remote_table 0.1.5
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/.document +5 -0
- data/.gitignore +5 -0
- data/LICENSE +20 -0
- data/README.rdoc +27 -0
- data/Rakefile +66 -0
- data/VERSION +1 -0
- data/lib/remote_table/file/csv.rb +45 -0
- data/lib/remote_table/file/fixed_width.rb +53 -0
- data/lib/remote_table/file/ods.rb +11 -0
- data/lib/remote_table/file/roo_spreadsheet.rb +30 -0
- data/lib/remote_table/file/xls.rb +11 -0
- data/lib/remote_table/file.rb +56 -0
- data/lib/remote_table/package.rb +84 -0
- data/lib/remote_table/request.rb +36 -0
- data/lib/remote_table/transform.rb +32 -0
- data/lib/remote_table.rb +51 -0
- data/remote_table.gemspec +74 -0
- data/test/remote_table_test.rb +101 -0
- data/test/test_helper.rb +10 -0
- metadata +117 -0
data/.document
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
Copyright (c) 2009 Brighter Planet
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
4
|
+
a copy of this software and associated documentation files (the
|
5
|
+
"Software"), to deal in the Software without restriction, including
|
6
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
7
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
8
|
+
permit persons to whom the Software is furnished to do so, subject to
|
9
|
+
the following conditions:
|
10
|
+
|
11
|
+
The above copyright notice and this permission notice shall be
|
12
|
+
included in all copies or substantial portions of the Software.
|
13
|
+
|
14
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
15
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
16
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
17
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
18
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
19
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
20
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.rdoc
ADDED
@@ -0,0 +1,27 @@
|
|
1
|
+
=remote_table
|
2
|
+
|
3
|
+
Remotely open and parse Excel XLS, ODS, CSV and fixed-width tables.
|
4
|
+
|
5
|
+
==Real-life usage
|
6
|
+
|
7
|
+
Used by data_miner (http://github.com/seamusabshere/data_miner)
|
8
|
+
|
9
|
+
==Example
|
10
|
+
|
11
|
+
Taken from <tt>#{GEMDIR}/test/remote_table_test.rb</tt>:
|
12
|
+
|
13
|
+
>> t = RemoteTable.new(:url => 'http://www.fueleconomy.gov/FEG/epadata/98guide6.zip', :filename => '98guide6.csv')
|
14
|
+
=> #<RemoteTable:0x359da50 @transform=#<RemoteTable::Transform:0x359d154 @select=nil, @reject=nil>, @file=#<RemoteTable::File:0x35970c4 @delimiter=nil, @headers=nil, @cut=nil, @filename="98guide6.csv", @skip=nil, @schema_name=nil, @crop=nil, @format=:csv, @trap=nil, @sheet=0, @schema=nil>, @package=#<RemoteTable::Package:0x359c538 @url="http://www.fueleconomy.gov/FEG/epadata/98guide6.zip", @filename="98guide6.csv", @compression=:zip, @packing=nil>, @request=#<RemoteTable::Request:0x3596bec @url="http://www.fueleconomy.gov/FEG/epadata/98guide6.zip">>
|
15
|
+
>> t.rows.first
|
16
|
+
=> {"cyl"=>"6", "eng dscr"=>"DOHC VTEC", "trans dscr"=>"2MODE CLKUP", "trans"=>"Auto(L4)", "cmb"=>"20", "2pv"=>nil, "carline name"=>"NSX", "displ"=>"3.0", "ucmb"=>"23.5311", "hpv"=>nil, "4pv"=>nil, "Class"=>"TWO SEATERS", "Manufacturer"=>"ACURA", "fl"=>"P", "2lv"=>nil, "G"=>nil, "hlv"=>nil, "drv"=>"R", "cty"=>"18", "ucty"=>"19.8733", "S"=>nil, "4lv"=>nil, "fcost"=>"1050", "T"=>nil, "hwy"=>"24", "uhwy"=>"30.3612"}
|
17
|
+
|
18
|
+
See the test file and also data_miner examples of custom parsers.
|
19
|
+
|
20
|
+
==Authors
|
21
|
+
|
22
|
+
* Seamus Abshere <seamus@abshere.net>
|
23
|
+
* Andy Rossmeissl <andy@rossmeissl.net>
|
24
|
+
|
25
|
+
== Copyright
|
26
|
+
|
27
|
+
Copyright (c) 2009 Brighter Planet. See LICENSE for details.
