From 8dfee1fd841955625d34391f67795407990b1907 Mon Sep 17 00:00:00 2001 From: Johan Van de Wauw Date: Fri, 31 Jan 2025 15:25:25 +0100 Subject: [PATCH] Remove chat link per room chat rooms per track now (upcoming commit) --- layouts/schedule/room.html | 8 - lib/helpers/pentabarf.rb | 1540 ------------------------------------ 2 files changed, 1548 deletions(-) delete mode 100644 lib/helpers/pentabarf.rb diff --git a/layouts/schedule/room.html b/layouts/schedule/room.html index eb6df22d..916f9ecc 100644 --- a/layouts/schedule/room.html +++ b/layouts/schedule/room.html @@ -32,14 +32,6 @@ <% end %> <% end %> -
  • - Chat: - <% if ! go_live || ! matrix_works %> - We've hit a snag. The Video only link still works! - <% else %> - Join the conversation! - <% end %> -

    • <% diff --git a/lib/helpers/pentabarf.rb b/lib/helpers/pentabarf.rb deleted file mode 100644 index 2fd8f4af..00000000 --- a/lib/helpers/pentabarf.rb +++ /dev/null @@ -1,1540 +0,0 @@ -# vim: set ts=2 sw=2 et ai ft=ruby: - -module Fosdem - class Pentabarf - require 'pg' - - def self.update(config) - conf = config.fetch :pentabarf - h = conf.fetch :host - p = conf.fetch :port - d = conf.fetch :database - u = conf.fetch :username - pw = conf.fetch :password, nil - #s = conf.fetch :schema - p = Pentabarf.new(h, p, u, pw, d) - begin - p.update(config) - ensure - p.close if p - end - end - - def initialize(hostaddress, port, user, password, database) - time_before = Time.now - log(:low, "connecting to Pentabarf database //#{hostaddress}:#{port}/#{database} with user #{user}") - begin - @db = PGconn.open( - host: hostaddress, - port: port, - user: user, - password: password, - dbname: database, - ) - log(:high, "connected to Pentabarf database #{hostaddress}:#{port}", Time.now - time_before) - rescue => e - raise "Failed to connect to the Pentabarf database //#{hostaddress}:#{port}/#{database} with user #{user}: #{e.message}" - end - end - - def close - if @db - @db.close - log(:high, "disconnected from Pentabarf database") - @db = nil - end - end - - def update(config) - total_time = start_time = Time.now - - $meta_extension = 'yaml' - $hash_extension = 'hash' - - # load the configuration from the site - conf = config.fetch :pentabarf - $thumb_width, $thumb_height = begin - t = conf.fetch :thumbnail - [t.fetch(:width), t.fetch(:height)] - end - - $photo_width, $photo_height = begin - t = conf.fetch :photo - [t.fetch(:width), t.fetch(:height)] - end - - $logo_width, $logo_height = begin - t = conf.fetch :eventlogo - [t.fetch(:width), t.fetch(:height)] - end - - $photo_export_root, $thumbnail_export_root, $attachment_export_root, $eventlogos_export_root = - begin - t = conf.fetch :export_roots - a = [:photos, :thumbnails, :attachments, :eventlogos].map { |f| t.fetch(f) } - $export_roots = a - a - end - - $meta_export_file = conf.fetch :meta_export_file - - run_internal(conf, total_time) - end - - private - - def run_internal(conf, total_time) - require 'yaml' - require 'time' - require 'fileutils' - - cid = begin - t = conf.fetch :conference_id - if t.to_s =~ /^\d+$/ - t.to_i - else - @db.exec('SELECT conference_id FROM conference WHERE acronym=$1', [t]) do |res| - fail "no conference found for acronym #{t}" if res.ntuples < 1 - fail "found more than one conference that matches acronym #{t}" if res.ntuples > 1 - - res.first['conference_id'].to_i - end - end - end - - $export_roots.each do |d| - FileUtils.mkdir_p(d) unless File.exists? d - end - - # crawl the current cache directory and keep a tree of all files - # that are in there in order to detect the files we need to delete - # after rendering changes; don't bother about directories, as we - # remove all empty directories at the end of the process - $cache_tree_before = begin - require 'find' - r = [] - $export_roots.each do |d| - Find.find(d) do |f| - r << f unless File.directory? f - end - end - r.sort - end - - # obviously, we haven't created/updated any files yet - $cache_tree_after = [] - - # some helpers - to_slug = lambda { |x| x.fetch('slug') } - - def yf(file, content, time_before = nil) - b = File.dirname(file) - FileUtils.mkdir_p(b) unless File.exists? b - - action = if File.exists?(file) - new_digest = sha(content) - old_digest = sha(IO.read(file)) - if new_digest != old_digest - :update - else - :identical - end - else - :create - end - - unless action == :identical - File.open(file, "wb") do |f| - f.write(content) - end - end - - $cache_tree_after << file - - Nanoc::CLI::Logger.instance.file(action == :identical ? :low : :high, action, file, time_before.nil? ? nil : Time.now - time_before) - - action - end - - def slug(o, attr) - slug = if o.has_key? 'slug' and o['slug'] - o.fetch('slug') - else - o[attr.to_s] - end - raise "slug: o[#{attr}] is nil" if o.nil? - - sanitized = begin - # Inflector takes care of replacing locale specific symbols to - # something "safe" for URLs without requiring symbol gibberish - require 'active_support/inflector/transliterate' - require 'active_support/inflector/methods' - # Teach Inflector's i18n backend about some extra characters - # we'd like it to transliterate for us. - I18n.config.available_locales = [:custom] - I18n.backend.store_translations(:custom, :i18n => { - :transliterate => { - :rule => { - "’" => "'" - } - } - }) - I18n.locale = :custom - ActiveSupport::Inflector - .transliterate(slug) - .downcase - .gsub(/[&]/, '_and_') - .gsub(/[\s\-\+\/,;:~=]+/, '_') - .gsub(/[\.!'"