kaplan 0.2.4 → 0.2.5

data/README.md

@@ -29,7 +29,7 @@ It's easy to add support for another framework or ORM, though.
 
 ## Usage
 
-Here are the Rake tasks.
+### Available Rake tasks
 
 * **kaplan:db:reset** - Resets the database corresponding to the given environment by re-creating it from the development schema and then re-seeding it. Accepts: `RAILS_ENV` (required, must not be "development").
 * **kaplan:db:seed** - Seeds the database corresponding to the given environment with bootstrap data. The tables that will be seeded are truncated first so you don't have to. Accepts: `RAILS_ENV` (optional, default is "development").
@@ -38,36 +38,61 @@ Here are the Rake tasks.
 * **kaplan:db:create** - Creates the database corresponding to the given environment. Accepts: `RAILS_ENV` (optional, default is "development").
 * **kaplan:db:drop** - Drops the database corresponding to the given environment. Accepts: `RAILS_ENV` (optional, default is "development").
 
-Seeding is a bit different than Rails's built-in seeding, so I'll talk a little bit about that. Basically, you can still keep seeds in `db/seeds.rb` like before, but if you want to break them into separate files, you can put them in `seeds/`, and Kaplan will look for them there too. In fact, the seed files don't have to be Ruby, they can also be YAML or text files. So a YAML seed file would look like:
+### Seeding
 
-    people:
+Seeding is a bit different than Rails's built-in seeding, so I'll talk a little bit about that. Basically, you can still keep seeds in `db/seeds.rb` like before, but if you want to break them into separate files, you can put them in `seeds/`, and Kaplan will look for them there too. You name the file after the table the data in the file should be inserted into -- so for instance, you could have a file `seeds/people.rb`, and this might look like:
+
+    Person.create!(
+      :first_name => "Joe",
+      :last_name => "Bloe",
+      :likes => "eating spaghetti, watching The Godfather"
+    )
+
+The seed files don't have to be Ruby, though; they can also be YAML or text files. So a YAML seed file would look like this:
+
+    Person:
+      -
+        first_name: Joe
+        last_name: Bloe
+        likes: "eating spaghetti, watching The Godfather"
+      -
+        first_name: Jill
+        last_name: Schmoe
+        likes: "biking, watching Family Guy"
+        
+For a YAML file, the key at the top indicates which model should be fed the records underneath. What if you have an STI situation, where more than one model shares a table? In this case, you keep the filename the same (`seeds/people.yml`), but you can list multiple keys:
+
+    Man:
       -
         first_name: Joe
         last_name: Bloe
         likes: eating spaghetti, watching The Godfather
+    Woman:
       -
         first_name: Jill
         last_name: Schmoe
         likes: biking, watching Family Guy
 
-and a text file would look like (note it looks kind of like CSV, but not quite, because you can't quote fields):
+A text file looks like this (note it looks kind of like CSV, but not quite, because you can't quote fields):
 
     # first name, last name, likes
     Joe  Bloe  eating spaghetti, watching The Godfather
     Jill  Schmoe  biking, watching Family Guy
 
-In addition, you can have environment-specific files. So if I have a directory structure that looks like:
+One thing about the naming of the files. What if you need to run your seeds in a certain order? Well, the files in the `seeds/` directory are sorted before being run, so you can simply prepend numbers to the name of the file. For instance, if I am seeding two tables, `cars` and `people`, but I want to make sure `people` is seeded before `cars`, I could name them `seeds/01_people.yml` and `seeds/02_cars.rb`.
+
+You can also keep seed files per environment. So if I have a directory structure that looks like:
 
     seeds/
-      people.txt
+      01_people.txt
       development/
-        cars.rb
-        venues.rb
+        02_cars.rb
+        03_venues.rb
       test/
-        cars.rb
-        car_types.yml
+        02_cars.rb
+        03_car_types.yml
 
-Then, when I run `rake kaplan:db:seed` in different environments, this is which tables get populated:
+If I were to run `rake kaplan:db:seed` in all three environments, this is which tables would get populated:
 
     +-------------+-------------------------+
     | Environment | Tables                  |
@@ -77,9 +102,13 @@ Then, when I run `rake kaplan:db:seed` in different environments, this is which
     | test        | people, cars, car_types |
     +-------------+-------------------------+
     
