parseinput 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/Rakefile +41 -0
- data/lib/parse/input.rb +140 -0
- data/setup.rb +1360 -0
- data/test/tc_reading.rb +70 -0
- data/test/tc_state.rb +142 -0
- data/test/ts_all.rb +11 -0
- metadata +47 -0
data/test/tc_reading.rb
ADDED
@@ -0,0 +1,70 @@
|
|
1
|
+
#!/usr/local/bin/ruby -w
|
2
|
+
|
3
|
+
# tc_reading.rb
|
4
|
+
#
|
5
|
+
# Created by James Edward Gray II on 2005-08-14.
|
6
|
+
# Copyright 2005 Gray Productions. All rights reserved.
|
7
|
+
|
8
|
+
require "test/unit"
|
9
|
+
|
10
|
+
require "parse/input"
|
11
|
+
|
12
|
+
class TestReading < Test::Unit::TestCase
|
13
|
+
def test_line_by_line_read
|
14
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
15
|
+
lines = File.readlines(path)
|
16
|
+
test = self
|
17
|
+
|
18
|
+
Parse::Input.new(path) do
|
19
|
+
read { |line| test.assert_equal(lines.shift, line) }
|
20
|
+
end
|
21
|
+
end
|
22
|
+
|
23
|
+
def test_simple_interface
|
24
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
25
|
+
lines = File.readlines(path)
|
26
|
+
test = self
|
27
|
+
|
28
|
+
input(path) do
|
29
|
+
read { |line| test.assert_equal(lines.shift, line) }
|
30
|
+
end
|
31
|
+
end
|
32
|
+
|
33
|
+
def test_paragraph_read
|
34
|
+
path = File.join(File.dirname(__FILE__), "five_paragraphs.txt")
|
35
|
+
paragraphs = File.readlines(path, "")
|
36
|
+
test = self
|
37
|
+
|
38
|
+
input(path, "") do
|
39
|
+
read { |paragraph| test.assert_equal(paragraphs.shift, paragraph) }
|
40
|
+
end
|
41
|
+
end
|
42
|
+
|
43
|
+
def test_restricted_reading
|
44
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
45
|
+
numbers = %w{two three}
|
46
|
+
test = self
|
47
|
+
|
48
|
+
input(path) do
|
49
|
+
read(/ (t\w+)\./) do |number|
|
50
|
+
test.assert_equal("This is line #{numbers.first}.\n", @read)
|
51
|
+
test.assert_equal(numbers.shift, number)
|
52
|
+
end
|
53
|
+
end
|
54
|
+
end
|
55
|
+
|
56
|
+
def test_data_saving_and_retrieving
|
57
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
58
|
+
|
59
|
+
data = input path do
|
60
|
+
read(/ (o\w+)\./) { |number| @number = number }
|
61
|
+
read { |line| (@lines ||= Array.new) << line }
|
62
|
+
end
|
63
|
+
|
64
|
+
assert_equal("one", data[:number])
|
65
|
+
assert_equal("one", data.number)
|
66
|
+
assert_equal(path, data.path)
|
67
|
+
assert_equal(5, data.lines.size)
|
68
|
+
assert_equal(File.readlines(path), data.lines)
