Noteworthy changes in this release

For this release Bert Frees has been hard at work to clean up the code base. He fixed a few bugs one of which was causing memory corruption. Others have contributed corrections to braille tables or even provided new tables. For example there are now tables for Tshivenda and for Russian contracted braille.

I'd like to thank everybody for helping to bring liblouis forward.

For a detailed list of all the changes refer to the list of closed issues.

Bug fixes

• Fix a bug in the uplow opcode thanks to Bert Frees.
• Fix a possible crash due to memory corruption thanks to Bert Frees.

Braille table improvements

• The Norwegian 6 dots braille tables now correctly handles signs like section sign ('§' and '§§'), as well as numbersign ('#') before numbers. In addition the letters a with acute, o with tilde, n with long right leg and thorn are redefined. Phrases with capitals only are know handled correctly. All thanks to Lars Bjørndal.
• Fix a problem with 'ç' in Northern Kurdish thank to Imam Kahraman and Christian Egli.
• Add support for back translation in Russian literary braille thanks to Andrey Yakuboy.
• In Russian literary braille, if the letters 'Ё' and 'Й' consist of two code points, they are displayed correctly.
• Add new Russian contracted (grade 1) table thanks to Andrey Yakuboy and Ekaterina Anisimova.
• Update to Urdu, mostly to cater for spelling variations found in source text. Thanks to Jake Kyle.
• New contracted braille table for Tshivenda created by Christo de Klerk at the request and under the auspices of the South African Braille Authority.
• Fix in Arabic computer braille table thanks to Ikrami Ahmad:
  • back-translate 256 to Latin full stop
• Fix translation of URLs and file paths in Polish grade 1 table, thanks to Łukasz Golonka.

Other changes

• Fix the build script for Microsoft nmake thanks to Leonard de Ruijter.

Deprecation notice

• None

Backwards incompatible changes

• None

Invisible changes

• Major refactoring of compileTranslationTable.c thanks to Bert Frees. Thanks to these changes a few bugs were found and fixed in the code and a few tables.

New, renamed or removed tables

New

• ru-ru-g1.ctb
• ve-za-g1.utb
• ve-za-g2.ctb

Renamed

None

Removed

None

Noteworthy changes in this release

This release brings support for many new languages: There is support for six new languages from South Africa, Northern Kurdish, Kazakh, Tatar, Yakut, Bulgarian literary braille and finally Khmer, Burmese and Vietnamese. Aside from that there are also major improvements to Afrikaans, Russian literary braille, Uzbek and Hebrew Computer Braille.

I'd like to thank everybody for helping to bring liblouis forward.

For a detailed list of all the changes refer to the list of closed issues.

Bug fixes

• Don't silently ignore the last line of a table when it doesn't end with a new line character. Thanks to Bert Frees.

Braille table improvements

• New contracted braille table for the isiXhosa and isiZulu languages created by Christo de Klerk and Laurent Cadet de Fontenay at the request and under the auspices of the South African Braille Authority.
• New contracted braille table for the Sesotho, Setswana and Sepedi languages created by Christo de Klerk at the request and under the auspices of the South African Braille Authority.
• New braille table for Northern Kurdish thanks to Imam Kahraman.
• Fix a problem with ']' in Finnish 8 dot braille thanks to Christian Egli. The real problem was an unintended continuation line.
• Major update to the grade 2 Afrikaans table. It now implements the new Afrikaans rules adopted and implemented recently by the South African Braille Authority. The rule changes simplify and rationalize the rules, especially those based on pronunciation. 571 of the list of 25912 test words were impacted by the changes and all have been corrected.
• Improvements to the Russian literary braille tables thanks to Andrey Yakuboy and Bert Frees:
  • Fixed a bug in ru-litbrl-detailed.utb that was causing a capital sign to be missing for Cyrillic uppercase letters after Latin letters.
  • Fixed a bug in ru-litbrl-detailed.utb that was causing a number sign to be missing in some cases.
  • Added more accented letters
• New tables for Kazakh uncontracted, Tatar uncontracted and Yakut uncontracted braille thanks to Andrey Yakuboy.
• Added more apostrophe symbols to English Computer Braille Code table thanks to BAUM Engineering.
• Small fix to Italian computer braille thanks to Simone Dal Maso
• New table for Bulgarian literary braille thanks to Румяна Каменска.
• Fixes to Uzbek grade 1 thanks to BAUM Engineering.
  • capital sign for Roman numbers
  • signs (math, parentheses and other)
  • sh and ch inside a word
  • g and o with different accent marks
• Major overhaul of the 8 dot Hebrew Computer Braille table thanks to Adi Kushnir
  • Fixed Hebrew input to properly work. It did not work at all before.
  • Added Russian support
  • Added special European characters
  • Fixed Arabic support.
  • Fixed some symbols to comply with the Israeli standard for computer Braille.
  • Changed display name from Israeli Multilingual to just indicate Hebrew.
• New tables for Khmer, Burmese (contracted and uncontracted) and Vietnamese (uncontracted, partially and fully contracted, as well as a variant for Southern Vietnam) thanks to Dang Hoai Phúc.

