Last active
March 11, 2016 00:40
-
-
Save bastosmichael/9db6e76bf14f108e2415 to your computer and use it in GitHub Desktop.
User information
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
facebook_ids = [ | |
"cirospizzeriabeerhousegaslamp", | |
"lavalenciahotel", | |
"tendergreensfood", | |
"craftpizzaco", | |
"tapiocaexpress", | |
"singlefinsd", | |
"thebakedbear", | |
"artofespresso", | |
"julianbakery", | |
"theduckdive", | |
"thefisherysd", | |
"westendpb", | |
"175450065799869", | |
"121921894540281", | |
"fatsalsdeli", | |
"146013036729", | |
"harryscoffeeshop", | |
"herringboneeats", | |
"aroicafe", | |
"thefrenchgourmet", | |
"goporkyland", | |
"carlsjr", | |
"sugarandscribebakery", | |
"pacificbeachsd", | |
"pacificbeachshoreclub", | |
"rimelsrestaurants", | |
"pbfishshop", | |
"seasons52", | |
"woodstockspb", | |
"crabcatcherrestaurant", | |
"theturquoisehome", | |
"lodgetorreypines", | |
"burgerlounge", | |
"cabocantinabrentwood", | |
"roysrestaurant", | |
"eatpuesto", | |
"thecottagelajolla", | |
"isabella.lovin", | |
"wahoosfishtaco", | |
"whisknladlerestaurant", | |
"chedithai", | |
"brocktonvilla", | |
"108276955864187", | |
"eddievsprimeseafood", | |
"manhattanoflajolla", | |
"dlushlife", | |
"biteofboston", | |
"thehake", | |
"georgesatthecove", | |
"kwikfill", | |
"thrusterslounge", | |
"45664163027", | |
"marineroom", | |
"sdsubookstore", | |
"followcpc", | |
"tacosurf", | |
"157031437646026", | |
"elixirespressobar", | |
"tower23", | |
"rubicondeli", | |
"schlotzskys", | |
"217993311551780", | |
"sunsettrocadero", | |
"tapenaderestaurant", | |
"carinositalian", | |
"boomerangsburgerjoint", | |
"birdrockcoffeeroasters", | |
"hiltonlajollatorreypines", | |
"theflamebroiler", | |
"bombaycompany", | |
"donovanssteakandchophouse", | |
"catamaranresort", | |
"flytpg", | |
"116208461740332", | |
"shakesmart", | |
"shakeawayland", | |
"noodlesandcompanyvalleywestmall", | |
"haagendazsus", | |
"wetzelspretzels", | |
"710beachclubpb", | |
"estancialajolla", | |
"brokenyolkcafe", | |
"firehousesd", | |
"ironpigalehouse", | |
"vacanteenservice", | |
"WFMlajolla", | |
"ljshoreshotel", | |
"perkybeanscoffee", | |
"glassdoorsd", | |
"doubletreedowntown", | |
"buddhagroove", | |
"teksalonandspa", | |
"heatherchastyles", | |
"sohofashioninc", | |
"platformsalon", | |
"sandiegostyleweddingsmagazine", | |
"styleseat", | |
"154255451263878", | |
"kutnbeautypicorivera", | |
"270510862980153", | |
"signaturehomestylesus", | |
"thestyleloungecarlsbad", | |
"127914767254061", | |
"homeadvisor", | |
"beauty.com", | |
"Hairsaloninmissionviejo" | |
].uniq | |
instagram_ids = [ | |
"barwestsd", | |
"lavalenciahotel", | |
"tendergreens", | |
"eatcraftpizza", | |
"singlefinsd", | |
"thebakedbear", | |
"julianbakery", | |
"theduckdive", | |
"westendpb", | |
"livingroomcafebyeplus", | |
"fatsalsdeli", | |
"harryscoffeeshop", | |
"herringboneeats", | |
"sugarandscribe", | |
"pacificbeachsandiego1969", | |
"pbshoreclub", | |
"rimelsrotisserie", | |
"seasons52", | |
"woodstockspizza", | |
"theturquoiseiris", | |
"lodgetorreypines", | |
"burgerlounge", | |
"cabocantina", | |
"eatpuesto", | |
"thecottagelajolla", | |
"isabella.sd", | |
"wahoosfishtaco", | |
"whisknladle", | |
"chedithaibistro", | |
"brocktonvilla", | |
"dirtybirdspb", | |
"bestwestern", | |
"eddievs_", | |
"dlushlife", | |
"thehakelj", | |
"treyfoshee", | |
"themeltingpot", | |
"crownpointcoffee", | |
"fredshb", | |
"prospectbar", | |
"elixirespressobar", | |
"jrdn.c", | |
"rubicondeli", | |
"schlotzskys", | |
"pbcantina", | |
"tapenade.restaurant", | |
"birdrockcoffeeroasters", | |
"theflamebroiler", | |
"catamaranresortsandiego", | |
"russoshouseofpizza", | |
"shakesmart", | |
"noodlescompany", | |
"haagendazs_us", | |
"710beachclub", | |
"firehousepb", | |
"ironpigalehouse", | |
"wfmlajolla", | |
"perkybeanscoffee", | |
"starbucks", | |
"teksalonandspa", | |
"heatherchastyles", | |
"shopsohostyle", | |
"bridgenott_stylist", | |
"kutnbeautypicorivera", | |
"luigis_barbers", | |
"hairaffairvh", | |
"stylesbarberspa", | |
"Instylebyingrid" | |
].uniq | |
handles = (facebook_ids.map {|id| facebook(id) }.flatten + instagram_ids.map {|id| instagram(id) }.flatten).compact.uniq | |
CanonicalUser.where(handle: handles).where('id_klout_handle IS NULL').pluck(:handle).with_progress.each {|handle| Crawler::Klout.perform_async handle } | |
CanonicalUser.where(handle: handles).where('id_klout_handle IS NULL').pluck(:handle).with_progress.each {|handle| begin test = Crawler::Klout.new; test.perform handle rescue ap handle end } | |
CanonicalUser.where(handle: handles).where('first_name IS NULL or last_name IS NULL').find_each.with_progress.each do |u| | |
begin | |
if object = FacebookUser.find_by(handle: u.handle) | |
u.first_name = object.first unless u.first_name | |
u.last_name = object.last unless u.last_name | |
u.middle_name = object.middle unless u.middle_name | |
u.id_fb_handle = object.handle unless u.id_fb_handle | |
u.save | |
end | |
rescue | |
ap u | |
end | |
end | |
CanonicalUser.where(handle: handles).where('first_name IS NULL or last_name IS NULL').find_each.with_progress.each do |u| | |
begin | |
if object = InstagramUser.find_by(handle: u.handle) | |
u.first_name = object.name.split(' ').first unless u.first_name | |
u.last_name = object.name.split(' ').last unless object.name.split(' ').count == 1 || u.last_name | |
u.id_inst_handle = object.handle unless u.id_inst_handle | |
u.save | |
end | |
rescue | |
ap u | |
end | |
end | |
require 'csv' | |
CSV.open("/home/michael/users_test2.csv", "wb") do |csv| | |
csv << CanonicalUser.attribute_names | |
CanonicalUser.where(handle: handles).where('first_name IS NOT NULL').with_progress.each do |user| | |
csv << user.attributes.values | |
end | |
end |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment