News for hackers. Stuff that matters.
Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by sigil on Friday January 05, 2018 @02:59PM (#16081173)
Equivalent sort(1) invocations for your examples:

    sort -k2 -k1 -k3 contacts.tsv
    sort -k1 -k2 -k3 contacts.tsv
This assumes TSV input, but there are plenty of reasons to prefer that to CSV. If I'm working from CSV sources I usually convert to TSV first thing in my shell pipeline.
Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by feelin_googley on Friday January 05, 2018 @08:19PM (#16083232)
When sort is used on really large files, it will automatically attempt to use disk, putting temp files in TMPDIR. This can be really slow.

To overcome the slowdown of disk I/O, perhaps a workaround could be to use mfs or tmpfs, maybe something like:

   mkdir /dir
   mount -t tmpfs tmpfs /dir
   TMPDIR=/dir sort -k2 -k1 -k3 contacts.tsv
   TMPDIR=/dir sort -k1 -k2 -k3 contacts.tsv
Personally, I gave up on sort for large files and use k/kdb+. I suspect it is faster for sorting than sort or the Go libraries, but I could be wrong.
Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by sigil on Saturday January 06, 2018 @01:50PM (#16086791)
For a dataset larger than physical memory, using a memory filesystem like tmpfs for the merge stage will either swap (|tmpfs| < |ram|) or deadlock (|tmpfs| >= |ram|).

Instead, your best bet in that case is to give sort as much physical memory as you can spare:

    sort -S 95% -k1 huge.tsv
Extra disk I/O is inevitable since your dataset doesn't fit in memory. At least during a merge sort your disk reads will be O(N) and sequentially ordered.

Note: in the special case that your dataset is slightly larger than physical memory, splitting it up in advance such that one of the `sort -m` input files lives on a tmpfs should indeed be faster.

Other things to check out if you need Very Fast Large Sorts:

- Use `sort --parallel=N` to use multiple cores. By default it only uses 1.

- Use `sort --batch-size=NMERGE` to increase the number of files merged at once. Otherwise you may be doing more mergesort stages than are necessary.

Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by johnweldon on Friday January 05, 2018 @03:02PM (#16081188)
Thanks - I've used sort quite a bit, and I like it. I wrote this partly to just fulfil my desire to sort by named fields rather than column indexes.
Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by sigil on Friday January 05, 2018 @03:13PM (#16081269)
I know the annoyance you're talking about, but I think you're better off wrapping sort(1) with something translates from column names to indexes. Among the reasons: sortcsv buffers all input into memory [1], while sort(1) uses a divide-and-conquer merge sort to avoid this.

[1] https://github.com/johnweldon/sortcsv/blob/55818bd8e5f9feecc...

Re:Show HN: Command Line Tool to Sort CSV and TSV Files by M... by johnweldon on Friday January 05, 2018 @03:16PM (#16081290)
That's a good point. I may consider a stand alone utility to wrap this[1] code that converts names to indexes.

[1] https://github.com/johnweldon/sortcsv/blob/master/main.go#L1...