I had to stop submitting ISC SANS posts because they similarly have a canonical link in their head which HN, for reasons I'm not sure about, use to 'correct' the URL, errantly.
Funny how exploiting uninitialized variables is evil in (nonstrict) JavaScript but good in Awk. I agree that’s true, mind you, I just can’t pinpoint what about the languages’ designs makes it work in one case but not the other.
Doesn’t awk set null for all uninitialized variables. Does that solve the problem for you? We can use that feature to find lines with same values which form a group.
Guaranteed initialisation rules I would imagine. A lot of languages don’t do this, or assign null or something. For a language with a narrow scope of application you can this but in more general purpose languages it makes less sense.
Note Awk does have a special value for uninitialized variables—its distinguishing feature is that it compares equal to both zero and empty string simultaneously. (No official way to spell that value, but you can spell it e.g. “undefined” if you never initialize a variable called “undefined”.) It’s really really close to old JS in that respect, and no wonder, given that Awk was an explicit inspiration for JS. But somehow it works for Awk and not for JS, even at the hundred-line scale that’s common and reasonable for both languages.
I love awk as a language / framework. If it got an uplift to make it useful for more complex problems it would be an absolute winner for lots of basic data processing tasks.
I’ve often written incantations many lines long and even broken out to actually writing awk scripts from time to time.
Once the penny drops with it it’s great fun but it’s absolutely useless once your problems get to any degree of sophistication.
I typically move into Python at this stage. Perl and Ruby are probably a more elegant fit here but those aren’t rows I want to have.
In this day and age, awk really needs CSV (RFC-4180) support and better semantic scoping and library support.
I’d also think it would be neat as an embedded language for various data processing platforms but if we haven’t seen it yet I doubt we will ever see it.
EDIT support for file formats beyond plain text would also be a winner.
I wanted to like awk and tried really hard but in the end was disappointed by what I see as unnecessary complications or limitations in the language. For example, it has first class support for regexes but only for matching. You can’t do ‘s/foo/bar’. I also found string manipulation to be cumbersome with the string functions. I would have expected a string processing language to have better primitives for this. And function arguments/variables are just a mess, it’s hard to understand how they came up with that design. It’s also quirky and unintuitive in some places you would not expect. Take the non-working example from the article:
awk -v FS=';' -v OFS=',' 1
I expect this to change the change the separator in the output. Period. The “efficiency” argument for why it doesn’t work just doesn’t cut it for me. First, it’s very simple to do a one time comparison of FS and OFS, if they are different then you know you know you _have_ to perform the change, because the user is asking you! If I do this in reality and it doesn’t work I just switch over to sed or perl and call it a day.
All in all, perl -eP is a better awk. And for data processing I switched to miller. It has it’s idiosyncrasies as well but it’s much better for working with structured records.
Seems like the link should be https://backreference.org/2010/02/10/idiomatic-awk/ instead of https://backreference.org/index.html
Also, (2010).
Yes - but I am sure I sent the first link.
I'm not sure what has happened then.
I think that it is this in the head:
I had to stop submitting ISC SANS posts because they similarly have a canonical link in their head which HN, for reasons I'm not sure about, use to 'correct' the URL, errantly.Technically an empty awk program is an implementation of cat(1) (before it came back from berkeley waving flags, anyway).
Of course no awk will run an empty program, so the article's '1' or '"a"' or other truthy value is required.
Funny how exploiting uninitialized variables is evil in (nonstrict) JavaScript but good in Awk. I agree that’s true, mind you, I just can’t pinpoint what about the languages’ designs makes it work in one case but not the other.
Not sure if it was intended, but it helps to keep CLI one-liners concise. For example:
Doesn’t awk set null for all uninitialized variables. Does that solve the problem for you? We can use that feature to find lines with same values which form a group.
Guaranteed initialisation rules I would imagine. A lot of languages don’t do this, or assign null or something. For a language with a narrow scope of application you can this but in more general purpose languages it makes less sense.
Note Awk does have a special value for uninitialized variables—its distinguishing feature is that it compares equal to both zero and empty string simultaneously. (No official way to spell that value, but you can spell it e.g. “undefined” if you never initialize a variable called “undefined”.) It’s really really close to old JS in that respect, and no wonder, given that Awk was an explicit inspiration for JS. But somehow it works for Awk and not for JS, even at the hundred-line scale that’s common and reasonable for both languages.
I never considered that awk was a precursor to JavaScript but I can certainly see that
I love awk as a language / framework. If it got an uplift to make it useful for more complex problems it would be an absolute winner for lots of basic data processing tasks.
I’ve often written incantations many lines long and even broken out to actually writing awk scripts from time to time.
Once the penny drops with it it’s great fun but it’s absolutely useless once your problems get to any degree of sophistication.
I typically move into Python at this stage. Perl and Ruby are probably a more elegant fit here but those aren’t rows I want to have.
In this day and age, awk really needs CSV (RFC-4180) support and better semantic scoping and library support.
I’d also think it would be neat as an embedded language for various data processing platforms but if we haven’t seen it yet I doubt we will ever see it.
EDIT support for file formats beyond plain text would also be a winner.
Good news! Not only are there AWK implementations (for instance GoAWK) that can use CSV, but even the one true AWK does do CSV and Unicode nowadays!
https://arstechnica.com/gadgets/2022/08/unix-legend-who-owes...
I wanted to like awk and tried really hard but in the end was disappointed by what I see as unnecessary complications or limitations in the language. For example, it has first class support for regexes but only for matching. You can’t do ‘s/foo/bar’. I also found string manipulation to be cumbersome with the string functions. I would have expected a string processing language to have better primitives for this. And function arguments/variables are just a mess, it’s hard to understand how they came up with that design. It’s also quirky and unintuitive in some places you would not expect. Take the non-working example from the article:
I expect this to change the change the separator in the output. Period. The “efficiency” argument for why it doesn’t work just doesn’t cut it for me. First, it’s very simple to do a one time comparison of FS and OFS, if they are different then you know you know you _have_ to perform the change, because the user is asking you! If I do this in reality and it doesn’t work I just switch over to sed or perl and call it a day.All in all, perl -eP is a better awk. And for data processing I switched to miller. It has it’s idiosyncrasies as well but it’s much better for working with structured records.
Awk does support that kind of pattern matching for replacement...?