run_ical 0.5.0 → 0.6.0

data/tasks/higdon_programs.yml

@@ -0,0 +1,179 @@
+# Source manifest for `rake data:refresh`.
+#
+# One entry per training program. `class` is the public API name passed to
+# RunIcal::Generator.generate(program:) -- never rename an existing one.
+# `weeks` is asserted against the scraped table so a silent site change fails loudly.
+---
+- class: Novice5K
+  file: novice_5_k
+  description: "5-K : Novice"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/5k-training/novice-5k/
+- class: Intermediate5K
+  file: intermediate_5_k
+  description: "5-K : Intermediate"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/5k-training/intermediate-5k/
+- class: Advanced5K
+  file: advanced_5_k
+  description: "5-K : Advanced"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/5k-training/advanced-5k/
+- class: Walkers5K
+  file: walkers_5_k
+  description: "5-K - Walkers"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/5k-training/walkers-5k/
+
+- class: Novice8K
+  file: novice_8_k
+  description: "8-K - Novice"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/8k-training/novice-8k/
+- class: Intermediate8K
+  file: intermediate_8_k
+  description: "8-K - Intermediate"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/8k-training/intermediate-8k/
+- class: Advanced8K
+  file: advanced_8_k
+  description: "8-K - Advanced"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/8k-training/advanced-8k/
+
+- class: Novice10K
+  file: novice_10_k
+  description: "10-K - Novice"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/10k-training/novice-10k/
+- class: Intermediate10K
+  file: intermediate_10_k
+  description: "10-K - Intermediate"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/10k-training/intermediate-10k/
+- class: Advanced10K
+  file: advanced_10_k
+  description: "10-K - Advanced"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/10k-training/advanced-10k/
+- class: Walking10K
+  file: walking_10_k
+  description: "10-K - Walking"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/10k-training/walkers-10k/
+
+- class: Novice15K
+  file: novice_15_k
+  description: "15-K - Novice"
+  weeks: 10
+  url: https://www.halhigdon.com/training-programs/15k-10-mile-training/novice-15k-10-mile/
+- class: Intermediate15K
+  file: intermediate_15_k
+  description: "15-K - Intermediate"
+  weeks: 10
+  url: https://www.halhigdon.com/training-programs/15k-10-mile-training/intermediate-15k-10-mile/
+- class: Advanced15K
+  file: advanced_15_k
+  description: "15-K - Advanced"
+  weeks: 10
+  url: https://www.halhigdon.com/training-programs/15k-10-mile-training/advanced-15k-10-mile/
+
+- class: HalfMarathonNovice1
+  file: half_marathon_novice1
+  description: "Half Marathon - Novice 1"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/novice-1-half-marathon/
+- class: HalfMarathonNovice2
+  file: half_marathon_novice2
+  description: "Half Marathon - Novice 2"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/novice-2-half-marathon/
+- class: HalfMarathonIntermediate1
+  file: half_marathon_intermediate1
+  description: "Half Marathon - Intermediate 1"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/intermediate-1-half-marathon/
+- class: HalfMarathonIntermediate2
+  file: half_marathon_intermediate2
+  description: "Half Marathon - Intermediate 2"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/intermediate-2-half-marathon/
+- class: HalfMarathonAdvanced
+  file: half_marathon_advanced
+  description: "Half Marathon - Advanced"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/advanced-half-marathon/
+- class: HalfMarathonHM3
+  file: half_marathon_hm3
+  description: "Half Marathon - HM3"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/half-marathon-3/
+- class: HalfMarathonWalkers
+  file: half_marathon_walkers
+  description: "Half Marathon - Walkers"
+  weeks: 12
+  url: https://www.halhigdon.com/training-programs/half-marathon-training/walkers-half-marathon/
+
+- class: MarathonNovice1
+  file: marathon_novice1
+  description: "Marathon : Novice 1"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/novice-1-marathon/
+- class: MarathonNovice2
+  file: marathon_novice2
+  description: "Marathon : Novice 2"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/novice-2-marathon/
+- class: MarathonIntermediate1
+  file: marathon_intermediate1
+  description: "Marathon : Intermediate 1"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/intermediate-1-marathon/
+- class: MarathonIntermediate2
+  file: marathon_intermediate2
+  description: "Marathon : Intermediate 2"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/intermediate-2-marathon/
+- class: MarathonAdvanced1
+  file: marathon_advanced1
+  description: "Marathon : Advanced 1"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/advanced-1-marathon/
+- class: MarathonAdvanced2
+  file: marathon_advanced2
+  description: "Marathon : Advanced 2"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/advanced-2-marathon/
+- class: MarathonNoviceSupreme
+  file: marathon_novice_supreme
+  description: "Marathon : Novice Supreme"
+  weeks: 30
+  url: https://www.halhigdon.com/training-programs/marathon-training/novice-supreme/
+- class: MarathonPersonalBest
+  file: marathon_personal_best
+  description: "Marathon : Personal Best"
+  weeks: 30
+  url: https://www.halhigdon.com/training-programs/marathon-training/personal-best/
+- class: MarathonSenior
+  file: marathon_senior
+  description: "Marathon : Senior"
+  weeks: 8
+  url: https://www.halhigdon.com/training-programs/marathon-training/senior/
+- class: Marathon3
+  file: marathon_3
+  description: "Marathon : Marathon 3"
+  weeks: 24
+  url: https://www.halhigdon.com/training-programs/marathon-training/marathon-3/
+- class: MarathonDopeyChallenge
+  file: marathon_dopey_challenge
+  description: "Marathon : Dopey Challenge"
+  weeks: 18
+  url: https://www.halhigdon.com/training-programs/marathon-training/dopey-challenge/
+# Boston is run on Patriots' Day (a Monday), so the final table row holds a single
+# Monday cell instead of a full week. The scraper truncates to 12 full weeks + race day.
