Easier contraction typing (#289)

* removed isPunctiationPartOfWords hack and created a new contraction input method that allows typing just anything, instead of predefined list coming from the dictionary

* updated the common compound words and contractions in Bulgarian, Dutch, English and French

* removed some non-sense and rarely used English words

* fixed crashing when trying to find words with apostrophes in the database

* fixed a crash when trying to capitalize single character strings

* improved dictionary validation at build time: spaces are now disallowed

CONTRIBUTING.md

@@ -54,11 +54,9 @@ To support a new language one needs to:
   - `locale` contains the language and the country codes (e.g. "en-US", "es-AR", "it-IT"). Refer to the list of [supported locales in Java](https://www.oracle.com/java/technologies/javase/jdk8-jre8-suported-locales.html#util-text).
   - `dictionaryFile` is the name of the dictionary in `assets/` folder.
   - `characterMap` contains the letters and punctuation marks associated with each key.
-  - Set `isPunctuationPartOfWords` to `true`, if the dictionary contains words with apostrophes or dashes, such as: `it's`, `you'll`, `a'tje` or `п'ят`. This will allow using 1-key for typing them (they will appear as suggestions). `false` will enable faster typing when apostrophes or other punctuation are not part of the words (no such words will be suggested).
   - `abcString` _(optional)_. A custom string to display in ABC mode. By default, the first three letters on 2-key are used (e.g. "ABC" or "АБВ"). Set this if the first letters of the alphabet are _not_ on 2-key, like in Hebrew, or if a different string makes more sense.
-  - `hasUpperCase` _(optional)_ set to `false` when the language has no upper- and lowercase letters. For example: Arabic, Hebrew, East Asian
+  - `hasUpperCase` _(optional)_ set to `false` when the language has no upper- and lowercase letters. For example: Arabic, Hebrew, East Asian languages, and so on. The default is `true`.
-    languages, and so on. The default is `true`.
+  - `name` _(optional)_ is automatically generated and equals the native name of the language (e.g. "English", "Deutsch", "Українська"). However, sometimes, the automatically selected name may be ambiguous. For example, both Portuguese in Portugal and Brazil will default to "Português", so assigning "Português brasileiro" would make it clear it's the language used in Brazil.
-  - `name` _(optional)_ defaults to the native name of the language (e.g. "English", "Deutsch", "Українська"). Useful to set when the default is ambiguous. For example, both Portuguese in Portugal and Brazil will default to "Português", so assigning "Português brasileiro" would make it clear it's the language used in Brazil.
 - Finally, add the new language to the list in `LanguageCollection.java`. You only need to add it in one place, in the constructor. Please, be nice and maintain the alphabetical order.
 
 

assets/bg-utf8.csv

@@ -490,6 +490,7 @@
 пия	170
 плу
 плю
+по-
 под	204
 пое
 пои
@@ -2246,6 +2247,7 @@
 наем	167
 нает	137
 наех
+най-
 назе
 нали	184
 нане	90

assets/fr-utf8.csv

@@ -1,3 +1,26 @@
+c'
+ç'
+d'
+j'
+l'
+m'
+n'
+s'
+t'
+y'
+qu'
+-ce
+-ci
+-il
+-la
+-toi
+-elle
+-nous
+-vous
+jusqu'
+lorsqu'
+puisqu'
+quoiqu'
 abaissa	59
 abaissable
 abaissables

