cassava 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +17 -0
- data/.rspec +1 -0
- data/Gemfile +4 -0
- data/LICENSE.txt +22 -0
- data/README.md +42 -0
- data/Rakefile +16 -0
- data/bin/cassava +13 -0
- data/cassava.gemspec +24 -0
- data/lib/cassava.rb +6 -0
- data/lib/cassava/document.rb +388 -0
- data/lib/cassava/main.rb +199 -0
- data/lib/cassava/version.rb +3 -0
- data/spec/cast_strings_spec.rb +103 -0
- data/spec/spec_helper.rb +1 -0
- metadata +116 -0
data/.gitignore
ADDED
data/.rspec
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
--color -f d
|
data/Gemfile
ADDED
data/LICENSE.txt
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
Copyright (c) 2012 Kurt Stephens
|
2
|
+
|
3
|
+
MIT License
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
6
|
+
a copy of this software and associated documentation files (the
|
7
|
+
"Software"), to deal in the Software without restriction, including
|
8
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
9
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
10
|
+
permit persons to whom the Software is furnished to do so, subject to
|
11
|
+
the following conditions:
|
12
|
+
|
13
|
+
The above copyright notice and this permission notice shall be
|
14
|
+
included in all copies or substantial portions of the Software.
|
15
|
+
|
16
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
17
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
18
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
19
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
20
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
21
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
22
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.md
ADDED
@@ -0,0 +1,42 @@
|
|
1
|
+
# Cassava
|
2
|
+
|
3
|
+
A command-line CSV tool.
|
4
|
+
|
5
|
+
## Installation
|
6
|
+
|
7
|
+
Add this line to your application's Gemfile:
|
8
|
+
|
9
|
+
gem 'cassava'
|
10
|
+
|
11
|
+
And then execute:
|
12
|
+
|
13
|
+
$ bundle
|
14
|
+
|
15
|
+
Or install it yourself as:
|
16
|
+
|
17
|
+
$ gem install cassava
|
18
|
+
|
19
|
+
## Usage
|
20
|
+
|
21
|
+
# Concat mulitple CSVs, while merging columns:
|
22
|
+
$ cassava cat first.csv second.csv
|
23
|
+
|
24
|
+
# Cut columns out of result; '-' is used to pipe results to next command:
|
25
|
+
$ cassava cat first.csv second.csv - cut -c first_name,last_name
|
26
|
+
|
27
|
+
# Format as ASCII table.
|
28
|
+
$ cassava format first.csv
|
29
|
+
|
30
|
+
# Select where.
|
31
|
+
$ cassava cat *.csv - where last_name=Smith - format
|
32
|
+
|
33
|
+
# Sort by a column:
|
34
|
+
$ cassava cat *.csv - sort -by last_name,first_name - format
|
35
|
+
|
36
|
+
## Contributing
|
37
|
+
|
38
|
+
1. Fork it
|
39
|
+
2. Create your feature branch (`git checkout -b my-new-feature`)
|
40
|
+
3. Commit your changes (`git commit -am 'Add some feature'`)
|
41
|
+
4. Push to the branch (`git push origin my-new-feature`)
|
42
|
+
5. Create new Pull Request
|
data/Rakefile
ADDED
@@ -0,0 +1,16 @@
|
|
1
|
+
require "bundler/gem_tasks"
|
2
|
+
gem 'rspec'
|
3
|
+
require 'rspec/core/rake_task'
|
4
|
+
|
5
|
+
desc "Default => :test"
|
6
|
+
task :default => :test
|
7
|
+
|
8
|
+
desc "Run all tests"
|
9
|
+
task :test => [ :spec ]
|
10
|
+
|
11
|
+
desc "Run specs"
|
12
|
+
RSpec::Core::RakeTask.new(:spec) do |t|
|
13
|
+
t.pattern = "./spec/**/*_spec.rb" # don't need this, it's default.