Other changes

• Brilliant simplification of the table parser internals thanks to Bert Frees

New, renamed or removed tables

New

• kk.utb
• sah.utb
• tt.utb
• xh-za-g1.utb
• xh-za-g2.ctb
• zu-za-g1.utb
• zu-za-g2.ctb
• nso-za-g1.utb
• nso-za-g2.ctb
• sot-za-g2.ctb
• tsn-za-g2.ctb
• bg.utb
• kmr.tbl
• vi-vn-g0.utb
• vi-vn-g1.ctb
• vi-vn-g2.ctb
• vi-saigon-g1.ctb
• my-g1.utb
• my-g2.ctb
• km-g1.utb

Renamed

None

Removed

• he.ctb -> superseded by he-IL-comp8.utb
• vi-g1.ctb -> superceded by vi-vn-g1.ctb

Noteworthy changes in this release

This release stands in the tradition of small incremental improvements that make it another fine and solid release. Andrey Yakuboy put in a lot of work to improve Russian literary and computer braille. He also provided a new table for Belarusian uncontracted and computer braille. Aside from that there are numerous improvements in Braille tables such as French, Polish and Urdu. The documentation has been improved as has the table meta marking the tables whether they are geared towards forward or backwards translation. I'd like to thank everybody for helping to bring liblouis forward.

For a detailed list of all the changes refer to the list of closed issues.

Braille table improvements

• The backtranslation of æ in French 6 dot braille has been fixed thanks to André-Abush Clause.
• The definition of the vertical line symbol (|) in the Russian computer braille table was corrected thanks to Andrey Yakuboy.
• Improvements to the Russian literary braille thanks to Andrey Yakuboy:
  • Letters which are used in Old-Russian texts were added.
  • Arrows, integrals and other symbols were added.
  • the horizontal bar symbol (u+2015) was added and marked as a dash-symbol.
  • Now accent marks are put before a letter according to the Russian braille rules.
  • Many bugs with processing space-symbols were fixed.
  • Improvements to processing dialogs and direct speech.
  • Improvements to processing letters of different alphabets.
• New tables for Belarusian uncontracted and computer braille thanks to Andrey Yakuboy.
• Various fixes to Polish tables thanks to Łukasz Golonka:
  • Add em dash to both computer and literary Polish tables
  • Remove some definitions which made it impossible to back translate quoted strings from the Grade 1 table
  • Fix back translation of capital M and add tests for back translation of all Latin capital letters for Grade 1 table
• Update to Urdu thanks to Jake Kyle:
  • Revised definitions of SUPERSCRIPT ALEF and ARABIC LIGATURE ALLAH ISOLATED FORM.
  • Changed contractions of ٹھ, ان, اِن, بے, بی, اَور and اور.
  • Dots 3-3 used for repeated words instead of 6-6.

Other changes

• Fixed some valgrind issues thanks to André-Abush Clause.
• Improve the documentation of caps opcodes thanks to Bue Vester-Andersen.
• Add meta data to all tables indicating whether a table is geared towards forward or backwards translation or both. Again thanks to Bue Vester-Andersen.

New, renamed or removed tables

New

• bel.utb
• bel-comp.utb

Renamed

None

Removed

None

Noteworthy changes in this release

In the frenzy to ship the 3.16.0 release a little mishap slipped in. The test suite contains an error and consequently doesn't pass. It was promptly discovered and fixed by Samuel Thibault. So we present to you the real, hopefully fixed 3.16 release. Let's name it 3.16.1.

