class PDF::Reader::Buffer

An internal PDF::Reader class that mediates access to the underlying PDF File or IO Stream
###############################################################################

def eof?

is empty
returns true if the underlying IO object is at end and the internal buffer
############################################################################### 
def eof?
  if @buffer
    @buffer.empty? && @io.eof?
  else
    @io.eof?
  end
end

def find_first_xref_offset

table in the underlying IO stream.
objects in the file. This method attempts to locate the byte offset of the xref
The Xref table in a PDF file acts as an aid for finding the location of various
############################################################################### 
def find_first_xref_offset
  @io.seek(-1024, IO::SEEK_END) rescue seek(0)
  data = @io.read(1024)
  # the PDF 1.7 spec (section #3.4) says that EOL markers can be either \r, \n, or both.
  # To ensure we find the xref offset correctly, change all possible options to a 
  # standard format
  data = data.gsub("\r\n","\n").gsub("\n\r","\n").gsub("\r","\n")
  lines = data.split(/\n/).reverse
  eof_index = nil
  lines.each_with_index do |line, index|
    if line =~ /^%%EOF\r?$/
      eof_index = index
      break
    end
  end
  raise MalformedPDFError, "PDF does not contain EOF marker" if eof_index.nil?
  raise MalformedPDFError, "PDF EOF marker does not follow offset" if eof_index >= lines.size-1
  lines[eof_index+1].to_i
end

def head (chars, with_strip=true)

############################################################################### 
def head (chars, with_strip=true)
  val = @buffer[0, chars]
  @buffer = @buffer[chars .. -1] || ""
  @buffer.lstrip! if with_strip
  val
end

def initialize (io)

Creates a new buffer around the specified IO object
############################################################################### 
def initialize (io)
  @io = io
  @buffer = nil
end

def pos

############################################################################### 
def pos
  @io.pos
end

def raw

return the internal buffer used by this class when reading from the IO stream.
############################################################################### 
def raw
  @buffer
end

def read (length)

length should be a positive integer.

reads the requested number of bytes from the underlying IO stream.
############################################################################### 
def read (length)
  out = ""
  if @buffer and !@buffer.empty?
    out << head(length)
    length -= out.length
  end
  out << @io.read(length) if length > 0
  out
end

def read_until(bytes)

bytes - the bytes to search for.

Reads from the buffer until the specified token is found, or the end of the buffer
############################################################################### 
def read_until(bytes)
  out = ""
  size = bytes.size
   
  loop do
    out << @io.read(1)
    if out[-1 * size,size].eql?(bytes)
      out = out[0, out.size - size]
      seek(pos - size)
      break
    end
  end
  out
end

def ready_token (with_strip=true, skip_blanks=true)

This prepares the buffer for use by reading the next line of tokens into memory.
PDF files are processed by tokenising the content into a series of objects and commands.
############################################################################### 
def ready_token (with_strip=true, skip_blanks=true)
  while @buffer.nil? or @buffer.empty?
    @buffer = @io.readline
    @buffer.sub!(/%.*$/, '')
    @buffer.chomp!
    break unless skip_blanks
  end
  @buffer.lstrip! if with_strip
end

def seek (offset)

Seek to the requested byte in the IO stream.
############################################################################### 
def seek (offset)
  @io.seek(offset, IO::SEEK_SET)
  @buffer = nil
  self
end

def token

return the next token from the underlying IO stream
############################################################################### 
def token
  ready_token
  
  i = @buffer.index(/[\[\]()<>{}\s\/]/) || @buffer.size
  token_chars = 
    if i == 0 and @buffer[i,2] == "<<"    then 2
    elsif i == 0 and @buffer[i,2] == ">>" then 2
    elsif i == 0                          then 1
    else                                    i
    end
  strip_space = !(i == 0 and @buffer[0,1] == '(')
  head(token_chars, strip_space)
end