|
data/Rakefile
ADDED
@@ -0,0 +1,66 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'rake'
|
3
|
+
|
4
|
+
begin
|
5
|
+
require 'jeweler'
|
6
|
+
Jeweler::Tasks.new do |gem|
|
7
|
+
gem.name = "remote_table"
|
8
|
+
gem.summary = %Q{Remotely open and parse XLS, ODS, CSV and fixed-width tables.}
|
9
|
+
gem.description = %Q{Remotely open and parse Excel XLS, ODS, CSV and fixed-width tables.}
|
10
|
+
gem.email = "seamus@abshere.net"
|
11
|
+
gem.homepage = "http://github.com/seamusabshere/remote_table"
|
12
|
+
gem.authors = ["Seamus Abshere", "Andy Rossmeissl"]
|
13
|
+
%w{ activesupport fastercsv ryanwood-slither }.each { |name| gem.add_dependency name }
|
14
|
+
gem.add_dependency 'roo', '1.3.11'
|
15
|
+
gem.require_path = "lib"
|
16
|
+
gem.files.include %w(lib/remote_table) unless gem.files.empty? # seems to fail once it's in the wild
|
17
|
+
gem.rdoc_options << '--line-numbers' << '--inline-source'
|
18
|
+
gem.requirements << 'curl'
|
19
|
+
gem.rubyforge_project = "remotetable"
|
20
|
+
end
|
21
|
+
Jeweler::GemcutterTasks.new
|
22
|
+
Jeweler::RubyforgeTasks.new do |rubyforge|
|
23
|
+
rubyforge.doc_task = "rdoc"
|
24
|
+
end
|
25
|
+
rescue LoadError
|
26
|
+
puts "Jeweler (or a dependency) not available. Install it with: sudo gem install jeweler"
|
27
|
+
end
|
28
|
+
|
29
|
+
require 'rake/testtask'
|
30
|
+
Rake::TestTask.new(:test) do |test|
|
31
|
+
test.libs << 'lib' << 'test'
|
32
|
+
test.pattern = 'test/**/*_test.rb'
|
33
|
+
test.verbose = true
|
34
|
+
end
|
35
|
+
|
36
|
+
begin
|
37
|
+
require 'rcov/rcovtask'
|
38
|
+
Rcov::RcovTask.new do |test|
|
39
|
+
test.libs << 'test'
|
40
|
+
test.pattern = 'test/**/*_test.rb'
|
41
|
+
test.verbose = true
|
42
|
+
end
|
43
|
+
rescue LoadError
|
44
|
+
task :rcov do
|
45
|
+
abort "RCov is not available. In order to run rcov, you must: sudo gem install spicycode-rcov"
|
46
|
+
end
|
47
|
+
end
|
48
|
+
|
49
|
+
|
50
|
+
|
51
|
+
|
52
|
+
task :default => :test
|
53
|
+
|
54
|
+
require 'rake/rdoctask'
|
55
|
+
Rake::RDocTask.new do |rdoc|
|
56
|
+
if File.exist?('VERSION')
|
57
|
+
version = File.read('VERSION')
|
58
|
+
else
|
59
|
+
version = ""
|
60
|
+
end
|
61
|
+
|
62
|
+
rdoc.rdoc_dir = 'rdoc'
|
63
|
+
rdoc.title = "remote_table #{version}"
|
64
|
+
rdoc.rdoc_files.include('README*')
|
65
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
66
|
+
end
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.1.5
|
@@ -0,0 +1,45 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
module Csv
|
3
|
+
def each_row(&block)
|
4
|
+
skip_rows!
|
5
|
+
FasterCSV.parse(open(path), fastercsv_options) do |row|
|
6
|
+
if row.respond_to?(:fields) # it's a traditional fastercsv row hash
|
7
|
+
next if row.fields.compact.blank?
|
8
|
+
hash = HashWithIndifferentAccess.new(row.to_hash)
|
9
|
+
else # it's an array, which i think happens if you're using :headers => nil or :col_sep
|
10
|
+
next if row.compact.blank?
|
11
|
+
index = 0
|
12
|
+
hash = row.inject(ActiveSupport::OrderedHash.new) { |memo, element| memo[index] = element; index += 1; memo }
|
13
|
+
end
|
14
|
+
yield hash
|
15
|
+
end
|
16
|
+
ensure
|
17
|
+
restore_rows!
|
18
|
+
end
|
19
|
+
|
20
|
+
private
|
21
|
+
|
22
|
+
def fastercsv_options
|
23
|
+
fastercsv_options = { :skip_blanks => true } # ...and this will skip []
|
24
|
+
if headers == false
|
25
|
+
fastercsv_options.merge!(:headers => nil)
|
26
|
+
else
|
27
|
+
fastercsv_options.merge!(:headers => :first_row)
|
28
|
+
end
|
29
|
+
fastercsv_options.merge!(:col_sep => delimiter) if delimiter
|
30
|
+
fastercsv_options
|
31
|
+
end
|
32
|
+
|
33
|
+
def skip_rows!
|
34
|
+
return unless skip
|
35
|
+
original = "#{path}.original"
|
36
|
+
FileUtils.cp(path, original)
|
37
|
+
`cat #{original} | tail -n +#{skip + 1} > #{path}`
|
38
|
+
end
|
39
|
+
|
40
|
+
def restore_rows!
|
41
|
+
return unless skip
|
42
|
+
FileUtils.mv "#{path}.original", path
|
43
|
+
end
|
44
|
+
end
|
45
|
+
end
|
@@ -0,0 +1,53 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
module FixedWidth
|
3
|
+
def each_row(&block)
|
4
|
+
crop_rows!
|
5
|
+
skip_rows!
|
6
|
+
cut_columns!
