lp:~onboard/onboard/word-completion
Created by
marmuta
and last modified
- Get this branch:
- bzr branch lp:~onboard/onboard/word-completion
Members of
Onboard Devel Team
can upload to this branch. Log in for directions.
Branch merges
Propose for merging
No branches
dependent on this one.
- Onboard Devel Team: Pending (preview) requested
-
Diff: 111635 lines (+26524/-67988)124 files modified.bzrignore (+3/-0)
Onboard/Config.py (+38/-5)
Onboard/KeyCommon.py (+31/-12)
Onboard/KeyGtk.py (+146/-10)
Onboard/Keyboard.py (+308/-27)
Onboard/KeyboardGTK.py (+3/-2)
Onboard/KeyboardSVG.py (+95/-42)
Onboard/Layout.py (+17/-0)
Onboard/OnboardGtk.py (+5/-0)
Onboard/WordPredictor.py (+242/-0)
data/onboard.gschema.xml (+23/-2)
layouts/Full Keyboard-Alpha.svg (+1/-4)
layouts/Full Keyboard.onboard (+184/-170)
po/ace.po (+0/-842)
po/af.po (+264/-791)
po/am.po (+0/-867)
po/ar.po (+230/-781)
po/ast.po (+264/-866)
po/az.po (+0/-883)
po/be.po (+257/-797)
po/bg.po (+264/-848)
po/bn.po (+244/-875)
po/br.po (+264/-790)
po/bs.po (+242/-879)
po/ca.po (+267/-868)
po/ca@valencia.po (+0/-986)
po/cs.po (+240/-878)
po/cy.po (+0/-840)
po/da.po (+266/-801)
po/de.po (+270/-868)
po/el.po (+271/-806)
po/en_AU.po (+272/-868)
po/en_CA.po (+251/-813)
po/en_GB.po (+266/-862)
po/eo.po (+264/-860)
po/es.po (+269/-864)
po/et.po (+228/-769)
po/eu.po (+245/-783)
po/fi.po (+266/-854)
po/fil.po (+226/-722)
po/fo.po (+0/-847)
po/fr.po (+271/-871)
po/ga.po (+226/-724)
po/gl.po (+274/-871)
po/he.po (+272/-810)
po/hi.po (+256/-812)
po/hr.po (+231/-736)
po/hu.po (+262/-860)
po/hy.po (+0/-841)
po/id.po (+238/-824)
po/is.po (+231/-730)
po/it.po (+269/-865)
po/ja.po (+250/-856)
po/kk.po (+260/-808)
po/km.po (+0/-844)
po/kn.po (+231/-738)
po/ko.po (+246/-822)
po/ku.po (+226/-722)
po/ky.po (+0/-840)
po/lt.po (+241/-782)
po/lv.po (+263/-860)
po/ml.po (+233/-749)
po/mr.po (+0/-840)
po/ms.po (+268/-862)
po/my.po (+0/-840)
po/nb.po (+255/-815)
po/ne.po (+0/-885)
po/nl.po (+279/-876)
po/nn.po (+0/-846)
po/oc.po (+263/-864)
po/onboard.pot (+345/-569)
po/pl.po (+253/-882)
po/pms.po (+226/-722)
po/pt.po (+263/-833)
po/pt_BR.po (+277/-872)
po/ro.po (+270/-806)
po/ru.po (+260/-858)
po/si.po (+0/-846)
po/sk.po (+267/-802)
po/sl.po (+253/-880)
po/sn.po (+226/-722)
po/sq.po (+263/-862)
po/sr.po (+268/-874)
po/sv.po (+263/-859)
po/ta.po (+0/-861)
po/te.po (+0/-845)
po/th.po (+239/-766)
po/tl.po (+226/-722)
po/tr.po (+267/-862)
po/ug.po (+262/-860)
po/uk.po (+261/-799)
po/vi.po (+264/-800)
po/zh_CN.po (+262/-800)
po/zh_HK.po (+260/-856)
po/zh_TW.po (+259/-855)
prediction/gpredict (+456/-0)
prediction/makemodels (+233/-0)
prediction/pypredict/Makefile (+31/-0)
prediction/pypredict/README (+119/-0)
prediction/pypredict/__init__.py (+2/-0)
prediction/pypredict/analyze (+337/-0)
prediction/pypredict/entropy (+62/-0)
prediction/pypredict/ksr (+70/-0)
prediction/pypredict/lm.cpp (+409/-0)
prediction/pypredict/lm.h (+263/-0)
prediction/pypredict/lm_dynamic.cpp (+48/-0)
prediction/pypredict/lm_dynamic.h (+756/-0)
prediction/pypredict/lm_dynamic_cached.h (+471/-0)
prediction/pypredict/lm_dynamic_impl.h (+935/-0)
prediction/pypredict/lm_dynamic_kn.h (+393/-0)
prediction/pypredict/lm_merged.cpp (+223/-0)
prediction/pypredict/lm_merged.h (+130/-0)
prediction/pypredict/lm_python.cpp (+1781/-0)
prediction/pypredict/ngram-test (+252/-0)
prediction/pypredict/optimize (+217/-0)
prediction/pypredict/pool_allocator.cpp (+377/-0)
prediction/pypredict/predict (+98/-0)
prediction/pypredict/pypredict.py (+357/-0)
prediction/pypredict/setup.py (+21/-0)
prediction/pypredict/split_corpus (+72/-0)
prediction/pypredict/test_pypredict.py (+295/-0)
prediction/pypredict/train (+66/-0)
prediction/test-client (+44/-0)
setup.py (+1/-0)
Related bugs
Related blueprints
Branch information
Recent revisions
- 516. By marmuta
-
Implement short term memory to bridge the gap between typing and learning.
Text changes are periodically transferred to a new non-persistent language
model. Needs more real world testing to determine if performance is sufficient. - 515. By marmuta
-
Remove text classifier and dependency to libexttextcat. There's no time for language detection and the classifier uselessly takes up CPU cycles and memory.
- 513. By marmuta
-
Experimental domain specific prediction at text begin.
Prepend begin-of-text markers to prediction context and learned text.
The first <order-1> tokens of every text now favor n-grams local to
the current text-domain. Over time, generic text entries, terminal and
firefox url-bar should show different suggestions.
Branch metadata
- Branch format:
- Branch format 7
- Repository format:
- Bazaar repository format 2a (needs bzr 1.16 or later)