Module: Tapsoob::Utils

Extended by:
Utils
Included in:
Utils
Defined in:
lib/tapsoob/utils.rb

Instance Method Summary collapse

Instance Method Details

#base64decode(data) ⇒ Object 



36
37
38
 
# File 'lib/tapsoob/utils.rb', line 36

def base64decode(data)
  data.unpack("m").first
end

#base64encode(data) ⇒ Object 



32
33
34
 
# File 'lib/tapsoob/utils.rb', line 32

def base64encode(data)
  [data].pack("m")
end

#bin(cmd) ⇒ Object 



19
20
21
22
 
# File 'lib/tapsoob/utils.rb', line 19

def bin(cmd)
  cmd = "#{cmd}.cmd" if windows?
  cmd
end

#calculate_chunksize(old_chunksize) ⇒ Object 



130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/tapsoob/utils.rb', line 130

def calculate_chunksize(old_chunksize)
  c = Tapsoob::Chunksize.new(old_chunksize)

  begin
    c.start_time = Time.now
    c.time_in_db = yield c
  rescue Errno::EPIPE
    c.retries += 1
    raise if c.retries > 2

    # we got disconnected, the chunksize could be too large
    # reset the chunksize based on the number of retries
    c.reset_chunksize
    retry
  end

  c.end_time = Time.now
  c.calc_new_chunksize
end

#checksum(data) ⇒ Object 



24
25
26
 
# File 'lib/tapsoob/utils.rb', line 24

def checksum(data)
  Zlib.crc32(data)
end

#encode_blobs(row, columns) ⇒ Object 



109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
 
# File 'lib/tapsoob/utils.rb', line 109

def encode_blobs(row, columns)
  # Encode columns known to be blobs
  columns.each do |c|
    if row[c].is_a?(Sequel::SQL::Blob)
      row[c] = base64encode(row[c]) unless row[c].nil?
    elsif !row[c].nil? && row[c].encoding == Encoding::ASCII_8BIT
      # Handle binary data that might not be wrapped in Sequel::SQL::Blob
      row[c] = base64encode(row[c])
    end
  end unless columns.size == 0

  # Also check all values for Sequel::SQL::Blob objects that might not be in the columns list
  row.each do |key, value|
    if value.is_a?(Sequel::SQL::Blob) && !columns.include?(key)
      row[key] = base64encode(value)
    end
  end

  row
end

#export_indexes(dump_path, table, index_data) ⇒ Object 



156
157
158
159
160
161
162
163
164
165
 
# File 'lib/tapsoob/utils.rb', line 156

def export_indexes(dump_path, table, index_data)
  # Use append-only writes to avoid O(n²) complexity
  # Each index_data is a migration string that should be stored as NDJSON
  index_file = File.join(dump_path, "indexes", "#{table}.json")

  File.open(index_file, 'a') do |file|
    # Store as NDJSON - each line is a JSON-encoded migration string
    file.write(JSON.generate(index_data) + "\n")
  end
end

#export_rows(dump_path, table, row_data) ⇒ Object 



167
168
169
170
171
172
173
174
175
 
# File 'lib/tapsoob/utils.rb', line 167

def export_rows(dump_path, table, row_data)
  # Use append-only writes to avoid O(n²) complexity
  # Store metadata separately and append data chunks as NDJSON
  data_file = File.join(dump_path, "data", "#{table}.json")

  File.open(data_file, 'a') do |file|
    file.write(JSON.generate(row_data) + "\n")
  end
end

#export_schema(dump_path, table, schema_data) ⇒ Object 



150
151
152
153
154
 
# File 'lib/tapsoob/utils.rb', line 150

def export_schema(dump_path, table, schema_data)
  File.open(File.join(dump_path, "schemas", "#{table}.rb"), 'w') do |file|
    file.write(schema_data)
  end
end

#format_data(db, data, opts = {}) ⇒ Object 



40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
 
# File 'lib/tapsoob/utils.rb', line 40

def format_data(db, data, opts = {})
  return {} if data.size == 0
  string_columns = opts[:string_columns] || []
  schema = opts[:schema] || []
  table = opts[:table]

  max_lengths = schema.inject({}) do |hash, (column, meta)|
    if meta[:db_type] =~ /^varchar\((\d+)\)/
      hash.update(column => $1.to_i)
    end
    hash
  end

  header = data[0].keys
  only_data = data.collect do |row|
    row = encode_blobs(row, string_columns)
    row.each do |column, data|
      if data.to_s.length > (max_lengths[column] || data.to_s.length)
        raise Tapsoob::InvalidData.new(<<-ERROR)
Detected data that exceeds the length limitation of its column. This is
generally due to the fact that SQLite does not enforce length restrictions.

