smarter_csv 1.9.0 → 1.9.2.pre01
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/CHANGELOG.md +13 -0
- data/Gemfile +6 -3
- data/README.md +5 -1
- data/Rakefile +20 -5
- data/ext/smarter_csv/extconf.rb +2 -2
- data/ext/smarter_csv/smarter_csv.c +8 -2
- data/lib/smarter_csv/options_processing.rb +84 -0
- data/lib/smarter_csv/smarter_csv.rb +556 -0
- data/lib/smarter_csv/version.rb +1 -1
- data/lib/smarter_csv.rb +27 -610
- data/smarter_csv.gemspec +2 -1
- metadata +8 -5
- /data/lib/{extensions → core_ext}/hash.rb +0 -0
data/lib/smarter_csv.rb
CHANGED
@@ -1,622 +1,39 @@
|
|
1
1
|
# frozen_string_literal: true
|
2
2
|
|
3
|
-
|
4
|
-
require_relative "smarter_csv/version"
|
3
|
+
require "core_ext/hash"
|
5
4
|
|
6
|
-
|
7
|
-
|
5
|
+
require "smarter_csv/version"
|
6
|
+
require "smarter_csv/options_processing"
|
8
7
|
|
9
|
-
|
10
|
-
|
11
|
-
|
12
|
-
|
13
|
-
|
14
|
-
class DuplicateHeaders < SmarterCSVException; end
|
15
|
-
class MissingKeys < SmarterCSVException; end # previously known as MissingHeaders
|
16
|
-
class NoColSepDetected < SmarterCSVException; end
|
17
|
-
class KeyMappingError < SmarterCSVException; end
|
18
|
-
|
19
|
-
# first parameter: filename or input object which responds to readline method
|
20
|
-
def SmarterCSV.process(input, options = {}, &block) # rubocop:disable Lint/UnusedMethodArgument
|
21
|
-
options = default_options.merge(options)
|
22
|
-
options[:invalid_byte_sequence] = '' if options[:invalid_byte_sequence].nil?
|
23
|
-
puts "SmarterCSV OPTIONS: #{options.inspect}" if options[:verbose]
|
24
|
-
validate_options!(options)
|
25
|
-
|
26
|
-
headerA = []
|
27
|
-
result = []
|
28
|
-
@file_line_count = 0
|
29
|
-
@csv_line_count = 0
|
30
|
-
has_rails = !!defined?(Rails)
|
31
|
-
begin
|
32
|
-
fh = input.respond_to?(:readline) ? input : File.open(input, "r:#{options[:file_encoding]}")
|
33
|
-
|
34
|
-
# auto-detect the row separator
|
35
|
-
options[:row_sep] = guess_line_ending(fh, options) if options[:row_sep]&.to_sym == :auto
|
36
|
-
# attempt to auto-detect column separator
|
37
|
-
options[:col_sep] = guess_column_separator(fh, options) if options[:col_sep]&.to_sym == :auto
|
38
|
-
|
39
|
-
if (options[:force_utf8] || options[:file_encoding] =~ /utf-8/i) && (fh.respond_to?(:external_encoding) && fh.external_encoding != Encoding.find('UTF-8') || fh.respond_to?(:encoding) && fh.encoding != Encoding.find('UTF-8'))
|
40
|
-
puts 'WARNING: you are trying to process UTF-8 input, but did not open the input with "b:utf-8" option. See README file "NOTES about File Encodings".'
|
41
|
-
end
|
42
|
-
|
43
|
-
skip_lines(fh, options)
|
44
|
-
|
45
|
-
headerA, header_size = process_headers(fh, options)
