpblog.internal/ sfdgth http://stronka.wpblog.internal/dupa drgfthj
ghj gjerioghj fhjdf https://stronka.wpblog.internal/
Line 1 second link remains unchanged, line 2 third link remains unchanged, I'm lost, how could I tell sed to replace everything that is matching?
Because the .* wildcard is greedy, i.e. it will consume as much as possible of the line.
The simplest solution by far is to not use a wildcard at all; then sed does precisely what you expect on the simple input you provided.
sed 's#http://#https://#g' test3.txt
(Nothing in this regex needs anything except bog-standard 1968 regex, so the -r option - or its Linux equivalent -E - is not necessary or useful here.)
If for some reason you want a wildcard, use one which doesn't match across URL boundaries. In your example data, spaces seem to separate distinct URLs, so we can match greedily as many non-space characters as possible:
sed -r 's#http(://[^ ]*\.wpblog\.internal)#https\1#g' test3.txt
(Notice also how we use \. to match literal dots.)
Modern regex dialects like Perl's have non-greedy wildcards, but even then, it's better to use a regex which actually means what you want.
Try below:
sed -r 's/\bhttp\b/https/g'
\b is used to set boundaries around "http"
Based on the replies I've replaced the greedy wildcard .*:
sed -E 's#http(://[a-zA-Z0-9.-]*\.wpblog\.internal)#https\1#g'
And it's working as it should now, thank you all!

Why is my sed multiline find-and-replace not working as expected?

I have a simple sed command that I am using to replace everything between (and including) //thistest.com-- and --thistest.com with nothing (remove the block all together):
sudo sed -i "s#//thistest\.com--.*--thistest\.com##g" my.file
The contents of my.file are:
//thistest.com--
zone "awebsite.com" {
type master;
file "some.stuff.com.hosts";
};
//--thistest.com
As I am using # as my delimiter for the regex, I don't need to escape the / characters. I am also properly (I think) escaping the . in .com. So I don't see exactly what is failing.
Why isn't the entire block being replaced?
You have two problems:
Sed doesn't do multiline pattern matches—at least, not the way you're expecting it to. However, you can use multiline addresses as an alternative.
Depending on your version of sed, you may need to escape alternate delimiters, especially if you aren't using them solely as part of a substitution expression.
So, the following will work with your posted corpus in both GNU and BSD flavors:
sed '\#^//thistest\.com--#, \#^//--thistest\.com# d' /tmp/corpus
Note that in this version, we tell sed to match all lines between (and including) the two patterns. The opening delimiter of each address pattern is properly escaped. The command has also been changed to d for delete instead of s for substitute, and some whitespace was added for readability.
I've also chosen to anchor the address patterns to the start of each line. You may or may not find that helpful with this specific corpus, but it's generally wise to do so when you can, and doesn't seem to hurt your use case.
# separation by line with 1 s//
sed -n -e 'H;${x;s#^\(.\)\(.*\)\1//thistest.com--.*\1//--thistest.com#\2#;p}' YourFile
# separation by line with address pattern
sed -e '\#//thistest.com--#,\#//--thistest.com# d' YourFile
# separation only by char (could be CR, CR/LF, ";" or "oneline") with s//
sed -n -e '1h;1!H;${x;s#//thistest.com--.*\1//--thistest.com##;p}' YourFile
Note:
assuming there is only 1 section thistest per file (if not, it remove anything between the first opening until the last closing section) for the use of s//
does not suite for huge file (load entire file into memory) with s//
sed using addresses pattern cannot select section on the same line, it search 1st pattern to start, and a following line to stop but very efficient on big file and/or multisection

