(**************************************************************************)
(* *)
(* The Sanskrit Heritage Platform *)
(* *)
(* Gérard Huet & Pawan Goyal *)
(* *)
(* ©2018 Institut National de Recherche en Informatique et en Automatique *)
(**************************************************************************)
(* Pada defines the allowed padas (Para, Atma or Ubha) for
a given combination of root, gana, and upasarga *)
(* It is used at conjugation computation time by Verbs, in order to generate
root forms for attested lexicalizations of root and gana (over all
possible upasarga usages) and at segmentation time, to filter out
by Dispatcher the non attested combinations of gana, pada and upasarga *)
(*i module Pada = struct i*)
type voices = (* permitted padas in present system *)
(* NB. These are distinctions within the active voice, as opposed to passive
("karma.ni prayoga"). Atma is called "middle" by Western grammarians. *)
[ Para (* parasmaipadin usage only - generated as Dictionary.Active *)
| Atma (* aatmanepadin usage only - generated as Dictionary.Middle *)
| Ubha (* ubhayapada admits both schemes - default *)
]
;
exception Unattested (* when a root/pada is attested only for some pvs *)
;
value voices_of = fun
(* Simplification: invariant when prefixing by preverbs *)
[ "ak.s" | "afg" | "aj" | "a.t" | "at" | "ad#1" | "an#2" | "am"
| "ard" | "av" | "az#2" | "as#1" | "as#2" | "aap" | "ifg" | "in" | "ind"
| "inv" | "il" | "i.s#2" | "iifkh" | "iir.s" | "uk.s" | "uc" | "ujjh" | "u~nch"
| "und" | "umbh" | "u.s" | ".rc#1" | ".rdh" | ".r.s" | "ej" | "kas" | "kiil"
| "ku.t" | "ku.n.th" | "kunth" | "kup" | "kul" | "kuuj" | "k.rt#1"
| "k.rz" | "krand" | "krii.d" | "kru~nc#1" | "krudh#1" | "kruz" | "klam"
| "klid" | "kliz" | "kvath" | "k.sar" | "k.sal" | "k.si" | "k.sii" | "k.su"
| "k.sudh#1" | "k.subh" | "k.svi.d" | "khaad" | "khid" | "khel" | "khyaa"
| "gaj" | "gad" | "garj" | "gard" | "gal" | "gaa#1" | "gaa#2" | "gu~nj"
| "gu.n.th" | "gup" | "gumph" | "g.rdh" | "g.rr#1" | "g.rr#2" | "granth"
| "grah" | "glai" | "ghas" | "ghu.s" | "gh.r" | "gh.r.s" | "ghraa" | "cakaas"
| "ca.t" | "cand" | "cam" | "car" | "cal" | "cit#1" | "cumb" | "chur"
| "ch.rd" | "jak.s" | "jap" | "jabh#2" | "jam" | "jalp" | "jas" | "jaag.r"
| "jinv" | "jiiv" | "jvar" | "jval" | "tak" | "tak.s" | "ta~nc"
| "tam" | "tarj" | "tup" | "tu.s" | "t.rp#1" | "t.r.s#1" | "t.rr" | "tyaj#1"
| "tras" | "tru.t" | "tvak.s" | "tsar" | "da.mz" | "dagh" | "dabh" | "dam#1"
| "dal" | "das" | "dah#1" | "daa#2" | "daa#3" | "diiv#1" | "du" | "du.s"
| "d.rp" | "d.rbh" | "d.rz#1" | "d.rh" | "d.rr" | "dhyaa" | "draa#1" | "dru#1"
| "druh#1" | "dham" | "dhaa#2" | "dhru" | "dhvan" | "dhv.r" | "na.t" | "nad"
| "nand" | "nam" | "nard" | "naz#1" | "nind" | "nu#1" | "n.rt" | "pa.t"
| "pat#1" | "path" | "paa#1" | "paa#2" | "pi#2" | "piz#1" | "pi.s" | "pu.t"
| "p.r#1" | "p.r.s" | "p.rr" | "praa#1" | "phal"
| "bal" | "b.rh#1" | "b.rh#2" | "bha~nj" | "bha.n" | "bha.s"
| "bhas" | "bhaa#1" | "bhii#1" | "bhuj#1" | "bhuu#1" | "bhuu.s" | "bhram"