#\?\{\}\*\^\%\$\|<>]+/, '') - .gsub(/[\(]/, '') - .gsub(/[\)]/, '') - .gsub(/__/, '_') - end - sanitized - end - - def slugify!(o, attr) - if o.nil? - nil - elsif o.is_a? Array - o.each { |x| slugify!(x, attr) } - o - else - o['slug'] = slug(o, attr) - o - end - end - - def markup(text) - return '' if text.nil? - - text = text.to_s.strip - return '' if text.empty? - - require 'bluecloth' - BlueCloth.new(text, :filter_html).to_html - end - - def model_value(key, value) - case value - when 't' - true - when 'f' - false - when /^\d+$/ - value.to_i - else - value - end - end - - def model(record, attributes = nil) - if record.nil? - nil - elsif record.is_a? Array - record.map { |x| model(x, attributes) } - elsif record.is_a? Hash - if attributes - h = {} - record.each do |k, v| - h[k.to_s] = model_value(k.to_sym, v) if attributes.include? k.to_sym - end - h - else - h = {} - record.each do |k, v| - h[k.to_s] = model_value(k.to_sym, v) - end - h - end - else - fail "unsupported record #{record.inspect} of class #{record.class}" - end - end - - def byid(ary, attr) - r = {} - ary.each { |i| r[i[attr.to_s]] = i } - r - end - - def yamldate(d) - d.strftime('%Y-%m-%d %T.000000000 %z') - end - - def yamltime(t) - t.strftime('%H:%M') - end - - log(:high, "rendering cache") - start_time = Time.now - - conference = model(@db.exec('SELECT * FROM conference WHERE conference_id=$1', [cid]) do |res| - fail "failed to find conference with conference_id=#{cid}" if res.ntuples < 1 - fail "found more than one conference with conference_id=#{cid}" if res.ntuples > 1 - - res.first - end) - event_time_offset = begin - dc = Time.parse(conference['day_change']) - if dc - dc.hour * 3600 + dc.min * 60 + dc.sec - else - 0 - end - end - conftz = begin - fail "conference has no timezone" unless conference.fetch('timezone') - - require 'tzinfo' - TZInfo::Timezone.get(conference.fetch('timezone')) - end - - def dblist(q, params = []) - l = [] - params = if params.is_a? Array - params - else - [params] - end - @db.exec(q, params) do |res| - res.each do |row| - l << row - end - end - l - end - - days = @db.exec(%q{ - SELECT * - FROM conference_day - WHERE conference_id=$1 - AND public=true - ORDER BY conference_day}, [cid]) { |res| - res.map { |row| - m = model row, [:conference_day_id, :name] - m['conference_day'] = Date.parse(row['conference_day']) - m - }.map { |x| slugify! x, :name } - } - day_by_day_id = byid days, :conference_day_id - - # decorate days with a title - days.each do |d| - d['title'] = d['name'] - end - - events = begin - time_before = Time.now - events = @db.exec(%q{ - SELECT * - FROM event - WHERE conference_id=$1 - AND event_state='accepted' - AND event_state_progress IN ('confirmed', 'reconfirmed') - AND event_type != 'movie' - AND start_time IS NOT NULL - AND (language IS NULL OR language='en') - AND public=true - ORDER BY event_id}, [cid]) { |res| - res - .select { |e| e['duration'] and e['conference_track_id'] and e['conference_room_id'] and e['conference_day_id'] } - .map { |e| - day = day_by_day_id.fetch(e['conference_day_id'].to_i) - - me = model e, [:event_id, :conference_id, :slug, :title, :subtitle, :conference_track_id, :event_type, :duration, :event_state, :event_state_progress, :language, :conference_room_id, :conference_day_id, :abstract, :description] - - me['start_time'], me['end_time'] = - begin - d = day.fetch('conference_day') - t = Time.parse(e.fetch('start_time')) + event_time_offset - dt = DateTime.new(d.year, d.month, d.day, t.hour, t.min, t.sec) - du = Time.parse(e['duration']) - [ - dt, - (dt + ((du.hour / 24.0) + (du.min / (24 * 60.0)) + (du.sec / (24 * 60 * 60.0)))) - ].map { |x| x.strftime('%H:%M') } - end - me['start_datetime'], me['end_datetime'] = - begin - d = day.fetch('conference_day') - ['start_time', 'end_time'].map do |x| - t = Time.parse(me[x]) - dt = DateTime.new(d.year, d.month, d.day, t.hour, t.min, t.sec) - offset = conftz.period_for_local(dt).utc_total_offset - yamldate DateTime.new(d.year, d.month, d.day, t.hour, t.min, t.sec, Rational(offset / 3600.0, 24)) - end - end - me['day'] = day.fetch('slug') - me['day_name'] = day.fetch('name') - #me['conference_day'] = day.fetch('conference_day_id') - - me - } - .map { |e| slugify! e, :title } - } - log(:high, "loaded #{events.size} events", Time.now - time_before) - events - end - - # validate events - begin - errors = {} - begin - events.each do |event| - err = [] - err << "missing slug" unless event.has_key? 'slug' - err << "accepted event without start_time" unless event.has_key? 'start_time' - err << "accepted event without duration" unless event.has_key? 