csv-utils 0.3.18 → 0.3.20
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.
- checksums.yaml +4 -4
- data/bin/csv-grep +91 -0
- data/csv-utils.gemspec +1 -1
- data/lib/csv_utils/csv_iterator.rb +18 -0
- metadata +4 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 762ef5d73b9c7995cb53fa62dcbd8341684b98a361da9dee744669865d1e3220
|
4
|
+
data.tar.gz: 240e15b3dafe12aba42d8a76fe3fe34a8c93dfe71c0975c68c2355466c441ef8
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 5bcf6704d8024d983147e27282fec6dc5ff23ce199e686a439d9a10cdfa689c9e5c22272580219990e3e29149277b766b91b3096073e691bd1733ee418f15159
|
7
|
+
data.tar.gz: 6cef1bfb34f1cdb0127820d99780f8795bc5fc463ad9daec0f6b0746b107d9543d2c982401e26663c4ac23d009f90287d4cc8a0cc76c7387675bf156952099d0
|
data/bin/csv-grep
ADDED
@@ -0,0 +1,91 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
require 'optparse'
|
4
|
+
|
5
|
+
options = {
|
6
|
+
search: nil,
|
7
|
+
exact_match: false,
|
8
|
+
headers: :first,
|
9
|
+
limit: nil,
|
10
|
+
search_regex_options: nil
|
11
|
+
}
|
12
|
+
OptionParser.new do |opts|
|
13
|
+
opts.banner = 'Usage: ' + File.basename(__FILE__) + ' [options] <csv file>'
|
14
|
+
|
15
|
+
opts.on('-h', '--help', 'Prints this help') do
|
16
|
+
puts opts
|
17
|
+
exit
|
18
|
+
end
|
19
|
+
|
20
|
+
opts.on('-s', '--search SEARCH', 'Search expression') do |v|
|
21
|
+
options[:search] = v.to_s
|
22
|
+
end
|
23
|
+
|
24
|
+
opts.on('-e', '--exact-match', 'Exact match') do
|
25
|
+
options[:exact_match] = true
|
26
|
+
end
|
27
|
+
|
28
|
+
# -c stands for column, since -h is used for help
|
29
|
+
opts.on('-c', '--headers HEADERS', 'Comma separated list of headers to search (default first column)') do |v|
|
30
|
+
options[:headers] = v == 'all' ? :all : v.split(',')
|
31
|
+
end
|
32
|
+
|
33
|
+
opts.on('-l', '--limit LIMIT', Integer, 'Limit the number of matches') do |v|
|
34
|
+
options[:limit] = v
|
35
|
+
end
|
36
|
+
|
37
|
+
opts.on('-i', '--ignore-case', 'Ignore case') do |v|
|
38
|
+
options[:search_regex_options] = Regexp::IGNORECASE
|
39
|
+
end
|
40
|
+
end.parse!
|
41
|
+
|
42
|
+
raise('no search specified') unless options[:search]
|
43
|
+
|
44
|
+
require 'csv-utils'
|
45
|
+
|
46
|
+
csv = CSVUtils::CSVIterator.new(ARGV[0])
|
47
|
+
|
48
|
+
search_regex =
|
49
|
+
if options[:exact_match]
|
50
|
+
Regexp.new('\A' + Regexp.escape(options[:search]) + '\z', options[:search_regex_options])
|
51
|
+
else
|
52
|
+
Regexp.new(options[:search], options[:search_regex_options])
|
53
|
+
end
|
54
|
+
|
55
|
+
headers =
|
56
|
+
case options[:headers]
|
57
|
+
when :first
|
58
|
+
csv.headers.first
|
59
|
+
when :all
|
60
|
+
csv.headers
|
61
|
+
else
|
62
|
+
options[:headers]
|
63
|
+
end
|
64
|
+
|
65
|
+
missing_headers = headers - csv.headers
|
66
|
+
raise("unknown headers #{headers.join(', ')}") unless missing_headers.empty?
|
67
|
+
|
68
|
+
matching_row_proc = proc do |row|
|
69
|
+
result = false
|
70
|
+
|
71
|
+
headers.each do |header|
|
72
|
+
next unless (val = row[header])
|
73
|
+
|
74
|
+
if search_regex.match?(val)
|
75
|
+
result = true
|
76
|
+
break
|
77
|
+
end
|
78
|
+
end
|
79
|
+
|
80
|
+
result
|
81
|
+
end
|
82
|
+
|
83
|
+
matches = 0
|
84
|
+
csv.each do |row|
|
85
|
+
next unless matching_row_proc.call(row)
|
86
|
+
|
87
|
+
matches += 1
|
88
|
+
print row.to_pretty_s + "\n"
|
89
|
+
|
90
|
+
break if options[:limit] && matches >= options[:limit]
|
91
|
+
end
|
data/csv-utils.gemspec
CHANGED
@@ -76,6 +76,24 @@ class CSVUtils::CSVIterator
|
|
76
76
|
cnt
|
77
77
|
end
|
78
78
|
|
79
|
+
def each_batch(batch_size = 1_000)
|
80
|
+
batch = []
|
81
|
+
|
82
|
+
process_batch_proc = Proc.new do
|
83
|
+
yield batch
|
84
|
+
batch = []
|
85
|
+
end
|
86
|
+
|
87
|
+
each do |row|
|
88
|
+
batch << row
|
89
|
+
process_batch_proc.call if batch.size >= batch_size
|
90
|
+
end
|
91
|
+
|
92
|
+
process_batch_proc.call if batch.size > 0
|
93
|
+
|
94
|
+
nil
|
95
|
+
end
|
96
|
+
|
79
97
|
private
|
80
98
|
|
81
99
|
def strip_bom!(col)
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: csv-utils
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.3.
|
4
|
+
version: 0.3.20
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Doug Youch
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2022-
|
11
|
+
date: 2022-11-14 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: inheritance-helper
|
@@ -32,6 +32,7 @@ executables:
|
|
32
32
|
- csv-duplicate-finder
|
33
33
|
- csv-explorer
|
34
34
|
- csv-find-error
|
35
|
+
- csv-grep
|
35
36
|
- csv-readline
|
36
37
|
- csv-validator
|
37
38
|
extensions: []
|
@@ -49,6 +50,7 @@ files:
|
|
49
50
|
- bin/csv-duplicate-finder
|
50
51
|
- bin/csv-explorer
|
51
52
|
- bin/csv-find-error
|
53
|
+
- bin/csv-grep
|
52
54
|
- bin/csv-readline
|
53
55
|
- bin/csv-validator
|
54
56
|
- csv-utils.gemspec
|