November 2025 - ntg-context

install modules in a Docker container
by Henning Hraban Ramm 05 Dec '25

05 Dec '25

I’ve not much experience with Docker and tried the context:lmtx container from Island of TeX. It downloads quickly and works well. But I need some additional modules. Installation with mtxrun seems to work, but ConTeXt can’t find the module afterwards. I’d guess that’s a Docker problem? Here’s the Dockerfile: FROM registry.gitlab.com/islandoftex/images/context:lmtx ENV TZ=Europe/Berlin USER root RUN apt update RUN apt-get install fonts-liberation fontconfig RUN fc-cache -f RUN cd /context/tex && mtxrun --script install-modules --install statistical-charts,tikz RUN mtxrun --generate RUN mtxrun --script fonts --reload --force USER jenkins Hraban

6 17

a bug in typoduc.lmt?
by Udi Fogiel 01 Dec '25

01 Dec '25

Hello, consider the following: -------------------------------------------- \setuppapersize[A4] \mainlanguage[arabic] \setupdirections[bidi=global, method=unicode] \setupbodyfontenvironment[default][em=italic] \definefontfamily [myfont][serif][Amiri] [features=arabic] \definefontfamily [myfont][mono] [Latin Modern Mono] \setupbodyfont [myfont] \setuppagenumbering[state=none] \starttext بى 19 az 29 تى 19 29 39 ثى 49-59 سى 69/79 شى 19.29 صى 39,49\stoptext -------------------------------------------- I'm not sure if I get the expected out put. is it a problem with my setup, my expectation, or a bug in typo-duc.lmt? I attached the out put I get (result.pdf), and what I expect to get (expected.pdf). thanks, Udi

2 3

XML/TEI parsing issue in LMTX: lxp.lom not found (small test case)
by Jean-Pierre Delange 01 Dec '25

01 Dec '25

Hello everyone, I am currently trying to build a minimal example showing how to process TEI-XML inside ConTeXt (see here for explanations : https://wiki.contextgarden.net/TEI_xml) To do this I created two small test files: 1. cicero-sample-tei.xml (a very small TEI fragment), which is : % cicero-sample-tei.xml <?xml version="1.0" encoding="UTF-8"?>  <TEI> <text> <body> <div type="edition" xml:lang="la"> <head>Exemple d'édition critique (Cicero)</head> <p> <persName ref="#cicero">Marcus Tullius Cicero</persName> in <placeName ref="#roma">Roma</placeName> <app> <lem wit="#A">virtutem magnam</lem> <rdg wit="#B">virtutem magni</rdg> </app> laudavit. </p> <note xml:lang="fr"> L'apparat signale une variation sur l'adjectif qualifiant <code>virtus</code>. </note> </div> <div type="translation" xml:lang="fr"> <head>Traduction française (exemple)</head> <p> Marcus Tullius Cicéron, à Rome, loua la vertu. </p> </div> </body> </text> </TEI> 2.**cicero-tei-demo.tex (a ConTeXt file using Lua to parse the XML), which is : % cicero-tei-demo.tex \setuppapersize[A5] \setupbodyfont[latin-modern] \starttext \startluacode local lom = require("lxp.lom") -- Fonctions utilitaires local function readfile(name) local f = assert(io.open(name,"r")) local c = f:read("*all") f:close() return c end local function find_child(node, tag) for _, child in ipairs(node) do if type(child) == "table" and child.tag == tag then return child end end return nil end local function find_children(node, tag) local t = {} for _, child in ipairs(node) do if type(child) == "table" and child.tag == tag then t[#t+1] = child end end return t end local function find_div_by_type(body, dtype) for _, child in ipairs(body) do if type(child) == "table" and child.tag == "div" and child.attr and child.attr.type == dtype then return child end end return nil end local function text_content(node) local parts = {} for _, item in ipairs(node) do if type(item) == "string" then parts[#parts+1] = item elseif type(item) == "table" and item[1] and type(item[1]) == "string" then parts[#parts+1] = item[1] end end return table.concat(parts, " ") end -- Index simples (collections de noms) local persons, places = {} local function add_unique(tbl, s) if not s or s == "" then return end tbl._seen = tbl._seen or {} if not tbl._seen[s] then tbl._seen[s] = true tbl[#tbl+1] = s end end -- Lecture du fichier TEI local xml_data = readfile("cicero-sample-tei.xml") local parsed = assert(lom.parse(xml_data)) local TEI = parsed[1] local text = find_child(TEI, "text") local body = find_child(text, "body") local edition_div = find_div_by_type(body, "edition") local head_edition = find_child(edition_div, "head") local p_edition = find_child(edition_div, "p") local extra_note = find_child(edition_div, "note") local trans_div = find_div_by_type(body, "translation") local head_trans = trans_div and find_child(trans_div, "head") local p_trans = trans_div and find_child(trans_div, "p") -- Traitement éditorial du paragraphe latin local function process_edition_paragraph(pnode) for _, item in ipairs(pnode) do if type(item) == "string" then context(item) elseif type(item) == "table" then if item.tag == "persName" then local n = text_content(item) add_unique(persons, n) context("\\sc{" .. n .. "} ") elseif item.tag == "placeName" then local n = text_content(item) add_unique(places, n) context("\\em{" .. n .. "} ") elseif item.tag == "app" then local lem = find_child(item, "lem") local rdgs = find_children(item, "rdg") context(text_content(lem)) if #rdgs > 0 then local note_parts = {} for _, r in ipairs(rdgs) do local wit = r.attr and r.attr.wit or "" local txt = text_content(r) if wit ~= "" then note_parts[#note_parts+1] = wit .. ": " .. txt else note_parts[#note_parts+1] = txt end end context.footnote(table.concat(note_parts, "; ")) end elseif item.tag == "note" then context.footnote(text_content(item)) else context(text_content(item)) end end end end -- Composition du document context.chapter(head_edition[1]) process_edition_paragraph(p_edition) context.par() if extra_note then context.footnote(extra_note[1]) end if head_trans and p_trans then context.blank() context.subject(head_trans[1]) context(p_trans[1]) context.par() end -- Mini-index context.blank() context.section("Personnes mentionnées") for _, p in ipairs(persons) do context.par() context(p) end context.blank() context.section("Lieux mentionnés") for _, l in ipairs(places) do context.par() context(l) end \stopluacode \stoptext The goal is simply to read the XML file, extract a few nodes, and print them in ConTeXt — nothing advanced, just a proof of concept to understand the workflow. However, when compiling the TeX file with LMTX, ConTeXt aborts immediately with the following error: |lua error: module'lxp.lom'notfound: no field package.preload['lxp.lom'] no file '.../lua/lxp/lom.lua'... | So the problem appears before any XML processing takes place: LMTX cannot load |lxp.lom|. I understand that LMTX uses a sandboxed Lua environment, so maybe |lxp| is not part of the allowed modules anymore. My question is therefore quite simple: Is there a recommended way (or example) to parse XML/TEI within LMTX today, given that |lxp.lom| is unavailable? Should one use the built-in ConTeXt XML tools instead (|lxml|, |xml.filter|, etc.)? Any advice or pointer to an up-to-date example would be very helpful. Thank you very much, JP