assets/nl-utf8.csv

@@ -1,14 +1,19 @@
-'s-Graveland	94
+Gravendeel	95
-'s-Gravendeel	95
+Gravenhaags
-'s-Gravenhaags
+Gravenhage	129
-'s-Gravenhage	129
+Gravenhagenaar
-'s-Gravenhagenaar
+Gravenmoer
-'s-Gravenmoer
+Gravenzande	103
-'s-Gravenzande	103
+Gravenzander
-'s-Gravenzander
+Gravenzands
-'s-Gravenzands
+Hertogenbosch	129
-'s-Hertogenbosch	129
+'k
-'t	180
+'s
+'t
+'s-
+m'n
+d'r
+da's
 A-attest
 A-attesten
 A-biljet
@@ -23,7 +28,6 @@ A-omroep
 A-omroepen
 A-selectie	92
 A-status	90
-A.D.
 AAA	109
 AAW	96
 AAW-voorziening
@@ -1928,7 +1932,6 @@ ayses
 aytekin
 ayten
 aytens
-az.
 azad
 azads
 azarkan
@@ -5025,7 +5028,6 @@ chilenen	91
 chili	110
 chimène
 chimènes
-chin.
 china	163
 china's	104
 chinees	119
@@ -5719,7 +5721,6 @@ d-day	98
 d-sluiting
 d-trein	80
 d-treinen
-d.v.
 ddt	99
 deet	99
 dna	144
@@ -10520,8 +10521,6 @@ gül	85
 güls
 h-bom
 h-bommen
-h.k.h.
-h.m.
 hrm	122
 htm	111
 html	127
@@ -15930,7 +15929,6 @@ l-profiel
 l-profielen
 l-vormig	86
 l-vormige	104
-l.b.
 lan	99
 lgpl
 lto	75
@@ -19174,7 +19172,6 @@ mexx'
 meyer	123
 meyerode
 meys	75
-mgr.
 mia	90
 mia's	72
 mica	97
@@ -19962,12 +19959,6 @@ mälarmeer
 mélanie	73
 mélanies
 münchen	135
-n.b.
-n.h.
-n.n.
-n.o.
-n.v.
-n.w.
 nam	121
 nap	95
 nasa	125
@@ -20249,7 +20240,6 @@ neals
 nebahat	81
 nebahats
 nebbeling
-ned.
 Neder-Betuwe	90
 Neder-Californië
 Neder-Congo
@@ -21093,7 +21083,6 @@ nyssa
 nyssa's
 nürnberg	102
 o-benen	78
-o.l.v.
 ocmw	132
 ocmw's	108
 OCMW-raad	88
@@ -21525,7 +21514,6 @@ opdorp	81
 opel	132
 opels	84
 opendocument
-openoffice.org
 opentaal	113
 opgeldenaken
 opgenoort
@@ -23029,7 +23017,6 @@ quitoos
 quitoër
 quix	72
 qureshi
-r.i.p.
 ram	115
 rdw	124
 ret	117
@@ -26750,8 +26737,6 @@ srananmans
 sranantongo	88
 srebrenica	117
 sri	78
-st.-eustatius
-st.-maarten
 staaks
 staals
 staarink
@@ -26847,7 +26832,6 @@ stavast	80
 stavele	76
 stavenga
 stavleu
-stct.
 steef	85
 steefs
 steegers
@@ -29997,8 +29981,6 @@ vänermeer
 vättermeer
 véronique	99
 véroniques
-w.
-w.v.str.
 wa	115
 WA-verzekering	93
 WAM-sticker
@@ -31611,8 +31593,6 @@ z-verpleegkundige
 z-verpleegkundigen
 z-vormig
 z-vormige
-z.h.
-z.k.h.
 zaaijer	82
 zaan	89
 zaandam	132
@@ -32158,13 +32138,6 @@ a-priorisch	79
 a-priorische
 a-prioristisch	79
 a-prioristische	81
-a.d.h.v.
-a.g.v.
-a.h.w.
-a.j.b.
-a.m.
-a.s.
-a.u.b.
 aagt	84
 aagtappel	80
 aagtappels	79
@@ -35328,7 +35301,6 @@ aanvuring	85
 aanvuurde	84
 aanvuurden	82
 aanvuurt	83
-aanw.
 aanwaai	80
 aanwaaide	80
 aanwaaiden	78
@@ -38653,7 +38625,6 @@ afasiepatiënten
 afat
 afatisch	80
 afatische	84
-afb.
 afbad	77
 afbak	85
 afbakbrood
@@ -38907,7 +38878,6 @@ afbuigmagneet	81
 afbuigt	94
 afcentiemen
 afchecken
-afd.
 afdaal	84
 afdaalde	95
 afdaalden	92
@@ -43068,7 +43038,6 @@ afwrijft	85
 afwrijven	92
 afwrijvingen	77
 afwringen	82
-afz.
 afzaag	84
 afzaagde	82
 afzaagden	84
@@ -45222,7 +45191,6 @@ amuzikaal
 amygdala	99
 amylose	72
 amyotrofie	80
-an.
 anaal	121
 anabaptisme	81
 anabaptist	81
@@ -47864,7 +47832,6 @@ aronskelken	83
 aroom	82
 arpeggio	89
 arpeggio's	81
-arr.
 arrangeer	82
 arrangeerde	93
 arrangeerden	82
@@ -50164,9 +50131,6 @@ aïssen	73
 b's
 b'tje
 b'tjes
-b.d.
-b.g.g.
-b.v.d.
 baad	88
 baadde	96
 baadden	89
@@ -53282,7 +53246,6 @@ bazuintje	82
 bazuintjes	81
 bbp	121
 bbp-groei	81
-bc.
 beaam	90
 beaamd	98
 beaamde	104
@@ -61835,7 +61798,6 @@ betrouwde	84
 betrouwden	81
 betrouwen	99
 betrouwt	83
-bett.
 betuig	99
 betuigd	111
 betuigde	106
@@ -64941,7 +64903,6 @@ bijkrijgt	86
 bijkwam	98
 bijkwamen	91
 bijl	129
-bijl.
 bijladen	88
 bijlag	87
 bijlage	157
@@ -65325,7 +65286,6 @@ bijtten	79
 bijtwond	77
 bijtwonden	82
 bijtwondje
-bijv.
 bijvak	95
 bijvakken	90
 bijvakprogramma	80
@@ -68420,7 +68380,6 @@ bluts	87
 blutsen	89
 blutst	81
 blutste	76
-blz.
 blèrde	82
 blèrden	81
 blèren	97
@@ -76207,10 +76166,8 @@ buy-outs
 buzzer	85
 buzzers	75
 bv'tjes
-bv.
 bvba	119
 bvba's
-bw.
 bye	108
 byebye
 byes
@@ -76249,9 +76206,6 @@ c'tjes
 c-pion
 c-sleutel
 c-sleutels
-c.q.
-c.s.
-ca.
 caban	82
 cabans	83
 cabaret	123
@@ -85268,10 +85222,6 @@ d's	84
 d'tje
 d'tjes
 d-pion
-d.d.
-d.i.
-d.m.v.
-d.w.z.
 db	132
 daad	149
 daadkracht	122
@@ -86782,7 +86732,6 @@ daze
 dazen	97
 dc	143
 dc's
-dd.
 de	255
 de-escalatie	77
 de-escalaties
@@ -89593,7 +89542,6 @@ dezulken	101
 deïsme	89
 deïst	86
 deïsten	86
-dhr.
 dia	123
 dia's	119
 dia-avond
@@ -92158,7 +92106,6 @@ dittografieën	77
 ditzelfde	118
 diuretica	93
 diureticum	93
-div.
 divalifeest
 divalioptocht
 divalioptochten
@@ -95906,9 +95853,7 @@ dozijnen	97
 dozijntje	82
 dozijntjes	82
 dpi	111
-dr.
 dra	115
-dra.
 draad	153
 draadachtig	83
 draadachtige	89
@@ -97578,8 +97523,6 @@ droste-effect
 drostendiensten	82
 drostschap	80
 drozen	79
-drs.
-drs.-titel
 drug	123
 drugbaron
 drugbeleid	93
@@ -98141,7 +98084,6 @@ druïdetempel
 dryade	92
 dryaden	84
 drycleaning
-ds.
 dtp	80
 dtp-prik
 dtp-prikken
@@ -99536,12 +99478,6 @@ e-pion
 e-snaar	78
 e-zine	98
 e-zines	77
-e.a.
-e.d.
-e.e.a.
-e.o.
-e.v.
-e.v.a.
 earlgreythee
 earnings
 eau-de-colognefles	80
@@ -103004,7 +102940,6 @@ enveloppenfinanciering
 enveloppetje
 environment	97
 environments	72
-enz.
 enzoverder	93
 enzovoort	144
 enzovoorts	133
@@ -103963,7 +103898,6 @@ etappezege	94
 etappezeges	76
 etatistisch
 etatistische
-etc.
 eten	181
 etend	97
 etende	101
@@ -104600,7 +104534,6 @@ evolutionistisch	83
 evolutionistische	82
 evoqueerde
 evoqueren	80
-evt.
 ex	149
 ex-kamerleden
 ex-kamerlid
@@ -104956,7 +104889,6 @@ excitatie	92
 excitaties
 exciteerde
 exciteren	83
-excl.
 exclamatie	79
 exclamaties	81
 exclave	91
@@ -105784,7 +105716,6 @@ f-sleutel	81
 f-sleutels	76
 fa	123
 fa's
-fa.
 faal	98
 faalangst	117
 faalangsten
@@ -106188,7 +106119,6 @@ falsifieert	82
 falsifiëren	84
 falsifiëring	78
 falsifiëringen
-fam.
 fameus	98
 fameust	77
 fameuze	122
@@ -107751,7 +107681,6 @@ fietszoektocht	75
 fietszoektochten
 fifties	93
 fiftyfifty	85
-fig.
 figaro	83
 figaro's
 figgelen	78
@@ -110574,7 +110503,6 @@ foxtrotte	80
 foxtrotten	80
 foyer	117
 foyers	93
-fr.
 fraai	148
 fraaie	152
 fraaier	107
@@ -111851,7 +111779,6 @@ g'tjes
 g-sleutel	81
 g-sleutels
 g-spot	76
-g.g.d.
 ga	183
 gaaf	141
 gaafheid	102
@@ -113444,7 +113371,6 @@ geavanceerder	101
 geavanceerdere	92
 geavondmaald	77
 geavontuurd	82
-geb.
 gebaad	98
 gebaald	81
 gebaand	103
@@ -119363,7 +119289,6 @@ geluwd	97
 geluwde	83
 gelyncht	97
 gelynchte	82
-gem.
 gemaaid	122
 gemaaide	100
 gemaakt	193
@@ -123739,7 +123664,6 @@ gesynthetiseerd	97
 gesynthetiseerde	86
 gesystematiseerd	87
 gesystematiseerde	85
-get.
 getaakt	81
 getaald	82
 getaand	89
@@ -144284,19 +144208,6 @@ i-bankieren
 i-grec	81
 i-grecs
 i-mode	83
-i.c.
-i.c.m.
-i.e.
-i.h.a.
-i.h.b.
-i.m.
-i.o.
-i.o.v.
-i.p.v.
-i.s.m.
-i.t.t.
-i.v.m.
-i.z.g.st.
 ipod	130
 ipods	95
 ia	133
@@ -145935,7 +145846,6 @@ inciviek	81
 incivieke
 incivieken
 incivisme	80
-incl.
 inclinatie	90
 inclinatiekompas	74
 inclinaties	82
@@ -147169,7 +147079,6 @@ infuuspompen	78
 infuuspompjes
 infuusvloeistof
 infuusvloeistoffen
-ing.
 ingaan	146
 ingaand	97
 ingaande	116
@@ -152012,7 +151921,6 @@ ionosferisch
 ionosferische
 ippon	94
 ippons
-ir.
 irenisch	84
 irenische	81
 iridium	90
@@ -153492,7 +153400,6 @@ jezuïetenstijl	74
 jezuïtisch	78
 jezuïtische	83
 jezuïtisme	77
-jhr.
 jicht	114
 jichtaanval	85
 jichtaanvallen	83
@@ -153528,8 +153435,6 @@ jip-en-janneketaal
 jippie	89
 jiujitsu	90
 jive	89
-jkvr.
-jl.
 jobaanbieding
 jobaanbiedingen	79
 jobaanbod
@@ -154017,7 +153922,6 @@ joyriders
 joyriding	88
 joystick	108
 joysticks	90
-jr.
 jubee	82
 jubel	97
 jubelde	96
@@ -154414,7 +154318,6 @@ juxtapositie	87
 k's	74
 k'tje
 k'tjes
-k.k.
 khz	114
 kj	94
 kw	131
@@ -178201,7 +178104,6 @@ librettoschrijver
 librije	86
 librijen	78
 librium	81
-lic.
 licentiaat	109
 licentiaatsdiploma
 licentiaatsdiploma's
@@ -183132,16 +183034,6 @@ m'etjes
 m's
 m-bankierde
 m-bankieren
-m.
-m.a.w.
-m.b.t.
-m.b.v.
-m.i.
-m.i.v.
-m.m.
-m.m.v.
-m.n.
-m.u.v.
 m/s
 ma's	92
 maag	150
@@ -186676,7 +186568,6 @@ mavoleerlingen
 mavoniveau
 mavoscholen
 mavoschool
-max.
 maxi	111
 maximaal	166
 maximale	158
@@ -190491,7 +190382,6 @@ meurt	77
 meute	118
 meuten	84
 meutes	86
-mevr.
 mevrouw	162
 mevrouwde	78
 mevrouwen	93
@@ -191969,7 +191859,6 @@ mimosa	91
 mimosa's
 min-maxcontract
 min-maxcontracten
-min.
 minacht	94
 minachten	98
 minachtend	106
@@ -193069,9 +192958,7 @@ mixtures
 mixtuur	83
 mkb	119
 ml	146
-mld.
 mlk-school
-mln.
 mm	167
 mnemotechniek	78
 mnemotechnisch	83
@@ -195711,7 +195598,6 @@ mozetta's
 mozzarella	116
 mp	98
 mp's
-mr.
 ms	120
 mts	88
 mu	111
@@ -196751,7 +196637,6 @@ muzikantesk	78
 muzikanteske
 muzisch	90
 muzische	101
-mw.
 mycelium	97
 mycologen	85
 mycologie	84
@@ -196829,9 +196714,6 @@ n'en
 n'etje	75
 n'etjes
 n's	72
-n.a.v.
-n.o.t.k.
-n.v.t.
 na	119
 na-aapt
 na-apen	93
@@ -201545,8 +201427,6 @@ niëlleerde	74
 niëlleren	78
 njet	91
 njonja	80
-nl.
-nl.openoffice.org
 nm	121
 no-claim	97
 no-claimkorting	92
@@ -201561,7 +201441,6 @@ no-nonsense	111
 no-nonsensebeleid
 no-nonsensepolitiek
 no-showstudent
-no.
 noachitisch
 noachitische
 nobele	120
@@ -202435,8 +202314,6 @@ novum	100
 nozem	91
 nozempje
 nozems	92
-nr.
-nrs.
 nu	138
 nuance	125
 nuanceer	83
@@ -202657,14 +202534,6 @@ nymfomanie	82
 o's	85
 o'tje
 o'tjes
-o.a.
-o.b.v.
-o.i.
-o.i.d.
-o.m.
-o.t.t.
-o.v.t.
-o.v.v.
 oase	126
 oasen	90
 oases	100
@@ -208408,7 +208277,6 @@ onfrisser	75
 onfrist	78
 onfunctioneel
 onfunctionele
-ong.
 ongaar
 ongaarne	99
 ongans	88
@@ -221699,12 +221567,6 @@ oölieten	81
 p's	78
 p'tje
 p'tjes
-p.
-p.a.
-p.m.
-p.o.
-p.p.
-p.w.
 ph	129
 ph's
 ph-neutraal
@@ -222284,7 +222146,6 @@ pafte	84
 paften	80
 pafzak	76
 pafzakken	75
-pag.
 pagaai	85
 pagaaide	81
 pagaaiden	77
@@ -230065,7 +229926,6 @@ plissérok
 plissérokken
 plissés	84
 plistoceen
-plm.
 ploeg	158
 ploegachtervolging
 ploegarts	72
@@ -230503,7 +230363,6 @@ pluvier	80
 pluvieren	81
 pluviometer	84
 pluviometers	82
-plv.
 pneumatiek	92
 pneumatisch	106
 pneumatische	117
@@ -236150,7 +236009,6 @@ proeverij	118
 proeverijen	106
 proevers	94
 prof	145
-prof.
 profaan	93
 profaanst
 profafdeling
@@ -238598,8 +238456,6 @@ pythons	90
 q's
 q'tje
 q'tjes
-q.e.d.
-q.q.
 qat	94
 qua	162
 quad	110
@@ -238768,7 +238624,6 @@ r'en
 r'etje
 r'etjes
 r's
-r.-k.
 ra	122
 ra's	88
 raad	168
@@ -241970,7 +241825,6 @@ recyclingbedrijf	86
 recyclingbedrijven	73
 recyclingproces
 red	141
-red.
 redacteur	138
 redacteuren	122
 redacteurs	105
@@ -245250,7 +245104,6 @@ resort	130
 resorts	112
 resource	107
 resources	118
-resp.
 respect	158
 respectabel	105
 respectabele	115
@@ -251073,8 +250926,6 @@ s'en	72
 s'je
 s'jes
 s's
-s.j.
-s.v.p.
 sa	121
 saai	142
 saaie	133
@@ -269364,7 +269215,6 @@ squasht	80
 squashte
 squaw	82
 squaws	81
-sr.
 sta	156
 sta-in-de-weg	94
 sta-in-de-wegs	80
@@ -279562,16 +279412,6 @@ t'tje
 t'tjes
 t-groep
 t-groepen
-t.a.v.
-t.b.v.
-t.g.v.
-t.h.t.
-t.h.v.
-t.n.v.
-t.o.v.
-t.w.
-t.w.v.
-t.z.t.
 taai	124
 taaie	121
 taaien	87
@@ -298781,16 +298621,6 @@ uzi's	81
 v's
 v'tje
 v'tjes
-v.
-v.chr.
-v.d.
-v.h.
-v.l.n.r.
-v.r.n.l.
-v.t.t.
-v.v.
-v.v.t.
-v.w.b.
 va	138
 va-banque	75
 va-et-vient	82
@@ -304041,7 +303871,6 @@ verfwerken
 verfwinkel	89
 verfwinkels	82
 verg	104
-verg.
 verga	109
 vergaan	132
 vergaand	108
@@ -312173,7 +312002,6 @@ vezelvlies
 vezelvliezen
 vezelwortels	77
 vezen	81
-vgl.
 via	193
 viaduct	126
 viaducten	111
@@ -315274,8 +315102,6 @@ vmbo-leerling
 vmbo-leerlingen	91
 vmbo-scholen	84
 vmbo-school	80
-vnl.
-vnw.
 vocaal	114
 vocabulaire	112
 vocabulaires	87
@@ -319604,7 +319430,6 @@ voorwoord	127
 voorwoorden	87
 voorworp	79
 voorworpen	77
-voorz.
 voorzaal	86
 voorzag	130
 voorzagen	111
@@ -321752,7 +321577,6 @@ vruchtzak	79
 vruchtzakje	74
 vruchtzakken
 vruchtzetting	94
-vs.
 vuig	89
 vuige	100
 vuiger	81
@@ -322255,8 +322079,6 @@ véél	132
 w's
 w'tje
 w'tjes
-w.o.
-w.v.t.t.k.
 waad	88
 waadbaar	78
 waadbare	80
@@ -333505,7 +333327,6 @@ wurmen	108
 wurmpje	83
 wurmpjes	81
 wurmt	92
-ww.
 www	174
 wyandotte
 wybertje	81
@@ -333633,11 +333454,6 @@ z'n	171
 z's
 z'tje
 z'tjes
-z.g.
-z.g.a.n.
-z.i.
-z.o.z.
-z.s.m.
 zaad	146
 zaadbakje	81
 zaadbakjes	81
@@ -336652,7 +336468,6 @@ zeverden	80
 zeverlap	78
 zeverlappen	79
 zevert	84
-zgn.
 zich	210
 zicht	162
 zichtas	77