|
14
|
+
# Put spec opts in a file named .rspec in root
|
15
|
+
end
|
16
|
+
|
data/bin/cassava
ADDED
@@ -0,0 +1,13 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
# -*- ruby -*-
|
3
|
+
$: << File.expand_path("../../lib", __FILE__)
|
4
|
+
if ARGV[0] == '-D'
|
5
|
+
ARGV.shift
|
6
|
+
$DEBUG = true
|
7
|
+
require 'rubygems'
|
8
|
+
gem 'ruby-debug'
|
9
|
+
require 'ruby-debug'
|
10
|
+
end
|
11
|
+
require 'cassava/main'
|
12
|
+
exit(Cassava::Main.new(ARGV).run!.exit_code)
|
13
|
+
|
data/cassava.gemspec
ADDED
@@ -0,0 +1,24 @@
|
|
1
|
+
# -*- encoding: utf-8 -*-
|
2
|
+
lib = File.expand_path('../lib', __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require 'cassava/version'
|
5
|
+
|
6
|
+
Gem::Specification.new do |gem|
|
7
|
+
gem.name = "cassava"
|
8
|
+
gem.version = Cassava::VERSION
|
9
|
+
gem.authors = ["Kurt Stephens"]
|
10
|
+
gem.email = ["ks.github@kurtstephens.com"]
|
11
|
+
gem.description = %q{A command-line CSV tool.}
|
12
|
+
gem.summary = %q{A command-line CSV tool.}
|
13
|
+
gem.homepage = "https://github.com/kstephens/cassava"
|
14
|
+
|
15
|
+
gem.files = `git ls-files`.split($/)
|
16
|
+
gem.executables = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
|
17
|
+
gem.test_files = gem.files.grep(%r{^(test|spec|features)/})
|
18
|
+
gem.require_paths = ["lib"]
|
19
|
+
|
20
|
+
gem.add_dependency 'terminal-table', '~> 1.4.5'
|
21
|
+
|
22
|
+
gem.add_development_dependency "rake", "~> 10.0.2"
|
23
|
+
gem.add_development_dependency "rspec", "~> 2.12.0"
|
24
|
+
end
|
data/lib/cassava.rb
ADDED
@@ -0,0 +1,388 @@
|
|
1
|
+
require 'csv'
|
2
|
+
|
3
|
+
module Cassava
|
4
|
+
class Document
|
5
|
+
attr_accessor :name, :columns, :column_offset, :offset_column
|
6
|
+
attr_accessor :rows
|
7
|
+
attr_accessor :debug
|
8
|
+
|
9
|
+
def initialize opts = nil
|
10
|
+
@opts = opts
|
11
|
+
@name = opts[:name] or raise ArgumentError, "name not specified"
|
12
|
+
@rows = [ ]
|
13
|
+
@index = { }
|
14
|
+
self.columns = opts[:columns] || [ ]
|
15
|
+
if x = opts[:rows]
|
16
|
+
append_row! x
|
17
|
+
end
|
18
|
+
end
|
19
|
+
|
20
|
+
def columns= x
|
21
|
+
@columns = x
|
22
|
+
update_columns!
|
23
|
+
end
|
24
|
+
|
25
|
+
def update_columns!
|
26
|
+
@ncols = @columns.size
|
27
|
+
@column_offset = { }
|
28
|
+
@offset_column = [ ]
|
29
|
+
i = -1
|
30
|
+
@columns.map! do | c |
|
31
|
+
i += 1
|
32
|
+
c = c.to_s
|
33
|
+
next if c.empty?
|
34
|
+
c = c.to_sym
|
35
|
+
@column_offset[c] = i
|
36
|
+
@offset_column[i] = c
|
37
|
+
c
|
38
|
+
end
|
39
|
+
@column_types = nil
|
40
|
+
@index.keep_if { | c, h | @column_offset[c] }
|
41
|
+
self
|
42
|
+
end
|
43
|
+
|
44
|
+
def add_column! c
|
45
|
+
c = c.to_sym
|
46
|
+
unless i = @column_offset[c]
|
47
|
+
i = @columns.size
|
48
|
+
@columns << c
|
49
|
+
update_columns!
|
50
|
+
end
|
51
|
+
i
|
52
|
+
end
|
53
|
+
|
54
|
+
def to_column_names! a
|
55
|
+
a.map! do | x |
|
56
|
+
case x
|
57
|
+
when Integer
|
58
|
+
c = @columns[x]
|
59
|
+
when String
|
60
|
+
if x == (i = x.to_i).to_s
|
61
|
+
c = @columns[i]
|
62
|
+
else
|
63
|
+
c = x.to_sym
|
64
|
+
end
|
65
|
+
when Symbol
|
66
|
+
c = x
|
67
|
+
else
|
68
|
+
raise TypeError, "#{x.inspect}"
|
69
|
+
end
|
70
|
+
raise TypeError, "#{x.inspect} => #{c.inspect}" unless c
|
71
|
+
c
|
72
|
+
end
|
73
|
+
a
|
74
|
+
end
|
75
|
+
|
76
|
+
def nrows
|
77
|
+
@rows.size
|
78
|
+
end
|
79
|
+
|
80
|
+
def empty_rows!
|
81
|
+
@rows = [ ]
|
82
|
+
@index.clear
|
83
|
+
@column_types = nil
|
84
|
+
self
|
85
|
+
end
|
86
|
+
|
87
|
+
def append_rows! rows
|
88
|
+
return self unless rows
|
89
|
+
if Document === rows
|
90
|
+
rows.columns.each { | c | add_column!(c) }
|
91
|
+
rows = rows.rows
|
92
|
+
end
|
93
|
+
rows.map!{ | r | array_to_row(r) }
|
94
|
+
row_i = @rows.size
|
95
|
+
@rows.concat(rows)
|
96
|
+
rows.each do | r |
|
97
|
+
@ncols = r.size if @ncols < r.size
|
98
|
+
r[:_row_i] = row_i
|
99
|
+
row_i += 1
|
100
|
+
end
|
101
|
+
@index.clear
|
102
|
+
@column_types = nil
|
103
|
+
self
|
104
|
+
end
|
105
|
+
|
106
|
+
def parse!