|
46
|
-
|
47
|
-
# in case we use chunking.. we'll need to set it up..
|
48
|
-
if !options[:chunk_size].nil? && options[:chunk_size].to_i > 0
|
49
|
-
use_chunks = true
|
50
|
-
chunk_size = options[:chunk_size].to_i
|
51
|
-
chunk_count = 0
|
52
|
-
chunk = []
|
53
|
-
else
|
54
|
-
use_chunks = false
|
55
|
-
end
|
56
|
-
|
57
|
-
# now on to processing all the rest of the lines in the CSV file:
|
58
|
-
until fh.eof? # we can't use fh.readlines() here, because this would read the whole file into memory at once, and eof => true
|
59
|
-
line = readline_with_counts(fh, options)
|
60
|
-
|
61
|
-
# replace invalid byte sequence in UTF-8 with question mark to avoid errors
|
62
|
-
line = line.force_encoding('utf-8').encode('utf-8', invalid: :replace, undef: :replace, replace: options[:invalid_byte_sequence]) if options[:force_utf8] || options[:file_encoding] !~ /utf-8/i
|
63
|
-
|
64
|
-
print "processing file line %10d, csv line %10d\r" % [@file_line_count, @csv_line_count] if options[:verbose]
|
65
|
-
|
66
|
-
next if options[:comment_regexp] && line =~ options[:comment_regexp] # ignore all comment lines if there are any
|
67
|
-
|
68
|
-
# cater for the quoted csv data containing the row separator carriage return character
|
69
|
-
# in which case the row data will be split across multiple lines (see the sample content in spec/fixtures/carriage_returns_rn.csv)
|
70
|
-
# by detecting the existence of an uneven number of quote characters
|
71
|
-
|
72
|
-
multiline = count_quote_chars(line, options[:quote_char]).odd? # should handle quote_char nil
|
73
|
-
while count_quote_chars(line, options[:quote_char]).odd? # should handle quote_char nil
|
74
|
-
next_line = fh.readline(options[:row_sep])
|
75
|
-
next_line = next_line.force_encoding('utf-8').encode('utf-8', invalid: :replace, undef: :replace, replace: options[:invalid_byte_sequence]) if options[:force_utf8] || options[:file_encoding] !~ /utf-8/i
|
76
|
-
line += next_line
|
77
|
-
@file_line_count += 1
|
78
|
-
end
|
79
|
-
print "\nline contains uneven number of quote chars so including content through file line %d\n" % @file_line_count if options[:verbose] && multiline
|
80
|
-
|
81
|
-
line.chomp!(options[:row_sep])
|
82
|
-
|
83
|
-
dataA, _data_size = parse(line, options, header_size)
|
84
|
-
|
85
|
-
dataA.map!{|x| x.strip} if options[:strip_whitespace]
|
86
|
-
|
87
|
-
# if all values are blank, then ignore this line
|
88
|
-
next if options[:remove_empty_hashes] && (dataA.empty? || blank?(dataA))
|
89
|
-
|
90
|
-
hash = Hash.zip(headerA, dataA) # from Facets of Ruby library
|
91
|
-
|
92
|
-
# make sure we delete any key/value pairs from the hash, which the user wanted to delete:
|
93
|
-
# Note: Ruby < 1.9 doesn't allow empty symbol literals!
|
94
|
-
hash.delete(nil)
|
95
|
-
hash.delete('')
|
96
|
-
eval('hash.delete(:"")') if RUBY_VERSION.to_f > 1.8
|
97
|
-
|
98
|
-
if options[:remove_empty_values] == true
|
99
|
-
hash.delete_if{|_k, v| has_rails ? v.blank? : blank?(v)}
|
100
|
-
end
|
101
|
-
|
102
|
-
hash.delete_if{|_k, v| !v.nil? && v =~ /^(0+|0+\.0+)$/} if options[:remove_zero_values] # values are Strings
|
103
|
-
hash.delete_if{|_k, v| v =~ options[:remove_values_matching]} if options[:remove_values_matching]
|
104
|
-
|
105
|
-
if options[:convert_values_to_numeric]
|
106
|
-
hash.each do |k, v|
|
107
|
-
# deal with the :only / :except options to :convert_values_to_numeric
|
108
|
-
next if only_or_except_limit_execution(options, :convert_values_to_numeric, k)
|
109
|
-
|
110
|
-
# convert if it's a numeric value:
|
111
|
-
case v
|
112
|
-
when /^[+-]?\d+\.\d+$/
|
113
|
-
hash[k] = v.to_f
|
114
|
-
when /^[+-]?\d+$/
|
115
|
-
hash[k] = v.to_i
|
116
|
-
end
|
117
|
-
end
|
118
|
-
end
|
119
|
-
|
120
|
-
if options[:value_converters]
|
121
|
-
hash.each do |k, v|
|
122
|
-
converter = options[:value_converters][k]
|
123
|
-
next unless converter
|
124
|
-
|
125
|
-
hash[k] = converter.convert(v)
|
126
|
-
end
|
127
|
-
end
|
128
|
-
|
129
|
-
next if options[:remove_empty_hashes] && hash.empty?