+- class: MarathonBostonBound
+  file: marathon_boston_bound
+  description: "Marathon : Boston Bound"
+  weeks: 13
+  partial_final_week: true
+  url: https://www.halhigdon.com/training-programs/marathon-training/boston-bound/

data/tasks/refresh_data.rake

@@ -0,0 +1,156 @@
+# frozen_string_literal: true
+
+require "net/http"
+require "uri"
+require "yaml"
+
+# Re-scrapes the training plans in lib/run_ical/data from halhigdon.com.
+#
+#   rake data:refresh            # every program in tasks/higdon_programs.yml
+#   rake data:refresh[Novice5K]  # just one
+#
+# Every page renders the schedule as a static table with a `Week | Mon..Sun` header
+# row, in miles first and kilometres second. We take the miles table, flatten it
+# row-major into one activity per day, and write a YAML sidecar plus a loader stub.
+module HigdonScraper
+  DATA_DIR = File.expand_path("../lib/run_ical/data", __dir__)
+  MANIFEST = File.expand_path("higdon_programs.yml", __dir__)
+  DAYS_PER_WEEK = 7
+
+  # The final cell of a plan is the race itself. Higdon isn't consistent about it:
+  # most read "<distance> Race", but Novice 10K ends on "10K Run", Walking 10K on
+  # "10K Walk", and the marathons on a bare "Marathon" / "Boston Marathon".
+  RACE = /\A(\d+K (Race|Run|Walk)|(Boston |Half )?Marathon)\z/i
+
+  class ScrapeError < StandardError; end
+
+  module_function
+
+  # Every program in the manifest, or the single one named.
+  def programs(only = nil)
+    all = YAML.safe_load_file(MANIFEST)
+    return all unless only
+
+    selected = all.select { |program| program["class"] == only }
+    raise ScrapeError, "unknown program #{only}" if selected.empty?
+
+    selected
+  end
+
+  def scrape(program)
+    rows = schedule_rows(Nokogiri::HTML(fetch(program["url"])), program)
+    unless rows.size == program["weeks"]
+      raise ScrapeError, "#{program['class']}: expected #{program['weeks']} weeks, got #{rows.size}"
+    end
+
+    activities = rows.flat_map { |row| week_activities(row) }
+    activities = trim_partial_final_week(activities) if program["partial_final_week"]
+    validate!(program, activities)
+    activities
+  end
+
+  def fetch(url)
+    uri = URI(url)
+    response = Net::HTTP.get_response(uri, "User-Agent" => "run_ical data refresh")
+    raise ScrapeError, "#{url}: HTTP #{response.code}" unless response.is_a?(Net::HTTPSuccess)
+
+    response.body
+  end
+
+  def schedule_rows(doc, program)
+    table = schedule_table(doc)
+    raise ScrapeError, "#{program['class']}: no schedule table found" unless table
+
+    table.css("tbody tr").map { |tr| tr.css("td").map { |td| cell_text(td) } }.reject(&:empty?)
+  end
+
+  # The first table whose header starts "Week | Mon" -- the walkers half marathon page
+  # leads with an unrelated pace legend, and the kilometre table always follows the
+  # miles one, so first-match gives us miles.
+  def schedule_table(doc)
+    doc.css("table").find do |candidate|
+      headers = candidate.css("th").map { |th| cell_text(th) }
+      headers[0] == "Week" && headers[1] == "Mon"
+    end
+  end
+
+  # Cells are taken by position, not by header: a few pages ship an empty <th> where
+  # "Wed" belongs. The 8K pages add a ninth "Miles" total column, and on the novice
+  # page the race sits in it rather than under Sunday -- pull it back onto race day.