build.gradle

@@ -86,29 +86,6 @@ def getReleaseVersion = { ->
 	return "${getVersionName()} (${getCurrentGitHash()})"
 }
 
-def isPunctuationInWordsAllowed (String dictionaryFile) {
-	boolean isAllowed = false
-
-	file("${project.projectDir}/src/io/github/sspanak/tt9/languages/definitions").listFiles().each { file ->
-		boolean isTheDefinitionFile = false
-		file.eachLine {line ->
-			if (line.contains(dictionaryFile)) {
-				isTheDefinitionFile = true
-			}
-		}
-
-		if (isTheDefinitionFile) {
-			file.eachLine {line ->
-				if (line.matches(".+?isPunctuationPartOfWords\\s*=\\s*true.+?")) {
-					isAllowed = true
-				}
-			}
-		}
-	}
-
-	return isAllowed
-}
-
 task validateDictionaries {
 	inputs.dir fileTree(dir:'assets', excludes:['dict.properties'])
 	outputs.file "${project.buildDir}/dict.validation.txt"
@@ -116,7 +93,6 @@ task validateDictionaries {
 	doLast {
 		final String CSV_DELIMITER = '	' // TAB
 		final GEOGRAPHICAL_NAME = ~"[A-Z]\\w+-[^\\n]+"
-		final PUNCTUATION_CHARS = ~".*?\\p{Punct}(?<!-).*?"
 
 		final MAX_ERRORS = 50
 		String errors = ""
@@ -131,7 +107,6 @@ task validateDictionaries {
 
 			println "Validating dictionary: " + file.name
 
-			def isPunctuationAllowed = isPunctuationInWordsAllowed(file.name)
 			def uniqueWords = [:]
 
 			int lineNumber = 0
@@ -151,6 +126,13 @@ task validateDictionaries {
 					return
 				}
 
+				if (line.contains(" ")) {
+					isFileValid = false
+					errorCount++
+					errors += "Dictionary '" + file.name + "' is invalid. Found space on line " + lineNumber + ". Make sure each word is on a new line. Phrases are not allowed.\n"
+					return
+				}
+
 				String[] parts = line.split(CSV_DELIMITER, 2)
 				String word = parts[0]
 				String frequency = parts.length > 1 ? parts[1] : ""
@@ -179,12 +161,6 @@ task validateDictionaries {
 					errors += "Dictionary '" + file.name + "' is invalid. Found a single letter: '" + word + "' on line " + lineNumber + ". Only uppercase single letters are allowed. The rest of the alphabet will be added automatically.\n"
 				}
 
-				if (!isPunctuationAllowed && word.matches(PUNCTUATION_CHARS)) {
-					isFileValid = false
-					errorCount++
-					errors += "Dictionary '" + file.name + "' is invalid. Found a punctuation mark in word: '" + word + "' on line " + lineNumber + ". Remove all punctuation characters when the language definition disallows them or update the definition.\n"
-				}
-
 				String uniqueWordKey = word ==~ GEOGRAPHICAL_NAME ? word : word.toLowerCase()
 				if (uniqueWords[uniqueWordKey] != null && uniqueWords[uniqueWordKey] == true) {
 					isFileValid = false
@@ -235,21 +211,12 @@ android {
 
 	defaultConfig {
 		minSdkVersion 19
+		//noinspection ExpiredTargetSdkVersion
 		targetSdk 30
 		versionCode getVersionCode()
 		versionName getVersionName()
 	}
 
