In Czech typography, „vlna" (literally “wave”) is a non-breaking space — the tilde ~ in Typst/LaTeX. This package inserts non-breaking spaces automatically, so that lines never break in the places where Czech typographic rules forbid it: after short prepositions, inside numbers and dates, between a number and its unit, in compound abbreviations, and more.

It implements the guideline of the Institute of the Czech Language (Internetová jazyková příručka, ?id=880) and mirrors the feature set of luavlna (Michal Hoftich) — every rule group has its own on/off switch.

Usage

#import "@preview/vlna:0.3.0": apply-vlna
#show: apply-vlna

That is all — the defaults follow the ÚJČ guideline. Debug mode outlines every glued block so you can see exactly what the package did:

#show: apply-vlna.with(debug: true)

What it does

• One-/two-letter words — k, s, v, z, a, i, but also na, po, je, se, … are bound to the following word. The first letter may be uppercase (a sentence may start with V lese…); the optional second letter must be lowercase, so acronyms like EU, AI, OSN are left alone.
• Titles / abbreviations before a name — Ing. arch. Novák, pplk. gšt. Svoboda, dr. h. c. Dvořák, viz tabulka, tzv. postup, s. 12, obr. 1, tab. 3.
• Titles after a name — Novák, Ph.D., Malý, CSc.
• Initials and abbreviated given names — M. J. Hegel, Ž. Zíbrt, Fr. Daneš, Ch. Novák.
• Numbers — digit grouping (2 500, 25,325 23), phone numbers (+420 800 123 987, 800 11 22 33), number + symbol (50 %, § 23, * 1921, † 2000), number + unit, abbreviation or currency (10 kg, 19 °C, 5 str., 1 000 000 Kč, 250 €).
• Ratios and scales — 1 : 50 000, 5 : 3, 10 : 2 = 5.
• Dates — day and month stay together, the year may separate: 21. 6. | 2024, 16. ledna | 1972.
• Compound abbreviations and codes — a. s., s. r. o., př. n. l., T. G. M., PS PČR, FF UK, ČSN 01 6910.
• Dashes — a spaced dash stays at the end of the line (slovo –); an unspaced range dash never breaks (1948–1989).
• Web and e-mail addresses — link() URLs get break opportunities exactly where ÚJČ recommends them (after /, before . and -, around @); social-media handles (@SenatCZ) never break after @.
• Chaining — runs like k s bratrem, o 2 500 Kč, od 21. 6. or plk. gšt. doc. Mgr. et Mgr. Ing. Libor Kutěj stay together as one block.
• Raw/code is skipped — nothing is glued inside raw, inline or block.

Options

Every rule group can be switched off (or on) independently:

Notes:

• units replaces the default list; extend it instead with units: default-units + ("µm", "dpt") (default-units is exported).
• number-word implements “číslo + název počítaného jevu” literally, which removes a lot of break opportunities — that is why it is off by default.
• The uppercase branch of abbreviations is limited to 2–4 letters per part, so ALL-CAPS headings are not glued word by word.
• links inserts zero-width spaces (U+200B) into the displayed address — the technique the ÚJČ note itself recommends. The link target (dest) is untouched and stays clickable.

Which short words are bound

By default, both one- and two-letter words are bound. The strict Czech rule requires this only for one-letter prepositions/conjunctions — binding all two-letter words is a stronger convention that can hurt line breaking in narrow layouts (mobile-friendly pages, multi-column text):

// Bind only one-letter words; `na`, `po`, `je`, `se`, … may end a line:
#show: apply-vlna.with(bind-two-letter-words: false)

// Bind exactly these words and nothing else (case-sensitive).
// A string is a set of single letters — this example binds one-letter
// prepositions, their uppercase forms, and uppercase A/I, while allowing
// lowercase `a`/`i` at the end of a line:
#show: apply-vlna.with(short-words: "vksuozVKSUOZAI")

// An array lists whole words, so two-letter words can be included:
#show: apply-vlna.with(short-words: ("v", "k", "s", "z", "a", "i", "na", "Na"))

short-words replaces the built-in short-word pattern entirely and takes precedence over bind-two-letter-words.

Per-language sets (luavlna \singlechars)

short-words also accepts a dictionary keyed by language code. The set is then chosen by the language of the surrounding text (text.lang), and — like in luavlna — languages not listed are not processed at all (neither short words nor title runs; the other rule groups stay language-independent):

#set text(lang: "cs")             // required — the default text.lang is "en"!
#show: apply-vlna.with(short-words: (
  cs: "AIiVvOoUuSsZzKk",          // luavlna's default set for Czech
  sk: "AIiVvOoUuSsZzKk",
))

To force one language’s set for the whole document regardless of text.lang (luavlna \preventsinglelang):

#show: apply-vlna.with(short-words: (cs: "AIiVvOoUuSsZzKk"), lang: "cs")

Initials (luavlna \compoundinitials)

Initials and abbreviated given names — an uppercase letter, optionally followed by a lowercase one, plus a period — are bound to the next word and chain with titles: M. J. Hegel, Fr. Daneš, Ing. B. Novák. Compound letters that count as one initial default to Ch (Ch. Novák):

#show: apply-vlna.with(compound-initials: ("Ch",))  // default
#show: apply-vlna.with(initials: false)             // turn initials off

Mid-document toggles (luavlna \preventsingleon/off)

A global show rule cannot be cancelled from inside the document, so the package provides state-based switches:

#import "@preview/vlna:0.3.0": apply-vlna, vlna-on, vlna-off, vlna-debug-on, vlna-debug-off
#show: apply-vlna

Tady se váže. #vlna-off() Tady ne. #vlna-on() A tady zase.
#vlna-debug-on() Orámuje zásahy odsud dál. #vlna-debug-off()

luavlna command mapping

How it works (and why jump-to-source is preserved)

A show rule that rebuilds text (it.text.replace(" ", "~")) produces new content whose source position points at the rule itself — so clicking a letter in the PDF jumps to the package code, not to your document. This package instead wraps the original matched content in #box(..). The box cannot break internally, yet the glyphs keep their original source span, so clicking them jumps to your document.

Compile-time cost is roughly 2–2.5× of the bare document and scales linearly (measured: a 300-page all-rules-firing document compiles in ~2.4 s vs. ~1.0 s without the package).

Limitations

• Inside a box, spaces do not stretch under justification and words are not hyphenated; a very long word right after a preposition can produce a looser line. In extremely narrow columns, a glued group wider than the line wraps at the column edge instead of gluing.
• Matches cannot cross styling boundaries: in k *lesu* the k is not bound, because the bold word is a separate element.
• Initials are a heuristic — a sentence ending with a bare capital (… vitamin C. Nový výzkum) is indistinguishable from an initial and gets bound too. Use initials: false if that bothers you.
• Text copied from a PDF may contain the invisible U+200B characters that links inserted, and jump-to-source is not preserved for the rewritten URL text (it is for everything else).

Implementation status

Changelog

See CHANGELOG.md.

License

MIT — © 2025–2026 iamanro.