|
107
|
+
# debugger if $DEBUG
|
108
|
+
csv = nil
|
109
|
+
if RUBY_VERSION =~ /^1\.8/
|
110
|
+
@rows = [ ]
|
111
|
+
csv = CSV.open(name, "rb", @opts[:col_sep]) do | r |
|
112
|
+
@rows << r
|
113
|
+
end
|
114
|
+
else
|
115
|
+
csv_opts = { }
|
116
|
+
csv_opts[:col_sep] = @opts[:col_sep] if @opts[:col_sep]
|
117
|
+
csv = CSV.open(name, "rb", csv_opts)
|
118
|
+
@rows = csv.read
|
119
|
+
end
|
120
|
+
|
121
|
+
@columns = @rows.shift if @columns.empty?
|
122
|
+
update_columns!
|
123
|
+
|
124
|
+
row_i = 0
|
125
|
+
@rows.map! do | r |
|
126
|
+
@ncols = r.size if @ncols < r.size
|
127
|
+
h = { :_row_i => (row_i += 1) }
|
128
|
+
@column_offset.each do | c, i |
|
129
|
+
h[c] = r[i] if c && i
|
130
|
+
end
|
131
|
+
h
|
132
|
+
end
|
133
|
+
# debugger
|
134
|
+
self
|
135
|
+
ensure
|
136
|
+
csv.close if csv
|
137
|
+
end
|
138
|
+
|
139
|
+
def index! c
|
140
|
+
c = @column[c] if Integer === c
|
141
|
+
unless ind = @index[c]
|
142
|
+
ind = { }
|
143
|
+
@rows.each do | r |
|
144
|
+
(ind[r[c]] ||= [ ]) << r
|
145
|
+
end
|
146
|
+
@index[c] = ind
|
147
|
+
end
|
148
|
+
ind
|
149
|
+
end
|
150
|
+
|
151
|
+
def get c, v
|
152
|
+
index!(c)[v]
|
153
|
+
end
|
154
|
+
|
155
|
+
def emit! file
|
156
|
+
out = nil
|
157
|
+
case file
|
158
|
+
when '/dev/stdout', '/dev/fd/0'
|
159
|
+
tmp_file = true
|
160
|
+
out = $stdout
|
161
|
+
when %r{^/dev/}
|
162
|
+
tmp_file = true
|
163
|
+
end
|
164
|
+
if tmp_file
|
165
|
+
tmp_file = "/tmp/cassava-#{$$}.csv"
|
166
|
+
# $stderr.puts " #{file} => #{tmp_file}"
|
167
|
+
file = tmp_file
|
168
|
+
end
|
169
|
+
|
170
|
+
_emit! file
|
171
|
+
|
172
|
+
if tmp_file
|
173
|
+
out ||= File.open(tmp, "w")
|
174
|
+
# $stderr.puts " #{tmp_file} => #{out}"
|
175
|
+
out.write(File.read(tmp_file))
|
176
|
+
end
|
177
|
+
ensure
|
178
|
+
File.unlink(tmp_file) if tmp_file
|
179
|
+
end
|
180
|
+
|
181
|
+
def _emit! file
|
182
|
+
CSV.open(file, "wb") do | out |
|
183
|
+
a = @offset_column.map do | c |
|
184
|
+
c && c.to_s
|
185
|
+
end
|
186
|
+
out << a
|
187
|
+
@rows.each do | r |
|
188
|
+
a = @offset_column.map do | i |
|
189
|
+
i && r[i]
|
190
|
+
end
|
191
|
+
out << a
|
192
|
+
end
|
193
|
+
end
|
194
|
+
self
|
195
|
+
end
|
196
|
+
alias :emit! :_emit!
|
197
|
+
|
198
|
+
def array_to_row a, columns = nil
|
199
|
+
if Array === a
|
200
|
+
columns ||= self.columns
|
201
|
+
h = { }
|
202
|
+
columns.each_with_index do | c, i |
|
203
|
+
h[c] = a[i]
|
204
|
+
end
|
205
|
+
a = h
|
206
|
+
end
|
207
|
+
a
|
208
|
+
end
|
209
|
+
|
210
|
+
def row_to_array r
|
211
|
+
unless Array === r
|
212
|
+
r = @offset_column.map do | c |
|
213
|
+
c && r[c]
|
214
|
+
end
|
215
|
+
end
|
216
|
+
r
|
217
|
+
end
|
218
|
+
|
219
|
+
def coerce_to_strings!