|
130
|
-
|
131
|
-
hash[:csv_line_number] = @csv_line_count if options[:with_line_numbers]
|
132
|
-
|
133
|
-
if use_chunks
|
134
|
-
chunk << hash # append temp result to chunk
|
135
|
-
|
136
|
-
if chunk.size >= chunk_size || fh.eof? # if chunk if full, or EOF reached
|
137
|
-
# do something with the chunk
|
138
|
-
if block_given?
|
139
|
-
yield chunk # do something with the hashes in the chunk in the block
|
140
|
-
else
|
141
|
-
result << chunk # not sure yet, why anybody would want to do this without a block
|
142
|
-
end
|
143
|
-
chunk_count += 1
|
144
|
-
chunk = [] # initialize for next chunk of data
|
145
|
-
else
|
146
|
-
|
147
|
-
# the last chunk may contain partial data, which also needs to be returned (BUG / ISSUE-18)
|
148
|
-
|
149
|
-
end
|
150
|
-
|
151
|
-
# while a chunk is being filled up we don't need to do anything else here
|
152
|
-
|
153
|
-
else # no chunk handling
|
154
|
-
if block_given?
|
155
|
-
yield [hash] # do something with the hash in the block (better to use chunking here)
|
156
|
-
else
|
157
|
-
result << hash
|
158
|
-
end
|
159
|
-
end
|
160
|
-
end
|
161
|
-
|
162
|
-
# print new line to retain last processing line message
|
163
|
-
print "\n" if options[:verbose]
|
164
|
-
|
165
|
-
# last chunk:
|
166
|
-
if !chunk.nil? && chunk.size > 0
|
167
|
-
# do something with the chunk
|
168
|
-
if block_given?
|
169
|
-
yield chunk # do something with the hashes in the chunk in the block
|
170
|
-
else
|
171
|
-
result << chunk # not sure yet, why anybody would want to do this without a block
|
172
|
-
end
|
173
|
-
chunk_count += 1
|
174
|
-
# chunk = [] # initialize for next chunk of data
|
175
|
-
end
|
176
|
-
ensure
|
177
|
-
fh.close if fh.respond_to?(:close)
|
178
|
-
end
|
179
|
-
if block_given?
|
180
|
-
chunk_count # when we do processing through a block we only care how many chunks we processed
|
8
|
+
case RUBY_ENGINE
|
9
|
+
when 'ruby'
|
10
|
+
begin
|
11
|
+
if `uname -s`.chomp == 'Darwin'
|
12
|
+
require 'smarter_csv/smarter_csv.bundle'
|
181
13
|
else
|
182
|
-
|
14
|
+
require_relative "smarter_csv/smarter_csv"
|
183
15
|
end
|
16
|
+
rescue Exception
|
17
|
+
# require_relative 'smarter_csv/smarter_csv'
|
184
18
|
end
|
19
|
+
# :nocov:
|
20
|
+
# when 'truffleruby'
|
21
|
+
# puts "\n\n truffleruby case in the load path | RUBY_ENGINE: #{RUBY_ENGINE} , #{RUBY_VERSION}\n\n"
|
22
|
+
# # this might not work - if you encounter problems, please contribute and create a PR
|
23
|
+
# # require 'truffleruby/smarter_csv'
|
24
|
+
else
|
25
|
+
puts <<-BLOCK_COMMENT
|
185
26
|
|
186
|
-
|
187
|
-
|
188
|
-
@has_acceleration ||= !!defined?(parse_csv_line_c)
|
189
|
-
end
|
190
|
-
|
191
|
-
def raw_header
|
192
|
-
@raw_header
|
193
|
-
end
|
194
|
-
|
195
|
-
def headers
|
196
|
-
@headers
|
197
|
-
end
|
198
|
-
|
199
|
-
# Counts the number of quote characters in a line, excluding escaped quotes.
|
200
|
-
def count_quote_chars(line, quote_char)
|
201
|
-
return 0 if line.nil? || quote_char.nil?