-Besides running seeds in a certain environment, you also have the option to run specific seed files, regardless of which environment they're stored under. So if for some reason we wanted to run the cars.rb seed in the development environment and the car_types.yml seed in the test environment, we could do that like this:
+What if you need to run a certain seed file even if you're not in that environment? Well, you can do that too. So if for some reason we wanted to run the cars.rb seed in the development environment and the car_types.yml seed in the test environment, we could do that like this:
+
+    rake db:seed ONLY=seeds/development/02_cars.rb,seeds/test/03_car_types.yml
+
+Note how we have to include the numbers before each file, because they are part of the file name. This would ordinarily be annoying, but the list you pass to `ONLY` (or, `SEED`, for compatibility) are actually patterns, so you can make them as short as you want provided the patterns do not yield ambiguous results. For instance, if we wanted to run the `venues` seed from the development environment and the `car_types` seed from the test environment, we could just say this:
 
-    rake db:seed ONLY=seeds/development/cars.rb,seeds/test/car_types.yml
+    rake db:seed ONLY=venues,car_types
 
 ## Author/License
 

data/lib/kaplan.rb

@@ -104,29 +104,45 @@ module Kaplan
     end
 
     def seeds(env, options={})
+      available_files = [
+        "#{project_root}/db/seeds.rb",
+        Dir["#{project_root}/seeds/*.{yml,yaml,rb,txt,csv}"],
+        Dir["#{project_root}/seeds/#{env}/*.{yml,yaml,rb,txt,csv}"]
+      ].flatten
+      available_files_with_shorter_path = available_files.map do |fn|
+        [ fn, fn.sub("#{project_root}/", "") ]
+      end
       if options[:only]
-        files = options[:only].
-          map {|path| Dir["#{project_root}/#{path}"] }.
-          flatten.
-          select {|file| File.file?(file) && file =~ /\.(yml|yaml|rb|txt|csv)/ }
-
-        if files.empty?
-          raise "Error in Kaplan.seeds: Couldn't find the seed files you specified."
+        files = options[:only].map do |pattern|
+          pattern = (pattern =~ %r{/}) ? Regexp.new(Regexp.escape(pattern)) : %r[/(?:\d+_)?#{pattern}]
+          possible_files = available_files_with_shorter_path.select {|_, sfn| sfn =~ pattern }
+          if possible_files.empty?
+            msg = "Error in Kaplan.seeds: <#{pattern}> doesn't match any files. Available files:\n"
+            available_files_with_shorter_path.map do |_, sfn|
+              msg << " - #{sfn}\n"
+            end
+            raise msg
+          end
+          if possible_files.size > 1
+            msg = "Error in Kaplan.seeds: <#{pattern}> matches more than one file. Available files:\n"
+            available_files_with_shorter_path.map do |_, sfn|
+              msg << " - #{sfn}\n"
+            end
+            raise msg
+          end
+          possible_files[0][0]
         end
       else
-        files = [
-          "#{project_root}/db/seeds.rb",
-          Dir["#{project_root}/seeds/*.{yml,yaml,rb,txt,csv}"],
-          Dir["#{project_root}/seeds/#{env}/*.{yml,yaml,rb,txt,csv}"]
-        ].flatten
+        files = available_files
       end
+      raise "Error in Kaplan.seeds: Got an empty set of seeds." if files.empty?
       files.sort.enum_for(:each_with_index).map do |filename, i|
         basename = ::File.basename(filename)
         # Ignore numbers at the beginning of the filename, as they are used for sorting
         basename =~ /^(?:\d+_)?(.+?)\.([^.]+)$/
         extension = $2.downcase
         # collection_name and model only applies to files within seeds/
-        if i > 0
+        if filename != "#{project_root}/db/seeds.rb"
           collection_name = $1
           model = collection_name.classify.constantize
         end

data/lib/kaplan/version.rb

@@ -1,3 +1,3 @@
 module Kaplan
-  VERSION = "0.2.4"
+  VERSION = "0.2.5"
 end

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: kaplan
 version: !ruby/object:Gem::Version 
-  hash: 31
+  hash: 29
   prerelease: false
   segments: 
   - 0
   - 2
-  - 4
-  version: 0.2.4
+  - 5
+  version: 0.2.5
 platform: ruby
 authors: 
 - Elliot Winkler
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-04-04 00:00:00 -06:00
+date: 2011-04-20 00:00:00 -06:00
 default_executable: 
 dependencies: []