Class: IOBlockReader::IOBlockReader
- Inherits:
-
Object
- Object
- IOBlockReader::IOBlockReader
- Defined in:
- lib/ioblockreader/ioblockreader.rb
Overview
Class giving a String-like interface over an IO, reading it by blocks. Very useful to access big files’ content as it was a String containing the whole file’s content.
Instance Method Summary collapse
-
#[](range) ⇒ Object
Get a subset of the data.
-
#index(token, offset = 0, max_size_regexp = 32) ⇒ Object
Perform a search of a token (or a list of tokens) in the IO.
-
#initialize(io, options = {}) ⇒ IOBlockReader
constructor
Constructor.
Constructor Details
#initialize(io, options = {}) ⇒ IOBlockReader
Constructor
- Parameters
-
io (IO): The IO object used to give the String interface
-
options (map<Symbol,Object>): Additional options:
-
:block_size (Fixnum): The block size in bytes used internally. [default = 268435456]
-
:blocks_in_memory (Fixnum): Maximal number of blocks in memory. If it is required to load more blocks than this value for a single operation, this value is ignored. [default = 2]
-
16 17 18 19 20 21 22 23 24 25 26 27 |
# File 'lib/ioblockreader/ioblockreader.rb', line 16 def initialize(io, = {}) # The underlying IO @io = io # Parse options @block_size = [:block_size] || 268435456 @blocks_in_memory = [:blocks_in_memory] || 2 # The blocks @blocks = [] # The last accessed block, used as a cache for quick [] access @cached_block = nil @cached_block_end_offset = nil end |
Instance Method Details
#[](range) ⇒ Object
Get a subset of the data. DO NOT USE NEGATIVE INDEXES.
- Parameters
-
range (Fixnum or Range): Range to extract
- Result
-
String: The resulting data
36 37 38 39 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 |
# File 'lib/ioblockreader/ioblockreader.rb', line 36 def [](range) #puts "[IOBlockReader] - [](#{range.inspect})" if (range.is_a?(Fixnum)) # Use the cache if possible return @cached_block.data[range - @cached_block.offset] if ((@cached_block != nil) and (range >= @cached_block.offset) and (range < @cached_block_end_offset)) #puts "[IOBlockReader] - [](#{range.inspect}) - Cache miss" # Separate this case for performance single_block_index, offset_in_block = range.divmod(@block_size) # First check if all blocks are already loaded if ((block = @blocks[single_block_index]) == nil) read_needed_blocks([single_block_index], single_block_index, single_block_index) block = @blocks[single_block_index] else block.touch end set_cache_block(block) return block.data[offset_in_block] else # Use the cache if possible return @cached_block.data[range.first - @cached_block.offset..range.last - @cached_block.offset] if ((@cached_block != nil) and (range.first >= @cached_block.offset) and (range.last < @cached_block_end_offset)) #puts "[IOBlockReader] - [](#{range.inspect}) - Cache miss" first_block_index, first_offset_in_block = range.first.divmod(@block_size) last_block_index, last_offset_in_block = range.last.divmod(@block_size) # First check if all blocks are already loaded if (first_block_index == last_block_index) if ((block = @blocks[first_block_index]) == nil) read_needed_blocks([first_block_index], first_block_index, last_block_index) block = @blocks[first_block_index] else block.touch end set_cache_block(block) return block.data[first_offset_in_block..last_offset_in_block] else # Get all indexes to be loaded indexes_needing_loading = [] (first_block_index..last_block_index).each do |block_index| if ((block = @blocks[block_index]) == nil) indexes_needing_loading << block_index else block.touch end end read_needed_blocks(indexes_needing_loading, first_block_index, last_block_index) if (!indexes_needing_loading.empty?) # Now read across the blocks result = @blocks[first_block_index].data[first_offset_in_block..-1].dup (first_block_index+1..last_block_index-1).each do |block_index| result.concat(@blocks[block_index].data) end result.concat(@blocks[last_block_index].data[0..last_offset_in_block]) # There are more chances that the last block will be accessed again. Cache this one. set_cache_block(@blocks[last_block_index]) return result end end end |
#index(token, offset = 0, max_size_regexp = 32) ⇒ Object
Perform a search of a token (or a list of tokens) in the IO. Warning: The token(s) to be found have to be smaller than the block size given to the constructor, otherwise they won’t be found (you’ve been warned!). If you really need to search for tokens bigger than block size, extract the data using [] operator first, and then use index on it ; it will however make a complete copy of the data in memory prior to searching tokens.