Noteworthy changes in this release

This release is mostly the work of Bert Frees. He has put in a lot of work and pushed some major improvements with regards to how emphasis is handled (and how this is documented). He fixed a crash when reading URLs using computer braille. He also worked with many contributors to get their improvements in. So we have better support for UEB, Dutch, Urdu, Malayalam, Arabic, Bashkir, Uzbek, Russian computer braille and for EBAE. I'd like to thank everybody for helping to bring liblouis forward.

For a detailed list of all the changes refer to the list of closed issues.

New features

• A new opcode noemphchars was added. This gives you more control over the placement and scope of various emphasis indicators. Thanks to Bert Frees.

Bug fixes

• Fix an endless loop when using compbrlAtCursor thanks to Bert Frees

Braille table improvements

• Improvements to US EBAE conformance with BANA emphasis guidelines thanks to Benetech.org
• Fixed a problem with apostrophe in Malayalam thanks to Jake Kyle
• Improvements to contraction use in UEB thanks to James Bowden. In particular the checkmark (U+2713) is added, the emdash (U+2014) is fixed, the right single quote used as apostrophe between letters is fixed and finally some more accented letters have been added.
• Improvements to Urdu Braille and Malayalam thanks to Jake Kyle
• Add characters for Bashkir and Uzbek languages to the Russian computer braille table. Thanks to Andrey Yakuboy.
• New table for Bashkir uncontracted braille thanks to Rustam Churagulov and Gabidullin Yunir.
• Updated < and > symbols in the Arabic tables thanks to Ikrami Ahmad.
• Improvements to Dutch thanks to Bert Frees.
  • Every word part in a capitalized compound word counts in the length of a passage.
  • begcapsphrase is allowed to start in a word preceded by punctuation.
• Improved translation of ligatures in UEB thanks to Bert Frees.
• Multiple improvements to the Russian literary braille thanks to Andrey Yakuboy and Bert Frees:
  • Many new symbols (punctuation, bullets, math symbols, etc) have been added.
  • The table now includes the international phonetic alphabet table.
  • Punctuation after digits and fractions is now marked with dot 6 to avoid ambiguities.
  • Other changes to conform better with Russian braille rules.
  • A new table that indicates capital letters was added. This is the new recommended Russian table for braille display users.
  • Removed the old ru-ru-g1.utb in favor of ru-litbrl.ctb and ru-litbrl-detailed.utb.

Other changes

• Support for Python 2 has been removed. The python bindings now only support Python 3. The deprecation notice was announced in Release 3.13 and the removal is finally done now.
• Improvements to the placement of emphasis and capital indicators.
• The documentation for the emphasis opcodes has been further improved thanks to Bert Frees.

Deprecation notice

• None

Backwards incompatible changes

• emphmodechars can and must now be set per emphasis class.

Invisible changes

• The emphasis and capitalization handling code has seen major streamlining, simplification and tidying thanks to Bert Frees.

New, renamed or removed tables

New

• ba.utb
• ru-litbrl-detailed.utb
• en-us-emphasis.uti

Renamed

None

Removed

• ru-ru-g1.utb

Noteworthy changes in this release

This release contains updates to the Afrikaans, Malayalam, Malay, Israeli, classical English and Portuguese tables. It also contains a new table for Coptic thanks to first-time contributor Ibraam Nasif. Behind the scenes there have been major changes to the cross-compilation tool-chain that will finally allow us to ship pre-built windows binaries that contain lou_checkyaml.exe.

For a detailed list of all the changes refer to the list of closed issues.

New features

None

Bug fixes

None

Braille table improvements

• New computer braille table for the Coptic language thanks to Ibraam Nasif from Coptic Bishopric of Youth
• Improvements to Afrikaans thanks to Christo de Klerk
  • Fixed correct handling of the ʼn character (0x149)
  • Prevent contraction of
    • "self" in "selfoon", "selfone",
    • "ui" in words and names like "quisling", "quidproquo", "quiëtis", "Quinten", "Quintus",
    • "er" in words like "konstituering".
• Improved table for Malayalam grade 1 thanks to Jake Kyle.
  • Fixed rules for Candrakkala character at the end of a word.
  • Fixed rules for letters ര മ ങ in certain contexts.
  • Improved back-translation.
• Improvements to grade 2 Malay braille thanks to Herbert Koh.
• Improvements to Israeli 6-dot thanks to BAUM Engineering.
  • Fixed number sign for Arabic digits.
  • Fixed Hebrew letters with dagesh.
• Added bullets to the classical English tables, thanks to Bue Vester-Andersen
• Improved Portuguese grade 1 thanks to Bue Vester-Andersen.
  • Fixed forward translation of &, <, >, `, ¶ and '.
  • Improved back-translation. The table is now tagged as going both forward and backward.
  • Added missing square root sign.

