Skip to content

Latest commit

 

History

History
110 lines (76 loc) · 3.41 KB

README.md

File metadata and controls

110 lines (76 loc) · 3.41 KB
Raw

buffer

buffer creates a read-only binary array from a string, and allows reads of data types from it. All functions throw errors in case something goes wrong.

Building

It's just one file, either add it to your project or build a loadable Lua module with:

$ gcc -std=c99 -O2 -Werror -Wall -Wpedantic -shared -fPIC -o buffer.so buffer.c

Usage

buffer.new(string)

buffer.new creates the buffer object, with the read position set to 0:

buffer.new(
    value,    -- String whose contents will be available for reading via the buffer.
)

Example:

local buffer = require 'buffer'

local buf = assert(buffer.new('abcd'))

Returns the created buffer, or nil plus an error message in case of errors.

buffer.read(type)

Reads a data type from the buffer, starting with the byte at the current read position. Allowed data types are:

  • sb: Signed byte
  • ub: Unsigned byte
  • swl: Signed word (16 bits) in little-endian
  • swb: Signed word in big-endian
  • uwl: Unsigned word in little-endian
  • uwb: Unsigned word in big-endian
  • sdl: Signed double word (32 bits) in little-endian
  • sdb: Signed double word in big-endian
  • udl: Unsigned double word in little-endian
  • udb: Unsigned double word in big-endian
  • sql: Signed quad word (64 bits) in little-endian
  • sqb: Signed quad word in big-endian
  • uql: Unsigned quad word in little-endian
  • uqb: Unsigned quad word in big-endian
  • fl: Float in little-endian
  • fb: Float in big-endian
  • dl: Double in little-endian
  • db: Double in big-endian
  • sw: Signed word in the current machine's endian mode
  • uw: Unsigned word in the current machine's endian mode
  • sd: Signed double word in the current machine's endian mode
  • ud: Unsigned double word in the current machine's endian mode
  • sq: Signed quad word in the current machine's endian mode
  • uq: Unsigned quad word in the current machine's endian mode
  • f: Float in the current machine's endian mode
  • d: Double in the current machine's endian mode
  • l: Reads an entire line terminated with the ASCII character \n'
  • A number: Returns a string with that many characters

The read position is incremented according to the type read.

Example:

local buffer = require 'buffer'

local buf = assert(buffer.new('\x12\x34\x56\x78'))
print(string.format('%x', assert(buf:read('udb')))) -- prints 12345678

Returns the read result, or nil plus an error message in case of errors.

buffer.seek(offset, whence)

Changes the current read position. The way offset is applied depends on the whence parameter:

  • "set" (default if omitted): The new read position will be the value of offset
  • "cur": The new read position will be offset added to the current read position
  • "end": The new read position will be the size of the buffer subtracted from offset

Returns the new absolute read position, or nil plus an error message in case of errors.

buffer.tell

Returns the current read position.

buffer.size

Returns the buffer size.

buffer.sub(begin, size)

Returns a piece of a buffer as a new buffer, starting at the original buffer begin position, with size bytes, or nil plus an error message in case of errors.

Changelog

  • 1.0.0
    • First public release

License

MIT, enjoy.