csv-utils 0.3.18 → 0.3.20

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: cfcbc7a92377a1b5d4dec01847defb5a4b2ac76c3dc65f40ec113c9d2085d2c6
-  data.tar.gz: bff2e91650ff90feb08e960ab3356d70486759aca7e4ddf54020081d31f2ce18
+  metadata.gz: 762ef5d73b9c7995cb53fa62dcbd8341684b98a361da9dee744669865d1e3220
+  data.tar.gz: 240e15b3dafe12aba42d8a76fe3fe34a8c93dfe71c0975c68c2355466c441ef8
 SHA512:
-  metadata.gz: 7d7db74ed5c52d058fabe0498f9404915c398826be471b8407f839a9eec89c64d797cc9b9f8f1751b89ccd1b80c14165190ec50f595e942f9f0fbc151f544985
-  data.tar.gz: 90578cb216ba3d5260304a8cd9302bcae3e8b8e892592f7adc75db7536652822e0c729e57a074a99d77472751bfa0c92e2470be0bd7c68393ece4a782e93eb2b
+  metadata.gz: 5bcf6704d8024d983147e27282fec6dc5ff23ce199e686a439d9a10cdfa689c9e5c22272580219990e3e29149277b766b91b3096073e691bd1733ee418f15159
+  data.tar.gz: 6cef1bfb34f1cdb0127820d99780f8795bc5fc463ad9daec0f6b0746b107d9543d2c982401e26663c4ac23d009f90287d4cc8a0cc76c7387675bf156952099d0

data/bin/csv-grep

@@ -0,0 +1,91 @@
+#!/usr/bin/env ruby
+
+require 'optparse'
+
+options = {
+  search: nil,
+  exact_match: false,
+  headers: :first,
+  limit: nil,
+  search_regex_options: nil
+}
+OptionParser.new do |opts|
+  opts.banner = 'Usage: ' + File.basename(__FILE__) + ' [options] <csv file>'
+
+  opts.on('-h', '--help', 'Prints this help') do
+    puts opts
+    exit
+  end
+
+  opts.on('-s', '--search SEARCH', 'Search expression') do |v|
+    options[:search] = v.to_s
+  end
+
+  opts.on('-e', '--exact-match', 'Exact match') do
+    options[:exact_match] = true
+  end
+
+  # -c stands for column, since -h is used for help
+  opts.on('-c', '--headers HEADERS', 'Comma separated list of headers to search (default first column)') do |v|
+    options[:headers] = v == 'all' ? :all : v.split(',')
+  end
+
+  opts.on('-l', '--limit LIMIT', Integer, 'Limit the number of matches') do |v|
+    options[:limit] = v
+  end
+
+  opts.on('-i', '--ignore-case', 'Ignore case') do |v|
+    options[:search_regex_options] = Regexp::IGNORECASE
+  end
+end.parse!
+
+raise('no search specified') unless options[:search]
+
+require 'csv-utils'
+
+csv = CSVUtils::CSVIterator.new(ARGV[0])
+
+search_regex =
+  if options[:exact_match]
+    Regexp.new('\A' + Regexp.escape(options[:search]) + '\z', options[:search_regex_options])
+  else
+    Regexp.new(options[:search], options[:search_regex_options])
+  end
+
+headers =
+  case options[:headers]
+  when :first
+    csv.headers.first
+  when :all
+    csv.headers
+  else
+    options[:headers]
+  end
+
+missing_headers = headers - csv.headers
+raise("unknown headers #{headers.join(', ')}") unless missing_headers.empty?
+
+matching_row_proc = proc do |row|
+  result = false
+
+  headers.each do |header|
+    next unless (val = row[header])
+
+    if search_regex.match?(val)
+      result = true
+      break
+    end
+  end
+
+  result
+end
+
+matches = 0
+csv.each do |row|
+  next unless matching_row_proc.call(row)
+
+  matches += 1
+  print row.to_pretty_s + "\n"
+
+  break if options[:limit] && matches >= options[:limit]
+end

data/csv-utils.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name        = 'csv-utils'
-  s.version     = '0.3.18'
+  s.version     = '0.3.20'
   s.licenses    = ['MIT']
   s.summary     = 'CSV Utils'
   s.description = 'Tools for debugging malformed CSV files'

data/lib/csv_utils/csv_iterator.rb

@@ -76,6 +76,24 @@ class CSVUtils::CSVIterator
     cnt
   end
 
+  def each_batch(batch_size = 1_000)
+    batch = []
+
+    process_batch_proc = Proc.new do
+      yield batch
+      batch = []
+    end
+
+    each do |row|
+      batch << row
+      process_batch_proc.call if batch.size >= batch_size
+    end
+
+    process_batch_proc.call if batch.size > 0
+
+    nil
+  end
+
   private
 
   def strip_bom!(col)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: csv-utils
 version: !ruby/object:Gem::Version
-  version: 0.3.18
+  version: 0.3.20
 platform: ruby
 authors:
 - Doug Youch
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2022-10-18 00:00:00.000000000 Z
+date: 2022-11-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: inheritance-helper
@@ -32,6 +32,7 @@ executables:
 - csv-duplicate-finder
 - csv-explorer
 - csv-find-error
+- csv-grep
 - csv-readline
 - csv-validator
 extensions: []
@@ -49,6 +50,7 @@ files:
 - bin/csv-duplicate-finder
 - bin/csv-explorer
 - bin/csv-find-error
+- bin/csv-grep
 - bin/csv-readline
 - bin/csv-validator
 - csv-utils.gemspec