Using the shell

The SDSI shell, `sdsi2sh', is a very useful SDSI tool. The shell allows you to create, sign, and issue certificates. If you're at home typing commands or writing scripts to do things, sdsi2sh is for you. The command environment is rather flexible: named variables are used with all commands, and command results are returned in new named variables.

You start the shell by typing `sdsi2sh'. The shell reads your public key out of your principal file `~/.sdsi/your-principal.sdsi2', and then tries to find the corresponding private key - in your PGP private keyring, if you made a PGP key, or in your SSH identity file, if you made an SSH key. sdsi2sh prompts you for the passphrase so it can unlock your private key, so you can use it to sign certificates.

Once it has your private key, sdsi2sh presents a simple prompt:

sdsi2sh>

From this prompt, you issue commands. You can see the list of commands by typing `help'. The commands currently available are:

sdsi2sh> help

command summary:

help or ?        - displays help
define           - defines a new S-expression variable
define-string    - defines a new bytestring variable
print            - prints a variable
width            - sets the display width
dollar           - sets the next $ temporary number
hash             - hashes an object
sign             - signs an object
cache            - adds an object to the cache
search           - compose a cache search

You can get a little more detailed help on an individual command by doing `help command'. For example, to find out more about the `sign' command, used for signing objects:

sdsi2sh> help sign
help for sign:

usage: sign <object> [<include-key-as-hash-flag>] - returns your 
  signature on <object>.  if <include-key-as-hash-flag> is given,
  includes the verification key as a hash instead of outright.

In the shell, all commands take their arguments - with the exception of their integer arguments - through named variables. SDSI objects are never given literally on the command line. Variables are introduced with the define command, and can be displayed with the print command. Here are examples of how variables are created and displayed:

sdsi2sh> define foo
(foo a b c)
sdsi2sh> print foo
(foo a b c)
sdsi2sh> define bar
(bar x y z)
sdsi2sh> print bar
(bar x y z)
sdsi2sh> define foo
(zoop zoop)
sdsi2sh> print foo
(zoop zoop)
sdsi2sh> 

See how redefining a variable that is already defined gives no error, and the new value simply replaces the old value of the variable.

There is one predefined variable in sdsi2sh: the speaker variable is initially defined to be your SDSI public key. You can redefine speaker if you want, however.

Once variables have been defined, you can give them to commands. For example, to sign an object, you put it in a variable and then use that variable with the sign command:

sdsi2sh> define foo
(zoop zoop)
sdsi2sh> sign foo
(signature
 (hash md5 |nsjklw4yb2cqKAh2e68Aug==|)
 (public-key
  rsa-pkcs1-md5
  (e #11#)
  (n
   |AM30TzJanfRxoZUu0zgniBGv/C82xzvna1BjB0BFbkSELk9Ltn7biJ/q
   6FHN5YCO78EiVM3/MEZegNS8ZXDy7HU=|))
 |AKVs9CwfqUsj8zkayoglT1N/cAm/i0zlWwoG2QyCo0p4dwx7PqYQdF0yG5
 /59xEylMQBuPM2ZIae3oOvFfGhWAA=|)
that returned value has been named $0.

Note how the result of the sign is both displayed and assigned to a new named variable, $0. A new "dollar temporary" is defined to hold the result of any command that returns some value, to allow the use of that value in subsequent commands.

The next command's result will go into $1, and so forth. You can set what the next dollar temporary will be by using the dollar command: `dollar 0' starts the dollar temporaries over at zero again.

Command summary

TBD

Go to the first, previous, next, last section, table of contents.