Other changes

None

Deprecation notice

None

Backwards incompatible changes

None

Invisible changes

• Major rework of the cross-compilation tool-chain. There are now Docker images that should allow for local builds of binaries for 32 and 64 bit. These include the libyaml library so the binaries contain lou_checkyaml.exe. This will allow table creators to test their changes locally under windows.

New, renamed or removed tables

New

• cop-eg-comp8.utb

Renamed

None

Removed

None

Noteworthy changes in this release

This release contains major updates to Arabic, Dutch 8-dot computer braille, German, Russian computer braille, Ukrainian computer braille and Uzbek. Also there are new tables for Israeli multi-language Hebrew/Arabic/English braille and Malay. Aside from that there have been many code cleanups and bug fixes.

For a detailed list of all the changes refer to the list of closed issues.

New features

• Add new test mode in lou_checkyaml for testing display tables. Thanks to Bert Frees.

Bug fixes

• Work around a bug that causes an endless loop in French uncontracted braille thanks to André-Abush Clause.

Braille table improvements

• Updates to the Russian computer braille table thanks to Andrey Yakuboy:
  • Fixed diacritics, Greek letters and Hebrew.
  • Added new Cyrillic letters.
  • Fixed several minor bugs.
• New table for Uzbek braille thanks to BAUM Engineering.
• New table for Ukrainian computer braille thanks to Andrey Yakuboy and Oleh Shpai.
• Improvements to Arabic thanks to Ikrami Ahmad and Mada, Qatar Assistive Technology Center.
  • Added a lot of new symbols (algebra, geometry, Greek, etc.) to the grade 1 table.
  • Changed some symbols for less conflicts with grade 2.
  • Implemented letter sign / grade 1 indicator to cancel numbers and inhibit contractions when in grade 2.
  • Updated 8-dot computer braille table for better (easier to memorize) presentation of basic math symbols and punctuation marks.
  • Improved back-translation.
• Improvement to German thanks to Bue Vester-Andersen.
  • Removed superfluous letter sign before certain lesser-used accented letters.
• Final version of Dutch 8-dot computer braille table thanks to Leonard de Ruijter.
• New Israeli multi-language Hebrew/Arabic/English tables for both uncontracted and computer braille. Thanks to BAUM Engineering and Erez Kugler from TSR Gaash.
• New table for grade 2 Malay braille (Malaysia) thanks to Herbert Koh.

Other changes

• You can now use up to 256 noletsign characters in a table, thanks to Christian Egli.
• Major improvements to the documentation for the Standing Alone Sequences and their related opcodes such as seqdelimiter and also clarifications regarding the begcaps and endcaps opcodes thanks to Bue Vester-Andersen.

Deprecation notice

• The class opcode has been deprecated in favor of attribute, which has been enhanced to do everything class can do.

Backwards incompatible changes

None

Invisible changes

New, renamed or removed tables

New

• uz-g1.utb
• ar-ar-g1-core.uti
• ar-ar-math.uti
• he-IL.utb
• he-IL-comp8.utb

Renamed

None

Removed

• da-dk.dis: This table has long been superseded by da-dk-octobraille.dis.
• da-lt.ctb: This table was a step on the way to defining a Danish 8 dot standard. The table was referring to an old Danish Braille note-taker from the 80s and 90s, which is not used any more.
• fi1.ctb and fi2.ctb: These tables are obsolete. Use fi-fi-8dot.ctb instead.

Noteworthy changes in this release

A tremendous amount of work by lots of different people has gone into this release. Bert Frees for example added new opcodes, fixed long standing bugs (with emphasis and capitalization), made the documentation more clear and helped table contributors bringing in their improvements. Bue Vester-Andersen added major improvements to Danish and German braille. This release also contains much improved tables for Afrikaans, Russian computer braille, Urdu and Chinese.

On a personal note I'd like to say that this is the 20th release of liblouis shipped by the current release team Bert Frees and Christian Egli. I hope we can continue to do so.

For a detailed list of all the changes refer to the list of closed issues.