+  def week_activities(row)
+    week = row[1, DAYS_PER_WEEK]
+    total = row[DAYS_PER_WEEK + 1]
+    week[-1] = total if total&.match?(RACE)
+    week
+  end
+
+  # Boston's last row is a lone Monday race day padded with six empty cells.
+  def trim_partial_final_week(activities)
+    race = activities.reverse.find { |activity| !activity.empty? }
+    activities.take(activities.size - DAYS_PER_WEEK) << race
+  end
+
+  def validate!(program, activities)
+    blank = activities.index(&:empty?)
+    raise ScrapeError, "#{program['class']}: empty activity at index #{blank}" if blank
+
+    expected = expected_size(program)
+    unless activities.size == expected
+      raise ScrapeError, "#{program['class']}: expected #{expected} days, got #{activities.size}"
+    end
+    return if activities.last.match?(RACE)
+
+    raise ScrapeError, "#{program['class']}: plan ends on #{activities.last.inspect}, not a race"
+  end
+
+  def expected_size(program)
+    weeks = program["weeks"]
+    return weeks * DAYS_PER_WEEK unless program["partial_final_week"]
+
+    ((weeks - 1) * DAYS_PER_WEEK) + 1
+  end
+
+  # Higdon's markup is littered with non-breaking spaces and hard line breaks.
+  def cell_text(node)
+    node.text.gsub(/[[:space:]]+/, " ").strip
+  end
+
+  def write(program, activities)
+    write_yaml(program, activities)
+    write_loader(program)
+  end
+
+  # Ruby's emitter quotes numeric-looking strings, which keeps the marathon plans'
+  # bare mileage cells ("6", "20") loading back as Strings.
+  def write_yaml(program, activities)
+    body = activities.map { |activity| { "activity" => activity } }
+    File.write(File.join(DATA_DIR, "#{program['file']}.yaml"), body.to_yaml)
+  end
+
+  def write_loader(program)
+    File.write(File.join(DATA_DIR, "#{program['file']}.rb"), <<~RUBY)
+      module RunIcal
+        module Data
+          class #{program['class']}
+            DESCRIPTION = #{program['description'].inspect}
+            DATA = YAML.safe_load_file(File.join(__dir__, '#{program['file']}.yaml'), symbolize_names: true)
+          end
+        end
+      end
+    RUBY
+  end
+end
+
+namespace :data do
+  desc "Re-scrape training plan data from halhigdon.com (optionally one program)"
+  task :refresh, [:program] do |_t, args|
+    require "nokogiri"
+
+    HigdonScraper.programs(args[:program]).each do |program|
+      activities = HigdonScraper.scrape(program)
+      HigdonScraper.write(program, activities)
+      puts format("%<name>-26s %<days>3d days  %<race>s",
+                  name: program["class"], days: activities.size, race: activities.last)
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: run_ical
 version: !ruby/object:Gem::Version
-  version: 0.5.0
+  version: 0.6.0
 platform: ruby
 authors:
 - Sean Cashin
@@ -10,19 +10,19 @@ cert_chain: []
 date: 1980-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: yaml
+  name: activesupport
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 5.0.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 5.0.0
 - !ruby/object:Gem::Dependency
   name: base64
   requirement: !ruby/object:Gem::Requirement
@@ -51,6 +51,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: icalendar
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.0.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.0.0
 - !ruby/object:Gem::Dependency
   name: logger
   requirement: !ruby/object:Gem::Requirement
@@ -94,103 +108,103 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: icalendar
+  name: yaml
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 2.0.0
+        version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 2.0.0
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: activesupport
+  name: irb
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 5.0.0
-  type: :runtime
+        version: '0'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 5.0.0
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: bundler
+  name: nokogiri
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '2.0'
+        version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '2.0'
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
 - !ruby/object:Gem::Dependency