-	// http://stackoverflow.com/a/19130098
-//	signingConfigs {
-//		release {
-//			storeFile file(System.getenv("KEYSTORE"))
-//			storePassword System.getenv("KEYSTORE_PASS")
-//			keyAlias System.getenv("KEY_ALIAS")
-//			keyPassword System.getenv("KEY_ALIAS_PASS")
-//		}
-//	}
-//
 	buildTypes {
 		debug { data ->
 			data.buildConfigField 'String', 'VERSION_FULL', "\"${getDebugVersion()}\""

src/io/github/sspanak/tt9/TextTools.java

@@ -0,0 +1,43 @@
+package io.github.sspanak.tt9;
+
+import java.util.regex.Pattern;
+
+public class TextTools {
+	private static final Pattern containsOtherThan1 = Pattern.compile("[02-9]");
+	private static final Pattern previousIsLetter = Pattern.compile("\\p{L}$");
+	private static final Pattern nextIsPunctuation = Pattern.compile("^\\p{Punct}");
+	private static final Pattern nextToWord = Pattern.compile("\\b$");
+	private static final Pattern startOfSentence = Pattern.compile("(?<!\\.)(^|[.?!¿¡])\\s*$");
+
+	public static boolean containsOtherThan1(String str) {
+		return str != null && containsOtherThan1.matcher(str).find();
+	}
+
+	public static boolean isNextToWord(String str) {
+		return str != null && nextToWord.matcher(str).find();
+	}
+
+	public static boolean isStartOfSentence(String str) {
+		return str != null && startOfSentence.matcher(str).find();
+	}
+
+	public static boolean nextIsPunctuation(String str) {
+		return str != null && nextIsPunctuation.matcher(str).find();
+	}
+
+	public static boolean previousIsLetter(String str) {
+		return str != null && previousIsLetter.matcher(str).find();
+	}
+
+	public static boolean startsWithWhitespace(String str) {
+		return str != null && !str.isEmpty() && (str.charAt(0) == ' ' || str.charAt(0) == '\n' || str.charAt(0) == '\t');
+	}
+
+	public static boolean startsWithNumber(String str) {
+		return str != null && !str.isEmpty() && (str.charAt(0) >= '0' && str.charAt(0) <= '9');
+	}
+
+	public static String removeNonLetters(String str) {
+		return str != null ? str.replaceAll("\\P{L}", "") : null;
+	}
+}

src/io/github/sspanak/tt9/db/DictionaryDb.java

@@ -191,24 +191,13 @@ public class DictionaryDb {
 
 				// In case the user has changed the text case, there would be no match.
 				// Try again with the lowercase equivalent.
-				String lowercaseWord = "";
 				if (affectedRows == 0) {
-					lowercaseWord = word.toLowerCase(language.getLocale());
+					String lowercaseWord = word.toLowerCase(language.getLocale());
 					affectedRows = getInstance().wordsDao().incrementFrequency(language.getId(), lowercaseWord, sequence);
 
 					Logger.d("incrementWordFrequency", "Attempting to increment frequency for lowercase variant: " + lowercaseWord);
 				}
 
-				// Some languages permit appending the punctuation to the end of the words, like so: "try,".
-				// But there are no such words in the dictionary, so try without the punctuation mark.
-				if (affectedRows == 0 && language.isPunctuationPartOfWords() && sequence.endsWith("1")) {
-					String truncatedWord = lowercaseWord.substring(0, word.length() - 1);
-					String truncatedSequence = sequence.substring(0, sequence.length() - 1);
-					affectedRows = getInstance().wordsDao().incrementFrequency(language.getId(), truncatedWord, truncatedSequence);
-
-					Logger.d("incrementWordFrequency", "Attempting to increment frequency with stripped punctuation: " + truncatedWord);
-				}
-
 				Logger.d("incrementWordFrequency", "Affected rows: " + affectedRows);
 			} catch (Exception e) {
 				Logger.e(

src/io/github/sspanak/tt9/db/migrations/DB11.java

@@ -0,0 +1,39 @@
+package io.github.sspanak.tt9.db.migrations;
+
+import androidx.annotation.NonNull;
+import androidx.room.migration.Migration;
+import androidx.sqlite.db.SupportSQLiteDatabase;
+
+import io.github.sspanak.tt9.Logger;
+import io.github.sspanak.tt9.languages.definitions.Dutch;
+import io.github.sspanak.tt9.languages.definitions.English;
+
+public class DB11 {
+	public static final Migration MIGRATION = new Migration(10, 11) {
+		@Override
+		public void migrate(@NonNull SupportSQLiteDatabase database) {
+			final String enWords = "'I''d','I''m','d''annunzio','I''ll','I''ve','prud''hon','an''t','bo''s''n','bo''s''ns','bo''sun','bo''suns','bos''n','bos''ns','br''er','ca''canny','could''ve','d''arezzo','d''estaing','e''en','e''er','fo''c''s''le','fo''c''s''les','fo''c''sle','fo''c''sles','ha''penny','he''d','he''ll','how''d','how''re','howe''er','it''d','it''ll','might''ve','must''ve','n''importe','ne''er','nor''easter','nor''wester','o''er','rec''d','sec''y','she''d','she''ll','should''ve','sou''wester','ta''en','that''d','that''ll','they''d','they''ll','they''re','they''ve','we''d','we''ll','we''re','we''ve','whate''er','whatsoe''er','whene''er','where''er','who''d','who''ll','who''re','who''ve','why''d','would''ve','you''d','you''ll','you''re','you''ve','Ch''in','L''Amour','L''Enfant','L''Oreal','L''Ouverture','T''ang','Xi''an'";
+			final String nlWords = "'''s-Graveland','''s-Gravendeel','''s-Gravenhaags','''s-Gravenhage','''s-Gravenhagenaar','''s-Gravenmoer','''s-Gravenzande','''s-Gravenzander','''s-Gravenzands','''s-Hertogenbosch','''t','A.D.','az.','chin.','d.v.','h.k.h.','h.m.','l.b.','mgr.','n.b.','n.h.','n.n.','n.o.','n.v.','n.w.','ned.','o.l.v.','openoffice.org','r.i.p.','st.-eustatius','st.-maarten','stct.','w.','w.v.str.','z.h.','z.k.h.','a.d.h.v.','a.g.v.','a.h.w.','a.j.b.','a.m.','a.s.','a.u.b.','aanw.','afb.','afd.','afz.','an.','arr.','b.d.','b.g.g.','b.v.d.','bc.','bett.','bijl.','bijv.','blz.','bv.','bw.','c.q.','c.s.','ca.','d.d.','d.i.','d.m.v.','d.w.z.','dd.','dhr.','div.','dr.','dra.','drs.','drs.-titel','ds.','e.a.','e.d.','e.e.a.','e.o.','e.v.','e.v.a.','enz.','etc.','evt.','excl.','fa.','fam.','fig.','fr.','g.g.d.','geb.','gem.','get.','i.c.','i.c.m.','i.e.','i.h.a.','i.h.b.','i.m.','i.o.','i.o.v.','i.p.v.','i.s.m.','i.t.t.','i.v.m.','i.z.g.st.','incl.','ing.','ir.','jhr.','jkvr.','jl.','jr.','k.k.','lic.','m.','m.a.w.','m.b.t.','m.b.v.','m.i.','m.i.v.','m.m.','m.m.v.','m.n.','m.u.v.','max.','mevr.','min.','mld.','mln.','mr.','mw.','n.a.v.','n.o.t.k.','n.v.t.','nl.','nl.openoffice.org','no.','nr.','nrs.','o.a.','o.b.v.','o.i.','o.i.d.','o.m.','o.t.t.','o.v.t.','o.v.v.','ong.','p.','p.a.','p.m.','p.o.','p.p.','p.w.','pag.','plm.','plv.','prof.','q.e.d.','q.q.','r.-k.','red.','resp.','s.j.','s.v.p.','sr.','t.a.v.','t.b.v.','t.g.v.','t.h.t.','t.h.v.','t.n.v.','t.o.v.','t.w.','t.w.v.','t.z.t.','v.','v.chr.','v.d.','v.h.','v.l.n.r.','v.r.n.l.','v.t.t.','v.v.','v.v.t.','v.w.b.','verg.','vgl.','vnl.','vnw.','voorz.','vs.','w.o.','w.v.t.t.k.','ww.','z.g.','z.g.a.n.','z.i.','z.o.z.','z.s.m.','zgn.'";
+
+			try {
+				database.beginTransaction();
+				database.execSQL(getDeleteEnglishSwordsQuery());
+				database.execSQL(getDeleteWordsQuery(new English().getId(), enWords));
+				database.execSQL(getDeleteWordsQuery(new Dutch().getId(), nlWords));
+				database.setTransactionSuccessful();
+			} catch (Exception e) {
+				Logger.e("Migrate to DB11", "Migration failed. " + e.getMessage());
+			} finally {
+				database.endTransaction();
+			}
+		}
+	};
+
+	private static String getDeleteEnglishSwordsQuery() {
+		return "DELETE FROM words WHERE lang=" + new English().getId() + " AND word LIKE '%''s'";
+	}
+
+	private static String getDeleteWordsQuery(int langId, String wordList) {
+		return "DELETE FROM words WHERE lang=" + langId + " AND word IN(" + wordList + ")";
+	}
+}

src/io/github/sspanak/tt9/db/room/TT9Room.java

@@ -8,12 +8,13 @@ import androidx.room.RoomDatabase;
 import androidx.sqlite.db.SimpleSQLiteQuery;
 
 import io.github.sspanak.tt9.db.migrations.DB10;
+import io.github.sspanak.tt9.db.migrations.DB11;
 import io.github.sspanak.tt9.db.migrations.DB6;
 import io.github.sspanak.tt9.db.migrations.DB7;
 import io.github.sspanak.tt9.db.migrations.DB8;
 import io.github.sspanak.tt9.db.migrations.DB9;
 
-@Database(version = 10, entities = Word.class, exportSchema = false)
+@Database(version = 11, entities = Word.class, exportSchema = false)
 public abstract class TT9Room extends RoomDatabase {
 	public abstract WordsDao wordsDao();
 
