The Loop

kaiser

00:03

@Rarst You got a job now? Where?

Matt D

02:34

sup

question

i had an upgrade forced on me (yay) and now, ?page=1 requests get transformed into /1/ ...

how the hell do i stop that from happening

GhostToast

sounds like you need to go to Settings->Permalinks and change it back to post_id or whatever

Matt D

well, the permalinks have always been set to post_name

even on UAT

it appears the behaviour has changed somewhere

GhostToast

plugin? still, revising permalinks page will often rebuild htaccess

unless something has happened which prevents htaccess from working or being writeable

Matt D

hmm

ill chat to the sysadmin

GhostToast

can you view/edit .htaccess?

Matt D

02:41

not atm

not on production

my money is still on the upgrade introducing something in wp that's changed the URL buildng.

GhostToast

i've always used /post-name/ so couldn't say - only thing i can think of is maybe the "default" being changed to pretty urls, but haven't heard of it

Matt D

the double whammy of having a post type called "page", and it being used in pagination

GhostToast

hehe. yea there are some post_type names to avoid

Matt D

didnt seem to be a problem on the previous version

i wasnt informed of the upgrade until the eventual "hey this doesnt work" email

sigh

GhostToast

saw you are in brisbane, you ever hear of Sennza

Matt D

02:45

nope

im in a digital agency, for the next few weeks at least. but never heard of them.

alright. lunchtime :) i've recommended we just rename the post variable

might be a problem on other sites too

thanks for the help

GhostToast

np goodluck

Sisir

07:21

is there any way to choose what columns WP should select when using WP_Query? say I only want ID and post_title not every column values..

Jared

Sisir check this page: codex.wordpress.org/Class_Reference/…

Sisir

i see.. I wish there was a filter for that..

Jared

Yeah, looks like the only other option is to return just the post IDs. You could then use get_the_title on each ID.

Rarst

08:31

picked up some WooCommerce work, #loopchat will disown me...

Tweeted by Rarst on February 4, 2014 at 11:11 AM

userabuser

@Rarst get out, and you have nerve to show your face here?

Rarst

well... I do own the room :)

bungeshea

@Rarst was that a threat?

Rarst

yes?

userabuser

WooRarst...

7

Rarst

08:42

:) I need my tea before @kaiser wakes up and sees that...

userabuser

@Rarst if you have any questions about WooCommerce plugin, ask them on main site.

Rarst

naturally, where else?

WooCommerce is now first priority in our scope

userabuser

@Rarst I want to devote all of my time helping their customers for free while they make a lot of money.

Rarst

08:58

that's the spirit

kaiser

10:46

@Rarst ...

Rarst

don't look at me like that! something got to pay for my WordCamp addiction!

kaiser

sounds more like a disease now

Anyway, we always were lacking an expert in Woo. Finally we got one and it's Rarst! (who can takeover whole woocommerce queue).

Rarst

nope

kaiser

11:02

If you stay with nope, I shall open a Meta question polling if people want you to adopt that tag.

kaiser

11:55

user187291

26.7k 4 43 85

I wouldn't have thought that someone with a generic user name, no avatar or profile information would write close to 800 answers.

userabuser

@kaiser someone who truly cares not for recognition...

Sven

Seems like the account is inactive :)

Sisir

@kaiser he must be an undercover ninja!

how do implement conception of A ∪ B using php array?

nvm

userabuser

@Sisir what you trying to do?

Sisir

12:12

I was trying to find union of two array values. Sounds like array_unique(array_merge($arr1, $arr2)) does the job

userabuser

is that bigcup? or union?

sorry posted msg timed out

kaiser

16:17

Sotchi Tweets from journalists. :D

Wyck

17:04

I would have stayed on one of the many cruise boats in the port

at least you know they work

Michael Ecklund

I'm a nub at regex, Is there a way to strip all HTML meta tags, in combination with this: preg_replace('/^<!DOCTYPE.+?>/' in the same line?

Wyck

@MichaelEcklund try an online regex debugger like regex101.com or debuggex.com

I also use this chrome.google.com/webstore/detail/regexp-tester-app/…

Michael Ecklund

Thanks for the links

toscho

@MichaelEcklund try '~<!DOCTYPE.+?>|<meta[^>]*>~'

Michael Ecklund

17:23

@toscho Perfect. Did the trick.

Thanks

Now I can return my HTML from DOM as a fragment. As seen here but he didn't remove <head></head> and <meta.

    $html = preg_replace(
        '~<!DOCTYPE.+?>|<meta[^>]*>~',
        '',
        str_replace(
            array(
                '<html>',
                '</html>',
                '<body>',
                '</body>',
                '<head>',
                '</head>'
            ),
            array('', '', '', '', '', ''),
            $dom->saveHTML()
        )
    );

toscho

uhm, that’s not ideal

$html = stripslashes( $post_data[ 'post_content' ] );
$doc  = new DOMDocument( '1.0', 'utf-8' );
$doc->loadHTML(
	// encoding to enforce UTF-8, the table to prevent additional markup
	'<?xml encoding="UTF-8"><table><tr><td>'
	. $html
	. '</td></tr></table>'
);
$xpath    = new DOMXPath( $doc );
// change the content, then:

$body = $doc->saveHTML( $doc->getElementsByTagName( 'td' )->item( 0 ) );

Michael Ecklund

18:09

@toscho I haven't learned xpath yet

Rarst

screw xpath so much. :) DomCrawler ftw

toscho

me too. didn’t stop me.