|
220
|
+
rows.each do | r |
|
221
|
+
r.each do | k, v |
|
222
|
+
r[k] = v.to_s unless String === v
|
223
|
+
end
|
224
|
+
end
|
225
|
+
self
|
226
|
+
end
|
227
|
+
|
228
|
+
def cast_strings rows
|
229
|
+
rows.each do | r |
|
230
|
+
r.each do | k, v |
|
231
|
+
next if v.nil?
|
232
|
+
old_v = v
|
233
|
+
v = v.to_s
|
234
|
+
if String === v
|
235
|
+
case v
|
236
|
+
when /\A[-+]?\d+\Z/
|
237
|
+
v = v.to_i
|
238
|
+
when /\A([-+]?([0-9]+\.[0-9]+|\.[0-9]+|[0-9]+\.)(e[-+]?\d+)?|[-+]?\d+e[-+]?\d+)\Z/i
|
239
|
+
v = v.to_f
|
240
|
+
end
|
241
|
+
# puts "old_v = #{old_v.inspect} => #{v.inspect}"
|
242
|
+
end
|
243
|
+
r[k] = v
|
244
|
+
end
|
245
|
+
end
|
246
|
+
rows
|
247
|
+
end
|
248
|
+
|
249
|
+
def cast_strings!
|
250
|
+
cast_strings @rows
|
251
|
+
@column_types = nil
|
252
|
+
self
|
253
|
+
end
|
254
|
+
|
255
|
+
def sort! by = nil
|
256
|
+
by ||= @columns
|
257
|
+
cast_strings!
|
258
|
+
# by = by.map { | x | column_offset[x] }
|
259
|
+
ct = { }
|
260
|
+
columns.each_with_index do | c, i |
|
261
|
+
ct[c] = column_types[i]
|
262
|
+
end
|
263
|
+
@rows.sort! do | a, b |
|
264
|
+
r = 0
|
265
|
+
by.each do | c |
|
266
|
+
av = a[c]
|
267
|
+
bv = b[c]
|
268
|
+
case
|
269
|
+
when av.nil? && bv.nil?
|
270
|
+
r = 0
|
271
|
+
when av.nil?
|
272
|
+
r = -1
|
273
|
+
when bv.nil?
|
274
|
+
r = 1
|
275
|
+
else
|
276
|
+
r = (av <=> bv rescue nil) || 0
|
277
|
+
end
|
278
|
+
break if r != 0
|
279
|
+
end
|
280
|
+
r
|
281
|
+
end
|
282
|
+
self
|
283
|
+
end
|
284
|
+
|
285
|
+
require 'pp'
|
286
|
+
def infer_column_types rows = self.rows
|
287
|
+
column_types = [ nil ] * @columns.size
|
288
|
+
ancestors_cache = { }
|
289
|
+
common_ancestor_cache = { }
|
290
|
+
rows.each do | r |
|
291
|
+
raise unless Hash === r
|
292
|
+
@columns.each_with_index do | k, i |
|
293
|
+
v = r[k]
|
294
|
+
next if v.nil?
|
295
|
+
ct = column_types[i]
|
296
|
+
vt = v.class
|
297
|
+
if ct.nil?
|
298
|
+
column_types[i] = vt
|
299
|
+
next
|
300
|
+
end
|
301
|
+
common_ancestor =
|
302
|
+
common_ancestor_cache[[ct, vt]] ||=
|
303
|
+
begin
|
304
|
+
ca =
|
305
|
+
ancestors_cache[ct] ||=
|
306
|
+
ct.ancestors.delete_if{|x| x.class == Module}
|
307
|
+
va =
|
308
|
+
ancestors_cache[vt] ||=
|
309
|
+
vt.ancestors.delete_if{|x| x.class == Module}
|
310
|
+
(ca & va).first || Object
|
311
|
+
end
|
312
|
+
if @debug && k == :float
|
313
|
+
pp [ :k, k, :v, v, :ct, ct, :vt, vt, :ca, ca, :va, va, :common_ancestor, common_ancestor ]