sed command over multiple lines not working

I am using sed to replace 14 different abbreviations like CA_23456, CB_scaffold34532,... with 'proper' names in a file and it works putting it all on one line.
acc=$1
sed -e 's/CA_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_arizonica/;s/CB_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_bakeri/;s/CM_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_macrocarpa/;s/CS_[A-Z]*[a-z]*[0-9]*/Cupressus_sempervirens/;s/CT_[A-Z]*[a-z]*[0-9]*/Cupressus_torulosa/;s/JD_[A-Z]*[a-z]*[0-9]*/Juniperus_drupacea/;s/JF_[A-Z]*[a-z]*[0-9]*/Juniperus_flaccida/;s/JI_[A-Z]*[a-z]*[0-9]*/Juniperus_indica/;s/JP_[A-Z]*[a-z]*[0-9]*/Juniperus_phoenicea/;s/JX_[A-Z]*[a-z]*[0-9]*/Juniperus_procera/;s/JS_[A-Z]*[a-z]*[0-9]*/Juniperus_scopulorum/;s/MD_[A-Z]*[a-z]*[0-9]*/Microbiota_decussata/;s/XN_[A-Z]*[a-z]*[0-9]*/Xanthocyparis_nootkatensis/;s/XV_[A-Z]*[a-z]*[0-9]*/Xanthocyparis_vietnamensis/' ${acc}.nex > ${acc}_replaced.nex
To make it more readable I'd like to have the command split over multiple lines using '\' (not all the replacements are shown for brevity)
acc=$1
sed -e 's/CA_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_arizonica/;\
s/CB_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_bakeri/;\
s/CM_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_macrocarpa/'\
${acc}.nex > ${acc}_replaced.nex
However, I get an error message: sed: -e expression #1, char 168: unterminated address regex. I have looked at the answers to similar problems on various webforums and tried various things (using 's/.../.../' on every line, leaving ';' out,....) but I can't get it to work. What am I doing wrong?
Drop the \ that escapes the newlines. (They are not actually doing it!, they are interpreted as wrong syntax by sed). However I would suggest to put it into a file and run it like this:
sed -f script.sed input
where script.sed looks like this:
s/CA_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_arizonica/
s/CB_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_bakeri/
s/CM_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_macrocarpa/
Remove the backslashes from the sed code.
Inside singly-quoted shell strings, backslashes are not needed to escape newlines and are not removed because they are not parsed as escape characters. This has the effect that sed sees them as part of its code, and it then expects to find an address regex with a different delimiter than / before the command ends at the next newline (similar to \,/home/, !d). This address regex does not appear (nor an associated command), and so sed complains about invalid code.
Apart from that: The semicolons in the sed code are no longer necessary when you terminate commands with newlines, and anything involving shell variables should be quoted to avoid splitting in case of whitespace.
In sum:
sed -e 's/CA_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_arizonica/
s/CB_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_bakeri/
s/CM_[A-Z]*[a-z]*[0-9]*/Hesperocyparis_macrocarpa/' \
"${acc}.nex" > "${acc}_replaced.nex"

Confining Substitution to Match Space Using sed?

