#번역툴: 이맥스 관련 패키지 도구 모음

#번역툴

TL;DR

Emacs 관련한 가장 훌륭한 번역 기술 자료를 담고 있다. 여기 자료를 기반으로 확장한다 [^fn:1].

emacs translation resources

translation packages

translate mode

https://github.com/rayw000/translate-mode/

paragraph-based
uses master
two file model
works with a pair of buffers
new, not a lot of features, but actively developed

org-translate

https://elpa.gnu.org/packages/org-translate.html

org-based
features segmentation (auto in source, manual in target)
single file source/target/glossary
author plans to do a full re-write as pure text mode and using a standard glossary format (TMX).
not really actively developed, but pretty well done.

emacs-cat

really only just enables highlighting sentence in other buffer and moving the highlight without switching buffers.

dictionaries

emacs-leo

https://github.com/mtenders/emacs-leo (updated version https://codeberg.org/martianh/emacs-leo)

online
interface to german focused online dictionaries

wordreference

https://codeberg.org/martianh/wordreference.el

online
interface to romance language focused dictionaries

reverso.el

https://github.com/SqrtMinusOne/reverso.el

online
includes translation, grammar, bilingual concordances, synonyms search.

sdcv-mode

https://repo.or.cz/sdcv.el.git

offline
star dict dictionaries
old and ugly (uses outline-mode but there are high-quality offline dictionaries available for sdcv, such as the Oxford English Dictionary, Le Littré Dictionnaire de la langue française, etc.
bi-gram phrases

etym.el

https://codeberg.org/martianh/etym.el

online
scrapes https://etymonline.com
quality etymological data
bare bones display still

thesauruses

mw-thesaurus

https://github.com/agzam/mw-thesaurus.el

merriam webster, open/free (non-commercial) API.
results formatted with org.

synosaurus

https://github.com/hpdeifel/synosaurus

uses wordnet (offline) and opendictionary (online) by default, can add others.
displays results as a popup, or ido or minibuffer.

wikionary-bro

https://github.com/agzam/wiktionary-bro.el

online
pretty rendered shr views
foldable results like org tree
open/free data from wikionary.

powerthesaurus

machine translation

go-translate

https://github.com/lorniu/go-translate

translation framework that supports various backends like google, bing, deepl.

google-translate

lingva.el

https://codeberg.org/martianh/lingva.el

access google translations via lingva.ml instances, without google tracking.

txl, deepl

https://github.com/tmalsburg/txl.el

requires API key, subscribers only
no longer offers a free tier for API access.

ob-translate

https://github.com/krisajenkins/ob-translate

translate text using org blocks
not updated in a long time

babel

https://github.com/juergenhoetzel/babel

handy helpers

hl-sentence

https://github.com/milkypostman/hl-sentence

highlight current sentence

hl-checkerboard

https://codeberg.org/martianh/hl-checkerboard

highlight every second sentence or clause

centered-cursor-mode

https://github.com/andre-r/centered-cursor-mode.el

scroll-all-mode

scrolls all (not just 2!) windows in a frame.

other handy packages

org-remark

https://github.com/nobiot/org-remark

annotations in separate org buffer/file.

annotate.el

https://github.com/bastibe/annotate.el/

in-buffer annotations as overlays
doesn't play with variable-pitch-mode
suited only for brief comments

guess-language.el

https://github.com/tmalsburg/guess-language.el

critic markup (`cm-mode`)

https://github.com/joostkremers/criticmarkup-emacs

track-changes functionality.
not recently updated but seems to work

emacs-sentence-navigation

https://github.com/noctuid/emacs-sentence-navigation

regexps for skipping abbreviations etc. when navigating (or highlighting) by sentence.