| "majj" | "ma.n.d" | "mad#1" | "manth" | "mah" | "maa#3" | "mi.s" | "mih"
| "miil" | "mu.s#1" | "muh" | "muurch" | "m.r.d" | "m.rz" | "mnaa" | "mre.d"
| "mlaa" | "mlecch" | "yabh" | "yam" | "yas" | "yaa#1" | "yu#2" | "ra.mh"
| "rak.s" | "ra.n" | "rad" | "radh" | "raa#1" | "raadh" | "ri.s" | "ru"
| "ruj#1" | "rudh#1" | "ru.s#1" | "ruh#1" | "lag" | "lafg" | "lap" | "lal"
| "las" | "laa" | "laa~nch" | "likh" | "liz" | "lu.n.th" | "lubh" | "lul"
| "vak.s" | "vac" | "vaj" | "va~nc" | "van" | "vam" | "valg" | "vaz" | "vas#1"
| "vaa#2" | "vas#4" | "vaa~nch" | "vid#1" | "vidh#1" | "vi.s#1" | "vii#1"
| "v.rj" | "v.r.s" | "v.rh" | "ven" | "vyac"| "vyadh" | "vraj" | "vrazc"
| "za.ms" | "zak" | "zam#1" | "zam#2" | "zal" | "zaz" | "zas" | "zaas"
| "zi.s" | "ziil" | "zuc#1" | "zudh" | "zumbh" | "zu.s" | "zuu" | "z.rr"
| "zcut#1" | "zram" | "zru" | "zli.s" | "zvas#1" | ".s.thiiv" | "sa~nj"
| "sad#1" | "sap#1" | "saa#1" | "sidh#1" | "sidh#2" | "siiv" | "sur" | "s.r"
| "s.rj#1" | "s.rp" | "skand" | "skhal" | "stan" | "stubh" | "sthag" | "snaa"
| "snih#1" | "snu" | "snuh#1" | "sp.r" | "sphal" | "sphu.t" | "sphur"
| "sm.r" | "sru" | "svan" | "svap" | "svar#1" | "svar#2" | "ha.th"
| "haa#1" | "hi#2" | "hi.ms" | "h.r.s" | "hras" | "hrii#1" | "hval"
| "maarg" (* root rather than nominal verb *)
(*| "viz#1" Atma needed for eg nivizate \Pan{1,3,17} *)
(*| "ji" Atma needed for eg vijayate paraajayate \Pan{1,3,19} *)
(*| "jyaa#1" Atma needed for jiiyate *)
(*| "kan" Atma needed for kaayamaana *)
(*| "gam" Atma needed for sa.mgacchate *)
(*| "van" Atma needed for vanute *)
(*| "mah" Atma needed for pft. maamahe *)
(*| "cit#1" Atma needed for pft. cikite *)
(*| "kaafk.s" | "han#1" occur also in Atma in BhG: kaafk.se hani.sye *)
(*| "has" Atma needed for hasate *)
(*| "a~nj" also Atma afkte | "naath" "praz" "sp.rz#1" idem *)
-> Para (* active only *)
| "az#1" | "aas#2" | "indh" | "iik.s" | "ii.d" | "iir" | "iiz#1" | "ii.s"
| "iih" | "edh" | "katth" | "kam" | "kamp" | "kaaz" | "kaas#1" | "kuu"
| "k.rp" | "k.lp" (* but Henry: {cak.lpur} "ils s'arrangèrent" *)
| "klav" | "k.sad" | "k.sam" | "galbh" | "gaah" | "gur" | "gha.t"
| "jabh#1" | "ju.s#1" | "j.rmbh" | ".damb" | ".dii" | "tandr" | "tij" | "trap"
| "trai" | "tvar" | "dak.s" | "day" | "diik.s" | "diip" | "d.r#1" | "dhii#1"
| "dhuk.s" | "pa.n" | "pad#1" | "pi~nj" | "p.r#2" | "pyaa" | "prath"
| "pru" | "plu" | "ba.mh" | "baadh" | "bha.n.d" | "bhand" | "bhaa.s"
| "bhuj#2" | "bhraaj" | "ma.mh" | "man" | "mand#1" | "yat#1" | "yudh#1"
| "rabh" | "ruc#1" | "lajj" | "lamb" | "lii" | "loc" | "vand" | "vas#2"
| "vaaz" | "vip" | "v.rdh#1" | "ve.s.t" | "vrii.d" | "zafk" | "zad"
| "zi~nj" | "zii#1" | "zrambh" | "zlaagh" | "zvit" | "sac" | "sev"
| "styaa" | "spand" | "spardh" | "spaz#1" | "sphaa" | "smi" | "sra.ms"
| "sva~nj" | "haa#2" | "hu.n.d" | "h.r#2" | "hnu" | "hraad" | "hlaad"
(*| "m.r" Ubha needed for non present tenses - see \Pan{1,3,61} for exact rule *)
(* DRP restriction: "dyut#1" *)
-> Atma (* "deponent" verbs: middle only *)