-  name: rspec
+  name: rdoc
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: irb
+  name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '3.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '3.0'
 - !ruby/object:Gem::Dependency
-  name: rdoc
+  name: rubocop
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '1.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '1.0'
 description: Uses Hal Higdon's plans for creating a calendar
 email:
 - scashin133@gmail.com
@@ -199,10 +213,10 @@ executables:
 extensions: []
 extra_rdoc_files: []
 files:
+- ".github/workflows/ci.yml"
 - ".gitignore"
 - ".rspec"
 - ".rubocop.yml"
-- ".travis.yml"
 - Gemfile
 - LICENSE.txt
 - README.md
@@ -212,30 +226,75 @@ files:
 - exe/run_ical
 - lib/run_ical.rb
 - lib/run_ical/data/advanced_10_k.rb
+- lib/run_ical/data/advanced_10_k.yaml
 - lib/run_ical/data/advanced_15_k.rb
+- lib/run_ical/data/advanced_15_k.yaml
 - lib/run_ical/data/advanced_5_k.rb
+- lib/run_ical/data/advanced_5_k.yaml
 - lib/run_ical/data/advanced_8_k.rb
+- lib/run_ical/data/advanced_8_k.yaml
 - lib/run_ical/data/half_marathon_advanced.rb
+- lib/run_ical/data/half_marathon_advanced.yaml
 - lib/run_ical/data/half_marathon_hm3.rb
+- lib/run_ical/data/half_marathon_hm3.yaml
 - lib/run_ical/data/half_marathon_intermediate1.rb
+- lib/run_ical/data/half_marathon_intermediate1.yaml
 - lib/run_ical/data/half_marathon_intermediate2.rb
+- lib/run_ical/data/half_marathon_intermediate2.yaml
 - lib/run_ical/data/half_marathon_novice1.rb
+- lib/run_ical/data/half_marathon_novice1.yaml
 - lib/run_ical/data/half_marathon_novice2.rb
+- lib/run_ical/data/half_marathon_novice2.yaml
+- lib/run_ical/data/half_marathon_walkers.rb
+- lib/run_ical/data/half_marathon_walkers.yaml
 - lib/run_ical/data/intermediate_10_k.rb
+- lib/run_ical/data/intermediate_10_k.yaml
 - lib/run_ical/data/intermediate_15_k.rb
+- lib/run_ical/data/intermediate_15_k.yaml
 - lib/run_ical/data/intermediate_5_k.rb
+- lib/run_ical/data/intermediate_5_k.yaml
 - lib/run_ical/data/intermediate_8_k.rb
+- lib/run_ical/data/intermediate_8_k.yaml
+- lib/run_ical/data/marathon_3.rb
+- lib/run_ical/data/marathon_3.yaml
+- lib/run_ical/data/marathon_advanced1.rb
+- lib/run_ical/data/marathon_advanced1.yaml
+- lib/run_ical/data/marathon_advanced2.rb
+- lib/run_ical/data/marathon_advanced2.yaml
+- lib/run_ical/data/marathon_boston_bound.rb
+- lib/run_ical/data/marathon_boston_bound.yaml
+- lib/run_ical/data/marathon_dopey_challenge.rb
+- lib/run_ical/data/marathon_dopey_challenge.yaml
+- lib/run_ical/data/marathon_intermediate1.rb
+- lib/run_ical/data/marathon_intermediate1.yaml
+- lib/run_ical/data/marathon_intermediate2.rb
+- lib/run_ical/data/marathon_intermediate2.yaml
 - lib/run_ical/data/marathon_novice1.rb
+- lib/run_ical/data/marathon_novice1.yaml
 - lib/run_ical/data/marathon_novice2.rb
 - lib/run_ical/data/marathon_novice2.yaml
+- lib/run_ical/data/marathon_novice_supreme.rb
+- lib/run_ical/data/marathon_novice_supreme.yaml
+- lib/run_ical/data/marathon_personal_best.rb
+- lib/run_ical/data/marathon_personal_best.yaml
+- lib/run_ical/data/marathon_senior.rb
+- lib/run_ical/data/marathon_senior.yaml
 - lib/run_ical/data/novice_10_k.rb
+- lib/run_ical/data/novice_10_k.yaml
 - lib/run_ical/data/novice_15_k.rb
+- lib/run_ical/data/novice_15_k.yaml
 - lib/run_ical/data/novice_5_k.rb
+- lib/run_ical/data/novice_5_k.yaml
 - lib/run_ical/data/novice_8_k.rb
+- lib/run_ical/data/novice_8_k.yaml
 - lib/run_ical/data/walkers_5_k.rb
+- lib/run_ical/data/walkers_5_k.yaml
 - lib/run_ical/data/walking_10_k.rb
+- lib/run_ical/data/walking_10_k.yaml
 - lib/run_ical/version.rb
 - run_ical.gemspec
+- tasks/higdon_programs.yml
+- tasks/refresh_data.rake
 homepage: https://github.com/scashin133/run_ical
 licenses:
 - MIT
@@ -247,14 +306,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '0'
+      version: '3.1'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.6.9
+rubygems_version: 4.0.16
 specification_version: 4
 summary: Generate an ical file for a race
 test_files: []

data/.travis.yml

@@ -1,5 +0,0 @@
-sudo: false
-language: ruby
-rvm:
-  - 2.3.5
-before_install: gem install bundler -v 1.16.0