|
7
|
+
a = Slither.parse(path, schema_name)
|
8
|
+
a[:rows].each { |row| yield HashWithIndifferentAccess.new(row) }
|
9
|
+
ensure
|
10
|
+
uncut_columns!
|
11
|
+
unskip_rows!
|
12
|
+
uncrop_rows!
|
13
|
+
end
|
14
|
+
|
15
|
+
private
|
16
|
+
|
17
|
+
def cut_columns!
|
18
|
+
return unless cut
|
19
|
+
original = "#{path}.uncut"
|
20
|
+
FileUtils.cp(path, original)
|
21
|
+
`cat #{original} | cut -c #{cut} > #{path}`
|
22
|
+
end
|
23
|
+
|
24
|
+
def uncut_columns!
|
25
|
+
return unless cut
|
26
|
+
FileUtils.mv "#{path}.uncut", path
|
27
|
+
end
|
28
|
+
|
29
|
+
def skip_rows!
|
30
|
+
return unless skip
|
31
|
+
original = "#{path}.unskipped"
|
32
|
+
FileUtils.cp(path, original)
|
33
|
+
`cat #{original} | tail -n +#{skip + 1} > #{path}`
|
34
|
+
end
|
35
|
+
|
36
|
+
def unskip_rows!
|
37
|
+
return unless skip
|
38
|
+
FileUtils.mv "#{path}.unskipped", path
|
39
|
+
end
|
40
|
+
|
41
|
+
def crop_rows!
|
42
|
+
return unless crop
|
43
|
+
original = "#{path}.uncropped"
|
44
|
+
FileUtils.cp(path, original)
|
45
|
+
`cat #{original} | tail -n +#{crop.first} | head -n #{crop.last - crop.first + 1} > #{path}`
|
46
|
+
end
|
47
|
+
|
48
|
+
def uncrop_rows!
|
49
|
+
return unless crop
|
50
|
+
FileUtils.mv "#{path}.uncropped", path
|
51
|
+
end
|
52
|
+
end
|
53
|
+
end
|
@@ -0,0 +1,30 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
module RooSpreadsheet
|
3
|
+
def each_row(&block)
|
4
|
+
headers = {}
|
5
|
+
oo = roo_klass.new(path, nil, :ignore)
|
6
|
+
oo.default_sheet = sheet.is_a?(Numeric) ? oo.sheets[sheet] : sheet
|
7
|
+
for col in (1..oo.last_column)
|
8
|
+
headers[col] = oo.cell(header_row, col)
|
9
|
+
headers[col] = oo.cell(header_row - 1, col) if headers[col].blank? # look up
|
10
|
+
end
|
11
|
+
first_data_row.upto(oo.last_row) do |row|
|
12
|
+
values = {}
|
13
|
+
for col in (1..oo.last_column)
|
14
|
+
values[headers[col]] = oo.cell(row, col).to_s.gsub(/<[^>]+>/, '').strip
|
15
|
+
end
|
16
|
+
yield HashWithIndifferentAccess.new(values)
|
17
|
+
end
|
18
|
+
end
|
19
|
+
|
20
|
+
private
|
21
|
+
|
22
|
+
def header_row
|
23
|
+
1 + skip.to_i
|
24
|
+
end
|
25
|
+
|
26
|
+
def first_data_row
|
27
|
+
1 + header_row
|
28
|
+
end
|
29
|
+
end
|
30
|
+
end
|
@@ -0,0 +1,56 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
class File
|
3
|
+
attr_accessor :filename, :format, :delimiter, :skip, :cut, :crop, :sheet, :headers, :schema, :schema_name, :trap
|
4
|
+
attr_accessor :path
|
5
|
+
|
6
|
+
def initialize(bus)
|
7
|
+
@filename = bus[:filename]
|
8
|
+
@format = bus[:format] || format_from_filename
|
9
|
+
@delimiter = bus[:delimiter]
|
10
|
+
@sheet = bus[:sheet] || 0
|
11
|
+
@skip = bus[:skip] # rows
|
12
|
+
@crop = bus[:crop] # rows
|
13
|
+
@cut = bus[:cut] # columns
|
14
|
+
@headers = bus[:headers]
|
15
|
+
@schema = bus[:schema]
|
16
|
+
@schema_name = bus[:schema_name]
|
17
|
+
@trap = bus[:trap]
|
18
|
+
extend "RemoteTable::#{format.to_s.camelcase}".constantize
|
19
|
+
end
|
20
|
+
|
21
|
+
def tabulate(path)
|
22
|
+
define_fixed_width_schema! if format == :fixed_width and schema.is_a?(Array) # TODO move to generic subclass callback
|
23
|
+
self.path = path
|
24
|
+
self
|
25
|
+
end
|
26
|
+
|
27
|
+
private
|
28
|
+
|
29
|
+
# doesn't support trap or spacer
|
30
|
+
def define_fixed_width_schema!
|
31
|
+
raise "can't define both schema_name and schema" if !schema_name.blank?