- Parameters
-
token (String, Regexp or list<Object>): Token to be found. Can be a list of tokens.
-
offset (Fixnum): Offset starting the search [optional = 0]
-
max_size_regexp (Fixnum): Maximal number of characters the match should take in case of a Regexp token. Ignored if token is a String. [optional = 32]
- Result
-
Fixnum: Index of the token (or the first one found from the given token list), or nil if none found.
-
Fixnum: In case token was an Array, return the index of the matching token in the array, or nil if none found.
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 |
# File 'lib/ioblockreader/ioblockreader.rb', line 103 def index(token, offset = 0, max_size_regexp = 32) #puts "[IOBlockReader] - index(#{token.inspect}, #{offset}, #{max_size_regexp})" # Separate the trivial algo for performance reasons current_block_index, offset_in_current_block = offset.divmod(@block_size) if ((current_block = @blocks[current_block_index]) == nil) read_needed_blocks([current_block_index], current_block_index, current_block_index) current_block = @blocks[current_block_index] else current_block.touch end index_in_block = nil index_matching_token = nil if (token_is_array = token.is_a?(Array)) token.each_with_index do |token2, idx| index_token2_in_block = current_block.data.index(token2, offset_in_current_block) if (index_token2_in_block != nil) and ((index_in_block == nil) or (index_token2_in_block < index_in_block)) index_in_block = index_token2_in_block index_matching_token = idx end end else index_in_block = current_block.data.index(token, offset_in_current_block) end if (index_in_block == nil) # We have to search further: across blocks # Compute the size of the token to be searched token_size = 0 if token_is_array token.each do |token2| if (token2.is_a?(String)) token_size = token2.size if (token2.size > token_size) else token_size = max_size_regexp if (max_size_regexp > token_size) end end elsif (token.is_a?(String)) token_size = token.size else token_size = max_size_regexp end # Loop on subsequent blocks to search for token result = nil while ((result == nil) and (!current_block.last_block?)) # Check that next block is loaded if ((next_block = @blocks[current_block_index+1]) == nil) read_needed_blocks([current_block_index+1], current_block_index+1, current_block_index+1) next_block = @blocks[current_block_index+1] else next_block.touch end # Get data across the 2 blocks: enough to search for token_size data only cross_data = current_block.data[1-token_size..-1] + next_block.data[0..token_size-2] if token_is_array token.each_with_index do |token2, idx| index_token2_in_block = cross_data.index(token2) if (index_token2_in_block != nil) and ((index_in_block == nil) or (index_token2_in_block < index_in_block)) index_in_block = index_token2_in_block index_matching_token = idx end end else index_in_block = cross_data.index(token) end if (index_in_block == nil) # Search in the next block if token_is_array token.each_with_index do |token2, idx| index_token2_in_block = next_block.data.index(token2) if (index_token2_in_block != nil) and ((index_in_block == nil) or (index_token2_in_block < index_in_block)) index_in_block = index_token2_in_block index_matching_token = idx end end else index_in_block = next_block.data.index(token) end if (index_in_block == nil) # Loop on the next block current_block_index += 1 current_block = next_block else result = next_block.offset + index_in_block end else result = next_block.offset - token_size + 1 + index_in_block end end if token_is_array return result, index_matching_token else return result end elsif token_is_array return current_block.offset + index_in_block, index_matching_token else return current_block.offset + index_in_block end end |