New features

• The nocross opcode has been made into a prefix opcode, which means that it can now be used in combination not just with the always opcode but also with other translation opcodes such as begword, e.g. nocross begword sh 146 for example. The old usage of the opcode no longer works, see Backwards incompatible changes below.
• Added a new opcode rependword, needed to implement Malay braille. Thanks to Bert Frees.
• Emphasis and capitals can now be marked correctly in more cases. Thanks to Bert Frees.
  • When begemphword (begcapsword) is defined but not endemphword (endcapsword), use emphletter (capsletter) to mark every character in an emphasized (capitalized) part that ends in the middle of a word.
  • When begemphword (begcapsword) and begemph (begcaps) are both not defined, use emphletter (capsletter) to mark every emphasized (uppercase) letter.
  • capsnocont has no influence anymore on whether capsletter is used for every uppercase letter or not, see Backwards incompatible changes below.

Bug fixes

• Fix a crash on sparc64 thanks to Samuel Thibault.
• Fixed a bug where the inputPos array was not monotonically increasing. This is a problem for language bindings that try to do hyphenation based on this information. Thanks to Bert Frees.
• A couple of small Coverity fixes thanks to David King.
• Fix an infinite loop involving multipass rules and backtranslation thanks to Bert Frees and Bue Vester-Andersen.
• Fix several problems in the Python bindings thanks to Leonard de Ruijter, Łukasz Golonka and André-Abush Clause:
  • Ensure that the mbcs encoding is always used to encode file path on Windows, especially when 'Unicode UTF-8' feature is enabled.
  • getTypeformForEmphClass was passed a decoded string on Python 3 instead of an encoded one.
  • The same applies to compileString. We encoded the input string, but never used that encoded one when passing it to liblouis.
• begemphword (begcapsword) and endemphword (endcapsword) are now used for single letters within a word when emphletter (capsletter) is not defined.
• For backward translation the returned input length is now equal to the initial input length (as it is for forward translation) thanks to Bert Frees.

Braille table improvements

• Improvements to Afrikaans thanks to Christo de Klerk
  • Fixed back translation of some lower words.
  • Fixed several contraction errors where contractions are based on pronunciation or occur where compound words join.
  • Eliminated grade 1 indicator where a single letter follows a single apostrophe to indicate an abbreviated word, e.g. 'k for ek.
  • Corrected the use of the grade 1 indicator in grade 2 with signs which would conflict with contractions, such as the trademark sign that would conflict with the contraction of tussen.
  • Fixed the back translation conflict between the UEB circle shape indicator and words containing $= being the contraction for &quot;alge&quot;.
• Improvements to Russian computer braille table thanks to Andrey Yakuboy.
  • The table is now based on the Unicode character set. A lot of new characters were added that aren't in KOI8-R, including Greek, Hebrew, accented letters and so on.
  • For special characters we now follow the definitions from other software (JAWS, TSS, DBT) instead of following the American standard.
• Improvements to Urdu Braille thanks to Jake Kyle
  • Changed definition of \x0624 (Waw with hamza above) from dots 1256 to dots 3-2456 following advice from proof reader.
  • Added 2 character versions of letter plus diactric (previously only the one character versions defined):
• Updates to the Danish Tables thanks to Bue Vester-Andersen:
  • Fixed back-translation for some JAWS Braille drivers, which deliver Unicode characters to Liblouis as input from a Braille keyboard.
  • Ensured proper use of letsign in connection with accented letters.
  • Re-arranged and strengthened tests to include all Unicode characters defined in the tables.
• Updates to the Chinese braille table (zh-tw.ctb) thanks to Bo-Cheng Jhan, Coscell Kao and Victor Cai.
• New experimental German tables for grade 0 (Basisschrift) and grade 1 (Vollschrift) that are more suitable for back-translation, thanks to Bue Vester-Andersen.
  • All capital letters are marked.
  • Accented letters are translated using de-accents-detailed.cti to make the translation as detailed as possible.
  • Apostrophes and single quotation marks are translated the same.

Other changes

• The documentation for the emphasis opcodes has been reworked and is now much more clear and accurate thanks to Bert Frees.

Deprecation notice

• Python 2 is no longer maintained since the beginning of this year. The current liblouis Python bindings work with both Python 2 and 3. However the support for Python 2 will be dropped in the next release (3.14) at the beginning of June 2020. If you have code that uses the liblouis Python bindings with Python 2 then please refer to the official porting guide for help with migrating it.

