Class: Job

Inherits:
Object
  • Object
show all
Defined in:
lib/job.rb

Overview

Job is a simplistic, atomic job processor that works off of the filesystem to get it’s list of tasks. It executes these tasks in order of time, oldest first. Each task is named and assigned a procedure to execute when it is encountered, but named tasks do not necessarily require execution at any point.

Example:

jt1 = JobTask.new do
    puts "This is job one"
    0
end

jt2 = JobTask.new do
    puts "This is job two"
    1
end

# jobs require a directory where they can place their files.
job = Job.new '/Users/erikh/tmp'

# naming a task
job.name_task jt1, :one
job.name_task jt2, :two

# adding jobs
job.add_job :two
# this is necessary in this example to create a delay of time
sleep 10
job.add_job :one

# run returns the results for further evaluation and
# short-circuiting.
job.run do |result|
    puts result
end

# Note that this does not clobber names
job.delete_all_jobs

job.add_job :one
sleep 10
job.add_job :two
# run_obliviously does not care about the result.
job.run_obliviously

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(jobdir) ⇒ Job

Creates a new Job object. Requires that the jobdir be defined – it does not necessarily have to exist already, it will be created.



96
97
98
99
100
101
 
# File 'lib/job.rb', line 96

def initialize(jobdir)
	fail "Must define jobdir" unless jobdir
	@jobdir = jobdir.dup
	@jobdir.freeze
	@jobnames = Hash.new
end

Instance Attribute Details

#jobdir ⇒ Object (readonly)

Returns the value of attribute jobdir.



90
91
92
 
# File 'lib/job.rb', line 90

def jobdir
  @jobdir
end

Instance Method Details

#add_job(name) ⇒ Object

Adds a job. This does not check to see if the job is named yet, so run will ignore this job if there is no named task to go along with it.



108
109
110
111
112
113
 
# File 'lib/job.rb', line 108

def add_job(name)
	name = @jobdir + '/' + name.to_s

	FileUtils.mkdir_p name
	FileUtils.touch name
end

#delete_all_jobs ⇒ Object

Deletes all jobs.



137
138
139
140
141
 
# File 'lib/job.rb', line 137

def delete_all_jobs
	Dir[@jobdir + '/*'].each do |name|
		FileUtils.rm_r(name)
	end
end

#delete_job_by_name(name) ⇒ Object

This deletes a job by its name.



129
130
131
 
# File 'lib/job.rb', line 129

def delete_job_by_name(name)
	FileUtils.rm_r(@jobdir + '/' + name.to_s)
end

#name_task(task, name) ⇒ Object

Names a task. Given a JobTask object and a name (string or symbol), it enters it into the task table.



120
121
122
123
 
# File 'lib/job.rb', line 120

def name_task(task, name)
	fail "Task must have an execute method" unless task.respond_to? :execute
	@jobnames[name] = task
end

#run ⇒ Object

Runs the tasks in order yielding for each result, allowing the user to back out if problems occur processing the job chain.



148
149
150
151
152
 
# File 'lib/job.rb', line 148

def run
	yield_tasks do |task|
		yield task.execute
	end
end

#run_obliviously ⇒ Object

This runs without checking the job results, but will return the result of the last task executed.



159
160
161
162
163
164
165
166
 
# File 'lib/job.rb', line 159

def run_obliviously
	result = nil
	yield_tasks do |task|
		result = task.execute
	end

	result
end