|
202
|
-
|
203
|
-
count = 0
|
204
|
-
previous_char = ''
|
27
|
+
-------------------------------------------------------------------------
|
28
|
+
RUBY_ENGINE: #{RUBY_ENGINE} , #{RUBY_VERSION}
|
205
29
|
|
206
|
-
|
207
|
-
count += 1 if char == quote_char && previous_char != '\\'
|
208
|
-
previous_char = char
|
209
|
-
end
|
30
|
+
Acceleration via C-Extension is currently not supported for #{RUBY_ENGINE}
|
210
31
|
|
211
|
-
|
212
|
-
|
213
|
-
|
214
|
-
protected
|
215
|
-
|
216
|
-
# NOTE: this is not called when "parse" methods are tested by themselves
|
217
|
-
def default_options
|
218
|
-
{
|
219
|
-
acceleration: true,
|
220
|
-
auto_row_sep_chars: 500,
|
221
|
-
chunk_size: nil,
|
222
|
-
col_sep: :auto, # was: ',',
|
223
|
-
comment_regexp: nil, # was: /\A#/,
|
224
|
-
convert_values_to_numeric: true,
|
225
|
-
downcase_header: true,
|
226
|
-
duplicate_header_suffix: nil,
|
227
|
-
file_encoding: 'utf-8',
|
228
|
-
force_simple_split: false,
|
229
|
-
force_utf8: false,
|
230
|
-
headers_in_file: true,
|
231
|
-
invalid_byte_sequence: '',
|
232
|
-
keep_original_headers: false,
|
233
|
-
key_mapping: nil,
|
234
|
-
quote_char: '"',
|
235
|
-
remove_empty_hashes: true,
|
236
|
-
remove_empty_values: true,
|
237
|
-
remove_unmapped_keys: false,
|
238
|
-
remove_values_matching: nil,
|
239
|
-
remove_zero_values: false,
|
240
|
-
required_headers: nil,
|
241
|
-
required_keys: nil,
|
242
|
-
row_sep: :auto, # was: $/,
|
243
|
-
silence_missing_keys: false,
|
244
|
-
skip_lines: nil,
|
245
|
-
strings_as_keys: false,
|
246
|
-
strip_chars_from_headers: nil,
|
247
|
-
strip_whitespace: true,
|
248
|
-
user_provided_headers: nil,
|
249
|
-
value_converters: nil,
|
250
|
-
verbose: false,
|
251
|
-
with_line_numbers: false,
|
252
|
-
}
|
253
|
-
end
|
254
|
-
|
255
|
-
def readline_with_counts(filehandle, options)
|
256
|
-
line = filehandle.readline(options[:row_sep])
|
257
|
-
@file_line_count += 1
|
258
|
-
@csv_line_count += 1
|
259
|
-
line = remove_bom(line) if @csv_line_count == 1
|
260
|
-
line
|
261
|
-
end
|
262
|
-
|
263
|
-
def skip_lines(filehandle, options)
|
264
|
-
return unless options[:skip_lines].to_i > 0
|
265
|
-
|
266
|
-
options[:skip_lines].to_i.times do
|
267
|
-
readline_with_counts(filehandle, options)
|
268
|
-
end
|
269
|
-
end
|
270
|
-
|
271
|
-
def rewind(filehandle)
|
272
|
-
@file_line_count = 0
|
273
|
-
@csv_line_count = 0
|
274
|
-
filehandle.rewind
|
275
|
-
end
|
276
|
-
|
277
|
-
###
|
278
|
-
### Thin wrapper around C-extension
|
279
|
-
###
|
280
|
-
def parse(line, options, header_size = nil)
|
281
|
-
# puts "SmarterCSV.parse OPTIONS: #{options[:acceleration]}" if options[:verbose]
|
282
|
-
|
283
|
-
if options[:acceleration] && has_acceleration?
|
284
|
-
# :nocov:
|
285
|
-
has_quotes = line =~ /#{options[:quote_char]}/
|
286
|
-
elements = parse_csv_line_c(line, options[:col_sep], options[:quote_char], header_size)
|
287
|
-
elements.map!{|x| cleanup_quotes(x, options[:quote_char])} if has_quotes
|
288
|
-
[elements, elements.size]
|
289
|
-
# :nocov:
|
290
|
-
else
|
291
|
-
# puts "WARNING: SmarterCSV is using un-accelerated parsing of lines. Check options[:acceleration]"
|
292
|
-
parse_csv_line_ruby(line, options, header_size)