|
32
|
+
self.schema_name = "autogenerated_#{filename.gsub(/[^a-z0-9_]/i, '')}".to_sym
|
33
|
+
self.trap ||= lambda { true }
|
34
|
+
Slither.define schema_name do |d|
|
35
|
+
d.rows do |row|
|
36
|
+
row.trap(&trap)
|
37
|
+
schema.each do |name, width, options|
|
38
|
+
if name == :spacer
|
39
|
+
row.spacer width
|
40
|
+
else
|
41
|
+
row.column name, width, options
|
42
|
+
end
|
43
|
+
end
|
44
|
+
end
|
45
|
+
end
|
46
|
+
end
|
47
|
+
|
48
|
+
def format_from_filename
|
49
|
+
extname = ::File.extname(filename).gsub('.', '')
|
50
|
+
return :csv if extname.blank?
|
51
|
+
format = [ :xls, :ods ].detect { |i| i == extname.to_sym }
|
52
|
+
format = :csv if format.blank?
|
53
|
+
format
|
54
|
+
end
|
55
|
+
end
|
56
|
+
end
|
@@ -0,0 +1,84 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
class Package
|
3
|
+
attr_accessor :url, :compression, :packing, :filename
|
4
|
+
|
5
|
+
def initialize(bus)
|
6
|
+
@url = bus[:url] or raise "need url"
|
7
|
+
@compression = bus[:compression] || compression_from_basename
|
8
|
+
@packing = bus[:packing] || packing_from_basename_and_compression
|
9
|
+
@filename = bus[:filename] || filename_from_basename_and_compression_and_packing
|
10
|
+
add_hints!(bus)
|
11
|
+
end
|
12
|
+
|
13
|
+
def add_hints!(hash)
|
14
|
+
hash[:filename] = filename unless hash.has_key?(:filename)
|
15
|
+
end
|
16
|
+
|
17
|
+
def stage(path)
|
18
|
+
decompress(path)
|
19
|
+
unpack(path)
|
20
|
+
identify(path)
|
21
|
+
file_path(path)
|
22
|
+
end
|
23
|
+
|
24
|
+
private
|
25
|
+
|
26
|
+
def decompress(path)
|
27
|
+
return unless compression
|
28
|
+
cmd, args = case compression
|
29
|
+
when :zip, :exe
|
30
|
+
["unzip", "-d #{::File.dirname(path)}"]
|
31
|
+
when :bz2
|
32
|
+
'bunzip2'
|
33
|
+
when :gz
|
34
|
+
'gunzip'
|
35
|
+
end
|
36
|
+
move_and_process path, compression, cmd, args
|
37
|
+
end
|
38
|
+
|
39
|
+
def unpack(path)
|
40
|
+
return unless packing
|
41
|
+
cmd, args = case packing
|
42
|
+
when :tar
|
43
|
+
['tar -xf', "-C #{::File.dirname(path)}"]
|
44
|
+
end
|
45
|
+
move_and_process path, packing, cmd, args
|
46
|
+
end
|
47
|
+
|
48
|
+
def move_and_process(path, extname, cmd, args)
|
49
|
+
`mv #{path} #{path}.#{extname} && #{cmd} #{path}.#{extname} #{args}`
|
50
|
+
end
|
51
|
+
|
52
|
+
# ex. A: 2007-01.csv.gz (compression not capable of storing multiple files)
|
53
|
+
# ex. B: 2007-01.tar.gz (packing)
|
54
|
+
# ex. C: 2007-01.zip (compression capable of storing multiple files)
|
55
|
+
# in C but not in the others, we can default to the basename of the package
|
56
|
+
# in order to do this we'll need to mv the uncompressed file on top of the original file
|
57
|
+
def identify(path)
|
58
|
+
::File.mv(path, file_path(path)) if !packing and [ nil, :bz2, :gz ].include?(compression)
|
59
|
+
end
|
60
|
+
|
61
|
+
def file_path(path)
|
62
|
+
::File.join(::File.dirname(path), filename)
|
63
|
+
end
|
64
|
+
|
65
|
+
def basename_parts
|
66
|
+
::File.basename(URI.parse(url).path).split('.').map(&:to_sym)
|
67
|
+
end
|
68
|
+
|
69
|
+
def compression_from_basename
|
70
|
+
[ :zip, :exe, :bz2, :gz ].detect { |i| i == basename_parts.last }
|
71
|
+
end
|
72
|
+
|
73
|
+
def packing_from_basename_and_compression
|
74
|
+
[ :tar ].detect { |i| i == ((basename_parts.last == compression) ? basename_parts[-2] : basename_parts.last) }
|
75
|
+
end
|
76
|
+
|
77
|
+
def filename_from_basename_and_compression_and_packing
|
78
|
+
ary = basename_parts
|
79
|
+
ary.pop if ary.last == compression
|
80
|
+
ary.pop if ary.last == packing
|
81
|
+
ary.join('.')