@@ -25,7 +26,8 @@ public abstract class TT9Room extends RoomDatabase {
 				new DB7().getMigration(context),
 				DB8.MIGRATION,
 				DB9.MIGRATION,
-				DB10.MIGRATION
+				DB10.MIGRATION,
+				DB11.MIGRATION
 			)
 			.build();
 	}
@@ -41,7 +43,7 @@ public abstract class TT9Room extends RoomDatabase {
 			" LIMIT " + limit;
 
 		if (word != null) {
-			sql = sql.replace("WHERE 1", "WHERE 1 AND word LIKE '" + word + "%'");
+			sql = sql.replace("WHERE 1", "WHERE 1 AND word LIKE '" + word.replace("'", "''") + "%'");
 		}
 
 		return new SimpleSQLiteQuery(sql);

src/io/github/sspanak/tt9/db/room/WordsDao.java

@@ -18,7 +18,7 @@ public interface WordsDao {
 	@Query("SELECT COUNT(id) FROM words WHERE :langId < 0 OR lang = :langId")
 	int count(int langId);
 
-	@Query("DELETE FROM words WHERE LANG IN(:langIds)")
+	@Query("DELETE FROM words WHERE lang IN(:langIds)")
 	void deleteByLanguage(ArrayList<Integer> langIds);
 
 	@Query("SELECT COUNT(id) FROM words WHERE lang = :langId AND word = :word")

src/io/github/sspanak/tt9/ime/TraditionalT9.java

@@ -262,18 +262,16 @@ public class TraditionalT9 extends KeyPadHandler {
 		cancelAutoAccept();
 		forceShowWindowIfHidden();
 
-		String currentWord = getComposingText();
+		// Automatically accept the previous word, when the next one is a space or punctuation,
-
-		// Automatically accept the current word, when the next one is a space or punctuation,
 		// instead of requiring "OK" before that.
-		if (mInputMode.shouldAcceptCurrentSuggestion(key, hold, repeat > 0)) {
+		// First pass, analyze the incoming key press and decide whether it could be the start of
+		// a new word.
+		if (mInputMode.shouldAcceptPreviousSuggestion(key)) {
 			autoCorrectSpace(acceptIncompleteSuggestion(), false, key);
-			currentWord = "";
 		}
 
 		// Auto-adjust the text case before each word, if the InputMode supports it.
-		// We don't do it too often, because it is somewhat resource-intensive.
+		if (getComposingText().isEmpty()) {
-		if (currentWord.length() == 0) {
 			mInputMode.determineNextWordTextCase(textField.isThereText(), textField.getTextBeforeCursor());
 		}
 
@@ -543,11 +541,11 @@ public class TraditionalT9 extends KeyPadHandler {
 	}
 
 	private void commitCurrentSuggestion(boolean entireSuggestion) {
-		if (!isSuggestionViewHidden() && currentInputConnection != null) {
+		if (!isSuggestionViewHidden()) {
 			if (entireSuggestion) {
 				textField.setComposingText(suggestionBar.getCurrentSuggestion());
 			}
-			currentInputConnection.finishComposingText();
+			textField.finishComposingText();
 		}
 
 		setSuggestions(null);
@@ -556,11 +554,8 @@ public class TraditionalT9 extends KeyPadHandler {
 
 	private void clearSuggestions() {
 		setSuggestions(null);
-
+		textField.setComposingText("");
-		if (currentInputConnection != null) {
+		textField.finishComposingText();
-			textField.setComposingText("");
-			currentInputConnection.finishComposingText();
-		}
 	}
 
 
@@ -570,10 +565,22 @@ public class TraditionalT9 extends KeyPadHandler {
 
 
 	private void handleSuggestions() {
+		// Automatically accept the previous word, without requiring OK. This is similar to what
+		// Second pass, analyze the available suggestions and decide if combining them with the
+		// last key press makes up a compound word like: (it)'s, (I)'ve, l'(oiseau), or it is
+		// just the end of a sentence, like: "word." or "another?"
+		if (mInputMode.shouldAcceptPreviousSuggestion()) {
+			String lastComposingText = getComposingText(mInputMode.getSequenceLength() - 1);
+			commitCurrentSuggestion(false);
+			mInputMode.onAcceptSuggestion(lastComposingText, true);
+			autoCorrectSpace(lastComposingText, false, -1);
+			mInputMode.determineNextWordTextCase(textField.isThereText(), textField.getTextBeforeCursor());
+		}
+
 		// key code "suggestions" take priority over words
 		if (mInputMode.getKeyCode() > 0) {
 			sendDownUpKeyEvents(mInputMode.getKeyCode());
-			mInputMode.onAcceptSuggestion("");
+			mInputMode.reset();
 			return;
 		}
 
@@ -605,16 +612,27 @@ public class TraditionalT9 extends KeyPadHandler {
 	}
 
 
-	private String getComposingText() {
+	private String getComposingText(int maxLength) {
+		if (maxLength == 0) {
+			return "";
+		}
+
+		maxLength = maxLength > 0 ? Math.min(maxLength, mInputMode.getSequenceLength()) : mInputMode.getSequenceLength();
+
 		String text = suggestionBar.getCurrentSuggestion();
-		if (text.length() > 0 && text.length() > mInputMode.getSequenceLength()) {
+		if (text.length() > 0 && text.length() > maxLength) {
-			text = text.substring(0, mInputMode.getSequenceLength());
+			text = text.substring(0, maxLength);
 		}
 
 		return text;
 	}
 
 
+	private String getComposingText() {
+		return getComposingText(-1);
+	}
+
+
 	private void refreshComposingText() {
 		textField.setComposingText(getComposingText());
 	}
@@ -729,11 +747,11 @@ public class TraditionalT9 extends KeyPadHandler {
 
 
 	private void showAddWord() {
-		if (currentInputConnection == null) {
+		if (shouldBeOff()) {
 			return;
 		}
 
-		currentInputConnection.finishComposingText();
+		textField.finishComposingText();
 		clearSuggestions();
 
 		UI.showAddWordDialog(this, mLanguage.getId(), textField.getSurroundingWord());

src/io/github/sspanak/tt9/ime/modes/InputMode.java

@@ -56,7 +56,8 @@ abstract public class InputMode {
 	abstract public boolean onOtherKey(int key);
 
 	// Suggestions
-	public void onAcceptSuggestion(@NonNull String suggestion) {}
+	public void onAcceptSuggestion(@NonNull String word) { onAcceptSuggestion(word, false); }
+	public void onAcceptSuggestion(@NonNull String word, boolean preserveWordList) {}
 
 	/**
 	 * loadSuggestions
@@ -97,7 +98,8 @@ abstract public class InputMode {
 	}
 
 	// Interaction with the IME. Return "true" if it should perform the respective action.
-	public boolean shouldAcceptCurrentSuggestion(int key, boolean hold, boolean repeat) { return false; }
+	public boolean shouldAcceptPreviousSuggestion() { return false; }
+	public boolean shouldAcceptPreviousSuggestion(int nextKey) { return false; }
 	public boolean shouldAddAutoSpace(InputType inputType, TextField textField, boolean isWordAcceptedManually, int nextKey) { return false; }
 	public boolean shouldDeletePrecedingSpace(InputType inputType) { return false; }
 	public boolean shouldSelectNextSuggestion() { return false; }

src/io/github/sspanak/tt9/ime/modes/ModeABC.java

@@ -70,7 +70,7 @@ public class ModeABC extends InputMode {
 	@Override final public boolean isABC() { return true; }
 	@Override public int getSequenceLength() { return 1; }
 
-	@Override public boolean shouldAcceptCurrentSuggestion(int key, boolean hold, boolean repeat) { return hold || !repeat; }
+	@Override public boolean shouldAcceptPreviousSuggestion() { return autoAcceptTimeout == 0 || !shouldSelectNextLetter; }
 	@Override public boolean shouldTrackUpDown() { return true; }
 	@Override public boolean shouldTrackLeftRight() { return true; }
 	@Override public boolean shouldSelectNextSuggestion() {

src/io/github/sspanak/tt9/ime/modes/ModePredictive.java

@@ -2,7 +2,10 @@ package io.github.sspanak.tt9.ime.modes;
 
 import androidx.annotation.NonNull;
 
+import java.util.ArrayList;
+
 import io.github.sspanak.tt9.Logger;
+import io.github.sspanak.tt9.TextTools;
 import io.github.sspanak.tt9.db.DictionaryDb;
 import io.github.sspanak.tt9.ime.helpers.InputType;
 import io.github.sspanak.tt9.ime.helpers.TextField;
@@ -19,7 +22,6 @@ public class ModePredictive extends InputMode {
 
 	private String digitSequence = "";
 	private String lastAcceptedWord = "";
-	private String lastAcceptedSequence = "";
 
 	// stem filter
 	private boolean isStemFuzzy = false;
@@ -123,6 +125,31 @@ public class ModePredictive extends InputMode {
 	}
 
