path: root/doc/context/third/enigma/enigma_manual.tex

%% If you’re looking for instructions concerning how to build this
%% manual, please for now drop me an email: ·phg42.2a at gmail dot com·
%% as you will need a certain mtx-script.
\defineitemgroup [credits]
\setupitemgroup  [credits] [
  % before={\startnarrower[middle]},
  % after=\stopnarrower,
  left=,
  right=,
  margin=,
  option=2,
]

\pushcatcodetable
\setcatcodetable\txtcatcodes
\useurl [chickenize]     [https://github.com/alt/chickenize]
\useurl [codebook]       [http://simonsingh.net/books/the-code-book/]
\useurl [key_procedure]  [http://users.telenet.be/d.rijmenants/en/enigmaproc.htm]
\useurl [luatex-web]     [http://www.luatex.org/]
\useurl [rotor_wirings]  [http://www.ellsbury.com/ultraenigmawirings.htm]
\useurl [wp:day_key]     [http://en.wikipedia.org/wiki/Cryptanalysis_of_the_Enigma#Key_setting]
        []               [\hyphenatedurl{http://en.wikipedia.org/wiki/Cryptanalysis_of_the_Enigma\#Key_setting}]
\useurl [wp:ring]        [http://en.wikipedia.org/wiki/Enigma_rotor_details#The_ring_setting]
        []               [\hyphenatedurl{http://en.wikipedia.org/wiki/Enigma_rotor_details\#The_ring_setting}]
\useurl [initialization] [http://w1tp.com/enigma/mewirg.htm]
\useurl [model-i]        [http://www.cryptomuseum.com/crypto/enigma/i/index.htm]
\useurl [ecollection]    [http://cryptocellar.org/enigma/]
\useurl [heinz_diss]     [http://rzbl04.biblio.etc.tu-bs.de:8080/docportal/receive/DocPortal_document_00001705]
\popcatcodetable

\startdocchapter[title=Introduction]
  This module implements an \emph{Enigma} cipher that is equivalent to
  the most widely distributed model: the Enigma~I (subtypes {\sc m1}
  {\sc m2} and {\sc m3}).%
  \footnote{%
    See \from[model-i] for a showcase.
  }
  Machines of this type had three rotors, a non-moving reflector and,
  as a novelty at the time of their introduction, a plugboard.
  The simulation accepts arbitrary configurations of these components,
  as well as the starting position of the rotors, and then processes
  text accordingly.
  Depending on the input, this yields the plaintext or ciphertext, as
  encryption and decryption are the same.

  The code is provided as a module (interface for \CONTEXT) as well as
  a package (\PLAIN, \LATEX).
  It is subject to the {\sc bsd} license, see below,
  \at{page}[license], for details.
\stopdocchapter

\startdocchapter[title=Usage]

\startdocsection[title=Loading the Module/Package]
  The intention is for the \modulename{Enigma} codebase to integrate
  with the three most popular (as of 2012) \TEX\ formats:
    \CONTEXT,
    \PLAIN, and
    \LATEX.
  If the user interface does not fully conform with the common practice
  of the latter two, please be lenient toward the author whose
  intuitions are for the most part informed by \CONTEXT.
  For this reason, a couple words concerning the interfaces will be
  necessary.
  The examples in this manual will be cycling through all three formats,
  but once you get the general idea of how it works, you will have no
  problem translating between coding styles.
  Those familiar with \CONTEXT\ might, therefore, skip the following
  paragraphs and continue directly with the next section on
  \at{page}[sec:opts].%
  
  The package is loaded as usual. For \PLAIN, issue a
  \type{\input{enigma}}.
  \LATEX-users need to place \type{\usepackage{enigma}} somewhere inside
  the preamble.
  (There are no package options.)
  From this point on, instructions for both formats are the same.