Backwards incompatible changes

• Since the nocross opcode has been made into a prefix opcode, it must now be used in combination with another opcode. For example nocross sh 146 must now be written as nocross always sh 146.
• The capsnocont opcode can not be used anymore to control whether capsletter is inserted for every uppercase letter. This now depends on whether begcaps and begcapsword are defined.

Invisible changes

New, renamed or removed tables

New

• de-g0-bidi-core.uti
• de-g0-bidi.utb
• de-g1-bidi-core.cti
• de-g1-bidi.ctb

Renamed

None

Removed

None

Noteworthy changes in this release

This release contains major updates to the UEB, Afrikaans, Chinese, Danish and Polish tables. Aside from that there have been many code cleanups, such as the elimination of many global vars and bug fixes such as an endless loop or a crash in lou_translate.

For a detailed list of all the changes refer to the list of closed issues.

New features

None

Bug fixes

• Fix a memory leak when several tables are specified and some of them that can't be resolved. Thanks to André-Abush Clause.
• Fix an endless loop with multipass rules where endReplace is smaller or equal to startMatch. Thanks to Bert Frees and Bue Vester-Andersen.

Braille table improvements

• Punctuation corrections in Ethiopic Braille thanks to Dr. Tamru E. Belay.
• Fixes to the Norwegian 8dot braille table in regards to 4 Sami characters, capital and small letters S and Z with caron. Thanks to Oddvar Øyan and Lars Bjørndal.
• Improvements to Afrikaans contracted braille thanks to Christo de Klerk
  • Over 50 cases have been corrected where braille rules were not correctly applied in words, mainly in those cases where contractions depend on pronunciation; for example, ui or ie must not be contracted in requiem.
  • Corrected contraction errors caused by start or end of input not properly taken into account.
  • Words are no longer contracted into lower word contractions when they are adjacent to lower punctuation, for example: &quot;hier.
  • Words are now contracted into their lower contractions when they are adjacent to upper punctuation, for example: (hier.
• New draft table for Dutch 8-dot computer braille thanks to Leonard de Ruijter.
• Updates to the Chinese braille table (zh-tw.ctb) thanks to Bo-Cheng Jhan, Coscell Kao, 特種兵, 黃偉豪, and Victor Cai.
• Fixes to Polish grade 1, thanks to Łukasz Golonka
  • Removes some unneeded midnum symbols from the Polish Grade 1.
  • Fixes some symbols which weren't defined according to the specification.
  • Makes it possible to type dot from a braille keyboard.
  • Adds Greek letters and some commonly used math operators to both the Grade 1 and the computer braille table.
• Fix several conflicts in fr-bfu-comp68.cti with regards to the IPA Unicode range. These had been especially noticed when compbrlAtCursor mode flag was used. Thanks to André-Abush Clause.
• Major Improvements to contraction use in UEB thanks to James Bowden.
• Updates to the Danish Tables thanks to Bue Vester-Andersen:
  • Added miscelaneous Unicode characters to 8 dots grade 1 and 2 (accented letters, punctuation, arrows and some math signs). Most of these characters have not been defined in the Danish Braille standard. This implementation is purely experimental, and the characters may be changed later.
  • Updated the 6 dots tables with more Unicode characters (no arrows or math signs).
  • Corrected a bug in 8 dots grade 2, which resulted in the &quot;var&quot; contraction not always being properly applied.

Other changes

• Make sure the log callback uses the same calling convention as all the other exported functions. Thanks to Leonard de Ruijter.

• Fix a problem with Non-ASCII characters in file paths in the Python bindings, thanks to André-Abush Clause.

• Eliminate some of the globals variables thanks to Bert Frees.

• The display and the translation are now separated at least internally, thanks to Bert Frees. As a reminder, there are two phases to a braille conversion:

  • translation: liblouis uses the rules in the translation table to convert characters to dots
  • display: display the dots as characters. Usually liblouis uses the characters defined in display rules (in display tables) but as a fallback it uses mappings defined in the translation table, e.g. letter rules.

• No longer install lou_compare, a tool that is used to run regression tests for UEB. As it is only run during testing it will no longer be installed by default on a users machine.

• Remove a hidden feature of lou_translate that would cause it to crash if passed an invalid file name. Thanks to Christian Egli.

• Raise an error if a dot pattern can not be displayed instead of silently ignoring it thanks to Bert Frees.

Deprecation notice

None

Backwards incompatible changes

None

Invisible changes

New, renamed or removed tables

New

• nl-comp8.utb

Renamed

None

Removed

None