How to apply regex on {} parameter in Linux in find command - regex

I want to change the file name on a certain type of files. It should run recursively.
I have almost got it and don't know how to work with the parameter {} (the absolute Path).
find $PWD -type f -name '*.jpg' -exec echo " {} " \;
For example, I want to change the extensions using reg. expressions and not the command rename etc. I need sometimes new name to pass it to a function, therefore rename is not applicable here.It should be possible to work with the parameter like in for-case with the parameter $each:
for each in /* do echo "${each\./\}.png" done
How can I apply regex on parameter {}, like here: "${each \ . / \ }.png"?

I found a workaround of the basename misbehaving when using it with find :
find $PWD -type f -name '*.jpg' -exec sh -c 'echo "$(basename "$0" .jpg).png"' {} \;
sh forces the following commands to be interpretated using your /bin/sh file. The -c option specifies arguments are passed as strings (here, your argument $0 is {}).
If you have the following files :
/home/username/image1.jpg
/home/username/Documents/image2.jpg
This will output :
image1.png
image2.png
EDIT
If you want to keep the full path, you can use this :
find $PWD -type f -name '*.jpg' -exec sh -c 'echo "${0%%.jpg}".png' {} \;
This will output :
/home/username/image1.png
/home/username/Documents/image2.png

Related

Strip unwanted characters from a series of text files

Hi I've got a list of csv files which need to be formatted properly by getting rid of some unwanted characters.
original:
9: ["2019-4-24",-7.101458109105941]
10: ["2019-5-6",-7.050609022950812]
100: ["2019-5-6",-7.050609022950812]
I'd like to modify as:
2019-4-24,-7.101458109105941
2019-5-6,-7.050609022950812
2019-5-6,-7.050609022950812
There are dozens of files in this format and I was thinking of writing a sed command which makes a series of null substitutions for all the files in directory, but these don't seem to work.
find ./ -type f -exec sed -i '' -e "s/^[[:space:]]*//" {} \;
find ./ -type f -exec sed -i '' -e "s/\[//" {} \;
find ./ -type f -exec sed -i '' -e "s/\]//" {} \;
Many thanks for suggestions.
I found this to work on my linux machine.
find ./ -type f -exec sed -i "s/^.\+\[//;s/\"//g;s/\]//" {} \;
Which, from what I gather is equivalent to the following in macOS:
find ./ -type f -exec sed -i '' "s/^.\+\[//;s/\"//g;s/\]//" {} \;
It comprises of 3 substitutions(separated by semicolon):
s/^.\+\[// deletes everything from the start to the "[" character.
s/\"//g deletes all occurences of the double quote character.
s/\]// deletes the final "]" at the end.
And please make a backup or something if you are going to use sed -i.

How to -find -exec cp files from a list containing a portion of the filename

I've been searching for a long time and can't find an answer that works. I have a list with partial filenames (the first few letters of the filenames). If I place the file names individually as follows it works:
find ~/directory/to/search -name "filename*" -print -exec cp '{}' ~/directory/to/copyto \;
If I try to include the list in this scenario it does not:
cat ~/directory/List.txt | while read line
do
echo "Text read from file - $line"
find ~/directory/to/search -name "$line*" -type f
done
neither does this:
cat ~/directory/List.txt | while read line
do
echo "Text read from file - $line"
find ~/directory/to/search -name "$line&*" -type f
done
Ultimately, I'd like to add:
-exec cp '{}' ~/directory/to/copy/to \;
And copy over all files matching the find criteria.
I've tried using grep but the files are huge so it would take forever. I tried using all sorts of combinations of find, xargs, cp, grep and regex as read in previous searches and no luck.
Is the only solution to write a long script with a bunch of if then statements? I've been using Linux but it would be cool to use it on mac as well.
Here is a crude attempt at getting away with a single find invocation.
predicates=()
or=''
while read -r line; do
predicates+=($or -name "$line*")
or="-o"
done < ~/directory/list.txt
find ~/directory/to/search -type f \( "${predicates[#]}" \) \
-exec cp -t ~/directory/to/copy/to {} +
The array functionality requires an extended shell (Bash, ksh, etc) with this functionality; it won't work with /bin/sh.
cp -t is a GNU extension; if you don't have that, maybe just use your original -exec cp {} dir \; though it will be less efficient. Some old versions of find also don't support -exec ... +.