  The interface provides two basic macros from which all functionality
  will be derived:
    \texmacro{defineenigma} and \texmacro{setupenigma}.
  Both are a kind of \emph{meta-macros}, meaning that they generate
  other macros which may then be employed to access the functionality of
  the package.
  As such they naturally belong inside the preamble (if you chose to use
  \modulename{Enigma} with \LATEX, that is).
  It follows that the correct and only meaningful order is to
  \texmacro{defineenigma} an enigma machine first and then
  \texmacro{setupenigma} it.
  The former takes a single, the latter a double argument.
  Thus, \type{\defineenigma{encrypt}} creates a new environment
  consisting of the macros \texmacro{startencrypt} and
  \texmacro{stopencrypt}.%
  \footnote{%
    \CONTEXT-users will have noticed that there is no simple macro
    \type{\encrypt{foo}}. The reason for this is that the callback
    which the module relies on operates on node-level.
    This means that for the Enigma encryption to have an effect it will
    have to process entire paragraphs.
    As encrypted passages are supposed to stand on their own, this
    small limitation should not have a severe impact on functionality.
    If you should, however, need a macro that works for smaller portions
    of text, please send a feature request to the maintainer
    (\ichdparameter{email}).
  }
  However, these are not usable right away, as we still have to set the
  initial state of the machine.
  This is achieved by the second meta-macro,
  \type{\setupenigma{encrypt}{<args>}}, where \type{<args>} is a
  placeholder for a list of \emph{assignments}, i.\,e. pairs of
  \emph{key=value} statements by means of which all further parameters
  are specified.
  The possible parameters are listed in the next section,
  examples of their effects will be given further below in the section
  on functionality (see \at{page}[sec:fun]).%
  \footnote{%
    If you grasp the concept of paired \type{\define<foo>} \endash\space
    \type{\setup<foo>} macros, then congratulations are in order:
    you qualify for migration from your current macro package to
    \CONTEXT.
  }
\stopdocsection
\startdocsection[title=Options Explained,reference=sec:opts]

At the moment, the \texmacro{setupenigma} macro in any format  accepts
the following parameters.
% \startpostponing[+2]%% messes up floats -- why?
  \placefigure[right]{Usage example for the \PLAIN\ format.}{
    % \framed[align=right,frameoffset=1em]{%
    \framed[align=right,frameoffset=1em,frame=off]{%
      \startlatextyping[before=,after=,]
      \input {enigma}
      %% Definition ··········· %%
      \defineenigma {encryption}
      %% Setup ················ %%
      \setupenigma  {encryption} {
        other_chars = no,
        day_key = I II III
                  01 01 01,
        rotor_setting = aaa,
        verbose = 1,
      }
      %% Usage ················ %%
      \startencryption
        aaaaa aaaaa aaaaa
        aaaaa aaaaa aaaaa 
      \stopencryption
      \startencryption
        Nobody in Poland is going
        to be able to read this,
        har har!
      \stopencryption
      \bye
      \stoplatextyping
    }
  }
% \stoppostponing
\startitemize
  \setuptolerance[tolerant]%% it’s crowded in here due to the float
  \let\olditem\item
  \let\item\undefined
  \def\item#1#2{% 1 name, 2 type
    \olditem
    \identifier{#1}%
    \reference[option:#1]{}%
    \space
    <\emph{#2}>%
    \hskip1em%
  }
  \item{other_chars}{boolean} How to handle non-encodable characters, i.\,e.
    glyphs outside the bare set of Latin letters; see below on
    \at{page}[explained:other_chars].
  \item{day_key}{string} Encryption key, i.\,e. a description of the initial
    setup of an Enigma machine: the reflector used, the choice and
    positions of the three rotors, the ring settings, and the plugboard
    wiring.
  \item{rotor_setting}{string} The initial rotor advancement.
  \item{verbose}{integer} Controls overall verbosity level (\emph{global}!).
\stopitemize

%% day key
To state the obvious, the value of \identifier{day_key} serves as the
\emph{day key} for encryption. An Enigma day key ordinarily consists of
(1) a list of the the rotor configuration,
(2) the ring settings, and
(3) the plugboard connections.\footnote{%
  For a description of the initialization process see
  \from[initialization].
}
Together these have the denotation \emph{day key}, because they are
meant to be supplied in special code books by central authority, one for
each day.\footnote{%
  Read about the historical directives for daily key renewal at
  \from[key_procedure].
  there are some {\sc pdf}s with images of \emph{Kenngruppenbücher} at
  \from[ecollection], section \emph{Enigma Messages and Keys}.
  Also, while you’re at it, don’t miss the explanation on Wikipedia:
  \from[wp:day_key].
}
In the \modulename{Enigma} setup, the day key starts
with a triple of Roman numerals ({\sc i} to {\sc v}) describing which
of the five rotors is located in which of the three slots.
(e.\,g. \type{I VI II}).\footnote{%
  For the individual wirings of the five rotors see
  \from[rotor_wirings], as well as the implementation below at
  \at{page}[listing:rotor_wiring].
}
Its next part is the ring setting, a triple of two-digit integers that
are the amount by which the internal wiring of each rotor has been
shifted (\type{03 23 11}). As the Enigma encrypts only the letters of
the Latin alphabet, sane values range from one (first position: no
shift) to twenty six.\footnote{%
  Consult \from[wp:ring] for an introduction into the ring mechanics.
}
The third section specifies which pairs of letters are substituted by
each other by means of plugboard connections (\type{NI CE GO LD ...}).
There can be zero to thirteen of these redirections, thus the presence
of this section is entirely optional.
Also part of the \identifier{day_key}, but not mentioned yet, is the
choice of the \emph{reflector}.
It may be specified as one of the three letters \type{A}, \type{B} and
\type{C} as the first item. If no reflector is requested explicitly, the
machine defaults to \type{B}, which is actually the only one of the
three models that had been in widespread use
(see below on \at{page}[listing:reflector] for the wirings).