|
314
|
+
end
|
315
|
+
# if Value's class is not a specialization of column class.
|
316
|
+
ct = common_ancestor
|
317
|
+
column_types[i] = ct
|
318
|
+
end
|
319
|
+
end
|
320
|
+
# pp columns.zip(column_types)
|
321
|
+
column_types
|
322
|
+
end
|
323
|
+
|
324
|
+
def column_types
|
325
|
+
unless @column_types
|
326
|
+
@column_types ||= infer_column_types
|
327
|
+
end
|
328
|
+
@column_types
|
329
|
+
end
|
330
|
+
|
331
|
+
def clone_rows rows = self.rows
|
332
|
+
rows.map { | r | r.dup }
|
333
|
+
end
|
334
|
+
|
335
|
+
# Format as ASCII table.
|
336
|
+
def to_text opts = { }
|
337
|
+
gem 'terminal-table'
|
338
|
+
require 'terminal-table'
|
339
|
+
|
340
|
+
table = Terminal::Table.new() do | table |
|
341
|
+
# t.title = self.name
|
342
|
+
s = table.style
|
343
|
+
s.border_x = s.border_y = s.border_i = ''
|
344
|
+
# s.border_i = '|'
|
345
|
+
s.padding_left = 0
|
346
|
+
s.padding_right = 1
|
347
|
+
|
348
|
+
table << self.columns.map{|c| { :value => c.to_s, :alignment => :center }}
|
349
|
+
|
350
|
+
# Convert rows to Arrays and handle nil, etc.
|
351
|
+
self.rows.each do | r |
|
352
|
+
r = self.row_to_array(r)
|
353
|
+
r.map! do | c |
|
354
|
+
c = case c
|
355
|
+
when nil
|
356
|
+
''
|
357
|
+
when Integer
|
358
|
+
thousands(c)
|
359
|
+
else
|
360
|
+
c
|
361
|
+
end
|
362
|
+
# c = "#{c} |"
|
363
|
+
end
|
364
|
+
table << r
|
365
|
+
end
|
366
|
+
|
367
|
+
# Align numeric columns to the left.
|
368
|
+
column_types = infer_column_types(cast_strings(clone_rows))
|
369
|
+
column_types.each_with_index do | type, ci |
|
370
|
+
if type && type.ancestors.include?(Numeric)
|
371
|
+
# puts " column #{ci} #{columns[ci]} #{t}"
|
372
|
+
table.align_column(ci, :right)
|
373
|
+
end
|
374
|
+
end
|
375
|
+
end
|
376
|
+
|
377
|
+
|
378
|
+
# Return formatted table.
|
379
|
+
table.to_s
|
380
|
+
end
|
381
|
+
|
382
|
+
def thousands x, sep = '_'
|
383
|
+
x && x.to_s.reverse!.gsub(/(\d{3})/, "\\1#{sep}").reverse!.sub(/^(\D|\A)#{sep}/, '')
|
384
|
+
end
|
385
|
+
end
|
386
|
+
end
|
387
|
+
|
388
|
+
|
data/lib/cassava/main.rb
ADDED
@@ -0,0 +1,199 @@
|
|
1
|
+
require 'cassava'
|
2
|
+
require 'cassava/document'
|
3
|
+
|
4
|
+
module Cassava
|
5
|
+
class Main
|
6
|
+
attr_accessor :progname
|
7
|
+
attr_accessor :args, :cmd, :opts, :exit_code
|
8
|
+
attr_accessor :result, :output
|
9
|
+
attr_accessor :by, :columns
|
10
|
+
|
11
|
+
def initialize args = nil
|
12
|
+
@progname = File.basename($0)
|
13
|
+
@args = (args || ARGV).dup
|
14
|
+
@opts = { }
|
15
|
+
@exit_code = 0
|
16
|
+
@output = "/dev/stdout"
|
17
|
+
@debug = true
|
18
|
+
@select_where = { }
|
19
|
+
end
|
20
|
+
|
21
|
+
def run!
|
22
|
+
cmds = [ ]
|
23
|
+
last_arg = nil
|
24
|
+
|
25
|
+
cmd = [ ]
|
26
|
+
args = @args.dup
|
27
|
+
while arg = args.shift
|
28
|
+
if arg == '-'
|
29
|
+
if last_arg
|
30
|
+
cmd << last_arg
|
31
|
+
last_arg = nil
|
32
|
+
end
|
33
|
+
last_arg = '--result'
|
34
|
+
cmds << cmd
|
35
|
+
cmd = [ ]
|
36
|
+
else
|
37
|
+
cmd << arg
|
38
|
+
end
|
39
|
+
end
|
40
|
+
if last_arg
|
41
|
+
cmd << last_arg
|
42
|
+
last_arg = nil
|
43
|
+
end
|
44
|
+
cmds << cmd
|
45
|
+
# pp cmds
|
46
|
+
|
47
|
+
# Run each command:
|
48
|
+
cmds.each do | cmd |
|
49
|
+
@args = cmd.dup
|
50
|
+
next_cmd!
|
51
|
+
end
|
52
|
+
|
53
|
+
result.emit!(@output) if result
|
54
|
+
self
|
55
|
+
rescue ::Exception => exc
|
56
|
+
$stderr.puts "#{progname}: ERROR: #{exc.inspect}"
|
57
|
+
$stderr.puts " #{exc.backtrace * "\n "}" if @debug || $DEBUG
|
58
|
+
@exit_code = 1
|
59
|
+
self
|
60
|
+
end
|
61
|
+
|
62
|
+
def next_cmd!
|
63
|
+
@cmd = @args.shift
|
64
|
+
sel = :"_#{cmd}!"
|
65
|
+
raise ArgumentError, "Invalid command: #{cmd.inspect}" unless respond_to?(sel)
|
66
|
+
send(sel)
|
67
|
+
end
|
68
|
+
|
69
|
+
def next_document!