|
293
|
-
end
|
294
|
-
end
|
295
|
-
|
296
|
-
# ------------------------------------------------------------------
|
297
|
-
# Ruby equivalent of the C-extension for parse_line
|
298
|
-
#
|
299
|
-
# parses a single line: either a CSV header and body line
|
300
|
-
# - quoting rules compared to RFC-4180 are somewhat relaxed
|
301
|
-
# - we are not assuming that quotes inside a fields need to be doubled
|
302
|
-
# - we are not assuming that all fields need to be quoted (0 is even)
|
303
|
-
# - works with multi-char col_sep
|
304
|
-
# - if header_size is given, only up to header_size fields are parsed
|
305
|
-
#
|
306
|
-
# We use header_size for parsing the body lines to make sure we always match the number of headers
|
307
|
-
# in case there are trailing col_sep characters in line
|
308
|
-
#
|
309
|
-
# Our convention is that empty fields are returned as empty strings, not as nil.
|
310
|
-
#
|
311
|
-
#
|
312
|
-
# the purpose of the max_size parameter is to handle a corner case where
|
313
|
-
# CSV lines contain more fields than the header.
|
314
|
-
# In which case the remaining fields in the line are ignored
|
315
|
-
#
|
316
|
-
def parse_csv_line_ruby(line, options, header_size = nil)
|
317
|
-
return [] if line.nil?
|
318
|
-
|
319
|
-
line_size = line.size
|
320
|
-
col_sep = options[:col_sep]
|
321
|
-
col_sep_size = col_sep.size
|
322
|
-
quote = options[:quote_char]
|
323
|
-
quote_count = 0
|
324
|
-
elements = []
|
325
|
-
start = 0
|
326
|
-
i = 0
|
327
|
-
|
328
|
-
previous_char = ''
|
329
|
-
while i < line_size
|
330
|
-
if line[i...i+col_sep_size] == col_sep && quote_count.even?
|
331
|
-
break if !header_size.nil? && elements.size >= header_size
|
332
|
-
|
333
|
-
elements << cleanup_quotes(line[start...i], quote)
|
334
|
-
previous_char = line[i]
|
335
|
-
i += col_sep.size
|
336
|
-
start = i
|
337
|
-
else
|
338
|
-
quote_count += 1 if line[i] == quote && previous_char != '\\'
|
339
|
-
previous_char = line[i]
|
340
|
-
i += 1
|
341
|
-
end
|
342
|
-
end
|
343
|
-
elements << cleanup_quotes(line[start..-1], quote) if header_size.nil? || elements.size < header_size
|
344
|
-
[elements, elements.size]
|
345
|
-
end
|
32
|
+
Please contribute and create a pull request if you need this
|
33
|
+
-------------------------------------------------------------------------
|
346
34
|
|
347
|
-
|
348
|
-
return field if field.nil?
|
349
|
-
|
350
|
-
# return if field !~ /#{quote}/ # this check can probably eliminated
|
351
|
-
|
352
|
-
if field.start_with?(quote) && field.end_with?(quote)
|
353
|
-
field.delete_prefix!(quote)
|
354
|
-
field.delete_suffix!(quote)
|
355
|
-
end
|
356
|
-
field.gsub!("#{quote}#{quote}", quote)
|
357
|
-
field
|
358
|
-
end
|
359
|
-
|
360
|
-
# SEE: https://github.com/rails/rails/blob/32015b6f369adc839c4f0955f2d9dce50c0b6123/activesupport/lib/active_support/core_ext/object/blank.rb#L121
|
361
|
-
# and in the future we might also include UTF-8 space characters: https://www.compart.com/en/unicode/category/Zs
|
362
|
-
BLANK_RE = /\A\s*\z/.freeze
|
363
|
-
|
364
|
-
def blank?(value)
|
365
|
-
case value
|
366
|
-
when String
|
367
|
-
value.empty? || BLANK_RE.match?(value)
|
368
|
-
|
369
|
-
when NilClass
|
370
|
-
true
|
371
|
-
|
372
|
-
when Array
|
373
|
-
value.empty? || value.inject(true){|result, x| result &&= elem_blank?(x)}
|
374
|
-
|
375
|
-
when Hash
|
376
|
-
value.empty? || value.values.inject(true){|result, x| result &&= elem_blank?(x)}
|
377
|
-
|
378
|
-
else
|
379
|
-
false
|
380
|
-
end
|
381
|
-
end
|
382
|
-
|
383
|
-
def elem_blank?(value)
|
384
|
-
case value
|
385
|
-
when String
|
386
|
-
value.empty? || BLANK_RE.match?(value)
|
387
|
-
|
388
|
-
when NilClass
|
389
|
-
true
|
390
|
-
|
391
|
-
else
|
392
|
-
false
|
393
|
-
end
|
394
|
-
end
|
395
|
-
|
396
|
-
# acts as a road-block to limit processing when iterating over all k/v pairs of a CSV-hash:
|
397
|
-
def only_or_except_limit_execution(options, option_name, key)
|
398
|
-
if options[option_name].is_a?(Hash)
|
399
|
-
if options[option_name].has_key?(:except)
|
400
|
-
return true if Array(options[option_name][:except]).include?(key)
|
401
|
-
elsif options[option_name].has_key?(:only)
|
402
|
-
return true unless Array(options[option_name][:only]).include?(key)