'duration' - if event['start_time'] - t = Time.parse(event['start_time']) - err << "event starts before 09:00: #{event['start_time']}" if t.hour < 9 - err << "event starts after 19:00: #{event['start_time']}" if t.hour >= 19 - else - err << "event has no start_time" - end - errors[%Q!event #{event['event_id']} ("#{event['title']}")!] = err unless err.empty? - end - end - - unless errors.empty? - msg = errors.map do |w, errlist| - errlist.map { |e| "#{w}: #{e}" }.join($/) - end.join($/) - fail "Invalid events in database:#{$/}#{msg}" - end - end - - events.each do |e| - # render the event markup (abstract and description) - [:abstract, :description].each do |a| - e[a.to_s] = markup(e.fetch(a.to_s)) - e["raw_#{a.to_s}"] = e.fetch(a.to_s) - end - - # add an empty attachments hash to each - e['attachments'] = [] - end - - event_by_event_id = begin - h = {} - events.each { |e| h[e['event_id']] = e } - h - end - - eventpersons = begin - time_before = Time.now - eventpersons = dblist(%q{ - SELECT * - FROM event_person - WHERE event_role IN ('moderator', 'speaker') - AND event_role_state IN ('confirmed', 'offer') - }) - .map { |ep| - %w(event_id person_id).each do |x| - ep[x] = ep[x].to_i - end - ep - } - .select { |ep| event_by_event_id.has_key? ep['event_id'] } - - log(:high, "loaded #{eventpersons.size} eventpersons", Time.now - time_before) - eventpersons - end - - eventpersons_by_event_id = begin - h = {} - eventpersons.each { |ep| h[ep['event_id']] = [] } - eventpersons.each { |ep| h[ep['event_id']] << ep } - h - end - eventpersons_by_person_id = begin - h = {} - eventpersons.each { |ep| h[ep['person_id']] = [] } - eventpersons.each { |ep| h[ep['person_id']] << ep } - h - end - - speakers = begin - time_before = Time.now - - list = @db.exec('SELECT * FROM person ORDER BY person_id') do |res| - res - .reject { |p| eventpersons_by_person_id.fetch(p['person_id'].to_i, []).empty? } - .map { |p| model(p, [:person_id, :title, :gender, :first_name, :last_name, :public_name, :nickname]) } - .map do |p| - name = if p['public_name'] - p['public_name'] - elsif p['first_name'] and p['last_name'] - "#{p['first_name']} #{p['last_name']}" - elsif p['nickname'] - p['nickname'] - else - %w(first_name last_name).map { |x| p[x] }.reject(&:nil?).join(' ') - end - p['name'] = name - p['title'] = name - - # - # Figure out where to sort speakers in alphabetised - # lists. Harder than it looks! - # - # - If a speaker lists a 'public name', we need to - # list it in preference to any other name fields - # (which may or may not exist). Unfortunately, - # we can't determine sort order easily from it, - # so we have to guess. - # - # - Ruby's hopelessly inadequate sorting features - # provide us with no reasonable way to specify - # custom collation rules in the face of multiple - # alphabets. We hack around this by replacing - # letters with letter pairs so their sort well. - # - # NOTE: sortname should _never_ be printed anywhere! - # - if p['public_name'] - # Assume that the last component of the speaker's - # public name is their last name. - n = p['public_name'].split(/\s+/) - sortname = "#{n.last} #{n.first}" - - # Deal with common exceptions (composed last names). - ['de', 'le', 'van de', 'van den', 'van', 'von'].map { |s| - Regexp.new("(.*)\s+(#{s}\s+.*)$", true)}.each do |r| - if p['public_name'] =~ r - sortname = "#{$2} #{$1}" - break - end - - # Deal with people who suffix their public_name with - # (nickname) or similar. - if sortname =~ /^\(/ - sortname = "#{n[-2]} #{n.first}" - end - end - # Be very evil and sort "roles" under "#" - ['Staff', 'Team'].map { |s| - Regexp.new("(.*\s#{s})$", false)}.each do |r| - if p['public_name'] =~ r - sortname = "# #{$1}" - break - end - end - elsif p['last_name'] - sortname = p['last_name'] - if p['first_name'] - sortname << ' ' << p['first_name'] - end - else - sortname = p['name'] - end - # Sort non-ASCII characters (more) correctly. - p['sortname'] = sortname.upcase - .gsub('é', 'EZ') - .gsub('ñ', 'NZ') - .gsub('ó', 'OZ') - - slugify! p, :name - end - end - log(:high, "loaded #{list.size} speakers", Time.now - time_before) - list - end - - speaker_by_person_id = byid speakers, :person_id - - # post-process speakers with conference_person - begin - # decorate speakers with conference_persons, but only with selected fields - # to avoid bleeding private information such as their email - @db.exec('SELECT * FROM conference_person ORDER BY conference_id DESC') do |res| - res.each do |cp| - %w(person_id conference_person_id conference_id).each do |x| - cp[x] = cp[x].to_i - end - - p = speaker_by_person_id[cp['person_id']] - # if we have no match (p) then it's not a speaker, and simply skip to the next - if p - unless p.has_key? 'abstract' or p.has_key? 'description' - # copy those attributes and convert the markup - %w(abstract description).each do |a| - p[a] = markup(cp[a]) - end - if cp['conference_id'] != cid - # mark as data from a previous conference - p['abstract_old'] = true - p['description_old'] = true - end - # Store the record this data came from - links come from here - p['conference_person_id'] = cp['conference_person_id'] - next - end - else - # it's not a speaker, ignore - end - end - end - end - - # post-process speakers with conference_person_links - # (now that we have person.conference_person_id) - # but fetch and cache all conference_person_link rows, - # it's faster - begin - cplinks = model(dblist('SELECT * FROM conference_person_link')) - # and hash them by conference_person_id - cplinks_by_cpid = begin - h = {} - cplinks.each { |l| h[l.fetch('conference_person_id')] = [] } - cplinks.each { |l| h[l.fetch('conference_person_id')] << l } - h - end - # now decorate the persons with their links - speakers.each do |p| - links = model(cplinks_by_cpid.fetch(p['conference_person_id'], []), [:url, :title, :rank]).reject { |l| l['url'] == 'http://' }.sort_by { |l| l['rank'] } - # post-process links by setting 'title' if not set - links.each do |l| - l['title'] = l.fetch('url') unless l['title'] - end - p['links'] = links - end - end - - # post-process events with event_link - begin - time_before = Time.now - - eventlinks = model(dblist(%q{ - SELECT * - FROM fosdem.view_event_link - WHERE conference_id=$1 - ORDER BY event_id}, [cid])) - eventlinks_by_event_id = begin - h = {} - eventlinks.each { |l| h[l['event_id']] = [] } - eventlinks.each { |l| h[l['event_id']] << l } - h - end - events.each do |e| - links = model(eventlinks_by_event_id.fetch(e['event_id'], []).sort_by { |l| [l['rank'].to_i, l['event_link_id'].to_i] }, [:url, :title, :rank]) - # post-process the links and set 'title' if not set - links.each do |l| - l['title'] = l.fetch('url') unless l['title'] - end - e['links'] = links - end - - log(:high, "loaded #{eventlinks.size} event links", Time.now - time_before) - end - - # post-process events with event_related - begin - time_before = Time.now - - # fetch all event_related into a cache, faster - eventrelated = model(dblist(%q{ - SELECT * - FROM event_related - ORDER BY event_id1, event_id2})).select { |r| event_by_event_id.has_key? r['event_id1'] } - h = begin - require 'set' - - h = {} - eventrelated.each do |r| - [ - [r['event_id1'], r['event_id2']], - [r['event_id2'], r['event_id1']], - ].each do |a| - h[a.first] = Set.new unless h.has_key? a.first - h[a.first] << a.last - end - end - h - end - - h.each do |from_id, to_ids| - from = event_by_event_id.fetch(from_id) - from['related'] = to_ids.map { |id| event_by_event_id.fetch(id)['slug'] } - end - - log(:high, "loaded #{eventrelated.size} event relations", Time.now - time_before) - end - - # tracks - tracks = begin - time_before = Time.now - - list = @db.exec(%q{ - SELECT * - FROM conference_track - WHERE conference_id=$1 - ORDER BY rank, conference_track_id}, [cid]) do |res| - res - .reject { |t| t['conference_track'] == 'Main Tracks' } - .reject { |t| t['conference_track'] == 'Main tracks' } - .map do |t| - # XXX: Disgusting hack for 2015 because I really don't have - # time to come up with anything more elegant than this! - # pph 20150109. Disabled agk 20160104 (for Security devroom). - # Re-enabled 20171222 for Community devroom with same name as main track. - #if t['conference_track'] == 'Community devroom' then - # t['name'] = t['conference_track'] - #else - t['name'] = t['conference_track'].gsub(/\s+(track|devroom)$/i, '') - #end - t['title'] = t['conference_track'] - t['type'] = case t['conference_track'] - when /\s+devroom$/i - 'devroom' - when /^lightning\s*talks?$/i - 'lightningtalk' - when /^keynotes?$/i - 'keynote' - when /^certification?$/i - 'certification' - when /^keysigning?$/i - 'keysigning' - when /^bof/i - 'bof' - when /workshop/i - 'workshop' - when /^stands/i - 'stand' - when /\s+stand/i - 'standtrack' - when /^infodesk/i - 'infodesk' - else - 'maintrack' - end - - slugify!(model(t, [:name, :title, :type, :conference_track_id, :conference_track, :rank, :conference_call_for_papers_url]), :name) - end - end - log(:high, "loaded #{list.size} tracks", Time.now - time_before) - list - end - - track_by_track_id = byid tracks, :conference_track_id - # decorate tracks with events - begin - events_by_track_id = begin - h = {} - events.each { |e| h[e['conference_track_id']] = [] } - events.each { |e| h[e['conference_track_id']] << e } - h - end - tracks.each do |t| - trackevents = events_by_track_id.fetch(t['conference_track_id'], []).sort_by { |e| e['start_datetime'] } - t['events'] = trackevents.map(&to_slug) - t['events_by_day'] = begin - h = {} - days.each { |d| h[d.fetch('slug')] = trackevents.select { |e| e['conference_day_id'] == d['conference_day_id'] }.map(&to_slug) } - h - end - end - end - # decorate events with track information - begin - events.each do |e| - t = track_by_track_id[e['conference_track_id']] - if t - e['track'] = t.fetch('slug') - e['track_name'] = t.fetch('name') - e['track_full_name'] = t['conference_track'] - e['track_rank'] = t['rank'] - e['type'] = t['type'] - end - end - end - - # rooms - rooms = begin - time_before = Time.now - rooms = slugify!(model(dblist(%q{ - SELECT DISTINCT r.* - FROM conference_room r - JOIN event using (conference_room_id) - WHERE r.conference_id=$1 - AND r.public=true - ORDER BY r.rank, r.conference_room_id}, [cid]), - [:conference_room_id, :conference_room, :size, :rank]), :conference_room) - log(:high, "loaded #{rooms.size} rooms", Time.now - time_before) - rooms - end - - room_by_room_id = byid rooms, :conference_room_id - - # decorate rooms with a title - # flag virtual rooms - rooms.each do |r| - r['virtual'] = (r['conference_room'] =~ /^[DIMS]/) ? true : false - if r['virtual'] - r['title'] = r['conference_room'] + " (online)" - else - r['title'] = r['conference_room'] - end - # decorate with chat and live video link - r['chat_link'] = 'https://chat.fosdem.org/#/room/#' + '2023-' + r['conference_room'].gsub(/\(/, '').gsub(/[\) ]/, '_').downcase + ':fosdem.org' - r['live_video_link'] = 'https://live.fosdem.org/watch/' + r['slug'].gsub(/_.