#microformats 2018-05-26
2018-05-26 UTC
[cleverdevil], [jgmac1106], [tantek], [Natris1979], tantek_, toomim, [dougbeal], tantek, barpthewire, KartikPrabhu, [chrisaldrich], tae and [quinnvinlove] joined the channel
# 
@jgmac1106 Okay @cogdog used this: https://html5up.net/ to build Dimensions. Pretty neat. 100% incompatible with microformats2 plug-in, but it is built using a http://skel.io framework so yeah for CSS Grid. Fun digging through this stuff. (https://jgregorymcverry.com/2911-2/) (twitter.com/_/status/1000356522323599360)
# @jgmac1106 Okay so looking at @cogdog’s dimension a bit more it meant to be a business card, a portfolio directory. I think an h-card and h-entry are going to be easier to add. Going to turn my attention to Hueman after that. Seems to leave some microformats kruff… https://jgregorymcverry.com/2913-2/ (twitter.com/_/status/1000360710474366976)
[kevinmarks], [kim_landwehr], Kyle-K, barpthewire, jalcine, [miklb], [cleverdevil] and [grantcodes] joined the channel; Kyle-K left the channel
# 
KartikPrabhu !tell [kevinmarks] Is it possible that I get permissions to close and label issues on the mf2py main repo?
# KartikPrabhu Zegnat: do you actively use mf2py as your parser?
# KartikPrabhu ok
# KartikPrabhu Zegnat: here is an issue in any case https://github.com/microformats/mf2py/issues/104 not related to actual parsing but some mf2py quirk
# KartikPrabhu more like an mf2py quirk while implementing backcompat
# KartikPrabhu yeah I don't speak PHP
# KartikPrabhu but php-mf2 also does class subsitutions
# KartikPrabhu yeah that seems reasonable. I haven't run into this in practice either
# KartikPrabhu maybe file bug after testing :P
# KartikPrabhu haha
[chrisaldrich] joined the channel
# 
gRegorLove Yeah, php-mf2 does similarly: http://pin13.net/mf2/?id=20180526223157419 though it does it on the properties too, not just the hentry
# KartikPrabhu gRegorLove: This is not for the parsed content.html Does php-mf2 change the original "DOMDocument" that the users gives it?
# gRegorLove Ah, let me check.
# KartikPrabhu I have fixed the content.html thing for mf2py very recently (yesterday)
# KartikPrabhu so in mf2py running parse a second time on the same "DOM" object gives wrong results
# gRegorLove I'll have to check how you did that
# KartikPrabhu gRegorLove: https://github.com/kartikprabhu/mf2py/blob/experimental/mf2py/backcompat.py#L49 basically I attach a copy of the original element if the replaced class has any "e-*". Then in the "e-*" parsing it uses the original instead of the modified one ;)
# gRegorLove Confirmed if you pass in a DOMDocument to php-mf2 it's changed after parsing.
# KartikPrabhu ok. don't know if that is good or bad yet. In php-mf2 it might not be an issue since the whole DOM is changed to mf2
# gRegorLove Including the children, though, so a second pass shouldn't (I think) cause problems
# KartikPrabhu I might have preemptively overoptimised :|
# gRegorLove Sounds like php-mf2 should definitely copy the DOMDocument and work on that
# KartikPrabhu right if the backcompat rule that "ignore mf1 if mf2 is present" is used then it won't be an issue in second pass
# KartikPrabhu I am also in favour of not changing things the user passes in explicitly
# KartikPrabhu in mf2py that might need some funny code since BeautifulSoup does not support "deepcopy" i.e. making a separate duplicate
# gRegorLove Yeah, I'm wondering if PHP is similar. I wouldn't have expected it to change the input DOMDocument since it's not by reference, but maybe it's a special case
# KartikPrabhu I suspect it is by reference since it is some object or something
# KartikPrabhu that is what happens in python
# KartikPrabhu (sometimes)
# gRegorLove Yeah, looks like object references don't contain the object, just the identifer so accessors can find it.
# KartikPrabhu yeah. OOPs!
# gRegorLove Appears to be simple fix in PHP, `$doc = clone $input;`
[cleverdevil] joined the channel
# KartikPrabhu oh hey! I can make deepcopies of the whole doc too! just not of the elements :P
# gRegorLove Huh, interesting
# KartikPrabhu It depends on which objects have "deepcopy" methods or something
# gRegorLove For PHP it's a shallow copy apparently.
[grantcodes] joined the channel
# KartikPrabhu hmm deepcopy works in my console but not in mf2py tests! WTF
KartikPrabhu joined the channel