nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

reading a folder & sorting

nipper

OK, building a press release index page. Customer wants it sorted by release date (though modification date by DCR may suffice).

Is there a way to order readdir ?

My current thought is to read each DCR, grab the release date from inside the DCR, drop it in a hash (with DCR name), sort the dates and process.

That being said is there an easier way to read a folder pre-sorted say by modification date ?

Find more posts tagged with

Comments

ISCBorisB

OK, building a press release index page. Customer wants it sorted by release date (though modification date by DCR may suffice).

Is there a way to order readdir ?

My current thought is to read each DCR, grab the release date from inside the DCR, drop it in a hash (with DCR name), sort the dates and process.

That being said is there an easier way to read a folder pre-sorted say by modification date ?

What platform? Often you can play with OS Commands to achieve that but it's not portable.
In windows for example (like in [post=100513]this[/post] one) /OD will give you "Oldest first" and /O-D "Oldest Last".

I do not beleive you can control readdir order directly, it depends of the File creation order

nipper

Cool, thanks Bo. I will try dir /A-D /B /O-D /TW

Now, if someone updates the perss release without changing the date/time then it will be out of order. So I am suggesting to the customer that I set it automatically onSave.

Andy

nipper

FYI, if you do this (and I like the idea) you need to change / back to \\

$mydir =~ tr|/|\\|;

my @files = map {chomp $_; $_;} qx(dir /B /A-D /ON $mydir);

and also make certain you do not cygwin loaded (with their own DIR command)

I had to rename to get my script to work

moonkhanna

Hi Andy,

another way would be to dirwalk the directories involved. Get Dates as per the real world logic and put into a hash.. using yyyymmdd_uniqueIdentifier format... Then sort the hash..

I would prefer this approach rather than depending on OS commands as it will offer much more flexibility in changing and maintaining business logic

Adam Stoller

OK, building a press release index page. Customer wants it sorted by release date (though modification date by DCR may suffice).

Is there a way to order readdir ?

My current thought is to read each DCR, grab the release date from inside the DCR, drop it in a hash (with DCR name), sort the dates and process.

That being said is there an easier way to read a folder pre-sorted say by modification date ?

Are you parsing the DCRs for any other information? Like title, author, etc.?

If so - then the hash idea is reasonable because you're going to have to pay the price for reading in and parsing each file anyway - so there's no need to add to the price by running a system, rather than builtin, command (as was suggested) -- and this way if someone modifies something to fix a typo or some such, it remains sorted by its specified release date rather than by the modification timestamp on the file.

If possible - I recommend getting the dates in yyyymmdd [HHMM[SS]] format so that you can sort them trivially both lexically and chronologically.

Depending on what you need, you could do this as hash of array references, or a hash of hash references. E.g.:#!/usr/bin/perl

my %hash1 = (
    20080101 => ["Harpo", "Chico", "Groucho", "Zeppo", "Gummo"],
    20080214 => ["Sleepy", "Doc", "Grumpy", "Bashful"],
    20080315 => ["Sneezy", "Happy", "Dopey"],
);

foreach my $k (sort keys %hash1) {
    print "$k\n\t" . join("\n\t", sort @{$hash1{$k}}) . "\n";
}

my %hash2 = (
    20080101 => [
        {title => "Hippy dippy dee", author => "Some One"},
        {title => "An apple climbed a tree", author => "Somebody Else"},
    ],
    20080214 => [
        {title => "He lost his pants", author => "Another Person"},
        {title => "And fell in a trance", author => "Some One"},
    ],
    20080315 => [
        {title => "Hippy dippy dee - whoops", author => "Guess Who"},
    ],
);

foreach my $k (sort keys %hash2) {
    print "$k\n";
    foreach my $t (sort {$a->{title} cmp $b->{title}} @{$hash2{$k}}) {
        print "\t$t->{title} ($t->{author})\n";
    }
}

[* The little ditty in hash2 was written by my brother about 40 years ago, when he was stuck in bed due to food poisoning - go figure :-)]

ISCBorisB

Hi Andy,

another way would be to dirwalk the directories involved. Get Dates as per the real world logic and put into a hash.. using yyyymmdd_uniqueIdentifier format... Then sort the hash..

I would prefer this approach rather than depending on OS commands as it will offer much more flexibility in changing and maintaining business logic

It's a reply like this one: "I can read to hash, sort and process. Is there another way?" - "Sure is! Read to hash, sort and process".
Why didn't you read the thread before chiming in? Here is the excerpt from the very first one:

...read each DCR, grab the release date[...], drop it in a hash[...] sort and process.
Is there a way to order readdir? is there an easier way to read a folder pre-sorted?

Do you have something to contribute to what has been really asked?

nipper

Are you parsing the DCRs for any other information? Like title, author, etc.?

If so - then the hash idea is reasonable because you're going to have to pay the price for reading in and parsing each file anyway - so there's no need to add to the price by running a system, rather than builtin, command (as was suggested) -- and this way if someone modifies something to fix a typo or some such, it remains sorted by its specified release date rather than by the modification timestamp on the file.

Your brother needs a life. Must run in the family.

Yes I am using iw_load_dcr later down the road. My current approach is

@files = map {dir...}

foreach @files
iw_load_dcr
build the HTML for that

You are suggesting I read & parse all DCRs into a hash, sort by date, and generate the HTML. Seems faster to run the TPL but certainly more complicated. I use the iwpt tags instead of creating a Perl TPL.

I will keep it as such initially & see if I really need to change.

Thanks

Andy

Adam Stoller

Your brother needs a life. Must run in the family.

Yes - it does run in the family (you should hear his version of "Home on the range" entitled "Home for the deranged" :-)

Yes I am using iw_load_dcr later down the road. My current approach is

@files = map {dir...}

foreach @files
iw_load_dcr
build the HTML for that

You are suggesting I read & parse all DCRs into a hash, sort by date, and generate the HTML. Seems faster to run the TPL but certainly more complicated. I use the iwpt tags instead of creating a Perl TPL.

I will keep it as such initially & see if I really need to change.

Thanks

Andy

You should be able to mesh the two approaches together - loop through the iwpt_load_dcr - getting the values - but instead of outputting them when you get them - load them into the hash as I suggested - then, loop through the hash (sorting the contents) and output the data the way you want it.

In cases like this (for me, almost any case) I think the Perl PT approach is called for - the XML tags are too limiting. How you'd do this in XSL I'll leave to someone else to figure out ...

moonkhanna

It's a reply like this one: "I can read to hash, sort and process. Is there another way?" - "Sure is! Read to hash, sort and process".
Why didn't you read the thread before chiming in? Here is the excerpt from the very first one:
Do you have something to contribute to what has been really asked?

Hi Bo,

Weirdo response.. You need a nice day ..