Initialization is not complete without a \identifier{rotor_setting}.
This is a triple of letters, each representing the initial state of
one rotor (e.\,g. \type{fkd}). Historically this was not part of the day
key but supposed to be chosen at random by the operating signal officer.

%%% other_chars
% The option \identifier{other_chars} determines how the machine -- in our
% Unicode-aware times -- should behave when it encounters a non-Latin
% letter. For the time being it can either remove them from the input
% (value \emph{no}) or pass them through unmodified (\emph{yes}). Note
% that this is independent of the \emph{preprocessing} of characters like
% punctuation and umlauts, which is done automatically.
% (See below \at{page}[listing:preproc] for the internal substitution
% list.)

Most documents don’t naturally adhere to the machine-imposed restriction
to the 26 letters of the Latin alphabet. The original encipherment
directives comprised substitution tables to compensate for a set of
intrinsic peculiarities of the German language, like umlauts and common
digraphs.  The \modulename{Enigma} simulation module strives to apply
these automatically but there is no guarantee of completeness.

However,\reference[explained:other_chars]{} the Enigma lacks means of
handling languages other than German.  When the substitution lookup
fails, there are two ways of proceeding: either to ignore the current
character or to pass it on to the output as if nothing happened. The
default behaviour is to drop alien letters and move on. If the user
intends to keep these foreign characters instead, E can achieve this by
setting the \identifier{other_chars} key in the \modulename{Enigma}
setup to the value \emph{true}. An example of how the result of both
methods may look, other things being equal, is given in below listing
(example for \CONTEXT; see the file \type{enigma-example-context.tex} in
the \type{doc/} subtree of your installation path).

\startcontexttyping
\usemodule [enigma]
\defineenigma [secretmessage]
\setupenigma  [secretmessage] [
  other_chars = yes,
  day_key = B V III II 12 03 01 GI JV KZ WM PU QY AD CN ET FL,
  rotor_setting = ben,
]

\defineenigma [othermessage] [secretmessage]
\setupenigma  [othermessage] [other_chars=wrong]

\starttext

\startsecretmessage
  føo bąr baž
\stopsecretmessage
\startothermessage
  føo bąr baž
\stopothermessage

\stoptext
\stopcontexttyping

Both methods have their disadvantages: if the user chooses to have the
unknown characters removed it might distort the decrypted text to
becoming illegible. Far more serious, however, are the consequences of
keeping them. As artefacts in the ciphertext they would convey
information about the structure of the plain text.

\stopdocsection

\startdocsection[title=Basic Functionality,reference=sec:fun]

Encrypt the text of your document using the script interface. For
a start try out the settings as given in below listing.

\starttyping
mtxrun --script mtx-t-enigma                    \
       --setup="day_key = B I II III 01 01 01,  \
                rotor_setting = xyz,            \
                verbose=0"                      \
       --text="Gentlemen don’t read each other’s mail, Mr. Turing\!"
\stoptyping

This will result in the thoroughly scrambled string
\typ{omribshpwfrfjovkntgqgi abbkhjpxmhdztapkatwrvf}.
Then, use the same settings you encrypted the text with in your
document.

\startcontexttyping
\usemodule[enigma]
\defineenigma [secretmessage]
\setupenigma  [secretmessage] [
  day_key = B I II III 01 01 01,
  rotor_setting = xyz,
  verbose=3,
]

\starttext

\startsecretmessage
  omribshpwfrfjovkntgqgiabbkhjpxmhdztapkatwrvf
\stopsecretmessage

\stoptext
\stopcontexttyping

If you compile this document with \CONTEXT, the plain text will
reappear.
Notice that punctuation is substituted with the letter “x” before
encryption and that spaces are omitted.