|
82
|
+
end
|
83
|
+
end
|
84
|
+
end
|
@@ -0,0 +1,36 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
class Request
|
3
|
+
attr_accessor :url, :post_data, :username, :password
|
4
|
+
|
5
|
+
# TODO: support post_data
|
6
|
+
# TODO: support HTTP basic auth
|
7
|
+
def initialize(bus)
|
8
|
+
@url = bus[:url] or raise "need url"
|
9
|
+
end
|
10
|
+
|
11
|
+
def download
|
12
|
+
path = ::File.join(staging_dir_path, 'REMOTE_TABLE_PACKAGE')
|
13
|
+
`curl --silent \"#{url_with_google_docs_handling}\" > #{path}`
|
14
|
+
path
|
15
|
+
end
|
16
|
+
|
17
|
+
private
|
18
|
+
|
19
|
+
def staging_dir_path
|
20
|
+
path = Tempfile.open(url.gsub(/[^a-z0-9]+/i, '_')).path
|
21
|
+
FileUtils.rm_f(path)
|
22
|
+
FileUtils.mkdir(path)
|
23
|
+
at_exit { FileUtils.rm_rf(path) }
|
24
|
+
path
|
25
|
+
end
|
26
|
+
|
27
|
+
def url_with_google_docs_handling
|
28
|
+
url = self.url
|
29
|
+
if url.include?('spreadsheets.google.com')
|
30
|
+
url = url.gsub(/\&output=.*(\&|\z)/, '')
|
31
|
+
url << "&output=csv"
|
32
|
+
end
|
33
|
+
url
|
34
|
+
end
|
35
|
+
end
|
36
|
+
end
|
@@ -0,0 +1,32 @@
|
|
1
|
+
class RemoteTable
|
2
|
+
class Transform
|
3
|
+
attr_accessor :select, :reject, :transform_class, :transform_options, :transform, :raw_table
|
4
|
+
|
5
|
+
def initialize(bus)
|
6
|
+
if transform_params = bus.delete(:transform)
|
7
|
+
@transform_class = transform_params.delete(:class)
|
8
|
+
@transform_options = transform_params
|
9
|
+
@transform = @transform_class.new(@transform_options)
|
10
|
+
@transform.add_hints!(bus)
|
11
|
+
end
|
12
|
+
@select = bus[:select]
|
13
|
+
@reject = bus[:reject]
|
14
|
+
end
|
15
|
+
|
16
|
+
def apply(raw_table)
|
17
|
+
self.raw_table = raw_table
|
18
|
+
self
|
19
|
+
end
|
20
|
+
|
21
|
+
def each_row(&block)
|
22
|
+
raw_table.each_row do |row|
|
23
|
+
virtual_rows = transform ? transform.apply(row) : row # allow transform.apply(row) to return multiple rows
|
24
|
+
Array.wrap(virtual_rows).each do |virtual_row|
|
25
|
+
next if select and !select.call(virtual_row)
|
26
|
+
next if reject and reject.call(virtual_row)
|
27
|
+
yield virtual_row
|
28
|
+
end
|
29
|
+
end
|
30
|
+
end
|
31
|
+
end
|
32
|
+
end
|
data/lib/remote_table.rb
ADDED
@@ -0,0 +1,51 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'activesupport'
|
3
|
+
require 'tempfile'
|
4
|
+
require 'fastercsv'
|
5
|
+
require 'slither'
|
6
|
+
require 'roo'
|
7
|
+
require 'remote_table/transform'
|
8
|
+
require 'remote_table/request'
|
9
|
+
require 'remote_table/package'
|
10
|
+
require 'remote_table/file'
|
11
|
+
require 'remote_table/file/csv'
|
12
|
+
require 'remote_table/file/fixed_width'
|
13
|
+
require 'remote_table/file/roo_spreadsheet'
|
14
|
+
require 'remote_table/file/ods'
|
15
|
+
require 'remote_table/file/xls'
|
16
|
+
|
17
|
+
class RemoteTable
|
18
|
+
attr_accessor :request, :package, :file, :transform
|
19
|
+
attr_accessor :table
|
20
|
+
|
21
|
+
def initialize(bus)
|
22
|
+
@transform = Transform.new(bus)
|
23
|
+
@package = Package.new(bus)
|
24
|
+
@request = Request.new(bus)
|
25
|
+
@file = File.new(bus)
|
26
|
+
end
|
27
|
+
|
28
|
+
def each_row
|
29
|
+
finish_table! unless table
|
30
|
+
table.each_row { |row| yield row }
|
31
|
+
end
|
32
|
+
|
33
|
+
def rows
|
34
|
+
cache_rows! if @_row_cache.nil?
|
35
|
+
@_row_cache
|
36
|
+
end
|
37
|
+
|
38
|
+
private
|
39
|
+
|
40
|
+
def finish_table!
|
41
|
+
package_path = request.download
|
42
|
+
file_path = package.stage(package_path)
|
43
|
+
raw_table = file.tabulate(file_path)
|
44
|
+
self.table = transform.apply(raw_table) # must return something that responds to each_row
|
45
|
+
end
|
46
|
+
|
47
|
+
def cache_rows!