desirable but non-existent features

segmentation

using CAT parsers for sophisticated segmentation/sentence-end functionality (break at normal sentence ending periods, but not at "Mr." or "etc.,")
- "Once the text is extracted, it needs to be segmented. Basic "no" segmentation usually means paragraph based segmentation. Paragraphs are defined differently depending on the original format (1, or 2 line breaks for a text file, a block tag for XML-based formats, etc.). // Fine-grained segmentation is obtained by using a set of native language based regex that includes break rules and no-break rules. A simple example is break after a "period followed by a space" but don't break after "Mr. " for English." (from the discussion above)
might it not be possible to convert OmegaT's Java segmentation regexes into Elisp, then plug them in to our sentence-end or segmentation variables?
- the regexes are in pairs: before poss break and after poss break.
- segment.el has almost achieved this now.
OmegaT's segmentation info is in /OmegaT_ver.sion.num_Source/src/org/omegat/core/segmentation/
the default rules are stored as XML /OmegaT_ver.sion.num_Source/src/org/omegat/core/segmentation/defaultRules.srx.
Okapi's segmentation rules files are in this repo.

two-buffer scrolling/sync

we want this to be both rigorous and flexible. it has been done various times but it has never been done rigorously and flexibly (including assuming that buffers will lose sync and so making easy/fast ways to recover)
it should work with all the various ways people scroll and move point, including cua-mode, centered-cursor-mode, org-mode, and anything else that modifies common scrolling commands.

thoughts about building new tools

martianh: I have been thinking recently about what would be needed to have some decent translation functionality in emacs, and quickly came to the conclusion that developing a CAT tool is likely too much work for an individual working in a free software environment.

martianh: 저는 최근에 emacs 에서 적절한 번역 기능을 갖기 위해 필요한 것이 무엇인지 생각해 보았고, CAT 도구를 개발하는 것은 자유 소프트웨어 환경에서 일하는 개인에게 너무 많은 일이라는 결론에 빠르게 도달했습니다.

That said, there are various handy packages available. My thought for the moment was that if we agreed on a few common needs, we could each work on implementing them as separate modules, then see about integrating them as minor modes in an encompassing major mode, or similar. Someone could work on segementation, someone on glossaries, somone on UI/highlighting, and so on.

즉, 다양한 편리한 패키지를 사용할 수 있습니다. 잠시 동안 내 생각은 우리가 몇 가지 일반적인 요구 사항에 동의하면 개별 모듈로 구현하는 작업을 수행한 다음 이를 포괄하는 주요 모드 또는 유사 항목에서 보조 모드로 통합하는 방법을 볼 수 있다는 것입니다. 누군가는 세분화, 누군가는 용어집, 누군가는 UI/강조 표시 등에 대해 작업할 수 있습니다.

I also think that, at least for my needs, the main things that would be a big help are actually the pretty basic things about highlighting, easy paired-buffer navigation, and so on. Rather than full-fledged CAT tools like glossaries and text analysis.

나는 또한 적어도 내 필요에 따라 큰 도움이 될 주요 사항은 실제로 강조 표시, 쉬운 쌍 버퍼 탐색 등에 대한 매우 기본적인 사항이라고 생각합니다. 용어집 및 텍스트 분석과 같은 본격적인 CAT 도구가 아닙니다.

Do you have some more ideas yourself? Why not let's discuss some and maybe delegate a few simple tasks each?

아이디어가 더 있으신가요? 각각 몇 가지 간단한 작업을 논의하고 위임하는 것은 어떻습니까?

non-emacs free software translation resources

omegat

https://omegat.org/

free translation memory tool
fully featured (translation memories, segmentation, glossaries, document conversion/compatability, interface to machine translation)

apertium

https://apertium.org/index.eng.html#?dir=fra-epo&q=

GPL, public API
limited language pairs
rule-based (not neural)

okapi framework

https://www.okapiframework.org/

vast open framework for various translation tasks.
maybe we could write some basic interfaces to the various tools it includes

swordfish and friends

https://www.maxprograms.com/products/index.html

newer
is it open/free software? sort of, $120 per year after trial period. but open source.
good support from the dev
uses diff gloss format

other discussion

https://groups.google.com/g/gnu.emacs.help/c/Eb95WZ4iD6g

other emacs writing resources

🧠Junghanacs

Recent Notes

#마음챙김: #텍스트마스터 - 창조적 인간론

#조직모드 #템플릿 #샘플 - latex menuscript

#대시보드 #dashboard

#비둘기 - pigeon

#리처드파인만 #물리학자 #봉고