*/, '') - end - - # decorate rooms with events - begin - events_by_room_id = begin - h = {} - events.each { |e| h[e['conference_room_id']] = [] } - events.each { |e| h[e['conference_room_id']] << e } - h - end - rooms.each do |r| - roomevents = events_by_room_id.fetch(r['conference_room_id'], []).sort_by { |e| e['start_datetime'] } - r['events'] = roomevents.map(&to_slug) - # also set the events by day - r['events_by_day'] = begin - h = {} - days.each { |d| h[d.fetch('slug')] = roomevents.select { |e| e['conference_day_id'] == d['conference_day_id'] }.map(&to_slug) } - h - end - end - end - - # add room to events - begin - events.each do |e| - room = room_by_room_id.fetch(e.fetch('conference_room_id')) - e['room'] = room['slug'] - e['room_name'] = room['conference_room'] - e['room_rank'] = room['rank'] - e['virtual'] = room['virtual'] - end - end - - # Add video links for 2 days before until 30 minutes after each event - begin - video_link_count = 0 - events.each do |e| - next if e['room'] =~ /corridor/ - next if e['room'] =~ /level/ - next if e['room'] =~ /106/ - next if e['room'] =~ /2111/ - next if e['room'] =~ /3242/ - next if e['room'] =~ /3244/ - next if e['room'] =~ /147/ - next if e['track'] =~ /certification/ - # next if e['track'] =~ /bof/ - next if e['track'] =~ /infodesk/ - next if e['track'] =~ /workshop/ - next if e['track'] =~ /stand/ - - current_time = Time.now - - # Uncomment the following line to test what the site looks like at a specific time - # current_time = Time.parse('2022-02-05 10:15:00.000000000 +0100') - - event_start = Time.parse(e['start_datetime']) - event_end = Time.parse(e['end_datetime']) - next if event_start - current_time > 48 * 60 * 60 - next if current_time - event_end > 30 * 60 - - ll = {} - ll['title'] = 'Live video stream from the room (during event)' - ll['url'] = 'https://live.fosdem.org/watch/' << e['room'].gsub(/_.*/, '') - ll['rank'] = nil - e['links'] << ll - e['live_video_link'] = ll['url'] - video_link_count += 1 - end - log(:high, "added #{video_link_count} live video links") - end - -# TODO Only if virtual? - # Add chat room links - begin - chat_link_count = 0 - events.each do |e| - next if e['track'] =~ /certification/ - - current_time = Time.now - - # Uncomment the following line to test what the site looks like at a specific time - # current_time = Time.parse('2022-02-05 16:15:00.000000000 +0100') - - event_start = Time.parse(e['start_datetime']) - event_end = Time.parse(e['end_datetime']) - next if event_start - current_time > 48 * 60 * 60 - - ll = {} - lm = {} - lh = {} - li = {} - - chatroom_name = '2023-' + e['room_name'].gsub(/\(/, '').gsub(/[\) ]/, '_').downcase + ':fosdem.org' - - ll['title'] = 'Chat room (web)' - ll['url'] = 'https://chat.fosdem.org/#/room/#' + chatroom_name - ll['rank'] = nil - e['links'] << ll - e['chat_link'] = ll['url'] - - lm['title'] = 'Chat room (app)' - lm['url'] = 'https://matrix.to/#/#' + chatroom_name + '?web-instance[element.io]=chat.fosdem.org' - lm['rank'] = nil - e['links'] << lm - - chat_link_count += 2 - - # For talks, also show the private room afterwards - if ( e['room'] =~ /^[dkb]/ && current_time > event_end ) - chatroom_name = e['room_name'].gsub(/^../, '') + "-" - chatroom_name += e['slug'] + ':fosdem.org' - - lh['title'] = 'Hallway chat room (web)' - lh['url'] = 'https://chat.fosdem.org/#/room/#' + chatroom_name - lh['rank'] = nil - e['links'] << lh - - li['title'] = 'Hallway chat room (app)' - li['url'] = 'https://matrix.to/#/#' + chatroom_name + '?web-instance[element.io]=chat.fosdem.org' - li['rank'] = nil - e['links'] << li - - chat_link_count += 2 - end - - end - log(:high, "added #{chat_link_count} chat room links") - end - - # Add feedback links (FIXME Only after the event!) - begin - feedback_link_count = 0 - events.each do |e| - ll = {} - ll['title'] = 'Submit feedback' - ll['url'] = "https://submission.fosdem.org/feedback/#{e['event_id']}.php" - ll['rank'] = nil - e['links'] << ll - feedback_link_count += 1 - end - log(:high, "added #{feedback_link_count} feedback links") - end - - # add rooms to tracks - tracks.each do |t| - trackevents = events_by_track_id.fetch(t['conference_track_id'], []).sort_by { |e| e['start_datetime'] } - trackrooms = trackevents.map { |e| e['conference_room_id'] }.uniq.map { |id| room_by_room_id[id] }.sort_by { |r| [r['rank'], r['conference_room_id']] } - t['rooms'] = trackrooms.map(&to_slug) - - t['events_per_room_per_day'] = begin - by_day_hash = {} - days.each { |d| by_day_hash[d.fetch('slug')] = {} } - trackevents.group_by { |e| e['conference_day_id'] }.each do |day_id, track_events_per_day| - day_name = day_by_day_id.fetch(day_id).fetch('slug') - track_events_per_day.group_by { |e| e['conference_room_id'] }.each do |room_id, track_events_per_room| - room_slug = room_by_room_id.fetch(room_id).fetch('slug') - by_room_hash = by_day_hash.fetch(day_name) - by_room_hash[room_slug] = [] unless by_room_hash.has_key? room_slug - track_events_per_room.each { |e| by_room_hash[room_slug] << e.fetch('slug') } - end - end - by_day_hash - end - end - - # cross-post-process events and speakers - begin - events.each do |e| - event_speakers = eventpersons_by_event_id.fetch(e['event_id'], []).map { |ep| speaker_by_person_id[ep['person_id']] }.reject { |s| s.nil? }.sort_by { |p| p['person_id'] }.uniq { |p| p['person_id'] } - - e['speakers'] = event_speakers.map(&to_slug) - end - speakers.each do |s| - speaker_events = eventpersons_by_person_id.fetch(s['person_id'], []).map { |ep| event_by_event_id[ep['event_id']] }.reject { |e| e.nil? }.sort_by { |e| e['event_id'] }.uniq { |e| e['event_id'] } - s['events'] = speaker_events.map(&to_slug) - s['events_by_day'] = begin - h = {} - days.each { |d| h[d.fetch('slug')] = speaker_events.select { |e| e['conference_day_id'] == d['conference_day_id'] }.map(&to_slug) } - h - end - end - end - - # compute start_time and end_time for various schedule items - begin - events_per_day = begin - h = {} - days.each do |d| - cdi = d.fetch('conference_day_id') - h[d.fetch('slug')] = events.select { |e| e.fetch('conference_day_id') == cdi } - end - h - end - - days.each do |d| - earliest = events_per_day.fetch(d['slug']).sort_by { |e| e.fetch 'start_time' }.first - latest = events_per_day.fetch(d['slug']).sort_by { |e| e.fetch 'end_time' }.last - d['start_time'] = earliest.fetch('start_time') if earliest - d['end_time'] = latest.fetch('end_time') if latest - end - - { - conference_track_id: tracks, - conference_room_id: rooms, - }.each do |a, list| - k = a.to_s - #raise "empty list? (#{a})" if list.empty? - list.each do |item| - kv = item.fetch(k) - ['start_time', 'end_time'].each { |x| item[x] = {} } - days.each do |d| - matching_events = events_per_day.fetch(d['slug']).select { |e| e.fetch(k) == kv } - earliest = matching_events.sort_by { |e| e.fetch 'start_time' }.first - latest = matching_events.sort_by { |e| e.fetch 'end_time' }.last - item['start_time'][d['slug']] = earliest ? earliest.fetch('start_time') : nil - item['end_time'][d['slug']] = latest ? latest.fetch('end_time') : nil - end - end - end - end - - # compute start_time_index and end_time_index where applicable - begin - # compute time slot interval in minutes from timeslot_duration on the conference object - tsim = begin - t = Time.parse(conference.fetch('timeslot_duration')) - raise "conference :timeslot_duration has seconds" unless t.sec == 0 - raise "conference :timeslot_duration is less than 5 minutes" unless t.min >= 5 - - t.min + t.hour * 60 - end - - [ - events, - days, - ].each do |list| - list.each do |item| - ['start_time', 'end_time'].each do |a| - time = item[a] - # if the attribute isn't there, it just means there are no scheduled - # events for that day yet - if time - time = Time.parse(time) - fail "time with granularity != 5 min: #{time} for item #{item.inspect}" if time.min % tsim != 0 - - index = (time.hour * 60 + time.min) / tsim - - item[a + '_index'] = index - end - end - end - end - - # same here, except that the start_time, end_time attributes are - # hashes (key = day slug) - [tracks, rooms].each do |list| - list.each do |item| - ['start_time', 'end_time'].each do |a| - item[a + '_index'] = {} - item.fetch(a).each do |dayslug, time| - unless time.nil? - time = Time.parse(time) - raise "time with granularity != 5 min: #{time} for item #{item.inspect}" if time.min % tsim != 0 - - index = (time.hour * 60 + time.min) / tsim - item[a + '_index'][dayslug] = index - else - item[a + '_index'][dayslug] = nil - end - end - end - end - end - end - - # event attachments - before_attachments = Time.now - begin - log(:high, 'processing attachments') - - t = Time.now - to_export = [] - # *MUCH* faster to load them all rather then querying them one by one - # (load all = 2.19s, one by one = 12.35s) - counter = 0 - @db.exec(%q{ - SELECT event_attachment_id, mime_type, title, pages, event_id, attachment_type, filename, data_hash - FROM fosdem.view_event_attachment - WHERE conference_id=$1}, [cid]) do |res| - res - .map { |a| - h = a['data_hash'] - a = model(a) - a['data_hash'] = h - a - }.each do |a| - counter += 1 - - event = event_by_event_id[a['event_id']] - next unless event - - # attachment filename sanitization - d, f, b, ext = sanitize_filename a['filename'] - fail "sanitized basename for attachment with event_attachment_id=#{a['event_attachment_id']} still contains a dot: #{b}" if b =~ /\./ - - mf = "#{b}.#{$meta_extension}" - - filename, meta_filename = [f, mf] - .map { |x| File.join([d, event['slug'], a['attachment_type'], a['event_attachment_id'].to_s, x].reject(&:nil?)) } - file, meta_file = [filename, meta_filename].map { |x| "#{$attachment_export_root}/#{x}".gsub(%r{/+}, '/') } - - type = a['attachment_type'] or 'attachment' - - meta = { - 'file' => file, - 'filename' => f, - 'type' => type, - 'event_id' => event['event_id'], - 'event_slug' => event['slug'], - 'id' => a['event_attachment_id'], - 'identifier' => "/schedule/event/#{event['slug']}/attachments/#{type}/#{a['event_attachment_id']}/#{b}/", - } - meta['mime'] = a['mime_type'] if a['mime_type'] - meta['title'] = a['title'] if a['title'] - meta['pages'] = a['pages'].to_i if a['pages'] - - $cache_tree_after << meta_file if File.exists? meta_file - - needs_export = - if all_exist(file, meta_file) - digest = md5_file(file) - if a.fetch('data_hash').downcase == digest.downcase - # file exists and is the same - #Nanoc::CLI::Logger.instance.file(:low, :identical, file, 0) - meta['size'] = File.size(file) - $cache_tree_after << file - false - else - # attachment file exists, but the content differs - true - end - else - # attachment file does not exist - true - end - - if needs_export - to_export << { - file: file, - event_attachment_id: a['event_attachment_id'], - event_id: event['event_id'], - meta: meta, - meta_file: meta_file - } - end - - event['attachments'] << meta - end - end - log(:high, "loaded #{counter} attachment hashes, #{to_export.size} need to be exported", Time.now - t) - - log(:high, "exporting #{to_export.size} modified or missing attachments") - t = Time.now - to_export.each do |todo| - tt = Time.now - a = begin - @db.exec(%q{ - SELECT data - FROM event_attachment - WHERE event_attachment_id=$1}, [todo[:event_attachment_id]]) do |res| - fail "failed to find event_attachment with event_attachment_id=#{todo[:event_attachment_id]}" if res.ntuples < 1 - fail "found more than one event_attachment with event_attachment_id=#{todo[:event_attachment_id]}" if res.ntuples > 1 - - res.first - end - end - yf(todo[:file], PGconn.unescape_bytea(a['data']), tt) - File.open(todo[:meta_file], "wb") do |f| - f.write(todo[:meta].to_yaml) - end - $cache_tree_after << todo[:file] - $cache_tree_after << todo[:meta_file] - end - log(:high, "exported #{to_export.size} modified or missing attachments", Time.now - t) - - # decorate all events with the attachment sizes - events.each do |event| - event['attachments'].each do |a| - unless a['size'] - a['size'] = File.size(a.fetch('file')) - end - end - end - end - after_attachments = Time.now - start_time += (after_attachments - before_attachments) - - # add logos to events - before_logos = Time.now - begin - counter = 0 - log(:high, 'processing event logos') - to_export = [] - t = Time.now - @db.exec(%q{ - SELECT event_id, mime_type, md5(image) AS image_hash, octet_length(image) AS image_length - FROM event_image - ORDER BY event_id}) do |res| - res - .map { |i| model(i) } - .each do |i| - counter += 1 - next if i.fetch('image_length') < 1 - - event = event_by_event_id[i['event_id']] - next unless event - - extension = mime_to_extension(i.fetch('mime_type')) - _, n = sanitize_filename event['slug'] - filename, meta_filename, hash_filename = [extension, $meta_extension, $hash_extension].map { |x| "#{n}.#{x}" } - file, meta_file, hash_file = [filename, meta_filename, hash_filename].map { |x| File.join($eventlogos_export_root, x) } - - id = "/schedule/event/#{event['slug']}/logo/" - meta = { - 'file' => file, - 'filename' => filename, - 'event_id' => event['event_id'], - 'event_slug' => event['slug'], - 'identifier' => id, - 'mime' => i.fetch('mime_type'), - } - - event['logo'] = { - 'identifier' => id, - 'mime' => i.fetch('mime_type'), - } - - file_hash = - if all_exist(file, meta_file, hash_file) - IO.read(hash_file) - else - nil - end - - needs_export = - if file_hash - i['image_hash'].downcase != file_hash - else - true - end - - if needs_export - to_export << { - file: file, - event_id: i['event_id'], - slug: event['slug'], - meta_file: meta_file, - meta: meta, - hash_file: hash_file, - hash: i['image_hash'].downcase - } - else - $cache_tree_after << file << meta_file << hash_file - end - end - end - log(:high, "loaded #{counter} event logo hashes, #{to_export.size} need to be exported", Time.now - t) - - before_export = Time.now - to_export.each do |todo| - t = Time.now - i = @db.exec(%q{ - SELECT image - FROM event_image - WHERE event_id=$1}, [todo[:event_id]]) do |res| - fail "failed to find event_image for event_id=#{todo[:event_id]}" if res.ntuples < 1 - fail "found more than one event_image for event_id=#{todo[:event_id]}" if res.ntuples > 1 - - res.first - end - image = nil - begin - i['image'] = PGconn.unescape_bytea(i['image']) - require 'rmagick' - image = Magick::Image.from_blob(i['image']).first - - # Unset dates to ensure consistent checksums of the image file - image['date:create'] = nil - image['date:modify'] = nil - - blob, width, height = if image.rows > $logo_width or image.columns > $logo_height - [image.resize_to_fill($logo_width, $logo_height).to_blob, $logo_width, $logo_height] - else - [i['image'], image.columns, image.rows] - end - - yf(todo[:file], blob, t) - - File.open(todo[:hash_file], 'wb') do |f| - f.write(todo[:hash]) - end - $cache_tree_after << todo[:hash_file] - - meta = todo[:meta].clone - meta[:height] = height - meta[:width] = width - File.open(todo[:meta_file], 'wb') do |f| - f.write(meta.to_yaml) - end - $cache_tree_after << todo[:meta_file] - ensure - image.destroy! if image - end - end - log(:high, "exported #{to_export.size} event logos", Time.now - before_export) - end - after_logos = Time.now - start_time += (after_logos - before_logos) - - # add photos to speakers - before_photos = Time.now - begin - counter = 0 - log(:high, 'processing photos') - t = Time.now - to_export = [] - @db.exec(%q{ - SELECT person_id, mime_type, image_hash, image_length - FROM fosdem.view_person_image - WHERE conference_id=$1 AND public=true - ORDER BY person_id}, [cid]) do |res| - res.each do |i| - %w(person_id image_length).each { |x| i[x] = i[x].to_i } - - next if i.fetch('image_length') < 1 - - counter += 1 - speaker = speaker_by_person_id[i['person_id']] - next unless speaker - - extension = mime_to_extension(i.fetch('mime_type')) - _, n = sanitize_filename speaker['slug'] - filename, thumb_filename, meta_filename, thumb_meta_filename, hash_filename = - [extension, extension, $meta_extension, $meta_extension, $hash_extension].map { |x| "#{n}.