|
48
|
+
@_row_cache = []
|
49
|
+
each_row { |row| @_row_cache << row }
|
50
|
+
end
|
51
|
+
end
|
@@ -0,0 +1,74 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE DIRECTLY
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run the gemspec command
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{remote_table}
|
8
|
+
s.version = "0.1.5"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Seamus Abshere", "Andy Rossmeissl"]
|
12
|
+
s.date = %q{2009-11-02}
|
13
|
+
s.description = %q{Remotely open and parse Excel XLS, ODS, CSV and fixed-width tables.}
|
14
|
+
s.email = %q{seamus@abshere.net}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
".gitignore",
|
22
|
+
"LICENSE",
|
23
|
+
"README.rdoc",
|
24
|
+
"Rakefile",
|
25
|
+
"VERSION",
|
26
|
+
"lib/remote_table.rb",
|
27
|
+
"lib/remote_table/file.rb",
|
28
|
+
"lib/remote_table/file/csv.rb",
|
29
|
+
"lib/remote_table/file/fixed_width.rb",
|
30
|
+
"lib/remote_table/file/ods.rb",
|
31
|
+
"lib/remote_table/file/roo_spreadsheet.rb",
|
32
|
+
"lib/remote_table/file/xls.rb",
|
33
|
+
"lib/remote_table/package.rb",
|
34
|
+
"lib/remote_table/request.rb",
|
35
|
+
"lib/remote_table/transform.rb",
|
36
|
+
"remote_table.gemspec",
|
37
|
+
"test/remote_table_test.rb",
|
38
|
+
"test/test_helper.rb"
|
39
|
+
]
|
40
|
+
s.homepage = %q{http://github.com/seamusabshere/remote_table}
|
41
|
+
s.rdoc_options = ["--charset=UTF-8", "--line-numbers", "--inline-source"]
|
42
|
+
s.require_paths = ["lib"]
|
43
|
+
s.requirements = ["curl"]
|
44
|
+
s.rubyforge_project = %q{remotetable}
|
45
|
+
s.rubygems_version = %q{1.3.5}
|
46
|
+
s.summary = %q{Remotely open and parse XLS, ODS, CSV and fixed-width tables.}
|
47
|
+
s.test_files = [
|
48
|
+
"test/remote_table_test.rb",
|
49
|
+
"test/test_helper.rb"
|
50
|
+
]
|
51
|
+
|
52
|
+
if s.respond_to? :specification_version then
|
53
|
+
current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
|
54
|
+
s.specification_version = 3
|
55
|
+
|
56
|
+
if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
|
57
|
+
s.add_runtime_dependency(%q<activesupport>, [">= 0"])
|
58
|
+
s.add_runtime_dependency(%q<fastercsv>, [">= 0"])
|
59
|
+
s.add_runtime_dependency(%q<ryanwood-slither>, [">= 0"])
|
60
|
+
s.add_runtime_dependency(%q<roo>, ["= 1.3.11"])
|
61
|
+
else
|
62
|
+
s.add_dependency(%q<activesupport>, [">= 0"])
|
63
|
+
s.add_dependency(%q<fastercsv>, [">= 0"])
|
64
|
+
s.add_dependency(%q<ryanwood-slither>, [">= 0"])
|
65
|
+
s.add_dependency(%q<roo>, ["= 1.3.11"])
|
66
|
+
end
|
67
|
+
else
|
68
|
+
s.add_dependency(%q<activesupport>, [">= 0"])
|
69
|
+
s.add_dependency(%q<fastercsv>, [">= 0"])
|
70
|
+
s.add_dependency(%q<ryanwood-slither>, [">= 0"])
|
71
|
+
s.add_dependency(%q<roo>, ["= 1.3.11"])
|
72
|
+
end
|
73
|
+
end
|
74
|
+
|
@@ -0,0 +1,101 @@
|
|
1
|
+
require 'test_helper'
|
2
|
+
|
3
|
+
class FuelOilParser
|
4
|
+
def initialize(options = {})
|
5
|
+
# nothing
|
6
|
+
end
|
7
|
+
def add_hints!(bus)
|
8
|
+
bus[:sheet] = 'Data 1'
|
9
|
+
bus[:skip] = 2
|
10
|
+
bus[:select] = lambda { |row| row[:year] > 1989 }
|
11
|
+
end
|
12
|
+
def apply(row)
|
13
|
+
virtual_rows = []
|
14
|
+
row.keys.grep(/(.*) Residual Fuel Oil/) do |location_column_name|
|
15
|
+
next if (cost = row[location_column_name]).blank? or (date = row['Date']).blank?
|
16
|
+
if $1.starts_with?('U.S.')
