I have filenames in a directory like:
ACCT_GA12345_2015-01-10.xml
ACCT_GA12345_2015-01-09.xml
ACCT_GDC789g_2015-01-09.xml
ACCT_GDC567g_2015-01-09.xml
ACCT_GDC567g_2015-01-08.xml
ACCT_GCC7894_2015-01-01.xml
ACCT_GCC7894_2015-01-02.xml
ACCT_GAC7884_2015-02-01.xml
ACCT_GAC7884_2015-01-01.xml
I want to have only the latest file in the folder. The latest file can be found using only the file name (NOT the date stamp). For example ACCT 12345 has files from 1/10 & 1/09. I need to delete 1/09 file and have only 1/10 file, for ACCT 789g there is only one file so I have to have that file, and ACCT 567g the latest file is 1/09 so I have to remove 1/08 and have 1/09. So the combination for latest file should be ACCT & Max date for that ACCT.
I would need the final list of files as:
ACCT_GA12345_2015-01-10.xml
ACCT_GDC789g_2015-01-09.xml
ACCT_GDC567g_2015-01-09.xml
ACCT_GCC7894_2015-01-02.xml
ACCT_GAC7884_2015-02-01.xml
Can someone help me with this command in unix? Any help is appreciated
I'd do something like this.... to test start with ls command, when you get what you want to delete, then do rm.
ls ACCT_{GDC,GA1}*-{09,10}.xml
this will list any GDC or GA1 files that end in 09 or 10. You can play with combinations and different values until you have the right set of files showing that you want deleted. once you to just change ls to rm and you should be golden.
With some more info I could help you out. To test this out I did:
touch ACCT_{GDC,GA1}_{01..10}_{05..10}.xml
this will make 56 different dummy files with different combinations. Make a directory, run this command, and get your hands dirty. That is the best way to learn linux cli. Also 65% of commands you need, you will learn, understand, use then never use again...so learn how to teach yourself how to use man pages and setup a spot to play around in.
Related
I was trying to store multiple diff outputs in one .patch file to keep versioning in one file instead of running multiple
diff -u f1 f2 > f1.patch
commands. Preferably I'd keep running
diff -u[other params?] f1 f2 >> f1.patch
to have one file containing all changes which would allow me to later on run patch on those files to have a f1 file available in any given moment.
Unfortunately patch fails with file generated in such manner. It only seems to apply first patch from the file and then quits with error.
My question: is that possible with diff and patch? And if so, how?
Thank you in advance.
I really would like to know how to change directory to the next file, using 2 dots (/..).
I can go back to previous directory but how to do something like that to go to the next directory WITHOUT THE NAME OF THE FILE PLEASE!
I know it is possbile with the name of the folder but my answer is different, please let me know.
What O.S are you using? In linux ubuntu/Debian distributions (and probably in windows, not sure), you can enter the first child directory by simply typingcd */, this will change your current directory to the first child directory available in alphabetical order, example:
-/home
-alan
-desktop
-music
-alex
-documents
-pictures
-lisa
Assuming your in /home, if you write cd */ in the terminal you will be in /home/alan, because -alan is the first folder/directory available.
I am working in UNIX and trying to write the following commands. I am receiving a source file daily whose filename is in the format :
ONSITE_EXTR_ONSITE_EXTR_20170707.
Since I am receiving a file daily, the file name would change based on the current date, so ONSITE_EXTR_ONSITE_EXTR_20170708, ONSITE_EXTR_ONSITE_EXTR_20170709 etc. I need to strip the date out of the filename and rename it to ONSITE_EXTR_ONSITE_EXTR. After I have finished whatever data reading and processing I need to do, I need to change the file name back to ONSITE_EXTR_ONSITE_EXTR_20170707 for example. So since the file is being delivered daily, I cant hard code the date in whatever commands I write. Any help would be greatly appreciated
Depending on your toolchain, this may be as simple as running:
$ mv ONSITE_EXTR_ONSITE_EXTR_$(date +%Y%m%d) ONSITE_EXTR_ONSITE_EXTR
... before running the rest of your script, assuming you're using a Bash-like shell.
Having said that, you can just drop in ONSITE_EXTR_ONSITE_EXTR_$(date +%Y%m%d) into your script when trying to access your file instead.
