XML/TEI parsing issue in LMTX: lxp.lom not found (small test case)

29 Nov 2025

      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"?>
<!-- Exemple simplifié sans espace de noms TEI -->
<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:

Any advice or pointer to an up-to-date example would be very helpful.

Thank you very much,
JP

Jean-Pierre Delange

Henning Hraban Ramm

Jean-Pierre Delange

Jean-Pierre Delange

Jean-Pierre Delange

Hans Hagen

Jean-Pierre Delange

Hans Hagen

denismaier＠mailbox.org

Jean-Pierre Delange

tags

participants (4)