|
403
|
-
end
|
404
|
-
end
|
405
|
-
false
|
406
|
-
end
|
407
|
-
|
408
|
-
# If file has headers, then guesses column separator from headers.
|
409
|
-
# Otherwise guesses column separator from contents.
|
410
|
-
# Raises exception if none is found.
|
411
|
-
def guess_column_separator(filehandle, options)
|
412
|
-
skip_lines(filehandle, options)
|
413
|
-
|
414
|
-
delimiters = [',', "\t", ';', ':', '|']
|
415
|
-
|
416
|
-
line = nil
|
417
|
-
has_header = options[:headers_in_file]
|
418
|
-
candidates = Hash.new(0)
|
419
|
-
count = has_header ? 1 : 5
|
420
|
-
count.times do
|
421
|
-
line = readline_with_counts(filehandle, options)
|
422
|
-
delimiters.each do |d|
|
423
|
-
candidates[d] += line.scan(d).count
|
424
|
-
end
|
425
|
-
rescue EOFError # short files
|
426
|
-
break
|
427
|
-
end
|
428
|
-
rewind(filehandle)
|
429
|
-
|
430
|
-
if candidates.values.max == 0
|
431
|
-
# if the header only contains
|
432
|
-
return ',' if line.chomp(options[:row_sep]) =~ /^\w+$/
|
433
|
-
|
434
|
-
raise SmarterCSV::NoColSepDetected
|
435
|
-
end
|
436
|
-
|
437
|
-
candidates.key(candidates.values.max)
|
438
|
-
end
|
439
|
-
|
440
|
-
# limitation: this currently reads the whole file in before making a decision
|
441
|
-
def guess_line_ending(filehandle, options)
|
442
|
-
counts = {"\n" => 0, "\r" => 0, "\r\n" => 0}
|
443
|
-
quoted_char = false
|
444
|
-
|
445
|
-
# count how many of the pre-defined line-endings we find
|
446
|
-
# ignoring those contained within quote characters
|
447
|
-
last_char = nil
|
448
|
-
lines = 0
|
449
|
-
filehandle.each_char do |c|
|
450
|
-
quoted_char = !quoted_char if c == options[:quote_char]
|
451
|
-
next if quoted_char
|
452
|
-
|
453
|
-
if last_char == "\r"
|
454
|
-
if c == "\n"
|
455
|
-
counts["\r\n"] += 1
|
456
|
-
else
|
457
|
-
counts["\r"] += 1 # \r are counted after they appeared
|
458
|
-
end
|
459
|
-
elsif c == "\n"
|
460
|
-
counts["\n"] += 1
|
461
|
-
end
|
462
|
-
last_char = c
|
463
|
-
lines += 1
|
464
|
-
break if options[:auto_row_sep_chars] && options[:auto_row_sep_chars] > 0 && lines >= options[:auto_row_sep_chars]
|
465
|
-
end
|
466
|
-
rewind(filehandle)
|
467
|
-
|
468
|
-
counts["\r"] += 1 if last_char == "\r"
|
469
|
-
# find the most frequent key/value pair:
|
470
|
-
most_frequent_key, _count = counts.max_by{|_, v| v}
|
471
|
-
most_frequent_key
|
472
|
-
end
|
473
|
-
|
474
|
-
def process_headers(filehandle, options)