#{x}" } - - file, meta_file, hash_file = [filename, meta_filename, hash_filename].map { |x| File.join($photo_export_root, x) } - thumb_file, thumb_meta_file = [thumb_filename, thumb_meta_filename].map { |x| File.join($thumbnail_export_root, x) } - - id, thumb_id = ['photo', 'thumbnail'].map { |x| "/schedule/speaker/#{speaker['slug']}/#{x}/" } - meta, thumb_meta = [id, thumb_id].map do |x| - { - 'identifier' => x, - 'file' => file, - 'filename' => filename, - 'person_id' => speaker['person_id'], - 'speaker_slug' => speaker['slug'], - 'mime' => i.fetch('mime_type'), - } - end - - { photo: meta, thumbnail: thumb_meta }.each do |k, m| - speaker[k.to_s] = { - 'mime' => i.fetch('mime_type'), - 'identifier' => m.fetch('identifier'), - } - end - - file_hash = - if all_exist(file, hash_file, thumb_file, meta_file, thumb_meta_file) - IO.read(hash_file) - else - nil - end - - needs_export = - if file_hash - i['image_hash'].downcase != file_hash - else - true - end - - if needs_export - to_export << { - file: file, - person_id: i['person_id'], - slug: speaker['slug'], - meta_file: meta_file, - meta: meta, - hash_file: hash_file, - hash: i['image_hash'].downcase, - thumb_file: thumb_file, - thumb_meta: thumb_meta, - thumb_meta_file: thumb_meta_file, - } - else - $cache_tree_after << file << thumb_file << hash_file << meta_file << thumb_meta_file - end #needs_export - end #each i - end #res - log(:high, "loaded #{counter} speaker photo hashes, #{to_export.size} need to be exported", Time.now - t) - - before_export = Time.now - to_export.each do |todo| - t = Time.now - i = @db.exec('SELECT * FROM person_image WHERE person_id=$1', [todo.fetch(:person_id)]) do |res| - fail "failed to find person_image for person_id=#{todo.fetch(:person_id)}" if res.ntuples < 1 - fail "found more than one person_image for person_id=#{todo.fetch(:person_id)}" if res.ntuples > 1 - - res.first - end - speaker = speaker_by_person_id.fetch(todo.fetch(:person_id)) - dt = Time.now - t - - image = nil - begin - i['image'] = PGconn.unescape_bytea(i['image']) - require 'rmagick' - image = Magick::Image.from_blob(i['image']).first - - # Unset dates to ensure consistent checksums of the image file - image['date:create'] = nil - image['date:modify'] = nil - - [ - { f: todo[:file], mw: $photo_width, mh: $photo_height, m: todo[:meta], mf: todo[:meta_file] }, - { f: todo[:thumb_file], mw: $thumb_width, mh: $thumb_height, m: todo[:thumb_meta], mf: todo[:thumb_meta_file] }, - ].each do |x| - t = Time.now - blob, width, height = if image.rows > x[:mw] or image.columns > x[:mh] - [image.resize_to_fill(x[:mw], x[:mh]).to_blob, x[:mw], x[:mw]] - else - [i['image'], image.columns, image.rows] - end - - yf(x[:f], blob, t) - - meta = begin - meta = x[:m].clone - meta['height'] = height - meta['width'] = width - meta - end - - File.open(x[:mf], 'wb') do |f| - f.write(meta.to_yaml) - end - $cache_tree_after << x[:mf] - end - - File.open(todo.fetch(:hash_file), 'wb') do |f| - f.write(todo.fetch(:hash)) - end - $cache_tree_after << todo.fetch(:hash_file) - ensure - image.destroy! if image - end - end - log(:high, "exported #{to_export.size} speaker photos", Time.now - before_export) - end - after_photos = Time.now - start_time += (after_photos - before_photos) - - tree = { - 'conference' => conference, - 'days' => hash_by(days, 'slug'), - 'rooms' => hash_by(rooms, 'slug'), - 'tracks' => hash_by(tracks, 'slug'), - 'events' => hash_by(events, 'slug'), - 'speakers' => hash_by(speakers, 'slug'), - } - yf($meta_export_file, tree.to_yaml, start_time) - - # now let's compute which files and dirs we need to remove - # because they were there before but have not been touched - # while rendering the cache this time - begin - outdated = begin - r = [] - h = {} - $cache_tree_after.each { |x| h[x] = 1 } - $cache_tree_before.each do |f| - r << f unless h[f] - end - r.sort - end - - require 'fileutils' - outdated.each do |f| - if File.directory? f - dirs << f - else - FileUtils.rm f - Nanoc::CLI::Logger.instance.file(:high, :delete, f) - end - end - end - - # now crawl for empty directories and remove those too - begin - # let's pack this routine into a function of its own, - # as we need to reiterate the process (see below) - def find_empty_dirs(dir) - list = [] - require 'find' - if File.exists? dir - Find.find(dir) do |d| - next if d == dir - next unless File.directory? d - - # see http://stackoverflow.com/questions/5059156/check-if-directory-is-empty-in-ruby - if (Dir.entries(d) - %w{. ..}).empty? - list << d - end - end - end - list - end - - # we have to iterate until we can't find any empty directories any more, - # as removing one empty directory can yield other empty directories as - # empty too - $export_roots.each do |export_root| - empty_dirs = find_empty_dirs(export_root) - until empty_dirs.empty? - empty_dirs.each do |d| - unless $export_roots.include? d - FileUtils.rmdir d - Nanoc::CLI::Logger.instance.file(:high, :delete, d) - end - end - empty_dirs = find_empty_dirs(export_root) - end - end - end - - log(:high, "cache rendered", Time.now - total_time) - ensure - close - end - - def sha(content) - require 'digest/sha2' - sha = Digest::SHA256.new - sha << content - sha.hexdigest.downcase - end - - def log(prio, message, duration = nil) - t = if duration - "[%2.2fs]" % duration - else - " " * 7 - end - - Nanoc::CLI::Logger.instance.log(prio, "%s%12s%s %s %s" % ["\e[36m", "pentabarf", "\e[0m", t, message]) - end - end -end -