Is there a way to substitute only within the match space using sed?
I.e. given the following line, is there a way to substitute only the "." chars that are contained within the matching single quotes and protect the "." chars that are not enclosed by single quotes?
Input:
'ECJ-4YF1H10.6Z' ! 'CAP' ! '10.0uF' ! 'TOL' ; MGCDC1008.S1 MGCDC1009.A2
Desired result:
'ECJ-4YF1H10-6Z' ! 'CAP' ! '10_0uF' ! 'TOL' ; MGCDC1008.S1 MGCDC1009.A2
Or is this just a job to which perl or awk might be better suited?
Thanks for your help,
Mark
Give the following a try which uses the divide-and-conquer technique:
sed "s/\('[^']*'\)/\n&\n/g;s/\(\n'[^.]*\)\.\([^']*Z'\)/\1-\2/g;s/\(\n'[^.]*\)\.\([^']*uF'\)/\1_\2/g;s/\n//g" inputfile
Explanation:
s/\('[^']*'\)/\n&\n/g - Add newlines before and after each pair of single quotes with their contents
s/\(\n'[^.]*\)\.\([^']*Z'\)/\1-\2/g - Using a newline and the single quotes to key on, replace the dot with a dash for strings that end in "Z"
s/\(\n'[^.]*\)\.\([^']*uF'\)/\1_\2/g - Using a newline and the single quotes to key on, replace the dot with a dash for strings that end in "uF"
s/\n//g - Remove the newlines added in the first step
You can restrict the command to acting only on certain lines:
sed "/foo/{s/\('[^']*'\)/\n&\n/g;s/\(\n'[^.]*\)\.\([^']*Z'\)/\1-\2/g;s/\(\n'[^.]*\)\.\([^']*uF'\)/\1_\2/g;s/\n//g}" inputfile
where you would substitute some regex in place of "foo".
Some versions of sed like to be spoon fed (instead of semicolons between commands, use -e):
sed -e "/foo/{s/\('[^']*'\)/\n&\n/g" -e "s/\(\n'[^.]*\)\.\([^']*Z'\)/\1-\2/g" -e "s/\(\n'[^.]*\)\.\([^']*uF'\)/\1_\2/g" -e "s/\n//g}" inputfile
$ cat phoo1234567_sedFix.sed
#! /bin/sed -f
/'[0-9][0-9]\.[0-9][a-zA-Z][a-zA-Z]'/s/'\([0-9][0-9]\)\.\([0-9][a-zA-Z][a-zA-Z]\)'/\1_\2/
This answers your specific question. If the pattern you need to fix isn't always like the example you provided, they you'll need multiple copies of this line, with reg-expressions modified to match your new change targets.
Note that the cmd is in 2 parts, "/'[0-9][0-9].[0-9][a-zA-Z][a-zA-Z]'/" says, must match lines with this pattern, while the trailing "s/'([0-9][0-9]).([0-9][a-zA-Z][a-zA-Z])'/\1_\2/", is the part that does the substitution. You can add a 'g' after the final '/' to make this substitution happen on all instances of this pattern in each line.
The \(\) pairs in match pattern get converted into the numbered buffers on the substitution side of the command (i.e. \1 \2). This is what gives sed power that awk doesn't have.
If your going to do much of this kind of work, I highly recommend O'Rielly's Sed And Awk book. The time spent going thru how sed works will be paid back many times.
I hope this helps.
P.S. as you appear to be a new user, if you get an answer that helps you please remember to mark it as accepted, or give it a + (or -) as a useful answer.
this is a job most suitable for awk or any language that supports breaking/splitting strings.
IMO, using sed for this task, which is regex based , while doable, is difficult to read and debug, hence not the most appropriate tool for the job. No offense to sed fanatics.
awk '{
for(i=1;i<=NF;i++) {
if ($i ~ /\047/ ){
gsub(".","_",$i)
}
}
}1' file
The above says for each field (field seperator by default is white space), check to see if there is a single quote, and if there is , substitute the "." to "_". This method is simple and doesn't need complicated regex.

Deleting multiline text from multiple files

I have a bunch of java files from which I want to remove the javadoc lines with the license [am changing it on my code].
The pattern I am looking for is
^\* \* ProjectName .* USA\.$
but matched across lines
Is there a way sed [or a commonly used editor in Windows/Linux] can do a search/replace for a multiline pattern?
Here's the appropriate reference point in my favorite sed tutorial.
Probably someone is still looking for such solution from time to time. Here is one.
Use awk to find the lines to be removed. Then use diff to remove the lines and let sed clean up.
awk "/^\* \* ProjectName /,/ USA\.$/" input.txt \
| diff - input.txt \
| sed -n -e"s/^> //p" \
>output.txt
A warning note: if the first pattern exist while the second does not, you will loose all text below the first pattern - so check that first.
Yes. Are you using sed, awk, perl, or something else to solve this problem?
Most regular expression tools allow you to specify multi-line patterns. Just be careful with regular expressions that are too greedy, or they'll match the code between comments if it exists.
Here's an example:
/\*(?:.|[\r\n])*?\*/
perl -0777ne 'print m!/\*(?:.|[\r\n])*?\*/!g;' <file>
Prints out all the comments run
together. The (?: notation must be
used for non-capturing parenthesis. /
does not have to be escaped because !
delimits the expression. -0777 is used
to enable slurp mode and -n enables
automatic reading.
(From: http://ostermiller.org/findcomment.html )