hexdump's format string is hard to use IMO, and the shipped version on Debian 6 and OS X doesn't support 64-bit integer. So I created this little tool in Go to replace hexdump for my own need.

Limitations:

• Does not support floating point number and string
• Does not have special print specifier to control where to print offset, etc.
• Others features that I do not use

In bprint, the binary data specification and how the they are going to be printed is specified separately using the -e and -p option.

• -e specifies binary field. Using the same syntax as Ruby's Array.unpack.
  • c, s, l, q stands for signed 8,16,32,64-bit integer
  • C, S, L, Q stands for unsigned 8,16,32,64-bit integer
  • A number following the type specifier repeats that specifier. For example, c4 is equivalent to cccc
• -p specifies how to print the binary data. It uses C printf style field specifier
  • %c, %d, %x, %o are supported, size and signess information is implicit from the binary field information
  • if not specified, defaults to %02x for each binary field
  • field can be followed by an optional seperator and count to repeat. For example, %2d-3# is eqivalent to %2d-%2d-%2d
• -f read binary and print format from file. 1st line for binary format, 2nd line for print format (optional). Command line option overrides spec in file
• -o print offset at the left most column
• -c print how many record has been read (right after offset column)
• --version print version information

Suppose each record in the binary file contains 2 byte and 1 64-bit integer, here's invocation to print it's content:

bprint -e 'c2q' -p '%x %d2#' bindata