| _ -> Ubha (* default *)
(* Attested Ubha (over all ga.nas) :
[ "a~nc" | "arh" | "i" | "i.s#1" | "uurj#1" | "uuh" | ".r" | ".rj"
| "ka.n.d" | "kal" | "ka.s" | "ku.t.t" | "ku.n.d" | "k.r#1" | "k.r#2"
| "kram" | "krii" | "k.san" | "k.sap#1" | "k.sal" | "k.sip" | "k.sud" | "khan"
| "gam" | "garh" | "guh" | "gras" | "gha.t.t" | "cat" | "carc" | "ci"
| "cint" | "cud" | "ce.s.t" | "cyu" | "chad#1" | "chand" | "chid#1" | "jan"
| "juu" | "j~naa#1" | "jyaa#1" | "jyut" | "ta.d" | "tan#1" | "tan#2"
| "tud#1" | "tul" | "t.rd" | "daaz#1" | "diz#1" | "dih" | "duh#1"
| "dev#1" | "draa#2" | "dvi.s#1" | "dhaa#1" | "dhaav#1" | "dhaav#2"
| "dhuu#1" | "dh.r" | "dhva.ms" | "nah" | "naath" | "nij" | "nii#1"
| "nud" | "pac" | "paz" | "pa.th" | "pii.d" | "pu.s#1" | "puu#1" | "puuj"
| "puuy" | "p.rth" | "prii" | "budh#1" | "bruu" | "bhak.s" | "bhaj" | "bharts"
| "bhaas#1" | "bhid#1" | "bh.r" | "bh.rjj" | "maa#4"
| "mi" | "mith" | "mil" | "mii" | "muc#1" | "mud#1" | "m.r" | "m.rj"
| "m.rdh" | "m.r.s" | "yaj#1" | "yaac" | "yu#1" | "yuj#1" | "rac"
| "ra~nj" | "ram" | "rah" | "raaj#1" | "ri" | "ric" | "rud#1" | "rudh#2"
| "lafgh" | "labh" | "la.s" | "lip" | "lih#1" | "lup" | "luu#1" | "vad"
| "vap#1" | "vap#2" | "val" | "vah#1" | "vaa#3" | "vic" | "vij" | "viij"
| "v.r#2" | "v.rt#1" | "vyath" | "vyaa" | "zap" | "zaa" | "zubh#1"
| "zyaa" | "zri" | "san#1" | "sah#1" | "sic" | "su#2" | "suud" | "stambh"
| "stu" | "st.rr" | "sthaa#1" | "sp.rz#1" | "sp.rh" | "syand" | "svad" | "had"
| "hikk" | "hu" | "huu" | "h.r#1" ] *)
(* + corr. "pa.th" | "sthaa#1" | "praz" | "k.rr" | "p.rc" | "bandh" *)
(* NB. "ah" "rip" "vadh" have no pr, "mand2" is fictitious *)
(* "iiz1" and "lii" allowed Para in future *)
]
;
(* List of roots that admit different padas for distinct ganas:
as2 1U 4P (* 4P Vedic - may overgenerate ? *)
i 1A 2P 4A 5P
.r 1U 3P 5P
kuc 1U 6P
k.r.s 1P 6U
ghuur.n 1A 6P
jan 4A 1U
j.rr 1U 4P
jyaa1 4A 9P
.damb 1A 10P (vi-)
tap 1P 4A
daa1 2P 1U 3U
draa2 2P 4U
dh.r.s 1U 5P
nij 2A 3U
pu.s1 4U 9P
budh1 1P 4A
bhra.mz 1A 4P
man 1U 4U 8A
maa1 3A 2P
mid 1A 4P 1OP
mii 9P 4A
m.r 4A other tenses P
m.rj 1U 2P 6U
m.rd1 9P 1U
ri 4A 9U
ric 4A 7P
rud1 2P 1U 6U
van 1P 8U
vid2 2A 6U 7A
v.r1 1P 5U
zaa 3U 4P
su2 1P 2P 5U
suu1 1P 6P 2A
stambh 1U 5P 9P
svid2 1A 4P
*)
(* More precise selection for present system *)
value voices_of_gana g root = match g with
[ 1 -> match root with
[ ".r" | "k.r.s" | "cur" | "tap" | "budh#1" | "van" | "v.r#1" | "su#2"
| "suu#1"
-> Para (* but ".r" Atma for pv sam \Pan{1,3,29} *)
| "i" | "gha.t.t" | "ghuur.n" | ".damb" | "bhra.mz" | "mid" | "mok.s"
| "lok" | "svid#2"
-> Atma
| "i.s#1" | "j.rr" | "daa#1" | "dh.r.s" | "as#2" | "kuc"
| "m.rj" | "m.rd#1" | "rud#1" | "stambh"
-> Ubha
| "kliiba" -> Atma (* denominative verb *)
| _ -> voices_of root (* man U (epic P) *)
]
| 2 -> match root with
[ "daa#1" | "dyaa" | "draa#2" | "maa#1" | "m.rj" | "rud#1" | "su#2"
-> Para
| "nij" | "vid#2" | "suu#1" -> Atma
| _ -> voices_of root
]
| 3 -> match root with
[ ".r" -> Para
| "maa#1" -> Atma
| "daa#1" | "nij" -> Ubha