|
69
|
+
end
|
70
|
+
end
|
data/test/tc_state.rb
ADDED
@@ -0,0 +1,142 @@
|
|
1
|
+
#!/usr/local/bin/ruby -w
|
2
|
+
|
3
|
+
# tc_state.rb
|
4
|
+
#
|
5
|
+
# Created by James Edward Gray II on 2005-08-14.
|
6
|
+
# Copyright 2005 Gray Productions. All rights reserved.
|
7
|
+
|
8
|
+
require "test/unit"
|
9
|
+
|
10
|
+
require "parse/input"
|
11
|
+
|
12
|
+
class TestState < Test::Unit::TestCase
|
13
|
+
def test_single_skips
|
14
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
15
|
+
lines = File.readlines(path).values_at(0..2, 4)
|
16
|
+
test = self
|
17
|
+
|
18
|
+
input(path) do
|
19
|
+
skip(?u)
|
20
|
+
|
21
|
+
read { |line| test.assert_equal(lines.shift, line) }
|
22
|
+
end
|
23
|
+
end
|
24
|
+
|
25
|
+
def test_skipping_range
|
26
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
27
|
+
lines = File.readlines(path).values_at(0, 3..4)
|
28
|
+
test = self
|
29
|
+
|
30
|
+
input(path) do
|
31
|
+
start_skipping_at("two")
|
32
|
+
stop_skipping_at("three")
|
33
|
+
|
34
|
+
read { |line| test.assert_equal(lines.shift, line) }
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
def test_searching_skipped
|
39
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
40
|
+
|
41
|
+
data = input(path) do
|
42
|
+
@state = :skip
|
43
|
+
stop_skipping_at("four")
|
44
|
+
find_in_skipped(/ou/) { |line| @found_in_skip = line }
|
45
|
+
|
46
|
+
read { |line| (@lines ||= Array.new) << line }
|
47
|
+
end
|
48
|
+
|
49
|
+
assert_equal("This is line four.\n", data.found_in_skip)
|
50
|
+
assert_equal(["This is line five.\n"], data.lines)
|
51
|
+
end
|
52
|
+
|
53
|
+
def test_stop
|
54
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
55
|
+
|
56
|
+
data = input(path) do
|
57
|
+
stop_at("three")
|
58
|
+
|
59
|
+
read { |line| @last_line = line }
|
60
|
+
end
|
61
|
+
|
62
|
+
assert_equal("This is line two.\n", data.last_line)
|
63
|
+
|
64
|
+
data = input(path) do
|
65
|
+
read do |line|
|
66
|
+
@last_line = line
|
67
|
+
|
68
|
+
@state = :stop if @read.index("three")
|
69
|
+
end
|
70
|
+
end
|
71
|
+
|
72
|
+
assert_equal("This is line three.\n", data.last_line)
|
73
|
+
end
|
74
|
+
|
75
|
+
def test_pre_and_post
|
76
|
+
path = File.join(File.dirname(__FILE__), "five_lines.txt")
|
77
|
+
|
78
|
+
data = input(path) do
|
79
|
+
@pre = @post = 0
|
80
|
+
|
81
|
+
pre do
|
82
|
+
@pre += 1
|
83
|
+
@post = 0
|
84
|
+
end
|
85
|
+
read { |line| (@lines ||= Array.new) << line }
|
86
|
+
read(/\w+\./) { |number| (@numbers ||= Array.new) << number }
|
87
|
+
post { @post += 1 }
|
88
|
+
end
|
89
|
+
|
90
|
+
assert_equal(5, data.pre)
|
91
|
+
assert_equal(1, data.post)
|
92
|
+
end
|
93
|
+
|
94
|
+
def test_complex
|
95
|
+
path = File.join(File.dirname(__FILE__), "ross_report.txt")
|
96
|
+
test = self
|
97
|
+
|
98
|
+
input(path) do
|
99
|
+
@state = :skip
|
100
|
+
start_skipping_at("\f")
|
101
|
+
stop_skipping_at(/\A-[- ]+-\Z/)
|
102
|
+
skip(/\A\s*\Z/)
|
103
|
+
skip(/--\Z/)
|
104
|
