-
Notifications
You must be signed in to change notification settings - Fork 15
/
Copy pathcamxes-exp-changelog.txt
154 lines (120 loc) · 8.69 KB
/
camxes-exp-changelog.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
1) implementation of foretought BE-clause
• {lo be mi broda} now parses.
2) A/JA/JOI merge, gijek implementation
• {brode gi je brodo} now parses.
• {brode na gi joi nai brodo} now parses.
• {na je nai gi mi klama gi mi citka} now parses.
• {broda gi .a brode .e brodo} now parses.
3) merging ZIhE with JA/JOI
• {da po'u do je nai poi drata} now parses.
4) GI TAG BO (suggested by tsani; implemented by selpa'i)
• {mi klama lo zarci gi mu'i bo .ei cpacu lo cidja} now parses.
5) brivla / cmevla merge
• {lo .cmevlas. broda ku co'e} now parses.
• {la .cmevlas. broda ku} now parses.
• {la .cmevlas. ku} now parses.
6) ZOhOI and MEhOI implemented.
7) NAhE BO SUMTI → NAhE SUMTI (the bo is no longer necessary)
(So far it has been only implemented outside MEX.)
• {na'e mi broda} now parses.
8) Recycling SOI (implementation of the new SOI).
9) TAG grammar simplification + assimilation of FA into TAG
• {tavla fi je fo lo jbobau} and {.i se fe bo ko'a broda} now parses.
10) Add of a digit preprocessing (e.g. 3→ci, 4→vo) and a [diacritic → upper case] preprocessing.
11) NAI/UI merge.
12) ZOI now fully implemented.
13) Experimental change from Tsani: empty BE clause + enabling linkargs before tanru_unit_1.
• {be be broda} now parses.
• {jai be broda} now parses.
14) Changes in VUhO's grammar: now the following relative clause is optional, and if present, adding a trailing JOI+SUMTI is now allowed.
• {mi je do vu'o .ei broda} now parses.
• {mi je do vu'o noi broda joi ko'a} now parses.
15) Allowing JA to play the same role as GIhA (yes, this is possible!)
• {mi klama vau je sanga} is now equivalent to {mi klama gi je sanga} and {mi klama gi'e sanga}
• {mi klama do je sanga} is now equivalent to {mi klama do gi je sanga} and {mi klama do gi'e sanga}
• now {mi klama lo pendo ku je sanga} = {mi klama lo pendo gi je sanga} = {mi klama lo pendo gi'e sanga}
16) Allowing CU without bridi-head, so that it is now a mere main-selbri prefix, and not anymore a "bridi-head terminator"
• {fi'i lo vitke cu se pluka .a'o} now parses.
• {sei kucli cu slabu xu do} now parses.
17) Elided terminators are now shown in upper case in the parser's output.
18) NOhOI–KUhOI now implemented.
• {do sanga no'oi pluka mi} (= "you're singing, which is pleasant to me") now parses.
======
2013-11-07: Replaced _node() calls with _node2() within several *_clause productions (e.g. VAU_clause), fixing a bug in the original camxes.js.peg, whereby indicators (e.g. UI_clause) didn't show up in the output when following such productions. We should fix this in the original Camxes too. :)
2013-12-27: Empty prenexes allowed.
2013-12-27: Minor fix in gihek_1 (thanks to Selpa'i)
2013-12-27: {uy} and {iy} added to BY (requested by Selpa'i)
2013-12-27: Cleanup of camxes_postprocessor.js and output modes (minimalistic modes renamed to "prettified" and deletion of fa'orma'o modes).
2013-12-28: Add of {ko'oi} and {si'au} to UI.
======
2014-01-13: Term grammar changed so that tags immediately followed by a selbri become part thereof.
This behavior is more in accordande with that of Jbofi'e, except that in Jbofi'e, terms within the bridi-tail of a subordinate bridi are handled differently so that a trailing tag can't be stolen from the subordinate clause by the following selbri, i.e. {lo nu broda pu brode} parses as {lo nu broda pu (ku) (kei) (ku) (cu) brode} in Jbofi'e, whereas it currently parses as {lo nu broda (kei) (ku) (cu) pu brode} in Camxes-Exp. Copying Jbofi'e's behavior in this regard would necessitate further tinkering and a significantly increased complexity in terms and termset grammars, and the resulting twofold handling of terms is probably not as intuitive as unconditional absorbtion into selbri anyway.
2014-01-13: Allowing the pattern {TAG TAGTERM JA TAG TAGTERM}: {ba ko'a je na ku} now parses.
2014-01-13: Minor node naming mistake fixed (PU_clause was misnamed "PA_clause" in the output of the original Camxes).
======
• 2014-01-26: A few node naming mistakes fixed.
• Fix: selbri tcita before gek_sentence (fix also added to the standard version of the grammar).
• Implementation of a change by Selpa'i on the "number" production ( of whose purpose I don't recall :p ).
• Allowing NA sequences in sentence fragments (now {.i na ja'a .i} parses).
• 2014-03-19: Implementation of JACU bridi tail connection (a new alternative to GIhA proposed by Selpa'i).
