Bash Substring multiple parameters

Question

I need to extract two parameters from each line of a svn log but I am not able to do it with grep.

My Svn log command, such as

svn log http://svn.apache.org/repos/asf/xerces/java/trunk/ | grep "^r[0-9]\+ | " | cut -c2-

Outputs the result in this format:

318150 | lehors | 2002-01-28 20:48:11 +0100 (Mon, 28 Jan 2002) | 2 lines
318149 | elena | 2002-01-28 20:46:33 +0100 (Mon, 28 Jan 2002) | 12 lines
318148 | lehors | 2002-01-28 20:33:36 +0100 (Mon, 28 Jan 2002) | 2 lines
318147 | lehors | 2002-01-28 20:22:51 +0100 (Mon, 28 Jan 2002) | 2 lines

How can I grep the release number (first parameter) and the date in this format?

318150 2002-01-28
318149 2002-01-28
318148 2002-01-28
318147 2002-01-28

Inian · Accepted Answer · 2017-02-17 09:03:43Z

2

Use a more robust Awk for this to pattern-match/extract from individual columns.

.. | awk 'BEGIN{FS="|"}{split($3,temp, " "); print $1,temp[1]}'
318150  2002-01-28
318149  2002-01-28
318148  2002-01-28
318147  2002-01-28

The .. | part represents the command to be included that produces the required output which is pipe-lined to Awk

The logic is pretty straight-forward, split input lines with de-limiter as | which is done by FS="|". Now $1 represents the first field you want, and for the second part, split the part $3 and use the split() function to split on delimiter, a white-space character and store it in array temp, so that it can be accessed as temp[1], the other space fields are present in the array from the next index on wards.

So ideally I guess it should be,

svn log http://svn.apache.org/repos/asf/xerces/java/trunk/ | \
   awk 'BEGIN{FS="|"}{split($3,temp, " "); print $1,temp[2]}'

Alternatively you could use GNU grep with its -E extended regular expression capabilities, but it is just not good enough to show the matching entries on same line like,

grep -oE '[[:digit:]]{4}-[[:digit:]]{2}-[[:digit:]]{2}' file

(and)

grep -oE '^[[:digit:]]{6}' file

but not together as I have used the -o flag to print the match only part.

edited Feb 17, 2017 at 9:03

answered Feb 17, 2017 at 8:50

Inian

87.1k15 gold badges167 silver badges183 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Zumo de Vidrio Over a year ago

Just realized that this will print the time and OP wants the date, so changing temp[2] by temp[1] will do it :)

Zumo de Vidrio · Accepted Answer · 2017-02-17 08:58:57Z

1

As your file is separated by a single space and you want to have the first and 5th columns, this is another solution by using cut:

cut -d' ' -f1,5 < svn_log_output_file

(or piping cut -d' ' -f1,5 to your command)

answered Feb 17, 2017 at 8:58

Zumo de Vidrio

2,1412 gold badges18 silver badges37 bronze badges

1 Comment

Inian Over a year ago

Also, you could do process substitution as, cut -d' ' -f1,5 < <(svn log http://svn.apache.org/repos/asf/xerces/java/trunk/)

Chem-man17 · Accepted Answer · 2017-02-17 09:05:31Z

1

A much simpler approach with multiple delimiters-

awk -F '[| ]' '{print $1, $7}' file

Where file contains the output you showed in your question.

Output-

318150 2002-01-28
318149 2002-01-28
318148 2002-01-28
318147 2002-01-28

Of course, you don't need to store in an intermediate file. You can do-

svn log http://svn.apache.org/repos/asf/xerces/java/trunk/ \
| grep "^r[0-9]\+ | " | cut -c2- | \
awk -F '[| ]' '{print $1, $7}'

answered Feb 17, 2017 at 9:05

Chem-man17

1,7901 gold badge14 silver badges27 bronze badges

Comments

Claes Wikner · Accepted Answer · 2017-02-17 09:23:36Z

0

awk '{print $1,$5}' file

318150 2002-01-28
318149 2002-01-28
318148 2002-01-28
318147 2002-01-28

answered Feb 17, 2017 at 9:23

Claes Wikner

1,5271 gold badge9 silver badges8 bronze badges

Collectives™ on Stack Overflow

Bash Substring multiple parameters

4 Answers 4

1 Comment

1 Comment

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

1 Comment

1 Comment

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related