+
|
105
|
+
find_in_skipped(/((?:Period|Week)\s+\d.+?)\s*\Z/) do |period|
|
106
|
+
test.assert_equal("Period 02/2002", period)
|
107
|
+
end
|
108
|
+
|
109
|
+
stop_at("*** Selection Criteria ***")
|
110
|
+
|
111
|
+
read do |line|
|
112
|
+
test.assert_match(/\A\s+(?:Sales|Cust|SA)|\A[-\w]+\s+/, line)
|
113
|
+
end
|
114
|
+
end
|
115
|
+
|
116
|
+
path = File.join(File.dirname(__FILE__), "car_ads.txt")
|
117
|
+
|
118
|
+
data = input(path, "") do
|
119
|
+
@state = :skip
|
120
|
+
stop_skipping_at("Save Ad")
|
121
|
+
skip(/\A\s*\Z/)
|
122
|
+
|
123
|
+
pre { @price = @miles = nil }
|
124
|
+
read(/\$([\d,]+\d)/) { |price| @price = price.delete(",").to_i }
|
125
|
+
read(/([\d,]*\d)\s*m/) { |miles| @miles = miles.delete(",").to_i }
|
126
|
+
|
127
|
+
read do |ad|
|
128
|
+
if @price and @price < 20_000 and @miles and @miles < 40_000
|
129
|
+
(@ads ||= Array.new) << ad.strip
|
130
|
+
end
|
131
|
+
end
|
132
|
+
end
|
133
|
+
|
134
|
+
assert_equal([<<END_AD.strip], data.ads)
|
135
|
+
2003 Chrysler Town & Country LX
|
136
|
+
$16,990, green, 21,488 mi, air, pw, power locks, ps, power mirrors,
|
137
|
+
dual air bags, keyless entry, intermittent wipers, rear defroster, alloy,
|
138
|
+
pb, abs, cruise, am/fm stereo, CD, cassette, tinted glass
|
139
|
+
VIN:2C4GP44363R153238, Stock No:C153238, CALL DAN PERKINS AT 1-800-432-6326
|
140
|
+
END_AD
|
141
|
+
end
|
142
|
+
end
|
data/test/ts_all.rb
ADDED
metadata
ADDED
@@ -0,0 +1,47 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
rubygems_version: 0.8.11
|
3
|
+
specification_version: 1
|
4
|
+
name: parseinput
|
5
|
+
version: !ruby/object:Gem::Version
|
6
|
+
version: 0.0.1
|
7
|
+
date: 2005-08-18 00:00:00 -05:00
|
8
|
+
summary: Parse Input is a chain-saw tool for data mining.
|
9
|
+
require_paths:
|
10
|
+
- lib
|
11
|
+
email: james@grayproductions.net
|
12
|
+
homepage: http://input.rubyforge.org
|
13
|
+
rubyforge_project: input
|
14
|
+
description: "Parse Input is a library that aids in parsing generic input with Ruby. This
|
15
|
+
isn't intended to be a full-blown parser, but instead a chain-saw tool for data
|
16
|
+
mining arbitrary inputs quickly and easily."
|
17
|
+
autorequire: parse/input
|
18
|
+
default_executable:
|
19
|
+
bindir: bin
|
20
|
+
has_rdoc: false
|
21
|
+
required_ruby_version: !ruby/object:Gem::Version::Requirement
|
22
|
+
requirements:
|
23
|
+
-
|
24
|
+
- ">"
|
25
|
+
- !ruby/object:Gem::Version
|
26
|
+
version: 0.0.0
|
27
|
+
version:
|
28
|
+
platform: ruby
|
29
|
+
signing_key:
|
30
|
+
cert_chain:
|
31
|
+
authors:
|
32
|
+
- James Edward Gray II
|
33
|
+
files:
|
34
|
+
- lib/parse/input.rb
|
35
|
+
- test/tc_reading.rb
|
36
|
+
- test/tc_state.rb
|
37
|
+
- test/ts_all.rb
|
38
|
+
- Rakefile
|
39
|
+
- setup.rb
|
40
|
+
test_files:
|
41
|
+
- test/ts_all.rb
|
42
|
+
rdoc_options: []
|
43
|
+
extra_rdoc_files: []
|
44
|
+
executables: []
|
45
|
+
extensions: []
|
46
|
+
requirements: []
|
47
|
+
dependencies: []
|