• Implementation of forethought relative clause connections (both on NOI and NOhOI). Now {lo plise ge noi se citka gi poi xunre} parses.
• Implementation of FIhOI-FIhAU (bridi —adverbial— relative clause).
• 2014-04-02: Tag sequences fixed, so that {lo prenu pe pu zi} has {pu zi} treated back as a single term.
======
======
140403: Fix in JACU
140404: Fixing {JA TAG bo cu} - Thank you Selpahi!
140428: Making bare mex a legal quantifier (without vei); BO optional in NAhE+BO+operand
Implementation of GAJA
gaJA implemented; previous change in _quantifier_ reverted.
140501: The "gek-subsentence selbri tcita fix" was breaking {tag gi... gi...}, so I removed it.
140513: Allowing semi-vowels after diphthongs
140518: Adding cmavo - xai xo'o xoi xu'e
140529: Allowing NA without KU as a term
140626: Update camxes morphology to v119 (October 2009)
Adding the missing most frequent experimental cmavo.
Merging FIhOI and new-SOI.
Fixing NA selbritcita, plus allowing FREE within NAKU.
140628: Allowing {lo mi .e do broda}.
140630: The experimental change for allowing {lo mi .e do broda} is breaking the grammar ( {lo pa broda} doesn't parse anymore ), so I'm reverting this change.
140718: [patch from laMudri for the preprocessor] "ʃɛʒə" => "cejy" and related letters
140815: lagleki: {mi'ai} added
140816: Fixes: #1 in camxes-exp, the tag simplification resulted in VIhA being ungrammatical; #2 fixing an ordered choice mistake in space_interval, as well as simplifying statement_2 (standard camxes); Thanks alanpost. :)
Folding two orded choices into one in sumti_tail_1 (Thanks alanpost)
Combining two ordered choices into one in space_interval, plus a further fix of VIhA_clause in the same production (Thanks alanpost).
140820: Fix: {gau} was broken by the experimental gaJA connective. Postproc modified to render the ga_word selmaho properly.
Fixing the node outputs of Y and ybu.
140906: camxes-exp: Disallowing ViiV and VuuV vowel sequences.
140926: Adding experimental cmavo {sa'ei}, {mu'ei}, {poi'i} and {kai'u} to camxes-exp.
Implementing LOhOI-KUhAU in camxes-exp.
Fixing kai'u
141009: Fixing ku-less NA adverbial: a NA before a connective shouldn't be treated as an adverbial: it's part of the connective. :)
141017: Fix in JACU: {brode ja cu brodo vau mi} was broken.
141018: Fixing {lo me li pa moi} and reverting alanpost's optimization of the sumti_tail_1 production.
MAJOR CHANGE: MEX SIMPLIFICATION IMPLEMENTED (Thank you Selpahi!)
Camxes-exp: removing the experimental code for the internal grammar of quantifiers.
141026: Camxes-std: Fixing {vi bai broda} which was broken by the selbritcita absorbtion rules.
141030: Camxes-exp fix: The node names of bridi_tail_t1 and bridi_tail_t2 were wrong.
141106: CAMXES FIX: fu'ivla rafsi stress (thank you Xorxes!)
141223: Change "text" rule from node() to node_nonempty() to fix TO parsing (also in camxes-std)
142512: t o h a i / v o h a i / x o h a i / s u h e i / s u h a i added
======
150105: ban CGV strings anywhere per BPFK decision (http://mw.lojban.org/papri/BPFK_Section:_Ban_on_consonant-glide-vowel_strings )
150107: fix slinkuhi detection in extended rafsi (https://groups.google.com/forum/#!msg/bpfk-list/P0YKLOBDvNY/6l6LXBXb81sJ )
150116: (off only) show elided FEhU (this only included the source fix, not the compiled camxes.js)
150131-150201: (exp only) add lo'ai sa'ai le'ai as lo'u-like free modifier
150204: insert y into voiced-voiceless clusters (preprocessor)
150219: (exp only) add cmavo: CAI (mau'i, dau'i, me'ai), COI (a'oi), KE (pi'ai), PA (xei), UI (fu'au, a'ei), XI (te'ai)
150220: convert · to . (preprocessor)
150324: complete the FEhU fix from 150116
150325-150506: various additions to the grammar, later removed from here and spun off as altatufa (http://github.com/lagleki/glekitufa ):
* sentences and subsentences without selbri
* postpositions
* split nouns (fa ti fa je ta, fa ti fa noi nei)
* suffixed articles
* {ze'ei}
* sumti place detection
Some of these also survived into morfologi (https://github.com/mezohe/ilmentufa/tree/morfologi )
150525: (off and exp) update to BPFK morphology 137 (allow glides after diphthongs and rafsi ending in Gy), and fix stressed {cmavrfy}-shaped rafsi
(exp only) fix {to'e TAG SELBRI} (parse as selbritcita, not sumti), allow {cu TERMS SELBRI}
(see http://mw.lojban.org/papri/nu_curmi_lo_nu_lo_sumti_cu_jbini_zo_cu_joi_lo_selbri )