append epoch date at the beginning of a file in bash

I have a list of 20 files, 10 of them already have 1970-01-01- at the beginning of the name and 10 does not ( the remaining ones all start with a small letter ) .
So my task was to rename those files that do not have the epoch date in the beginning with the epoch date too. Using bash, the below code works, but I could not solve it using a regular expression for example using rename. I had to extract the basename and then further mv. An elegant solution would be just use one pipe instead of two.
Works
find ./ -regex './[a-z].*' | xargs -I {} basename {} | xargs -I {} mv {} 1970-01-01-{}
Hence looking for a solution with just one xargs or -exec?
You can just use a single rename command:
rename -n 's/^([a-z])/1970-01-01-$1/' *
Assuming you're operating on all the files present in current directory.
Note that -n flag (dry run) will only show intended actions by rename command but won't really rename any files.
If you want to combine with find then use:
find . -type f -maxdepth 1 -name '[a-z]*.txt' -execdir rename -n 's/^/1970-01-01-/' {} +
I always prefer readable code over short code.
r() {
base=$(basename "$1")
dir=$(dirname "$1")
if [[ "$base" =~ ^1970-01-01- ]]
then
: "ignore, already has correct prefix"
else
echo mv "$1" "$dir/1970-01-01-$base"
fi
}
export -f r
find . -type f -exec bash -c 'r {}' \;
This also just prints out what would have been done (for testing). Remove the echo before the mv to have to real thing.
Mind that the mv will overwrite existing files (if there is a ./a/b/c and an ./a/b/1970-01-01-c already). Use option -i to mv to be save from this.

Bash script find/sed not working

I have the following line in my bash script:
find . -name "*.html" -print |
xargs sed -i 's/http\:\/\/version2\.staging\.myname\.com//g'
and it's giving me the following error:
sed: 1: "./instant/index. ...": invalid command code .
What I'm trying to do is replace any occurrence of http://version2.staging.myname.com with /. How do you do it?
Usually I use something like:
find . -name "*.html" -exec sed -i 's|http://version2\.staging\.myname\.com/|/|g' '{}' ';'
To test this out, you can first insert an echo statement
find . -name "*.html" -exec echo sed -i 's|http://version2\.staging\.myname\.com/|/|g' '{}' ';'
... that will tell you if the output will be what you expect. I always recommend doing a dry-run with echo first before any mass update. Also you can use | as an alternate regex delimiter to avoid using as many `/' in the paths.
For OSX try this:
find . -name "*.html" -exec sed -i.bak 's#http://version2\.staging\.myname\.com##g' '{}' \; -print
I think you may be using a Mac (and now I see a comment that you are on an iMac). On Mac OS X, the sed -i option requires an argument. That makes sense of your error message. The sed command is interpreting your s/...//g command as the suffix to use for the back up file; it is then trying to interpret the first file name as the sed script, and fortunately, that is not working.
Additionally, you can avoid most of the escaping issues by using some character other than / as the delimiter for s///. Also, it is generally better (especially on Macs where file paths often end up with spaces in them) to avoid xargs and use -exec in find, along with the + option to do what xargs does — namely group many file names into one command invocation.
This leads to:
find . -name "*.html" -type f \
-exec sed -i .bak -e 's%http://version2.staging.myname.com%%g' {} +
(NB: strictly, that will map http://version2-staging*myname#com to / too; if you're really worried about that, by all means escape the dots in the URL.)
If you want to get rid of the .bak files afterwards:
find . -name '*.bak' -type f -exec rm -f {} +

changing folder name recursively using regex (shell)

So I want to change the name of a specific folder recursively. However, that folder isn't always at the same depth or position. I want to change the folder name from variables to constant.
So the variables folders might be located at depth 2, and/or 3, and/or 4, and/or 5, 6, etc... I do not know that
It might be
/var/me/variables/.../.../
or
/var/me/..../..../.../variables/...
or
/var/variables/..../variables/.../../variables/
What I want again is, WHEREVER there is a folder called variables, change its name to constant
I did the following code, but it doesn't work
find var -type d -exec echo `echo "{}" | sed 's/variables/constant/g'` \;
any help would be appreciated.
Thank you!!
This is fun! Here are my two cents:
find . -depth -type d -name variables -execdir mv -T {} constant \;
Rationale:
-depth avoids changing a path find later descends into; probably can be omitted.
-execdir avoids the need to play games with entire paths, so we can operate only on the directory basename
passing the -T option to mv makes it bail if a directory constant should already exist
You actually need to call mv, not echo:
find var -type d -name "*variables*" -exec mv {} `echo "{}" | sed 's/variables/constant/g'` \;
The above assumes you have directories with with string "variables" in names. If you are after directories which are called precisely "variables", then it could be a bit simpler:
find var -type d -name variables -exec mv {} constant \;
Try this find command:
find var -name "*variables*" -type d -exec bash -c 'mv "$1" "${1//variables/constant}"' - '{}' \;
Try using find with rename
find /var -type d -name 'variables' -type d -exec rename 's/variables/constant/' {} \;