|
475
|
-
@raw_header = nil
|
476
|
-
@headers = nil
|
477
|
-
if options[:headers_in_file] # extract the header line
|
478
|
-
# process the header line in the CSV file..
|
479
|
-
# the first line of a CSV file contains the header .. it might be commented out, so we need to read it anyhow
|
480
|
-
header = readline_with_counts(filehandle, options)
|
481
|
-
@raw_header = header
|
482
|
-
|
483
|
-
header = header.force_encoding('utf-8').encode('utf-8', invalid: :replace, undef: :replace, replace: options[:invalid_byte_sequence]) if options[:force_utf8] || options[:file_encoding] !~ /utf-8/i
|
484
|
-
header = header.sub(options[:comment_regexp], '') if options[:comment_regexp]
|
485
|
-
header = header.chomp(options[:row_sep])
|
486
|
-
|
487
|
-
header = header.gsub(options[:strip_chars_from_headers], '') if options[:strip_chars_from_headers]
|
488
|
-
|
489
|
-
file_headerA, file_header_size = parse(header, options)
|
490
|
-
|
491
|
-
file_headerA.map!{|x| x.gsub(%r/#{options[:quote_char]}/, '')}
|
492
|
-
file_headerA.map!{|x| x.strip} if options[:strip_whitespace]
|
493
|
-
|
494
|
-
unless options[:keep_original_headers]
|
495
|
-
file_headerA.map!{|x| x.gsub(/\s+|-+/, '_')}
|
496
|
-
file_headerA.map!{|x| x.downcase} if options[:downcase_header]
|
497
|
-
end
|
498
|
-
else
|
499
|
-
raise SmarterCSV::IncorrectOption, "ERROR: If :headers_in_file is set to false, you have to provide :user_provided_headers" unless options[:user_provided_headers]
|
500
|
-
end
|
501
|
-
if options[:user_provided_headers] && options[:user_provided_headers].class == Array && !options[:user_provided_headers].empty?
|
502
|
-
# use user-provided headers
|
503
|
-
headerA = options[:user_provided_headers]
|
504
|
-
if defined?(file_header_size) && !file_header_size.nil?
|
505
|
-
if headerA.size != file_header_size
|
506
|
-
raise SmarterCSV::HeaderSizeMismatch, "ERROR: :user_provided_headers defines #{headerA.size} headers != CSV-file has #{file_header_size} headers"
|
507
|
-
else
|
508
|
-
# we could print out the mapping of file_headerA to headerA here
|
509
|
-
end
|
510
|
-
end
|
511
|
-
else
|
512
|
-
headerA = file_headerA
|
513
|
-
end
|
514
|
-
|
515
|
-
# detect duplicate headers and disambiguate
|
516
|
-
headerA = process_duplicate_headers(headerA, options) if options[:duplicate_header_suffix]
|
517
|
-
header_size = headerA.size # used for splitting lines
|
518
|
-
|
519
|
-
headerA.map!{|x| x.to_sym } unless options[:strings_as_keys] || options[:keep_original_headers]
|
520
|
-
|
521
|
-
unless options[:user_provided_headers] # wouldn't make sense to re-map user provided headers
|
522
|
-
key_mappingH = options[:key_mapping]
|
523
|
-
|
524
|
-
# do some key mapping on the keys in the file header
|
525
|
-
# if you want to completely delete a key, then map it to nil or to ''
|
526
|
-
if !key_mappingH.nil? && key_mappingH.class == Hash && key_mappingH.keys.size > 0
|
527
|
-
# if silence_missing_keys are not set, raise error if missing header
|
528
|
-
missing_keys = key_mappingH.keys - headerA
|
529
|
-
# if the user passes a list of speciffic mapped keys that are optional
|
530
|
-
missing_keys -= options[:silence_missing_keys] if options[:silence_missing_keys].is_a?(Array)
|
531
|
-
|
532
|
-
unless missing_keys.empty? || options[:silence_missing_keys] == true
|
533
|
-
raise SmarterCSV::KeyMappingError, "ERROR: can not map headers: #{missing_keys.join(', ')}"
|
534
|
-
end
|
535
|
-
|
536
|
-
headerA.map!{|x| key_mappingH.has_key?(x) ? (key_mappingH[x].nil? ? nil : key_mappingH[x]) : (options[:remove_unmapped_keys] ? nil : x)}
|
537
|
-
end
|
538
|
-
end
|
539
|
-
|
540
|
-
# header_validations
|
541
|
-
duplicate_headers = []
|
542
|
-
headerA.compact.each do |k|
|
543
|
-
duplicate_headers << k if headerA.select{|x| x == k}.size > 1
|
544
|
-
end
|
545
|
-
|
546
|
-
unless options[:user_provided_headers] || duplicate_headers.empty?
|
547
|
-
raise SmarterCSV::DuplicateHeaders, "ERROR: duplicate headers: #{duplicate_headers.join(',')}"
|
548
|
-
end
|
549
|
-
|
550
|
-
# deprecate required_headers
|
551
|
-
unless options[:required_headers].nil?
|
552
|
-
puts "DEPRECATION WARNING: please use 'required_keys' instead of 'required_headers'"
|
553
|
-
if options[:required_keys].nil?