This is all assuming the script's run the same day and in the same time zone as the file is downloaded.
If you were using bash and you had the file name in a variable, you could do:
IN="ONSITE_EXTR_ONSITE_EXTR_20170707"
echo ${IN:0:23}
to give ONSITE_EXTR_ONSITE_EXTR
Googling gives all sorts of guides here...
Based on the count from some other file, I need to rename all the files extensions.
Ex: If the count is 10 and there are 5 files exists, I need to rename all the files as below.
from File_1.txt to File_11.txt,
from File_2.txt to File_12.txt,
from File_3.txt to File_13.txt,
from File_4.txt to File_14.txt,
from File_5.txt to File_15.txt
Can I use one unix command to do this, appreciate your help on this.
Regards,
NPK
With standard UNIX you'd need more than one command, e. g.
count=10
for file in File_*.txt
do augend=`echo $file|sed 's/File_\(.*\).txt/\1/'`
mv $file File_`expr $augend + $count`.txt
done
But if you have a system with this rename available, you can
rename 's/File_(.*).txt/"File_".($1+$ENV{count}).".txt"/e' File_*.txt
(assuming count has been exported to the environment) or
rename 's/File_(.*).txt/"File_".($1+'$count').".txt"/e' File_*.txt
as well.
I have a slew of CSS files to go through where someone just grunted through making alterations to various core stylesheets on a number of subsites. Obviously if the original developer had had some foresight they would have just included a master stylesheet and overridden the necessary elements…
I first started off with comm thinking that it might do the trick, but quickly found that it needed to receive a sorted input file.
I then switched over to diff and have gotten down to the following through some reading and research:
diff --unchanged-group-format="## %dn,%df%c'\012'%<" --old-group-format='' --new-group-format='' --changed-group-format='' file_1.css file_2.css
The previous obviously is almost there, but:
A) I need to grep out the ## lines (which should be fine, right? At first glance this appears right, but does diff throw in any other unexpected lines that need to be yanked?) and then
B) I need to create two more files that first is the leftover unique lines from file_1.css and then the leftover unique lines of file_2.css.
Obviously the first "in common" file will go into an include folder and then be included into the two latter created files as a #import url("common.css");
I am thinking that the following simple alteration will create the latter two files to which I'm referring:
diff --unchanged-group-format='' --old-group-format="## %dn,%df%c'\012'%<" --new-group-format='' --changed-group-format='' file_1.css file_2.css
diff --unchanged-group-format='' --old-group-format='' --new-group-format="## %dn,%df%c'\012'%<" file_1.css file_2.css
Sample files:
file 1: https://gist.github.com/c13843972c47b5037704
file 2: https://gist.github.com/fff39eae386e8969dc10
So for example, upon executing a test of the following:
diff --unchanged-group-format="## %dn,%df%c'\012'%<" --old-group-format='' --new-group-format='' --changed-group-format='' file_1.css file_2.css | egrep -v "^##\d*" > common.css
diff --unchanged-group-format='' --old-group-format="## %dn,%df%c'\012'%<" --new-group-format='' --changed-group-format='' file_1.css file_2.css | egrep -v "^##\d*" > old.css
And then searching for body with egrep "^body" *css, it yielded only a body in common.css and none in old.css, whereas it showed that there were two different entries in file_1.css and file_2.css. So obviously this methodology is flawed.
How would one about creating these two files that would ultimately become the common include and the override files?
#ylluminate, you have a couple of options:
use BeyondCompare to visually verify the differences. It does a fantastic job comparing similar files. It allows saving common lines/left only lines/right only lines. Only downside is it is interactive and if you have a lot of files, will take some time. On the positive side, it looks like you want to build trust first by testing it out a few times.
Add formatting text for --changed-group-format and capture modified code (and the old code as your command does it now). You need to run one more comparison to get what is in new code but not in old code. Downside here is the validation is going to be hard.
Saving all the lines in a database table and comparing columns is another option. Take care to store old and new line numbers. Downsides are the data lines need to be unique, blank lines will be chopped off.
I would go with option 1 if i have less than 50 files.
Hope this helps.
PS: I am not associated with BeyondCompare in any way. just a happy user of the software