|
17
|
+
locatable = "united_states (Country)"
|
18
|
+
elsif $1.include?('PADD')
|
19
|
+
/\(PADD (.*)\)/.match($1)
|
20
|
+
next if $1 == '1' # skip PADD 1 because we always prefer subdistricts
|
21
|
+
locatable = "#{$1} (PetroleumAdministrationForDefenseDistrict)"
|
22
|
+
else
|
23
|
+
locatable = "#{$1} (State)"
|
24
|
+
end
|
25
|
+
date = Time.parse(date)
|
26
|
+
virtual_rows << HashWithIndifferentAccess.new(
|
27
|
+
:locatable => locatable,
|
28
|
+
:cost => cost,
|
29
|
+
:year => date.year,
|
30
|
+
:month => date.month
|
31
|
+
)
|
32
|
+
end
|
33
|
+
virtual_rows
|
34
|
+
end
|
35
|
+
end
|
36
|
+
|
37
|
+
class RemoteTableTest < Test::Unit::TestCase
|
38
|
+
should "open an XLS inside a zip file" do
|
39
|
+
t = RemoteTable.new(:url => 'http://www.fueleconomy.gov/FEG/epadata/02data.zip', :filename => 'guide_jan28.xls')
|
40
|
+
assert_equal 'ACURA', t.rows.first['Manufacturer']
|
41
|
+
assert_equal 'NSX', t.rows.first['carline name']
|
42
|
+
assert_equal 'VOLVO', t.rows.last['Manufacturer']
|
43
|
+
assert_equal 'V70 XC AWD', t.rows.last['carline name']
|
44
|
+
end
|
45
|
+
|
46
|
+
should "have indifferent hash access" do
|
47
|
+
t = RemoteTable.new(:url => 'http://www.fueleconomy.gov/FEG/epadata/02data.zip', :filename => 'guide_jan28.xls')
|
48
|
+
assert_equal 'ACURA', t.rows.first['Manufacturer'.to_sym]
|
49
|
+
assert_equal 'NSX', t.rows.first['carline name'.to_sym]
|
50
|
+
assert_equal 'VOLVO', t.rows.last['Manufacturer'.to_sym]
|
51
|
+
assert_equal 'V70 XC AWD', t.rows.last['carline name'.to_sym]
|
52
|
+
end
|
53
|
+
|
54
|
+
should "open a Google Docs url" do
|
55
|
+
t = RemoteTable.new(:url => 'http://spreadsheets.google.com/pub?key=t5HM1KbaRngmTUbntg8JwPA')
|
56
|
+
assert_equal 'Gulf Coast', t.rows.first['PAD district name']
|
57
|
+
assert_equal 'AL', t.rows.first['State']
|
58
|
+
assert_equal 'Rocky Mountain', t.rows.last['PAD district name']
|
59
|
+
assert_equal 'WY', t.rows.last['State']
|
60
|
+
end
|
61
|
+
|
62
|
+
should "open an ODS" do
|
63
|
+
t = RemoteTable.new(:url => 'http://static.brighterplanet.com/science/profiler/footprint_model.ods', :sheet => 'Export')
|
64
|
+
assert_equal 'automobiles', t.rows.first['component']
|
65
|
+
assert_equal 2005.0, t.rows.first['period'].to_f
|
66
|
+
end
|
67
|
+
|
68
|
+
should "open a CSV inside a zip file" do
|
69
|
+
t = RemoteTable.new(:url => 'http://www.fueleconomy.gov/FEG/epadata/98guide6.zip', :filename => '98guide6.csv')
|
70
|
+
assert_equal 'ACURA', t.rows.first['Manufacturer']
|
71
|
+
assert_equal 'NSX', t.rows.first['carline name']
|
72
|
+
assert_equal 'TOYOTA', t.rows.last['Manufacturer']
|
73
|
+
assert_equal 'RAV4 SOFT TOP 4WD', t.rows.last['carline name']
|
74
|
+
end
|
75
|
+
|
76
|
+
should "open a fixed-width file with an inline schema inside a zip file" do
|
77
|
+
t = RemoteTable.new(:url => 'http://www.fueleconomy.gov/FEG/epadata/00data.zip',
|
78
|
+
:filename => 'Gd6-dsc.txt',
|
79
|
+
:format => :fixed_width,
|
80
|
+
:crop => 21..26, # inclusive
|
81
|
+
:cut => '2-',
|
82
|
+
:select => lambda { |row| /\A[A-Z]/.match row[:code] },
|
83
|
+
:schema => [[ :code, 2, { :type => :string } ],
|
84
|
+
[ :spacer, 2 ],
|
85
|
+
[ :name, 52, { :type => :string } ]])
|
86
|
+
assert_equal 'regular grade gasoline (octane number of 87)', t.rows.first['name']
|
87
|
+
assert_equal 'R', t.rows.first['code']
|
88
|
+
assert_equal 'electricity', t.rows.last['name']
|
89
|
+
assert_equal 'El', t.rows.last['code']
|
90
|
+
end
|
91
|
+
|
92
|
+
should "open an XLS with a parser" do
|
93
|
+
ma_1990_01 = {"month"=>1, "cost"=>"54.0", "locatable"=>"Massachusetts (State)", "year"=>1990}
|
94
|
+
ga_1990_01 = {"month"=>1, "cost"=>"50.7", "locatable"=>"Georgia (State)", "year"=>1990}
|
95
|
+
|
96
|
+