Table : #{table}
Column : #{column}
Type : #{schema.detect{|s| s.first == column}.last[:db_type]}
Data : #{data}
        ERROR
      end

      # Type conversion
      row[column] = data.strftime('%Y-%m-%d %H:%M:%S') if data.is_a?(Time)
    end
    header.collect { |h| row[h] }
  end

  res = { table_name: table, header: header, data: only_data }

  # Add types if schema isn't empty
  db.extension :schema_dumper # Add schema dumper extension in case it hasn't been added until now
  res[:types] = schema.map do |c|
    case db.column_schema_to_ruby_type(c.last)[:type].to_s
    when "BigDecimal"
      "float"
    when "Bignum"
      "integer"
    when "File"
      "blob"
    when "TrueClass"
      "boolean"
    else
      db.column_schema_to_ruby_type(c.last)[:type].to_s.downcase
    end
  end unless schema.empty?

  res
end

#incorrect_blobs(db, table) ⇒ Object

mysql text and blobs fields are handled the same way internally this is not true for other databases so we must check if the field is actually text and manually convert it back to a string



100
101
102
103
104
105
106
107
 
# File 'lib/tapsoob/utils.rb', line 100

def incorrect_blobs(db, table)
  columns = []
  db.schema(table).each do |data|
    column, cdata = data
    columns << column if cdata[:type] == :blob
  end
  columns
end

#load_indexes(database_url, index) ⇒ Object 



189
190
191
 
# File 'lib/tapsoob/utils.rb', line 189

def load_indexes(database_url, index)
  Tapsoob::Schema.load_indexes(database_url, index)
end

#load_schema(dump_path, database_url_or_db, table) ⇒ Object 



177
178
179
180
181
182
183
184
185
186
187
 
# File 'lib/tapsoob/utils.rb', line 177

def load_schema(dump_path, database_url_or_db, table)
  schema_file = File.join(dump_path, "schemas", "#{table}.rb")
  schema_content = File.read(schema_file)

  # If we have a connection object, use it directly for better performance
  if database_url_or_db.is_a?(Sequel::Database)
    Tapsoob::Schema.load(database_url_or_db, schema_content)
  else
    schema_bin(:load, database_url_or_db, schema_file)
  end
end

#order_by(db, table) ⇒ Object 



210
211
212
213
214
215
216
217
218
 
# File 'lib/tapsoob/utils.rb', line 210

def order_by(db, table)
  pkey = primary_key(db, table)
  if pkey
    pkey.kind_of?(Array) ? pkey : [pkey.to_sym]
  else
    table = table.to_sym unless table.kind_of?(Sequel::SQL::Identifier)
    db[table].columns
  end
end

#primary_key(db, table) ⇒ Object 



200
201
202
 
# File 'lib/tapsoob/utils.rb', line 200

def primary_key(db, table)
  db.schema(table).select { |c| c[1][:primary_key] }.map { |c| c[0] }
end

#schema_bin(command, *args) ⇒ Object 



193
194
195
196
197
198
 
# File 'lib/tapsoob/utils.rb', line 193

def schema_bin(command, *args)
  require 'tapsoob/cli'
  subcommand = "schema"
  script = Tapsoob::CLI::Schema.new
  script.invoke(command, args.map { |a| "#{a}" })
end

#single_integer_primary_key(db, table) ⇒ Object 



204
205
206
207
208
 
# File 'lib/tapsoob/utils.rb', line 204

def single_integer_primary_key(db, table)
  table = table.to_sym unless table.kind_of?(Sequel::SQL::Identifier)
  keys = db.schema(table).select { |c| c[1][:primary_key] and c[1][:type] == :integer }
  not keys.nil? and keys.size == 1
end

#valid_data?(data, crc32) ⇒ Boolean

Returns:

  • (Boolean) 


28
29
30
 
# File 'lib/tapsoob/utils.rb', line 28

def valid_data?(data, crc32)
  Zlib.crc32(data) == crc32.to_i
end

#windows? ⇒ Boolean

Returns:

  • (Boolean) 


13
14
15
16
17
 
# File 'lib/tapsoob/utils.rb', line 13

def windows?
  return @windows if defined?(@windows)
  require 'rbconfig'
  @windows = !!(::RbConfig::CONFIG['host_os'] =~ /mswin|mingw/)
end