|
70
|
+
doc = nil
|
71
|
+
opts = { }
|
72
|
+
until args.empty?
|
73
|
+
case arg = args.shift
|
74
|
+
when '-cv'
|
75
|
+
k = args.shift.to_sym
|
76
|
+
v = args.shift
|
77
|
+
@select_where[k] = v
|
78
|
+
when /\A([^=]+)=(.*)\Z/
|
79
|
+
k = $1.to_sym
|
80
|
+
v = $2
|
81
|
+
@select_where[k] = v
|
82
|
+
when '--result'
|
83
|
+
doc = @result
|
84
|
+
break
|
85
|
+
when '-c'
|
86
|
+
self.columns = args.shift.split(/\s*,\s*|\s+/)
|
87
|
+
when '-by'
|
88
|
+
self.by = args.shift.split(/\s*,\s*|\s+/)
|
89
|
+
when '-o'
|
90
|
+
self.output = args.shift
|
91
|
+
when '-FS'
|
92
|
+
opts[:col_sep] = args.shift
|
93
|
+
when '-IGNORE'
|
94
|
+
opts[:ignore] = args.shift
|
95
|
+
when '-C'
|
96
|
+
opts[:columns] = args.shift.split(/\s*,\s*|\s+/)
|
97
|
+
else
|
98
|
+
opts[:name] = arg
|
99
|
+
doc = Document.new(opts)
|
100
|
+
doc.parse!
|
101
|
+
break
|
102
|
+
end
|
103
|
+
end
|
104
|
+
doc
|
105
|
+
end
|
106
|
+
|
107
|
+
def self.define_command name, doc, &blk
|
108
|
+
meth = :"_#{name}!"
|
109
|
+
define_method meth, &blk
|
110
|
+
@@commands << { :name => name, :doc => doc }
|
111
|
+
end
|
112
|
+
|
113
|
+
def _cat!
|
114
|
+
self.result = next_document!
|
115
|
+
until args.empty?
|
116
|
+
other = next_document!
|
117
|
+
result.append_rows!(other)
|
118
|
+
end
|
119
|
+
end
|
120
|
+
|
121
|
+
def _where!
|
122
|
+
@select_where = { }
|
123
|
+
self.result = next_document!
|
124
|
+
result.coerce_to_strings!
|
125
|
+
@select_where.each do | col, val |
|
126
|
+
new_result = result.dup
|
127
|
+
new_result.empty_rows!
|
128
|
+
found_rows = result.get(col, val)
|
129
|
+
# puts " #{col}=#{val} => #{found_rows.size} rows"
|
130
|
+
new_result.append_rows!(found_rows)
|
131
|
+
self.result = new_result
|
132
|
+
end
|
133
|
+
end
|
134
|
+
alias :_select! :_where!
|
135
|
+
|
136
|
+
def _cut!
|
137
|
+
self.columns = nil
|
138
|
+
self.result = next_document!
|
139
|
+
self.columns ||= [ result.columns[0] ]
|
140
|
+
# pp result.columns
|
141
|
+
result.to_column_names! columns
|
142
|
+
new_result = result.dup.empty_rows!
|
143
|
+
new_result.columns = columns & result.columns
|
144
|
+
# pp new_result.columns
|
145
|
+
# pp new_result.column_offset
|
146
|
+
rows = self.result.rows.map do | r |
|
147
|
+
r = r.dup
|
148
|
+
r.keep_if { | c | new_result.column_offset[c] }
|
149
|
+
# pp r
|
150
|
+
r
|
151
|
+
end
|
152
|
+
new_result.append_rows! rows
|
153
|
+
self.result = new_result
|
154
|
+
end
|
155
|
+
|
156
|
+
def _join!
|
157
|
+
self.result = next_document!
|
158
|
+
until args.empty?
|
159
|
+
left_key = args.shift.to_sym
|
160
|
+
right_key = args.shift.to_sym
|
161
|
+
right = next_document!
|
162
|
+
new_result = Document.new
|
163
|
+
result.columns.each do | c |
|
164
|
+
new_result.add_column!(c)
|
165
|
+
end
|
166
|
+
right.columns.each do | c |
|
167
|
+
new_result.add_column!(c)
|
168
|
+
end
|
169
|
+
result.rows.each do | lr |
|
170
|
+
rrows = right.get(right_key, lr[left_key])
|
171
|
+
unless rrows.empty?
|
172
|
+
r = lr.dup
|
173
|
+
rrows.each do | rr |
|
174
|
+
r.merge(rr)
|
175
|
+
end
|
176
|
+
new_result.rows << r
|
177
|
+
end
|
178
|
+
end
|
179
|
+
self.result = new_result
|
180
|
+
end
|
181
|
+
end
|
182
|
+
|
183
|
+
def _sort!
