/dev/chat

1:21 PM

```
mc@mc:~/practice/data$ awk -F'[< ]' '/<td class="headerCovTableEntryLo">/{print $3; exit}' input2
class="headerCovTableEntryLo">39.2
mc@mc:~/practice/data$ awk -F'[> ]' '/<td class="headerCovTableEntryLo">/{print $3; exit}' input2
39.2
```

@terdon This is I get from these commmads.

@PrabhjotSingh Ah, good, I was about to ask you to come into chat

@PrabhjotSingh AAAAAAaaaaaaaaaaaaaa! I see what you mean. Yes, I had a typo there (I had originally used -F'[<> ]' and then manually edited it and removed the > instead of removing the <).

I'm sorry, I was confused because you had suggested using just -F'>' which would not work, I hadn't realized you were telling me to use -F'[> ]' (so the '<' and a space)

Thanks!

@terdon I would be thankful to you if you tell me what '( )' exactly does. as in this answer.

@PrabhjotSingh I cannot, I'm afraid. The first time I saw that was in that exact answer, so I don't know anything more than you do. I was also quite intrigued, and I confirmed it works (and so does -F'[ ]') but I don't know exactly how or why.

We should ask @αғsнιη

@terdon Thanks I will ask. And thanks to you for help and responding.

Hey, I should be thanking you! You're the one who pointed out my error :) So thanks!

2:36 PM

@PrabhjotSingh that sets the awk field separator to single space character.

which default is one-or-more of Space/Tabs

-F'[ ]' == -F'( )' == -F' | ' or anything else that can end up field separator to a single space

when setting like that (regex mode), anything between two spaces act as a separate field by itself

see this example

echo 'a b  c  d    e' |awk -F'[ ]' '{ while(++i<=NF) print "[" $i "]"; gsub(FS, "[]") }1'
[a]
[b]
[]
[c]
[]
[d]
[]
[]
[]
[e]
a[]b[][]c[][]d[][][][]e

better

echo 'a b  c  d    e' |awk -F'[ ]' '{ while(++i<=NF) printf "[" $i "]" }'
[a][b][][c][][d][][][][e]

so by this way consecutive spaces preserve intact when awk wants to rebuild the $0 as when we update the content of a field as I did in my that answer

@αғsнιη But the cool thing is that doing it that way preserves the original separators, right? That's what I don't quite understand. I can see it works:

$ printf 'f1 f2   f3       f4\n' | awk '{$1="";print}'
 f2 f3 f4
$ printf 'f1 f2   f3       f4\n' | awk -F'( )' '{$1="";print}'
 f2   f3       f4

yes

That's the part I don't get. Why would that preserve them?

Ah! You are saying that consecutive FS characters are not joined together!

empty string between FS is also counts as a field

gotcha

$ printf 'f1 f2   f3       f4\n' | awk -F'( )' '{$1="";print NF}'
12
$ printf 'f1 f2   f3       f4\n' | awk '{$1="";print NF}'
4

A: awk: Preserve formatting of output when substituting field

2:50 PM

so an FS must be add back around empty string when rebuilding $0

and that finally lets line keep all intact

I mean rebuild based on the OFS, so there we have some empty string fields, so awk adds OFS for them too

BTW that is not the only answer I used that way :D

1

another trick is to force awk's field separator to use "single space" as field separator by defining it as rexed like -F'( )', and perform the modification on Use% column counting down from the end of the line to $(NF-1), since last two columns always separated with single space, so it's easy to...

and some others, that I couldn't find now : )

3:12 PM

Thanks @αғsнιη for explaining about beauty output.