4 9

SIL Gentium v7
by Henning Hraban Ramm 01 Dec '25

01 Dec '25

Did you see? As of June 2025, there’s a new version of the Gentium font family. The designers have enhanced kerning and hinting, oldstyle numbers and Greek, and you can use it as a variable font with a weight axis. https://software.sil.org/gentium/v7changes/ Hraban

3 2

export adds space on hyphenation
by Henning Hraban Ramm 30 Nov '25

30 Nov '25

In the XML export, spaces are inserted in words when there’s a hyphenation over a page break in the PDF output: """ \setuppapersize[A7,landscape] \setupbackend[export=yes] \starttext \dorecurse{50}{nothing verylongword verylong toolong nothing } \stoptext """ You’ll find one instance of “very long” in the export. Hraban

1 0

\unit with converted measures
by Henning Hraban Ramm 30 Nov '25

30 Nov '25

I made tables of unit conversions to explain ConTeXt’s units. Conversion works like this: \define[2]\Conv{\scratchdimen #1 \the\nodimen #2 \scratchdimen} \Conv{1dk}{mm} That gives me "2.263348mm", while I’d like "2,263348 mm". I tried to use \unit within the definition above or like \unit{\Conv{1dk}{mm}}, but the result is empty – probably an expansion problem? Complete not-working example: """ \mainlanguage[de] \setupunit[method=3,space=small,] \define[2]\Conv{\unit{\dimexpr\scratchdimen #1 \the\nodimen #2 \scratchdimen\relax}} \starttext \bTABLE \bTR\bTH Einheit \eTH\bTH mm \eTH\bTH pt \eTH\eTR \bTR\bTD 1 dk \eTD\bTD \Conv{1dk}{mm} \eTD\bTD \Conv{1dk}{pt} \eTD\eTR \bTR\bTD 1 es \eTD\bTD \Conv{1es}{mm} \eTD\bTD \Conv{1es}{pt} \eTD\eTR \bTR\bTD 1 ts \eTD\bTD \Conv{1ts}{mm} \eTD\bTD \Conv{1ts}{pt} \eTD\eTR \eTABLE \stoptext """ (Yes, I should have used a loop…) Hraban