|
184
|
+
self.by = nil
|
185
|
+
self.result = next_document!
|
186
|
+
by = self.by || args
|
187
|
+
by.map! { | c | c.to_sym }
|
188
|
+
result.sort!(by)
|
189
|
+
end
|
190
|
+
|
191
|
+
def _format!
|
192
|
+
rows = result.to_text
|
193
|
+
# $stderr.puts rows
|
194
|
+
rows = rows.split("\n").map { | r | { :_ => r } }
|
195
|
+
# require 'pp'; pp rows
|
196
|
+
self.result = Cassava::Document.new(:name => "#{@result.name}->format", :columns => [ :_ ], :rows => rows)
|
197
|
+
end
|
198
|
+
end
|
199
|
+
end
|
@@ -0,0 +1,103 @@
|
|
1
|
+
require 'spec_helper'
|
2
|
+
|
3
|
+
class Exception
|
4
|
+
alias :_message :message
|
5
|
+
def message
|
6
|
+
@message ||= _message
|
7
|
+
end
|
8
|
+
def message= msg
|
9
|
+
@message = msg
|
10
|
+
end
|
11
|
+
end
|
12
|
+
|
13
|
+
describe "Cassava::Document" do
|
14
|
+
attr_accessor :d
|
15
|
+
before do
|
16
|
+
self.d = Cassava::Document.new(:name => 'test', :columns => [ :nil, :string, :fixnum, :float, :integer, :numeric ])
|
17
|
+
rows = <<"END".split("\n").map{|r| r.strip.split(/,\s+/).map{|x| x == 'nil' ? nil : x }}
|
18
|
+
nil, nil, nil, nil, nil, nil
|
19
|
+
nil, string, nil, nil, nil, nil
|
20
|
+
nil, +, nil, nil, nil, nil
|
21
|
+
nil, -, nil, nil, nil, nil
|
22
|
+
nil, 234+, nil, nil, nil, nil
|
23
|
+
nil, 1234ws, nil, nil, nil, nil
|
24
|
+
nil, nil, 1234, nil, nil, nil
|
25
|
+
nil, nil, +1234, nil, nil, nil
|
26
|
+
nil, nil, -1234, nil, nil, nil
|
27
|
+
nil, nil, nil, .2, nil, nil
|
28
|
+
nil, nil, nil, -1.2, nil, nil
|
29
|
+
nil, nil, nil, -2., nil, nil
|
30
|
+
nil, nil, nil, 2e10, nil, nil
|
31
|
+
nil, nil, nil, 1e1, nil, nil
|
32
|
+
nil, nil, nil, 2E-2, nil, nil
|
33
|
+
nil, nil, nil, 3E+3, nil, nil
|
34
|
+
nil, nil, nil, nil, 1234, nil
|
35
|
+
nil, nil, nil, nil, 12341231234523452345345234, nil
|
36
|
+
nil, nil, nil, nil, nil, 1234
|
37
|
+
nil, nil, nil, nil, nil, 123412348102394812934
|
38
|
+
nil, nil, nil, nil, nil, 1234.2342
|
39
|
+
END
|
40
|
+
d.append_rows! rows
|
41
|
+
d.cast_strings!
|
42
|
+
end
|
43
|
+
|
44
|
+
def each_row
|
45
|
+
i = 0
|
46
|
+
d.rows.each do | r |
|
47
|
+
begin
|
48
|
+
yield r
|
49
|
+
rescue ::Exception => exc
|
50
|
+
exc.message = "in row #{i} : #{r.inspect} : #{exc.message}"
|
51
|
+
raise exc
|
52
|
+
end
|
53
|
+
i += 1
|
54
|
+
end
|
55
|
+
end
|
56
|
+
|
57
|
+
context "#cast_strings!" do
|
58
|
+
it "should not touch nil columns" do
|
59
|
+
each_row do | r |
|
60
|
+
r[0].should == nil
|
61
|
+
end
|
62
|
+
end
|
63
|
+
it "should not touch string only columns" do
|
64
|
+
each_row do | r |
|
65
|
+
next if r[1].nil?
|
66
|
+
r[1].class.should == String
|
67
|
+
end
|
68
|
+
end
|
69
|
+
it "should coerce integer columns" do
|
70
|
+
each_row do | r |
|
71
|
+
next if r[2].nil?
|
72
|
+
r[2].is_a?(Integer).should == true
|
73
|
+
end
|
74
|
+
end
|
75
|
+
it "should coerce float columns" do
|
76
|
+
each_row do | r |
|
77
|
+
next if r[3].nil?