 
+	/**
+	 * clearLastAcceptedWord
+	 * Removes the last accepted word from the suggestions list and the "digitSequence"
+	 * or stops silently, when there is nothing to do.
+	 */
+	private void clearLastAcceptedWord() {
+		if (
+			lastAcceptedWord.isEmpty()
+			|| suggestions.isEmpty()
+			|| !suggestions.get(0).toLowerCase(language.getLocale()).startsWith(lastAcceptedWord.toLowerCase(language.getLocale()))
+		) {
+			return;
+		}
+
+		int lastAcceptedWordLength = lastAcceptedWord.length();
+		digitSequence = digitSequence.length() > lastAcceptedWordLength ? digitSequence.substring(lastAcceptedWordLength) : "";
+
+		ArrayList<String> lastSuggestions = new ArrayList<>(suggestions);
+		suggestions.clear();
+		for (String s : lastSuggestions) {
+			suggestions.add(s.length() >= lastAcceptedWordLength ? s.substring(lastAcceptedWordLength) : "");
+		}
+	}
+
+
 	/**
 	 * clearWordStem
 	 * Do not filter the suggestions by the word set using "setWordStem()", use only the digit sequence.
@@ -156,15 +183,18 @@ public class ModePredictive extends InputMode {
 	 * Note that you need to manually get the suggestions again to obtain a filtered list.
 	 */
 	@Override
-	public boolean setWordStem(String wordStem, boolean exact) {
+	public boolean setWordStem(String newStem, boolean exact) {
-		if (language == null || wordStem == null || wordStem.length() < 1) {
+		String sanitizedStem = TextTools.removeNonLetters(newStem);
+		if (language == null || sanitizedStem == null || sanitizedStem.length() < 1) {
 			return false;
 		}
 
 		try {
-			digitSequence = language.getDigitSequenceForWord(wordStem);
+			// digitSequence = "the raw input", so that everything the user typed is preserved visually
+			// stem = "the sanitized input", because filtering by anything that is not a letter makes no sense
+			digitSequence = language.getDigitSequenceForWord(newStem);
+			stem = sanitizedStem.toLowerCase(language.getLocale());
 			isStemFuzzy = !exact;
-			stem = digitSequence.startsWith("0") || digitSequence.startsWith("1") ? "" : wordStem.toLowerCase(language.getLocale());
 
 			Logger.d("tt9/setWordStem", "Stem is now: " + stem + (isStemFuzzy ? " (fuzzy)" : ""));
 			return true;
@@ -172,7 +202,7 @@ public class ModePredictive extends InputMode {
 			isStemFuzzy = false;
 			stem = "";
 
-			Logger.w("tt9/setWordStem", "Ignoring invalid stem: " + wordStem + ". " + e.getMessage());
+			Logger.w("tt9/setWordStem", "Ignoring invalid stem: " + newStem + ". " + e.getMessage());
 			return false;
 		}
 	}
@@ -239,19 +269,26 @@ public class ModePredictive extends InputMode {
 
 	/**
 	 * onAcceptSuggestion
-	 * Bring this word up in the suggestions list next time.
+	 * Bring this word up in the suggestions list next time and if necessary preserves the suggestion list
+	 * with "currentWord" cleaned from them.
 	 */
 	@Override
-	public void onAcceptSuggestion(@NonNull String currentWord) {
+	public void onAcceptSuggestion(@NonNull String currentWord, boolean preserveWords) {
 		lastAcceptedWord = currentWord;
-		lastAcceptedSequence = digitSequence;
-		reset();
 
-		if (currentWord.length() == 0) {
+		if (preserveWords) {
+			clearLastAcceptedWord();
+		} else {
+			reset();
+		}
+		stem = "";
+
+		if (currentWord.isEmpty()) {
 			Logger.i("acceptCurrentSuggestion", "Current word is empty. Nothing to accept.");
 			return;
 		}
 
+		// increment the frequency of the given word
 		try {
 			String sequence = language.getDigitSequenceForWord(currentWord);
 
@@ -284,22 +321,34 @@ public class ModePredictive extends InputMode {
 
 
 	/**
-	 * shouldAcceptCurrentSuggestion
+	 * shouldAcceptPreviousSuggestion
 	 * In this mode, In addition to confirming the suggestion in the input field,
 	 * we also increase its' priority. This function determines whether we want to do all this or not.
 	 */
 	@Override
-	public boolean shouldAcceptCurrentSuggestion(int key, boolean hold, boolean repeat) {
+	public boolean shouldAcceptPreviousSuggestion(int nextKey) {
 		return
-			hold
+			!digitSequence.isEmpty() && (
-			// Quickly accept suggestions using "space" instead of pressing "ok" then "space"
+				(nextKey == 0 && digitSequence.charAt(digitSequence.length() - 1) != '0')
-			|| (key == 0 && !repeat)
+				|| (nextKey != 0 && digitSequence.charAt(digitSequence.length() - 1) == '0')
-			// Punctuation is considered "a word", so that we can increase the priority as needed
+			);
-			// Also, it must break the current word.
+	}
-			|| (!language.isPunctuationPartOfWords() && key == 1 && digitSequence.length() > 0 && !digitSequence.endsWith("1"))
+
-			// On the other hand, letters also "break" punctuation.
+
-			|| (!language.isPunctuationPartOfWords() && key != 1 && digitSequence.endsWith("1"))
+		/**
-			|| (digitSequence.endsWith("0") && key != 0);
+	 * shouldAcceptPreviousSuggestion
+	 * Variant for post suggestion load analysis.
+	 */
+	@Override
+	public boolean shouldAcceptPreviousSuggestion() {
+		return
+			(autoAcceptTimeout == 0 && !digitSequence.startsWith("0"))
+			|| (
+				!digitSequence.isEmpty()
+				&& !predictions.areThereDbWords()
+				&& digitSequence.contains("1")
+				&& TextTools.containsOtherThan1(digitSequence)
+			);
 	}
 
 
@@ -307,11 +356,10 @@ public class ModePredictive extends InputMode {
 	public boolean shouldAddAutoSpace(InputType inputType, TextField textField, boolean isWordAcceptedManually, int nextKey) {
 		return autoSpace
 			.setLastWord(lastAcceptedWord)
-			.setLastSequence(lastAcceptedSequence)
+			.setLastSequence()
 			.setInputType(inputType)
 			.setTextField(textField)
 			.shouldAddAutoSpace(isWordAcceptedManually, nextKey);
-
 	}
 
 
@@ -319,7 +367,6 @@ public class ModePredictive extends InputMode {
 	public boolean shouldDeletePrecedingSpace(InputType inputType) {
 		return autoSpace
 			.setLastWord(lastAcceptedWord)
-			.setLastSequence(lastAcceptedSequence)
 			.setInputType(inputType)
 			.setTextField(null)
 			.shouldDeletePrecedingSpace();

src/io/github/sspanak/tt9/ime/modes/helpers/AutoSpace.java

@@ -1,22 +1,16 @@
 package io.github.sspanak.tt9.ime.modes.helpers;
 
-import java.util.regex.Pattern;
+import io.github.sspanak.tt9.TextTools;
-
 import io.github.sspanak.tt9.ime.helpers.InputType;
 import io.github.sspanak.tt9.ime.helpers.TextField;
 import io.github.sspanak.tt9.preferences.SettingsStore;
 
 public class AutoSpace {
-	private final Pattern isNumber = Pattern.compile("\\s*\\d+\\s*");
-	private final Pattern nextIsLetter = Pattern.compile("^\\p{L}+");
-	private final Pattern nextIsPunctuation = Pattern.compile("^\\p{Punct}");
-
 	private final SettingsStore settings;
 
 	private InputType inputType;
 	private TextField textField;
 	private String lastWord;
-	private String lastSequence;
 
 	public AutoSpace(SettingsStore settingsStore) {
 		settings = settingsStore;
@@ -37,8 +31,7 @@ public class AutoSpace {
 		return this;
 	}
 
-	public AutoSpace setLastSequence(String lastSequence) {
+	public AutoSpace setLastSequence() {
-		this.lastSequence = lastSequence;
 		return this;
 	}
 
@@ -57,12 +50,11 @@ public class AutoSpace {
 		return
 			settings.getAutoSpace()
 			&& !inputType.isSpecialized()
-			&& !nextChars.startsWith(" ")
+			&& nextKey != 0
-			&& !isNumber.matcher(previousChars).find()
+			&& !TextTools.startsWithWhitespace(nextChars)
-			&& !nextIsPunctuation.matcher(nextChars).find()
 			&& (
-				shouldAddAfterPunctuation(previousChars, nextKey)
+				shouldAddAfterWord(isWordAcceptedManually, previousChars, nextChars, nextKey)
-				|| shouldAddAfterWord(isWordAcceptedManually, nextChars)
+				|| shouldAddAfterPunctuation(previousChars, nextChars, nextKey)
 			);
 	}
 
@@ -73,24 +65,23 @@ public class AutoSpace {
 	 * The rules are similar to the ones in the standard Android keyboard (with some exceptions,
 	 * because we are not using a QWERTY keyboard here).
 	 */
-	private boolean shouldAddAfterPunctuation(String previousChars, int nextKey) {
+	private boolean shouldAddAfterPunctuation(String previousChars, String nextChars, int nextKey) {
-		return
+		char previousChar = previousChars.isEmpty() ? 0 : previousChars.charAt(previousChars.length() - 1);
-			// no space after whitespace or special characters
-			!previousChars.endsWith(" ") && !previousChars.endsWith("\n") && !previousChars.endsWith("\t") // previous whitespace
-			&& !lastSequence.equals("0") // previous previous math/special char
-			&& nextKey != 0 // composing (upcoming) whitespace or special character
 