|
554
|
-
options[:required_keys] = options[:required_headers]
|
555
|
-
options[:required_headers] = nil
|
556
|
-
end
|
557
|
-
end
|
558
|
-
|
559
|
-
if options[:required_keys] && options[:required_keys].is_a?(Array)
|
560
|
-
missing_keys = []
|
561
|
-
options[:required_keys].each do |k|
|
562
|
-
missing_keys << k unless headerA.include?(k)
|
563
|
-
end
|
564
|
-
raise SmarterCSV::MissingKeys, "ERROR: missing attributes: #{missing_keys.join(',')}" unless missing_keys.empty?
|
565
|
-
end
|
566
|
-
|
567
|
-
@headers = headerA
|
568
|
-
[headerA, header_size]
|
569
|
-
end
|
570
|
-
|
571
|
-
def process_duplicate_headers(headers, options)
|
572
|
-
counts = Hash.new(0)
|
573
|
-
result = []
|
574
|
-
headers.each do |key|
|
575
|
-
counts[key] += 1
|
576
|
-
if counts[key] == 1
|
577
|
-
result << key
|
578
|
-
else
|
579
|
-
result << [key, options[:duplicate_header_suffix], counts[key]].join
|
580
|
-
end
|
581
|
-
end
|
582
|
-
result
|
583
|
-
end
|
584
|
-
|
585
|
-
private
|
586
|
-
|
587
|
-
UTF_32_BOM = %w[0 0 fe ff].freeze
|
588
|
-
UTF_32LE_BOM = %w[ff fe 0 0].freeze
|
589
|
-
UTF_8_BOM = %w[ef bb bf].freeze
|
590
|
-
UTF_16_BOM = %w[fe ff].freeze
|
591
|
-
UTF_16LE_BOM = %w[ff fe].freeze
|
592
|
-
|
593
|
-
def remove_bom(str)
|
594
|
-
str_as_hex = str.bytes.map{|x| x.to_s(16)}
|
595
|
-
# if string does not start with one of the bytes, there is no BOM
|
596
|
-
return str unless %w[ef fe ff 0].include?(str_as_hex[0])
|
597
|
-
|
598
|
-
return str.byteslice(4..-1) if [UTF_32_BOM, UTF_32LE_BOM].include?(str_as_hex[0..3])
|
599
|
-
return str.byteslice(3..-1) if str_as_hex[0..2] == UTF_8_BOM
|
600
|
-
return str.byteslice(2..-1) if [UTF_16_BOM, UTF_16LE_BOM].include?(str_as_hex[0..1])
|
601
|
-
|
602
|
-
puts "SmarterCSV found unhandled BOM! #{str.chars[0..7].inspect}"
|
603
|
-
str
|
604
|
-
end
|
605
|
-
|
606
|
-
def validate_options!(options)
|
607
|
-
keys = options.keys
|
608
|
-
errors = []
|
609
|
-
errors << "invalid row_sep" if keys.include?(:row_sep) && !option_valid?(options[:row_sep])
|
610
|
-
errors << "invalid col_sep" if keys.include?(:col_sep) && !option_valid?(options[:col_sep])
|
611
|
-
errors << "invalid quote_char" if keys.include?(:quote_char) && !option_valid?(options[:quote_char])
|
612
|
-
raise SmarterCSV::ValidationError, errors.inspect if errors.any?
|
613
|
-
end
|
614
|
-
|
615
|
-
def option_valid?(str)
|
616
|
-
return true if str.is_a?(Symbol) && str == :auto
|
617
|
-
return true if str.is_a?(String) && !str.empty?
|
618
|
-
|
619
|
-
false
|
620
|
-
end
|
621
|
-
end
|
35
|
+
BLOCK_COMMENT
|
622
36
|
end
|
37
|
+
# :nocov:
|
38
|
+
require "smarter_csv/smarter_csv"
|
39
|
+
|
data/smarter_csv.gemspec
CHANGED
@@ -31,8 +31,9 @@ Gem::Specification.new do |spec|
|
|
31
31
|
spec.test_files = spec.files.grep(%r{^(test|spec|features)/})
|
32
32
|
|
33
33
|
spec.executables = spec.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
|
34
|
-
spec.require_paths = [
|
34
|
+
spec.require_paths = %w[lib ext]
|
35
35
|
spec.extensions = ["ext/smarter_csv/extconf.rb"]
|
36
|
+
spec.files += Dir.glob("ext/smarter_csv/**/*")
|
36
37
|
|
37
38
|
spec.add_development_dependency "awesome_print"
|
38
39
|
spec.add_development_dependency "codecov"
|