t = RemoteTable.new(:url => 'http://tonto.eia.doe.gov/dnav/pet/xls/PET_PRI_RESID_A_EPPR_PTA_CPGAL_M.xls',
|
97
|
+
:transform => { :class => FuelOilParser })
|
98
|
+
assert_equal ma_1990_01, t.rows[0]
|
99
|
+
assert_equal ga_1990_01, t.rows[1]
|
100
|
+
end
|
101
|
+
end
|
data/test/test_helper.rb
ADDED
metadata
ADDED
@@ -0,0 +1,117 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: remote_table
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.1.5
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Seamus Abshere
|
8
|
+
- Andy Rossmeissl
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
|
13
|
+
date: 2009-11-02 00:00:00 -05:00
|
14
|
+
default_executable:
|
15
|
+
dependencies:
|
16
|
+
- !ruby/object:Gem::Dependency
|
17
|
+
name: activesupport
|
18
|
+
type: :runtime
|
19
|
+
version_requirement:
|
20
|
+
version_requirements: !ruby/object:Gem::Requirement
|
21
|
+
requirements:
|
22
|
+
- - ">="
|
23
|
+
- !ruby/object:Gem::Version
|
24
|
+
version: "0"
|
25
|
+
version:
|
26
|
+
- !ruby/object:Gem::Dependency
|
27
|
+
name: fastercsv
|
28
|
+
type: :runtime
|
29
|
+
version_requirement:
|
30
|
+
version_requirements: !ruby/object:Gem::Requirement
|
31
|
+
requirements:
|
32
|
+
- - ">="
|
33
|
+
- !ruby/object:Gem::Version
|
34
|
+
version: "0"
|
35
|
+
version:
|
36
|
+
- !ruby/object:Gem::Dependency
|
37
|
+
name: ryanwood-slither
|
38
|
+
type: :runtime
|
39
|
+
version_requirement:
|
40
|
+
version_requirements: !ruby/object:Gem::Requirement
|
41
|
+
requirements:
|
42
|
+
- - ">="
|
43
|
+
- !ruby/object:Gem::Version
|
44
|
+
version: "0"
|
45
|
+
version:
|
46
|
+
- !ruby/object:Gem::Dependency
|
47
|
+
name: roo
|
48
|
+
type: :runtime
|
49
|
+
version_requirement:
|
50
|
+
version_requirements: !ruby/object:Gem::Requirement
|
51
|
+
requirements:
|
52
|
+
- - "="
|
53
|
+
- !ruby/object:Gem::Version
|
54
|
+
version: 1.3.11
|
55
|
+
version:
|
56
|
+
description: Remotely open and parse Excel XLS, ODS, CSV and fixed-width tables.
|
57
|
+
email: seamus@abshere.net
|
58
|
+
executables: []
|
59
|
+
|
60
|
+
extensions: []
|
61
|
+
|
62
|
+
extra_rdoc_files:
|
63
|
+
- LICENSE
|
64
|
+
- README.rdoc
|
65
|
+
files:
|
66
|
+
- .document
|
67
|
+
- .gitignore
|
68
|
+
- LICENSE
|
69
|
+
- README.rdoc
|
70
|
+
- Rakefile
|
71
|
+
- VERSION
|
72
|
+
- lib/remote_table.rb
|
73
|
+
- lib/remote_table/file.rb
|
74
|
+
- lib/remote_table/file/csv.rb
|
75
|
+
- lib/remote_table/file/fixed_width.rb
|
76
|
+
- lib/remote_table/file/ods.rb
|
77
|
+
- lib/remote_table/file/roo_spreadsheet.rb
|
78
|
+
- lib/remote_table/file/xls.rb
|
79
|
+
- lib/remote_table/package.rb
|
80
|
+
- lib/remote_table/request.rb
|
81
|
+
- lib/remote_table/transform.rb
|
82
|
+
- remote_table.gemspec
|
83
|
+
- test/remote_table_test.rb
|
84
|
+
- test/test_helper.rb
|
85
|
+
has_rdoc: true
|
86
|
+
homepage: http://github.com/seamusabshere/remote_table
|
87
|
+
licenses: []
|
88
|
+
|
89
|
+
post_install_message:
|
90
|
+
rdoc_options:
|
91
|
+
- --charset=UTF-8
|
92
|
+
- --line-numbers
|
93
|
+
- --inline-source
|
94
|
+
require_paths:
|
95
|
+
- lib
|
96
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
97
|
+
requirements:
|
98
|
+
- - ">="
|
99
|
+
- !ruby/object:Gem::Version
|
100
|
+
version: "0"
|
101
|
+
version:
|
102
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
103
|
+
requirements:
|
104
|
+
- - ">="
|
105
|
+
- !ruby/object:Gem::Version
|
106
|
+
version: "0"
|
107
|
+
version:
|
108
|
+
requirements:
|
109
|
+
- curl
|
110
|
+
rubyforge_project: remotetable
|
111
|
+
rubygems_version: 1.3.5
|
112
|
+
signing_key:
|
113
|
+
specification_version: 3
|
114
|
+
summary: Remotely open and parse XLS, ODS, CSV and fixed-width tables.
|
115
|
+
test_files:
|
116
|
+
- test/remote_table_test.rb
|
117
|
+
- test/test_helper.rb
|