#microformats 2014-11-25

2014-11-25 UTC
behind_you, caseorganic, KartikPrabhu and netweb joined the channel
#
KartikPrabhu
kylewm: regarding: https://github.com/kartikprabhu/mf2py/issues/47 which is the troublesome property?
#
Loqi
KartikPrabhu: tantek left you a message 5 days ago: I just added <area> element support for microformats objects to the microformats2 parsing spec, per the concrete use-case of photo-area-tag(ging) markup in posts/replies as discussed in #indiewebcamp on 2014-11-19. Please review: http://microformats.org/wiki/index.php?title=microformats2-parsing&diff=64684&oldid=64589&rcid=101047
#
KartikPrabhu
kylewm: it looks fine to me on first look
tantek and KartikPrabhu joined the channel
#
tantek
checks the logs
#
tantek
how's it going KartikPrabhu ?
#
KartikPrabhu
quite busy and well :)
#
KartikPrabhu
reviweing the <area> change
#
KartikPrabhu
if i recall there was some parsing problems that ben_thatmustbeme mentioned
caseorganic joined the channel
#
tantek
yes he created a test case that revealed one old bug with the parser, and revealed one spec omission, which I fixed.
#
KartikPrabhu
ok cool. will add to mf2py unless someone has already done so
Rastus_Vernon, ben_thatmust and caseorganic joined the channel
#
kylewm
KartikPrabhu: [content][html] is the problematic one
#
Loqi
kylewm: tantek left you a message 5 days, 1 hour ago: I just added <area> element support for microformats objects to the microformats2 parsing spec, per the concrete use-case of photo-area-tag(ging) markup in posts/replies as discussed in #indiewebcamp on 2014-11-19. Please review: http://microformats.org/wiki/index.php?title=microformats2-parsing&diff=64684&oldid=64589&rcid=101047
#
KartikPrabhu
kylewm: what is the problem? reviewing currently
#
kylewm
.repost-entry is included in the content even though it is part of a HTML comment
#
kylewm
</div> <!-- .repost-body -->\n'
#
kylewm
</div> .repost-entry \n'
#
KartikPrabhu
aah I see. the last line is a comment that for some reason is being picked up
#
kylewm
KartikPrabhu: I'll take a look at the code, unless you want to
#
KartikPrabhu
go for it if you are on it. seems like some bs4 thing again
#
KartikPrabhu
kylewm: this is the line that is barfing: https://github.com/kartikprabhu/mf2py/blob/master/mf2py/parse_property.py#L178 bs4 seems to treat comments as child elements and so that line takes their text value and joins it
#
kylewm
oh my
#
kylewm
makes sense
#
KartikPrabhu
will have to dig through bs4 again <sigh>
#
tantek
what is bs4?
#
KartikPrabhu
your tanteking does not work here!
#
KartikPrabhu
beautiful soup is a python based wrapper for HTML parsers
#
KartikPrabhu
kylewm: bleh... i don't have mf2py on my new machine. maybe doing a direct change as : 'html': text_type(el) works?
#
KartikPrabhu
if not i'll look into it after dinner
#
kylewm
good detective work, KartikPrabhu, amazed you found it that quickly
#
kylewm
KartikPrabhu++
#
Loqi
KartikPrabhu has 79 karma
#
KartikPrabhu
huh these channels have combined karma! w00t!
#
Loqi
giggles
caseorganic and KartikPrabhu joined the channel
#
KartikPrabhu
kylewm: ok that does not work! :(
#
KartikPrabhu
using unicode(element) in bs4 also includes the outerHTML as <div class="e-content"> which is why I suppose I used that other stranger method
#
kylewm
KartikPrabhu: decode_contents seems to work, based on http://stackoverflow.com/a/18602241/682648
#
KartikPrabhu
yeah just found it does work. will test on others and push
#
kylewm
formatter="html" will convert unicode characters to html entities
#
kylewm
not sure if that's what we want or not
#
KartikPrabhu
yeah stuck to default formatter
#
KartikPrabhu
pushed to my fork also updated it from tommorris
#
KartikPrabhu
kylewm: could you do that travis thing?