3 10

Odd metapost/metafun behaviour
by Mason, Roger 29 Nov '25

29 Nov '25

Hello, This code: \startMPcode u = 2cm; % unit length; xmax = 1.5 ; ymax = 1.5 ; % axes parameters path circle; circle = fullcircle scaled 2u; draw circle withcolor blue; x = -1.0 ; % any value between -1 and 1 len = 3u; % tangent length t = (acos x)/45 ; % point-node of fullcircle corresponding to x pair I; I = point t of circle ; if (x <> 0.0) and (abs(x) <> 1.0): draw origin -- I ; fi ; % tangent on the upper part of circle pair v ; v = unitvector I rotated 90 ; draw I - 0.5len*v -- I + 0.5len*v withcolor red ; draw I withpen pencircle scaled 3bp; drawarrow (-xmax*u, 0) -- (xmax*u, 0) ; % horizontal axis drawarrow (0, -ymax*u) -- (0, ymax*u) ; % vertical axis label.bot(btex $x$ etex, (xmax*u, 0)) ; \stopMPcode Should draw a circle with a tangent at x = -1 parallel to y. However, the point of tangency is shown at approximately x = 1, y = 0.01 or so. I got the code from a metapost example: u = 2cm; % unit length; xmax = 1.5 ; ymax = 1.5 ; % axes parameters beginfig(1); path circle; circle = fullcircle scaled 2u; draw circle withcolor blue ; x = -1.0 ; % any value between -1 and 1 len = 3u; % tangent length t = (acos x)/45 ; % point-node of fullcircle corresponding to x pair I; I = point t of circle ; if (x <> 0) and (abs(x) <> 1): draw origin -- I ; fi ; % tangent on the upper part of circle pair v ; v = unitvector I rotated 90 ; draw I - 0.5len*v -- I + 0.5len*v withcolor red ; draw I withpen pencircle scaled 3bp; drawarrow (-xmax*u, 0) -- (xmax*u, 0) ; % horizontal axis drawarrow (0, -ymax*u) -- (0, ymax*u) ; % vertical axis label.bot(btex $x$ etex, (xmax*u, 0)) ; endfig; end; The original code generates the expected output when run using: mpost --mem=metafun --numbersystem="double" tangent.mp What do I need to get ConTeXt to compile this code corretly? Thanks, Roger mtx-context | ConTeXt Process Management 1.06 mtx-context | mtx-context | main context file: /usr/home/rmason/tex/texmf-context/tex/context/base/mkiv/context.mkiv mtx-context | current version: 2025.02.28 18:12 mtx-context | main context file: /usr/home/rmason/tex/texmf-context/tex/context/base/mkxl/context.mkxl mtx-context | current version: 2025.02.28 18:12

2 2

\writetolit
by Jim 29 Nov '25

29 Nov '25

Hi all, at https://wiki.contextgarden.net/Command/writetolist we see the example %%%% \definelist[Reprints][criterium=all] \starttext \section{Sec 1} \writetolist[Reprints]{1.}{List entry A}% \writetolist[Reprints]{}{With no number}% \subsection{Subsec 1} \writetolist[Reprints]{2.}{List entry B}% \completelist[Reprints] \stoptext %%%% and the TOC entries cause me to speculate that the intention of this example is that, in the TOC, "List Entry A" will be annotated with "1." (and "List Entry B" will have "2."). But when I run that example, neither of those entries have numbers. Can someone tell me whether (a) it's somehow a "me" problem, or (b) it's a ConTeXt problem? Thanks. Bonus question: is there a macro or counter or token list or ... which has the number of the current chapter (and section and subsection and ...)? Thanks ^ 2. Jim

2 5

Uppercase letters with diacritics in grid mode lead to blank line.
by gbsolisb＠gmail.com 27 Nov '25

27 Nov '25

Hello everyone, A spurious empty line appears whenever I use an uppercase letter with a diacritical mark in grid typesetting. Here's a minimal example: ``` \setuplayout[grid=yes] \starttext \dorecurse{30}{word } È \stoptext ``` In which one can see an empty line between the second and third lines of text. Is this expected? Thank you.

3 3

Adjusting kerning/spread in lmt_followtext
by Mojca Miklavec Groenhuis 27 Nov '25

27 Nov '25

Hi, When typesetting large text on small curves with lmt_followtext, the letters start overlapping. There is a parameter "spread=true", but that spreads the text to the full curve, while I would only like to slightly increase the kerning between individual characters. Example: \setupbodyfont [antykwa,25pt] \starttext \startMPpage draw lmt_followtext [ text = "{\bf this is somewhat squeezed}", path = fullcircle scaled 5cm rotated 90, trace = false, spread = false, reverse = false, ]; \stopMPpage \stoptext In theory luatex/metapost could probably calculate the amount of kerning adjustment needed to avoid overlap of the letters based on height of the letters and angle difference, but it would already be sufficient if I could add a fixed amount of additional kerning between the individual letters. Second question: is it possible to somehow combine lmt_outline and lmt_followtext? That is: how do I typeset outline letters on a curve? Thank you, Mojca PS: I'm slightly confused about the documentation. Some elements are strings, others are numeric, yet the default value seems to be boolean. Based on other examples it looks like autoscaleup should be string rather than numeric, and it would be really helpful to have a list of valid values ("yes", "max", ...). Though for some reason neither of them seemed to have any effect for me. spread string true trace numeric false reverse numeric false autoscaleup numeric no autoscaledown string no

2 3