-			// add space after the these
+		return
+			nextKey != 1
+			&& !TextTools.nextIsPunctuation(nextChars)
+			&& !TextTools.startsWithNumber(nextChars)
 			&& (
-				previousChars.endsWith(".")
+				previousChar == '.'
-				|| previousChars.endsWith(",")
+				|| previousChar == ','
-				|| previousChars.endsWith(";")
+				|| previousChar == ';'
-				|| previousChars.endsWith(":")
+				|| previousChar == ':'
-				|| previousChars.endsWith("!")
+				|| previousChar == '!'
-				|| previousChars.endsWith("?")
+				|| previousChar == '?'
-				|| previousChars.endsWith(")")
+				|| previousChar == ')'
-				|| previousChars.endsWith("]")
+				|| previousChar == ']'
-				|| previousChars.endsWith("%")
+				|| previousChar == '%'
 				|| previousChars.endsWith(" -")
 				|| previousChars.endsWith(" /")
 			);
@@ -98,15 +89,15 @@ public class AutoSpace {
 
 
 	/**
-	 * shouldAddAfterPunctuation
+	 * shouldAddAfterWord
 	 * Similar to "shouldAddAfterPunctuation()", but determines whether to add a space after words.
 	 */
-	private boolean shouldAddAfterWord(boolean isWordAcceptedManually, String nextChars) {
+	private boolean shouldAddAfterWord(boolean isWordAcceptedManually, String previousChars, String nextChars, int nextKey) {
 		return
-			// Do not add space when auto-accepting words, because it feels very confusing when typing.
+			isWordAcceptedManually // Do not add space when auto-accepting words, because it feels very confusing when typing.
-			isWordAcceptedManually
+			&& nextKey != 1
-			// Right before another word
+			&& nextChars.isEmpty()
-			&& !nextIsLetter.matcher(nextChars).find();
+			&& TextTools.previousIsLetter(previousChars);
 	}
 
 

src/io/github/sspanak/tt9/ime/modes/helpers/AutoTextCase.java

@@ -1,14 +1,11 @@
 package io.github.sspanak.tt9.ime.modes.helpers;
 
-import java.util.regex.Pattern;
+import io.github.sspanak.tt9.TextTools;
-
 import io.github.sspanak.tt9.ime.modes.InputMode;
 import io.github.sspanak.tt9.languages.Language;
 import io.github.sspanak.tt9.preferences.SettingsStore;
 
 public class AutoTextCase {
-	private final Pattern nextToWordRegex = Pattern.compile("\\b$");
-	private final Pattern startOfSentenceRegex = Pattern.compile("(?<!\\.)(^|[.?!¿¡])\\s*$");
 	private final SettingsStore settings;
 
 
@@ -68,11 +65,11 @@ public class AutoTextCase {
 		}
 
 		// start of sentence, excluding after "..."
-		if (startOfSentenceRegex.matcher(textBeforeCursor).find()) {
+		if (TextTools.isStartOfSentence(textBeforeCursor)) {
 			return InputMode.CASE_CAPITALIZE;
 		}
 
-		if (nextToWordRegex.matcher(textBeforeCursor).find()) {
+		if (TextTools.isNextToWord(textBeforeCursor)) {
 			return InputMode.CASE_LOWER;
 		}
 

src/io/github/sspanak/tt9/ime/modes/helpers/Predictions.java

@@ -23,6 +23,7 @@ public class Predictions {
 	private Runnable onWordsChanged = () -> {};
 
 	// data
+	private boolean areThereDbWords = false;
 	private final ArrayList<String> words = new ArrayList<>();
 
 	// punctuation/emoji
@@ -82,6 +83,10 @@ public class Predictions {
 		return words;
 	}
 
+	public boolean areThereDbWords() {
+		return areThereDbWords;
+	}
+
 
 	/**
 	 * suggestStem
@@ -89,7 +94,7 @@ public class Predictions {
 	 * the user has pressed X keys (otherwise, it makes no sense to add it).
 	 */
 	private void suggestStem() {
-		if (stem.length() > 0 && stem.length() == digitSequence.length()) {
+		if (!stem.isEmpty() && stem.length() == digitSequence.length()) {
 			words.add(stem);
 		}
 	}
@@ -114,7 +119,7 @@ public class Predictions {
 	 * sequence or loads the static ones.
 	 */
 	public void load() {
-		if (digitSequence == null || digitSequence.length() == 0) {
+		if (digitSequence == null || digitSequence.isEmpty()) {
 			words.clear();
 			onWordsChanged.run();
 			return;
@@ -124,7 +129,7 @@ public class Predictions {
 			onWordsChanged.run();
 		} else {
 			DictionaryDb.getWords(
-				this::onDbWords,
+				(words) -> onDbWords(words, true),
 				language,
 				digitSequence,
 				stem,
@@ -170,6 +175,23 @@ public class Predictions {
 		return true;
 	}
 
+	private void loadWithoutLeadingPunctuation() {
+		DictionaryDb.getWords(
+			(dbWords) -> {
+				char firstChar = inputWord.charAt(0);
+				for (int i = 0; i < dbWords.size(); i++) {
+					dbWords.set(i, firstChar + dbWords.get(i));
+				}
+				onDbWords(dbWords, false);
+			},
+			language,
+			digitSequence.substring(1),
+			stem.length() > 1 ? stem.substring(1) : "",
+			settings.getSuggestionsMin(),
+			settings.getSuggestionsMax()
+		);
+	}
+
 
 	/**
 	 * dbWordsHandler
@@ -177,8 +199,18 @@ public class Predictions {
 	 * they will be generated based on the "inputWord". After the word list is compiled, it notifies the
 	 * external handler it is now possible to use it with "getList()".
 	 */
-	private void onDbWords (ArrayList<String> dbWords) {
+	private void onDbWords(ArrayList<String> dbWords, boolean isRetryAllowed) {
-		if (dbWords.size() == 0 && digitSequence.length() > 0) {
+		// only the first round matters, the second one is just for getting the letters for a given key
+		areThereDbWords = !dbWords.isEmpty() && isRetryAllowed;
+
+		// If there were no database words for ",a", try getting the letters only (e.g. "a", "b", "c").
+		// We do this to display them in the correct order.
+		if (dbWords.isEmpty() && isRetryAllowed && digitSequence.length() == 2 && digitSequence.charAt(0) == '1') {
+			loadWithoutLeadingPunctuation();
+			return;
+		}
+
+		if (dbWords.isEmpty() && !digitSequence.isEmpty()) {
 			emptyDbWarning.emitOnce(language);
 			dbWords = generatePossibleCompletions(inputWord);
 		}
@@ -186,7 +218,7 @@ public class Predictions {
 		words.clear();
 		suggestStem();
 		suggestMissingWords(generatePossibleStemVariations(dbWords));
-		suggestMissingWords(dbWords);
+		suggestMissingWords(insertPunctuationCompletions(dbWords));
 
 		onWordsChanged.run();
 	}
@@ -206,20 +238,16 @@ public class Predictions {
 
 		// Make sure the displayed word and the digit sequence, we will be generating suggestions from,
 		// have the same length, to prevent visual discrepancies.
-		baseWord = (baseWord != null && baseWord.length() > 0) ? baseWord.substring(0, Math.min(digitSequence.length() - 1, baseWord.length())) : "";
+		baseWord = (baseWord != null && !baseWord.isEmpty()) ? baseWord.substring(0, Math.min(digitSequence.length() - 1, baseWord.length())) : "";
 
 		// append all letters for the last digit in the sequence (the last pressed key)
 		int lastSequenceDigit = digitSequence.charAt(digitSequence.length() - 1) - '0';
-		for (String keyLetter : language.getKeyCharacters(lastSequenceDigit)) {
+		for (String keyLetter : language.getKeyCharacters(lastSequenceDigit, false)) {
-			// let's skip numbers, because it's weird, for example:
+			generatedWords.add(baseWord + keyLetter);
-			// | weird | weire | weirf | weir2 |
-			if (keyLetter.charAt(0) < '0' || keyLetter.charAt(0) > '9') {
-				generatedWords.add(baseWord + keyLetter);
-			}
 		}
 
 		// if there are no letters for this key, just append the number
-		if (generatedWords.size() == 0) {
+		if (generatedWords.isEmpty()) {
 			generatedWords.add(baseWord + digitSequence.charAt(digitSequence.length() - 1));
 		}
 