#
kylewm
BeauitulfSoup is failing to install in Python 3.3
#
KartikPrabhu
judas priest
#
KartikPrabhu
this is new isn't it?
#
kylewm
yeah, know why Flask is a requirement of mf2py now?
#
kylewm
that seems odd
#
KartikPrabhu
that is tommorris setting up the heroku mf2py server
KartikPrabhu joined the channel
#
KartikPrabhu
kylewm: tommorris: maybe we should make a mf2py organisation on github have all the separate pieces there
#
KartikPrabhu
!tell tommorris: how about separating the Flask app from the mf2py core? seems overkill to have Flask as a mf2py requirement
#
Loqi
Ok, I'll tell them that when I see them next
#
kylewm
KartikPrabhu: tommorris: I don't think that's really necessary, requirements.txt is only for other developers and Heroku, setup.py is what actually determines the requirements in PyPI
#
KartikPrabhu
aah ok makes sense
#
KartikPrabhu
!tell tommorris nevermind previous tell. Sent PR to mf2py with a bug fix for e-* > html
#
Loqi
Ok, I'll tell them that when I see them next
#
KartikPrabhu
kylewm: any idea about he bs4 py3.3 conflict?
#
kylewm
KartikPrabhu: the last line of requirements.txt is mf2py
#
kylewm
it's installing an old version of itself as a dependency...
#
kylewm
and the old version depends on BS3
#
KartikPrabhu
cool. hopefully tommorris fixes that with my PR
#
KartikPrabhu
i have no real idea how this requirements stuff works :)
caseorganic joined the channel
#
kylewm
KartikPrabhu: you know you can run Travis builds on your own repo if you want right?
#
KartikPrabhu
yeah... but then I have to figure out how to do it :P
#
KartikPrabhu
that's why I'm... err... outsourcing it to you ;)
#
kylewm
haha that's 100% fine with me :)
caseorganic, KartikPrabhu and ChiefRA joined the channel
#
ChiefRA
tommorris any news related to hListing?
#
KartikPrabhu
ChiefRA: is hListing used a lot? I couldn't find a mf2 equivalent
#
KartikPrabhu
for instance hListing has no direct equivalent here: http://microformats.org/wiki/microformats2#v2_vocabularies
caseorganic, chiui, krendil and Erkan_Yilmaz joined the channel
#
tommorris
ChiefRA: haven't had a chance to look at it yet.
#
Loqi
tommorris: KartikPrabhu left you a message 3 hours, 55 minutes ago: how about separating the Flask app from the mf2py core? seems overkill to have Flask as a mf2py requirement
#
Loqi
tommorris: KartikPrabhu left you a message 3 hours, 48 minutes ago: nevermind previous tell. Sent PR to mf2py with a bug fix for e-* > html
#
tommorris
kylewm: yeah, that should be on a separate branch. ;)
#
tommorris
KartikPrabhu: ^^
#
KartikPrabhu
yes tommorris :) nice work on adding <area> stuff btw
#
tommorris
I was going to try and tidy up all the loose ends and release a new version to PyPI the other day. maybe i'll do that today. :)
#
tommorris
(the lack of internet access at the boyfriend's place was an issue to the previous attempt to release. heh.)
#
KartikPrabhu
ha! of course! but that gave us time to make new patches ;)
KartikPrabhu and caseorganic joined the channel
#
ChiefRA
KartikPrabhu the hListing is used, and we are in the process to create the mf2 version
#
ChiefRA
stay tunned :)
#
KartikPrabhu
ok cool! a good use case for hListing will prompt a mf2 upgrade and we can add it to mf2py
#
ChiefRA
KartikPrabhu that's what I was talking to tommorris, to create a mf2 parser for it and start using it.
#
KartikPrabhu
yup! if there a good candidate mf2 spec for it it cane be added to the backwards compatible feature of mf2py :)
eschnou, alanpearce, caseorganic, KartikPrabhu, adactio, TallTed, Musk, tantek, gRegor`, mko, chiui, kez, globbot, krendil, elux, caseorga_ and elux_ joined the channel