\useURL[kgbuch] [http://de.wikipedia.org/wiki/Kenngruppenbuch]
       []       [code book]
Now it’s certainly not wise to carry around the key to encrypted
documents as plain text within those documents. The keys will have to be
distributed via an independent channel, e.\,g. a \from[kgbuch].
Keys in general don’t have to be supplied inside the document. If there
is none specified, the module will interrupt the \TEX\ run and
\emph{ask} for user input. Suppose Alice wanted to send an encrypted
file to Bob and already generated the cipher text as follows:

\starttyping
mtxrun --script mtx-t-enigma                                    \
        --setup="day_key =B I IV V 22 07 10 AZ DG IE YJ QM CW,  \
                 rotor_setting = bar,                           \
                 verbose=0"                                     \
        --text="I have nothing to hide. From the NSA, that is."
\stoptyping

Alice would then include the result of this line in her \LATEX\ document
as follows:

\startlatextyping
\documentclass{scrartcl}
\usepackage{enigma}
\defineenigma{decryption}
%% Encryption key not given in the setup.
\setupenigma{decryption}{
  rotor_setting = bar,
  verbose       = 3,
}
\begin{document}

\startdecryption
usbatbwcaajhzgeyzkqskupzbmdhbdepccgeh
\stopdecryption

\end{document}
\stoplatextyping

She subsequently mails this file to Bob and conveys the key through a
secure channel. They only thing that will be left for Bob to do now, is
to enter the key at the prompt when compiling the document with
\LUALATEX.
\stopdocsection

\startdocsection[title=Uses and Abuses]

In \LUATEX, \identifier{callback}s may stack. This allows filtering the
input through many enigma machines successively. For instance, in the
following listing, two instances of the same machine are generated and
applied.

\startcontexttyping
\usemodule[enigma]              %% load the module
\defineenigma [secretmessage]   %% generate and
\setupenigma  [secretmessage] [ %% configure a machine
  day_key = B IV V II 01 01 01 AD CN ET FL GI JV KZ PU QY WX,
  rotor_setting = foo,
  verbose=3,
]

%% now, copy the first machine’s settings
\defineenigma [othermessage] [secretmessage]

%% here we go!
\starttext

\startothermessage  %% enable machine 1
\startsecretmessage %% enable machine 2 while no 1 is active
Encryption equals decryption.
\stopothermessage
\stopsecretmessage

\stoptext \endinput
\stopcontexttyping

\stopdocsection

\stopdocchapter

\startdocchapter[title=Metadata]
\startdocsection[title=License,reference=license]

© 2012 \emph{Philipp Gesang}. All rights reserved.

Redistribution and use in source and binary forms, with or
without modification, are permitted provided that the following
conditions are met:

\startitemize[n]
  \item Redistributions of source code must retain the above
    copyright notice, this list of conditions and the following
    disclaimer.
  \item Redistributions in binary form must reproduce the
    above copyright notice, this list of conditions and the
    following disclaimer in the documentation and/or other
    materials provided with the distribution.
\stopitemize

\begingroup
\setuptolerance [horizontal,strict]
\startalignment [right,nothyphenated]
\noindentation\startsmallcaps
  this software is provided by the copyright
  holder “as is” and any express or implied warranties,
  including, but not limited to, the implied warranties of
  merchantability and fitness for a particular purpose are
  disclaimed. in no event shall the copyright holder or
  contributors be liable for any direct, indirect, incidental,
  special, exemplary, or consequential damages (including, but
  not limited to, procurement of substitute goods or services;
  loss of use, data, or profits; or business interruption)
  however caused and on any theory of liability, whether in
  contract, strict liability, or tort (including negligence or
  otherwise) arising in any way out of the use of this software,
  even if advised of the possibility of such damage.
\stopsmallcaps\endgraf
\stopalignment
\endgroup
\stopdocsection

\startdocsection[title=Acknowledgements]
\startcredits
  \item The idea to implement the \emph{Enigma} cipher for \TEX\ came up
        while I was reading \emph{The Code Book} by Simon Singh.
        This work contains an excellent portrayal of the history of
        German military cryptography and Allied cryptanalysis before and
        during the Second World War.\par
        \from[codebook]
  \item A more detailed account from a historical-cryptological
        perspective is provided by Ulrich Heinz in his Dissertation (in
        German), which is freely available.
        Includes an interesting albeit speculative note on the
        effectiveness of the Soviet code breaking efforts (chapter
        seven).\par
        \from[heinz_diss]
  \item Also, the \modulename{Enigma} module drew lots of inspiration
        from Arno Trautmann’s \modulename{Chickenize} package, which
        remains the unsurpassed hands-on introduction to callback
        trickery.\par
        \from[chickenize]
  \item Finally, without \LUATEX, encryption on node-level would not
        have been possible.\par
        \from[luatex-web]
\stopcredits
\stopdocsection
\stopdocchapter