@@ -227,6 +255,46 @@ public class Predictions {
 	}
 
 
+	/**
+	 * insertPunctuationCompletions
+	 * When given: "you'", for example, this also generates all other 1-key alternatives, like:
+	 * "you.", "you?", "you!" and so on. The generated words will be inserted after the direct
+	 * database matches and before the fuzzy matches, as if they were direct matches with low frequency.
+	 * This is to preserve the sorting by length and frequency.
+	 */
+	private ArrayList<String> insertPunctuationCompletions(ArrayList<String> dbWords) {
+		if (!stem.isEmpty() || dbWords.isEmpty() || digitSequence.length() < 2 || !digitSequence.endsWith("1")) {
+			return dbWords;
+		}
+
+		ArrayList<String> complementedWords = new ArrayList<>();
+		int exactMatchLength = digitSequence.length();
+
+		// shortest database words (exact matches)
+		for (String w : dbWords) {
+			if (w.length() <= exactMatchLength) {
+				complementedWords.add(w);
+			}
+		}
+
+		// generated "exact matches"
+		for (String w : generatePossibleCompletions(dbWords.get(0))) {
+			if (!dbWords.contains(w) && !dbWords.contains(w.toLowerCase(language.getLocale()))) {
+				complementedWords.add(w);
+			}
+		}
+
+		// longer database words (fuzzy matches)
+		for (String w : dbWords) {
+			if (w.length() > exactMatchLength) {
+				complementedWords.add(w);
+			}
+		}
+
+		return complementedWords;
+	}
+
+
 	/**
 	 * generatePossibleStemVariations
 	 * Similar to generatePossibleCompletions(), but uses the current filter as a base word. This is
@@ -242,7 +310,7 @@ public class Predictions {
 	 */
 	private ArrayList<String> generatePossibleStemVariations(ArrayList<String> dbWords) {
 		ArrayList<String> variations = new ArrayList<>();
-		if (stem.length() == 0) {
+		if (stem.isEmpty()) {
 			return variations;
 		}
 

src/io/github/sspanak/tt9/languages/Language.java

@@ -18,7 +18,6 @@ public class Language {
 
 	// settings
 	protected boolean hasUpperCase = true;
-	protected boolean isPunctuationPartOfWords; // see the getter for more info
 
 	final public int getId() {
 		if (id == 0) {
@@ -60,24 +59,6 @@ public class Language {
 		return abcString;
 	}
 
-	/**
-	 * isPunctuationPartOfWords
-	 * This plays a role in Predictive mode only.
-	 *
-	 * Return "true", if you need to use the 1-key for typing words, such as:
-	 * "it's" (English), "a'tje" (Dutch), "п'ят" (Ukrainian).
-	 *
-	 * Return "false" also:
-	 * 		- hide words like the above from the suggestions.
-	 *		- 1-key would commit the current word, then display the punctuation list.
-	 * 			For example, pressing 1-key after "it" would accept "it" as a separate word,
-	 * 			then display only: | , | . | ! | ? | ...
-	 *
-	 * "false" is recommended when apostrophes or other punctuation are not part of the words,
-	 * because it would allow faster typing.
-	 */
-	final public boolean isPunctuationPartOfWords() { return isPunctuationPartOfWords; }
-
 
 	public boolean hasUpperCase() {
 		return hasUpperCase;
@@ -124,7 +105,21 @@ public class Language {
 	}
 
 	public String capitalize(String word) {
-		return word != null ? word.substring(0, 1).toUpperCase(locale) + word.substring(1).toLowerCase(locale) : null;
+		if (word == null) {
+			return null;
+		}
+
+		String capitalizedWord = "";
+
+		if (!word.isEmpty()) {
+			capitalizedWord += word.substring(0, 1).toUpperCase(locale);
+		}
+
+		if (word.length() > 1) {
+			capitalizedWord += word.substring(1).toLowerCase(locale);
+		}
+
+		return capitalizedWord;
 	}
 
 	public boolean isMixedCaseWord(String word) {

src/io/github/sspanak/tt9/languages/definitions/BrazilianPortuguese.java

@@ -11,8 +11,6 @@ public class BrazilianPortuguese extends English {
 		locale = new Locale("pt","BR");
 		dictionaryFile = "pt-BR-utf8.csv";
 
-		isPunctuationPartOfWords = true;
-
 		characterMap.get(2).addAll(Arrays.asList("ç", "á", "â", "ã", "à"));
 		characterMap.get(3).addAll(Arrays.asList("é", "ê", "è"));
 		characterMap.get(4).add("í");

src/io/github/sspanak/tt9/languages/definitions/Bulgarian.java

@@ -12,8 +12,6 @@ public class Bulgarian extends Language {
 		locale = new Locale("bg","BG");
 		dictionaryFile = "bg-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap = new ArrayList<>(Arrays.asList(
 			Characters.Special, // 0
 			Characters.Sentence, // 1

src/io/github/sspanak/tt9/languages/definitions/Dutch.java

@@ -10,8 +10,6 @@ public class Dutch extends English {
 		locale = new Locale("nl","NL");
 		dictionaryFile = "nl-utf8.csv";
 
-		isPunctuationPartOfWords = true;
-
 		characterMap.get(2).addAll(Arrays.asList("à", "ä", "ç"));
 		characterMap.get(3).addAll(Arrays.asList("é", "è", "ê", "ë"));
 		characterMap.get(4).addAll(Arrays.asList("î", "ï"));

src/io/github/sspanak/tt9/languages/definitions/English.java

@@ -12,8 +12,6 @@ public class English extends Language {
 		locale = Locale.ENGLISH;
 		dictionaryFile = "en-utf8.csv";
 
-		isPunctuationPartOfWords = true;
-
 		characterMap = new ArrayList<>(Arrays.asList(
 			Characters.Special, // 0
 			Characters.Sentence, // 1

src/io/github/sspanak/tt9/languages/definitions/Finnish.java

@@ -10,8 +10,6 @@ public class Finnish extends English {
 		locale = new Locale("fi","FI");
 		dictionaryFile = "fi-utf8.csv";
 
-		isPunctuationPartOfWords = true;
-
 		characterMap.get(2).addAll(Arrays.asList("ä", "å"));
 		characterMap.get(6).add("ö");
 	}

src/io/github/sspanak/tt9/languages/definitions/French.java

@@ -10,8 +10,6 @@ public class French extends English {
 		locale = Locale.FRENCH;
 		dictionaryFile = "fr-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).addAll(Arrays.asList("à", "â", "æ", "ç"));
 		characterMap.get(3).addAll(Arrays.asList("é", "è", "ê", "ë"));
 		characterMap.get(4).addAll(Arrays.asList("î", "ï"));

src/io/github/sspanak/tt9/languages/definitions/German.java

@@ -9,8 +9,6 @@ public class German extends English {
 		locale = Locale.GERMAN;
 		dictionaryFile = "de-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).add("ä");
 		characterMap.get(6).add("ö");
 		characterMap.get(7).add("ß");

src/io/github/sspanak/tt9/languages/definitions/Hebrew.java

@@ -14,7 +14,6 @@ public class Hebrew extends Language {
 		abcString = "אבג";
 
 		hasUpperCase = false;
-		isPunctuationPartOfWords = true;
 
 		characterMap = new ArrayList<>(Arrays.asList(
 			Characters.Special, // 0

src/io/github/sspanak/tt9/languages/definitions/Italian.java

@@ -10,8 +10,6 @@ public class Italian extends English {
 		locale = Locale.ITALIAN;
 		dictionaryFile = "it-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).add("à");
 		characterMap.get(3).addAll(Arrays.asList("é", "è"));
 		characterMap.get(4).addAll(Arrays.asList("ì", "í", "î"));

src/io/github/sspanak/tt9/languages/definitions/Norwegian.java

@@ -10,8 +10,6 @@ public class Norwegian extends English {
 		locale = new Locale("nb","NO");
 		dictionaryFile = "nb-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).addAll(Arrays.asList("æ", "å"));
 		characterMap.get(3).addAll(Arrays.asList("é", "è"));
 		characterMap.get(6).addAll(Arrays.asList("ø", "ó", "ò", "ô"));

src/io/github/sspanak/tt9/languages/definitions/Polish.java

@@ -10,8 +10,6 @@ public class Polish extends English {
 		locale = new Locale("pl","PL");
 		dictionaryFile = "pl-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).addAll(Arrays.asList("ą", "ć"));
 		characterMap.get(3).add("ę");
 		characterMap.get(5).add("ł");

src/io/github/sspanak/tt9/languages/definitions/Russian.java

@@ -12,8 +12,6 @@ public class Russian extends Language {
 		locale = new Locale("ru","RU");
 		dictionaryFile = "ru-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap = new ArrayList<>(Arrays.asList(
 			Characters.Special, // 0
 			Characters.Sentence, // 1

src/io/github/sspanak/tt9/languages/definitions/Spanish.java

@@ -13,8 +13,6 @@ public class Spanish extends English {
 		locale = new Locale("es", "ES");
 		dictionaryFile = "es-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.set(1, new ArrayList<>(Characters.Sentence));
 		characterMap.get(1).addAll(Arrays.asList("¡", "¿"));
 

src/io/github/sspanak/tt9/languages/definitions/Swedish.java

@@ -10,8 +10,6 @@ public class Swedish extends English {
 		locale = new Locale("sv","SE");
 		dictionaryFile = "sv-utf8.csv";
 
-		isPunctuationPartOfWords = false;
-
 		characterMap.get(2).addAll(Arrays.asList("å", "ä"));
 		characterMap.get(3).add("é");
 		characterMap.get(6).add("ö");

src/io/github/sspanak/tt9/languages/definitions/Ukrainian.java

@@ -12,8 +12,6 @@ public class Ukrainian extends Language {
 		locale = new Locale("uk","UA");
 		dictionaryFile = "uk-utf8.csv";
 
-		isPunctuationPartOfWords = true;
-
 		characterMap = new ArrayList<>(Arrays.asList(
 			Characters.Special, // 0
 			Characters.Sentence, // 1

src/io/github/sspanak/tt9/languages/definitions/Yiddish.java

@@ -8,6 +8,5 @@ public class Yiddish extends Hebrew {
 
 		locale = new Locale("ji","JI");
 		dictionaryFile = "ji-utf8.csv";
-		isPunctuationPartOfWords = true;
 	}
 }