3

I know how to insert a single text to specific interval. But now my problem is, I want to insert different texts/words saved in an add.txt file to a specific intervals of another data.txt file.

I want to insert first word from add.txt to a specific position of data.txt, then add second word from add.txt to next specific position and so on.

My data.txt contain two columns, but the inserted word must appear as a merged row. Please see the example below of what I need.

add.txt

2001-01-01 00:00:00 42 1
2001-01-02 00:00:00 42 1
2001-01-03 00:00:00 42 1
2001-01-04 00:00:00 42 1
2001-01-05 00:00:00 42 1

data.txt

  -500      11.822788
  -400      12.006394
  -350      12.287062
  -300      12.793395
  -500      11.823597
  -400      12.008012
  -350      12.287062
  -300      12.794204
  -500      11.826023
  -400      12.011247
  -350      12.291915
  -300      12.800675
  -500      11.827641
  -400      12.013674
  -350      12.295959
  -300      12.805528
  -500      11.830067
  -400      12.016100
  -350      12.300003
  -300      12.811998

I want

2001-01-01 00:00:00 42 1 
 -500      11.822788
  -400      12.006394
  -350      12.287062
  -300      12.793395
2001-01-02 00:00:00 42 1
  -500      11.823597
  -400      12.008012
  -350      12.287062
  -300      12.794204
2001-01-03 00:00:00 42 1
  -500      11.826023
  -400      12.011247
  -350      12.291915
  -300      12.800675
2001-01-04 00:00:00 42 1
  -500      11.827641
  -400      12.013674
  -350      12.295959
  -300      12.805528
2001-01-04 00:00:00 42 1
  -500      11.830067
  -400      12.016100
  -350      12.300003
  -300      12.811998

I am looking for a simplest solution using awk, sed or something.

Improve this question

3 Answers 3

Reset to default
7 
$ awk '(FNR-1)%4 == 0 { getline add <"add.txt"; print add }; 1' data.txt
2001-01-01 00:00:00 42 1
  -500      11.822788
  -400      12.006394
  -350      12.287062
  -300      12.793395
2001-01-02 00:00:00 42 1
  -500      11.823597
  -400      12.008012
  -350      12.287062
  -300      12.794204
2001-01-03 00:00:00 42 1
  -500      11.826023
  -400      12.011247
  -350      12.291915
  -300      12.800675
2001-01-04 00:00:00 42 1
  -500      11.827641
  -400      12.013674
  -350      12.295959
  -300      12.805528
2001-01-05 00:00:00 42 1
  -500      11.830067
  -400      12.016100
  -350      12.300003
  -300      12.811998

This uses awk to read and output every line of the data.txt file. Before outputting any 4th line, a line is read and outputted from the add.txt file.

No check is made to verify that the data read from add.txt is correctly read (if the file is too short, the above code would repeat the last line).

Using paste:

$ paste -d '\n' add.txt - - - - <data.txt
2001-01-01 00:00:00 42 1
  -500      11.822788
  -400      12.006394
  -350      12.287062
  -300      12.793395
2001-01-02 00:00:00 42 1
  -500      11.823597
  -400      12.008012
  -350      12.287062
  -300      12.794204
2001-01-03 00:00:00 42 1
  -500      11.826023
  -400      12.011247
  -350      12.291915
  -300      12.800675
2001-01-04 00:00:00 42 1
  -500      11.827641
  -400      12.013674
  -350      12.295959
  -300      12.805528
2001-01-05 00:00:00 42 1
  -500      11.830067
  -400      12.016100
  -350      12.300003
  -300      12.811998

Here, I ask paste to create records with a line from add.txt as the first field, followed by four lines from data.txt as the next four fields. With -d '\n' I set the character to use as a field delimiter to a newline character.

Improve this answer
4
  • ♦ wow! that did the job. Many thanks and much appreciated. Commented Aug 25, 2020 at 7:59
  • 1
    The paste option seems more handy. But what should do if my intervals are large, say, 500? Is there any alternative for dashes ---- in this case? Commented Aug 25, 2020 at 10:53
  • 1
    @linux_lover Not really. The paste variation is really only useful for smaller intervals like the one in your question. There is no way to "abbreviate" the column specification with paste. Commented Aug 25, 2020 at 11:05
  • That make sense. Thanks for info. Commented Aug 25, 2020 at 11:08
3

If GNU sed is an option, especially for those on MacOS, we can run the following sed code. The R command and the ~ command, which are gnu additions are the key here.

The ~ will make sed look at every fourth line specially, for others it will pass through to stdout. The R shall take one line from add.txt and mix it with the data.txt line.

$ sed '
    1~4!b
    R add.txt
    $!N
' data.txt
Improve this answer
1
  • Great! I tested this and came out well as I needed. Thanks for the solution using sed command. Commented Aug 26, 2020 at 11:03
2

IMHO the simplest approach is just to read add.txt into an array and then print a line from the array every 4 lines of data.txt:

$ awk 'NR==FNR{a[NR]=$0; next} FNR%4==1{print a[++c]} 1' add.txt data.txt
2001-01-01 00:00:00 42 1
  -500      11.822788
  -400      12.006394
  -350      12.287062
  -300      12.793395
2001-01-02 00:00:00 42 1
  -500      11.823597
  -400      12.008012
  -350      12.287062
  -300      12.794204
2001-01-03 00:00:00 42 1
  -500      11.826023
  -400      12.011247
  -350      12.291915
  -300      12.800675
2001-01-04 00:00:00 42 1
  -500      11.827641
  -400      12.013674
  -350      12.295959
  -300      12.805528
2001-01-05 00:00:00 42 1
  -500      11.830067
  -400      12.016100
  -350      12.300003
  -300      12.811998
Improve this answer
5
  • You might have to contend with using quite a bit more memory than in Kusalananda's solution in case file add.txt is very large. Kind of a corner case though. ;-) Commented Aug 26, 2020 at 6:14
  • 1
    @Ed Morton your solution is also working smoothly. Many thanks. Commented Aug 26, 2020 at 10:46
  • @Cbhihe Thanks for pointing that out. I am asking this for a clarity: So, you are saying that the solution given by Kusalananda and Ed Morton will not work with very large add.txt?Then what will be the solution that will even work for larger add.txt? Because in near-future I will work with larger add.txt kind-of files. Commented Aug 26, 2020 at 10:49
  • 1
    @linux_lover: I never said or meant to say that kusalananda's or, for that matter, EdMorton's solution would not work for large files. I only do say that the latter would need more memory for large add.txt files (because in EdMorton's solution the whole file os parsed into his array). For that to be an issue you would probably have to run up add.txt file in the 10s of GB. So i do not believe it is relevant. In a nutshell, both solutions are fine and Kulasananda is likely to use less memory when executing. Is all. Commented Aug 26, 2020 at 12:48
  • linux_lover you asked for the simplest solution, not the most robust, or portable or efficient or anything else solution - careful what you wish for! When you talk about a "large" file in the comments above, "large" isn't a problem, "absolutely massive" would be a problem. As @Cbhihe says it'd have to be 10s of gigabytes before it'd be a problem, if your file is just a few million lines then it'd work just fine. There are other differences between how the scripts would handle cases such as add.txt not having as many lines as 1/4 the lines in data.txt and there would be efficiency differences. Commented Aug 26, 2020 at 14:16

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.