|
78
|
+
r[3].is_a?(Float).should == true
|
79
|
+
end
|
80
|
+
end
|
81
|
+
end
|
82
|
+
|
83
|
+
context "#infer_types!" do
|
84
|
+
it "should infer nil for all nils" do
|
85
|
+
d.column_types[0].should == nil
|
86
|
+
end
|
87
|
+
it "should infer String for String" do
|
88
|
+
d.column_types[1].should == String
|
89
|
+
end
|
90
|
+
it "should infer Fixum for Fixnum" do
|
91
|
+
d.column_types[2].should == Fixnum
|
92
|
+
end
|
93
|
+
it "should infer Float for all Float" do
|
94
|
+
d.column_types[3].should == Float
|
95
|
+
end
|
96
|
+
it "should infer Integer for Fixnum and Bignum" do
|
97
|
+
d.column_types[4].should == Integer
|
98
|
+
end
|
99
|
+
it "should infer Numeric for Fixnum, Bignum, Float" do
|
100
|
+
d.column_types[5].should == Numeric
|
101
|
+
end
|
102
|
+
end
|
103
|
+
end
|
data/spec/spec_helper.rb
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require 'cassava/main'
|
metadata
ADDED
@@ -0,0 +1,116 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: cassava
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Kurt Stephens
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2012-12-15 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: terminal-table
|
16
|
+
requirement: !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ~>
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: 1.4.5
|
22
|
+
type: :runtime
|
23
|
+
prerelease: false
|
24
|
+
version_requirements: !ruby/object:Gem::Requirement
|
25
|
+
none: false
|
26
|
+
requirements:
|
27
|
+
- - ~>
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: 1.4.5
|
30
|
+
- !ruby/object:Gem::Dependency
|
31
|
+
name: rake
|
32
|
+
requirement: !ruby/object:Gem::Requirement
|
33
|
+
none: false
|
34
|
+
requirements:
|
35
|
+
- - ~>
|
36
|
+
- !ruby/object:Gem::Version
|
37
|
+
version: 10.0.2
|
38
|
+
type: :development
|
39
|
+
prerelease: false
|
40
|
+
version_requirements: !ruby/object:Gem::Requirement
|
41
|
+
none: false
|
42
|
+
requirements:
|
43
|
+
- - ~>
|
44
|
+
- !ruby/object:Gem::Version
|
45
|
+
version: 10.0.2
|
46
|
+
- !ruby/object:Gem::Dependency
|
47
|
+
name: rspec
|
48
|
+
requirement: !ruby/object:Gem::Requirement
|
49
|
+
none: false
|
50
|
+
requirements:
|
51
|
+
- - ~>
|
52
|
+
- !ruby/object:Gem::Version
|
53
|
+
version: 2.12.0
|
54
|
+
type: :development
|
55
|
+
prerelease: false
|
56
|
+
version_requirements: !ruby/object:Gem::Requirement
|
57
|
+
none: false
|
58
|
+
requirements:
|
59
|
+
- - ~>
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: 2.12.0
|
62
|
+
description: A command-line CSV tool.
|
63
|
+
email:
|
64
|
+
- ks.github@kurtstephens.com
|
65
|
+
executables:
|
66
|
+
- cassava
|
67
|
+
extensions: []
|
68
|
+
extra_rdoc_files: []
|
69
|
+
files:
|
70
|
+
- .gitignore
|
71
|
+
- .rspec
|
72
|
+
- Gemfile
|
73
|
+
- LICENSE.txt
|
74
|
+
- README.md
|
75
|
+
- Rakefile
|
76
|
+
- bin/cassava
|
77
|
+
- cassava.gemspec
|
78
|
+
- lib/cassava.rb
|
79
|
+
- lib/cassava/document.rb
|
80
|
+
- lib/cassava/main.rb
|
81
|
+
- lib/cassava/version.rb
|
82
|
+
- spec/cast_strings_spec.rb
|
83
|
+
- spec/spec_helper.rb
|
84
|
+
homepage: https://github.com/kstephens/cassava
|
85
|
+
licenses: []
|
86
|
+
post_install_message:
|
87
|
+
rdoc_options: []
|
88
|
+
require_paths:
|
89
|
+
- lib
|
90
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
91
|
+
none: false
|
92
|
+
requirements:
|
93
|
+
- - ! '>='
|
94
|
+
- !ruby/object:Gem::Version
|
95
|
+
version: '0'
|
96
|
+
segments:
|
97
|
+
- 0
|
98
|
+
hash: -1985077025104642675
|
99
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
100
|
+
none: false
|
101
|
+
requirements:
|
102
|
+
- - ! '>='
|
103
|
+
- !ruby/object:Gem::Version
|
104
|
+
version: '0'
|
105
|
+
segments:
|
106
|
+
- 0
|
107
|
+
hash: -1985077025104642675
|
108
|
+
requirements: []
|
109
|
+
rubyforge_project:
|
110
|
+
rubygems_version: 1.8.23
|
111
|
+
signing_key:
|
112
|
+
specification_version: 3
|
113
|
+
summary: A command-line CSV tool.
|
114
|
+
test_files:
|
115
|
+
- spec/cast_strings_spec.rb
|
116
|
+
- spec/spec_helper.rb
|