10 releases (2 stable)
1.1.0 | Oct 3, 2023 |
---|---|
1.0.0 | Nov 19, 2022 |
0.9.1 | Nov 12, 2022 |
0.8.0 | Nov 8, 2022 |
0.3.0 | Apr 1, 2021 |
#663 in Text processing
23 downloads per month
245KB
4.5K
SLoC
Tengwar
Automated conversion of Latin text into Tengwar codepoints in the Unicode Private Use Area. Primarily targets the Classical Quenya mode, but also supports the Sindarin modes of Beleriand and Gondor. This allows for typesetting text in the various Elvish languages of J.R.R. Tolkien.
The codepoints used are the ones defined by the Free Tengwar Font Project, based on the mapping in the ConScript Unicode Registry, and will only correctly render if a font is installed that includes them. For an image example that does not require any fonts to be installed, see the bottom of the page.
This project provides both an executable program and a software library. The rest of this page describes the executable program. For information on using it as a library instead, refer to the library documentation.
Installation
With Cargo installed, the following command will build and install from crates.io automatically:
cargo install tengwar
Usage
When run directly, reads all command-line arguments and prints them back in Tengwar, separated by spaces:
> tengwar "Elen síla lúmenn' omentielvo :"
> tengwar "Utúlie'n aurë!, Aiya Eldalië ar Atanatári , utúlie'n aurë!:"
If no arguments are provided, reads from Standard Input and prints transliteration to Standard Output line by line. All input is assumed to be UTF-8 encoded, and output will also be UTF-8 encoded.
A short summary of the program can be seen by running tengwar -h
, and much more detailed help text is available with tengwar --help
.
Modes
Alternative "Modes" can be selected via command-line switch:
> tengwar --beleriand "Ennyn Durin aran Moria :. pedo mellon a minno :"
> tengwar --gondor "Ennyn Durin aran Moria :. pedo mellon a minno :"
The full list of available modes can be viewed with tengwar -h
.
Numerals
Numbers are parsed quite differently from alphabetical text. This is because in Tengwar, the least significant digit is placed on the left, and numbers typically use a Base 12 system. The requirement to parse a number as a whole does, however, allow for some formatting to be specified inline. Plain numbers are simple enough:
> tengwar "9"
The default presentation is in duodecimal, Base 12, as is the custom of the Elves:
> tengwar "22"
Prefixing a number with #
will force a number to be presented in decimal, Base 10:
> tengwar "#22"
Suffixing a number with @
will append a suffix to specify an Ordinal number — for instance, "second" instead of "two":
> tengwar "Laurelin nánë i 2@ alda"
Suffixing a number between 1 and 24 with #
will eschew the typical numerals entirely, and instead use the regular tengwar as indices for a sequence.
This usage was described by Christopher Tolkien, and published in Quettar #1.
This is analogous to using the Latin alphabet to enumerate a list (a.
, b.
, etc).
> tengwar "1#︙ Telperion
2#︙ Laurelin"
Ligation
Two forms of ligation are currently supported.
The --ligate-all
option will enable both of them at the same time.
Short Carriers
The short carrier () has an alternate "ligating" form (), which extends to the right, connecting to the next character.
When invoked with --ligate-short
(or -s
), this program will use the ligating form where appropriate.
Unlike with zero-width joiners (where lack of font support is unlikely to be a problem), this variant is an entirely different character, and may not be provided at all by some fonts.
Zero-Width Joiners
In certain typefaces, the Zero-Width Joiner may be used to form ligatures of Tengwar.
When invoked with the --ligate-zwj
(or -z
) switch on the command line, this program will insert joiners into the output text between certain characters, based on various overly convoluted rules.
These rules are based on the ligature behavior of Tengwar Telcontar, as well as some degree of personal taste.
For typefaces that do not support these ligatures, the presence of the joining characters should not affect the rendering; However, it does increase the number of bytes in the output string by approximately 15%.
Note on punctuation
Tengwar punctuation is essentially impossible to reliably extrapolate from punctuated Latin text. It is therefore probably best to not even try, and instead to punctuate the input text with the output in mind.
This program does convert punctuation marks into Unicode codepoints, but whitespace is passed through verbatim, neither added nor subtracted¹.
As of this writing at version 1.0.0
, punctuation is processed as can be seen above and in the following table:
Input | Output |
---|---|
' , . , , , · |
|
: , ; |
|
⁝ , ︙ |
|
⁘ , ⁛ , … |
|
⸭ |
|
- |
|
= |
|
? |
|
! |
|
‖ * |
|
( , [ , “ |
|
) , ] , ” , „ |
|
The Pipe character (|
) is also converted to
, but it cannot be included in this table due to a technical limitation of the Markdown format.
It may be important to note that the Tilde (~
) is not converted, despite strongly resembling the punctuation character
.
This is because the Tilde represents a specific type of whitespace in LaTeX, and running within LaTeX is the motivating use case for the creation of this program.
The Tilde is left alone so that it can safely pass through and be read by LaTeX.
¹ While this program does not affect whitespace, it can only work with the discrete arguments provided to it by the environment running it.
Most environments will split your input at spaces and provide the surrounding words as unconnected values;
This program cannot distinguish between tengwar asdf qwert
and tengwar asdf<TAB>qwert
.
It is therefore highly recommended to enclose all of your input text in double quotes.
Features
Cargo Features allow very powerful changes to the behavior of a program to be baked in at compile-time.
csur
The ConScript Unicode Registry defines a block of codepoints for the Tengwar, and these are the codepoints used by this program. However, the Tengwar are in the (very slow) process of being added to the official Unicode standard, in the Supplemental Multilingual Plane. At that time, this program will switch to using those codepoints instead.
The csur
feature does not currently do anything, but when the Tengwar are finally encoded, it will cause this program to continue using the CSUR block.
The feature is provided now, ahead of time, so that anyone who wants to may specify the feature in an update system — or, if using this program as a library, in their Cargo.toml
.
This essentially allows you to opt out of the eventual switch.
dots-standard
Several codepoints defined in the CSUR Tengwar block are punctuation marks composed of dots.
Since Unicode already defines many dot punctuation characters, it may be preferable to use those wherever possible.
Compiling this program with --features "dots-standard"
will cause it to use standard Unicode characters with the same arrangements, instead of the ones in the Tengwar block.
Usage in LaTeX
The following is a minimum working example using XeLaTeX.
It uses the Tengwar Telcontar font with the Graphite renderer, and the given command will produce main.pdf
as its output.
% main.tex
\documentclass[12pt]{article}
\usepackage{fontspec}
\defaultfontfeatures{Renderer=Graphite}
\newfontfamily\TengTelc[WordSpace=2]{Tengwar Telcontar}
\newcommand{\Tengwar}[2][]{\input{|tengwar #1 -- '#2'}\unskip}
\newcommand{\Classical}[2][\TengTelc]{{#1\Tengwar[-Qrsz]{#2}}}
\newcommand{\Beleriand}[2][\TengTelc]{{#1\Tengwar[-Brsz]{#2}}}
\newcommand{\Gondor}[2][\TengTelc]{{#1\Tengwar[-Grsz]{#2}}}
\begin{document}
% "The king of Doriath, Elu Thingol, was slain by Dwarves from Nogrod."
\Classical{i aran Lestanórëo ,}
\Beleriand{“Elu Thingol„}
\Classical{, nánë nahtana ló Casari Návaróto ::}
\end{document}
xelatex -interaction=nonstopmode --shell-escape main.tex
Dependencies
~4–13MB
~188K SLoC