Michael Ecklund

I spent all day yesterday figuring out how to manipulate my HTML using nodes. It would be super easy to just manipulate and instantly output, but I'm manipulating, and storing so that it can possibly be manipulated again before finally ready for output.

Symfony has only been around since '05? Took me until '13 to even hear about it.

Is this a decent beginners scraper class?

Rarst

DomCrawler

wasn't that big enough hint? :)

Michael Ecklund

lol

Rarst

18:18

DomCrawler

Michael Ecklund

Everything on Symfony just looks super complex

Rarst

you are looking at it wrong. don't go looking at framework, go look at individual components you need.

Michael Ecklund

Yeah but don't you needs to know the basics about that framework in order to use it's components?

toscho

Symfony is more a library than a framework

Rarst

no, no, no. all components just work by themselves.

Michael Ecklund

18:22

DomCrawler Documentation ... 404

Rarst

something weird with link there symfony.com/doc/current/components/dom_crawler.html

Michael Ecklund

It looks pretty easy to use.

Rarst

yay, it's sinking in... :)

kaiser

@MichaelEcklund symfony.com/doc/master/components/dom_crawler.html

Rarst

@kaiser 2.0 is old

better :)

kaiser

18:36

already corrected :)

boscho

18:50

What shell do you Windows devs use for git and other CLI tools? CMD, Powershell, Cygwin shells, etc?

Mostly out of curiosity, but also because I'm done with writing whack-$%* batch scripts.

kaiser

Git/GNU Bash

Rarst

cmd

kaiser

Powershell isn't too bad either as it can be turned to fullscreen has auto horizontal scrolling, etc.

But most bash scripts don't work there, font/readability sucks, etc.

cmd is for ~~DOS nerds~~ Rarst

boscho

@kaiser I just discovered the Powershell ISE - Microsoft apparently finally decided that usability isn't a passing fad. I used to be all for cmd, but lately I've been running into all sorts of weird crap with my $PATH. Sometimes I'd have access to git and other tools, sometimes I inexplicably wouldn't.

kaiser

well. my problem currently seems to narrow down to node.js having a wrong Reg entry. Reinstall didn't fix that. Grunt unable to find Gruntfile.js in a folder and bower doesn't recognize git/has ERNOGIT. Powershell makes bower magically work, but the problem with grunt remains.

Wyck

19:01

I use mobaxterm on win7 but it sucks on win8

kaiser

don't remember why I dropped mobaxterm on win7...

maybe because the name sucks :D

Wyck

lol

kaiser

node.js CLI is ok too. same goes for CLI that ships with ruby.

boscho

@Wyck I used mobaxterm for a while, but it felt too heavy for a term-emu to me. Might have to revisit it, though. Currently using ConEmu as my actual interface.

kaiser

ah yeah, there's one thing that sucks with Node.js/Ruby/Git bash/CLI: You have a very limited history. So if you're dumping/logging your commits for e.g. you mostly don't see the whole history.

boscho

19:06

@kaiser I ran into a bunch of nonsense like that, and ended up pre-loading a batch script into every cmd instance via some reg-key in order to actually properly set up my $PATH and all of my aliases, initialize utilities, etc. But the batch has become unmanageable.

Wyck

ya I cant use it on this machine, it's annoying, I use powershell when I have too

boscho

@kaiser not a fan of the PHP CLI though?

kaiser

php cli...?

boscho

on second though, not the same(?). Never actually played with writing too many CLI PHP scripts - not all that sure about the capabilities. I may be confused

Rarst

PHP scripts run in normal console. there is whacky console interpreter in PHP but no one really uses it

kaiser

19:12

My ~/Users/my-name/.bashrc that makes the CLI suck less.

@Rarst wasn't there even a php.net site for that one? need to look that up.

Rarst

that one what?

kaiser

php CLI somewhat something

nope. confused that with PHP built in server that I just found out about.

PHP 5.4 has a built in local server: php -S localhost:8000 - Details: http://www.php.net/manual/en/features.commandline.webserver.php

Tweeted by unserkaiser on January 31, 2014 at 10:37 PM

Rarst

PHP: Interactive shell

kaiser

aaaaand

Windows and PATH... POWERSHELL! for the rescue. Still some npm stuff doesn't recognize git or main files. Aaargh! 2 days lost!

Tweeted by unserkaiser on January 31, 2014 at 6:55 PM

@Rarst ah, that one. I wouldn't name it CLI, but still :)

boscho

@Rarst ah-hah! I knew I wasn't (entirely) crazy!

Michael Ecklund

22:44

Anyone have trouble updating multi-site?

The initial upgrade to 3.8.1 was fine, but then when I click "Upgrade Network" ... my connection eventually just times out.

Never used to do that.

"Your server may not be able to connect to sites running on it. Error message: Operation timed out after 120000 milliseconds with 0 bytes received"

Transcript for