| _ -> voices_of root
]
| 4 -> match root with
[ "as#2" | "j.rr" | "bhra.mz" | "mid" | "zaa"
| "svid#2" -> Para
| "i" | "jan" | "jyaa#1" | "tap" | "draa#2" | "budh#1" | "mii" | "ri"
| "ric" | "m.r" -> Atma
| "pu.s#1" (* | "raadh" Bergaigne vedic *) -> Ubha
| _ -> voices_of root
]
| 5 -> match root with
[ "i" | ".r" | "dh.r.s" | "raadh" | "stambh" -> Para
| "v.r#1" | "su#2" -> Ubha
| _ -> voices_of root
]
| 6 -> match root with
[ "kuc" | "ghuur.n" | "suu#1" -> Para
| "k.r.s" | "m.rj" | "rud#1" | "vid#2" -> Ubha
| _ -> voices_of root
]
| 7 -> match root with
[ "vid#2" -> Atma
| "ric" -> Para
| _ -> voices_of root
]
| 8 -> match root with
[ "man" -> Atma
| _ -> voices_of root (* van Ubha *)
]
| 9 -> match root with
[ "jyaa#1" | "pu.s#1" | "mii" | "m.rd#1" | "ri" -> Para
| _ -> voices_of root
]
| 10 -> match root with
[ "gha.t.t" | ".damb" | "mid" | "mok.s" | "lak.s" | "lok" | "stambh"
-> Para
| "arth" -> Atma
| _ -> voices_of root (* other denominatives will take Ubha as default *)
]
| _ -> voices_of root
]
;
(* Refining with potential preverb *)
value voices_of_pv upasarga gana = fun (* gana only used for "tap" "i" *)
(* Paninian requirements *)
[ "zru" | ".r" | "gam" | "svar" | "vid#1" (* | "praz" *) ->
if upasarga = "sam" then Atma else Para (* \Pan{1,3,29} *)
(* "praz" used in Atma with aa- but also without pv in epics (MW) *)
| "car" -> if upasarga = "sam" then Ubha else Para (* \Pan{1,3,54} *)
| "viz#1" -> if upasarga = "ni" then Atma else Para (* \Pan{1,3,17} *)
| "huu" -> match upasarga with
[ "ni" | "sam" | "upa" | "vi" -> Atma (* \Pan{1,3,30} *)
| "aa" -> Ubha (* \Pan{1,3,31} *)
| _ -> Para
]
| "yam" -> match upasarga with
[ "aa" | "upa" -> Ubha
| _ -> Para (* \Pan{1,3,28} and \Pan{1,3,56} *)
]
| "vah#1" -> if upasarga = "pra" then Para else Ubha (* \Pan{1,3,81} *)
| "vad" -> match upasarga with
[ "anu" -> Ubha (* \Pan{1,3,49} *)
| "apa" -> Atma (* \Pan{1,3,73} *)
| _ -> Para
]
| "g.rr#1" -> match upasarga with
[ "ava" -> Atma (* \Pan{1,3,51} *)
| "sam" -> Ubha (* \Pan{1,3,52} *)
| _ -> Para
]
| "ji" -> match upasarga with
[ "vi" | "paraa" -> Atma (* \Pan{1,3,19} *)
| _ -> Ubha (* was Para but "satyam eva jayate" *)
]
| "krii.d" -> match upasarga with
[ "aa" | "anu" | "pari" -> Atma (* \Pan{1,3,21} *)
| "sam" -> Ubha (* \Pan{1,3,21} vaartikaa *)
| _ -> Para
]
| "m.rz" -> if upasarga = "pari" then Para else Ubha (* \Pan{1,3,82} *)
| "tap" when gana = 1 -> match upasarga with
[ "ut" | "vi" -> Ubha
| _ -> Para (* \Pan{1,3,27} *)
]
| "i" when gana = 2 -> match upasarga with
[ "adhi" -> Ubha
| _ -> Para
]
| "zii#1" -> if upasarga = "sam" then Ubha else Atma
| "krii" -> match upasarga with
[ "vi" | "pari" | "ava" -> Atma
| _ -> Para (* \Pan{1,3,18} *)
]
(* Next three equivalent to marking "unused" in lexicon *)
| "ta~nc" | "saa#1" | "zam#2" | "zal" (* also "khyaa" ? *) ->
match upasarga with
[ "" -> raise Unattested
| _ -> Para
]
| "loc" | "zrambh" | "hnu" -> match upasarga with
[ "" -> raise Unattested
| _ -> Atma
]
| ".damb" -> match upasarga with
[ "vi" -> Ubha
| _ -> raise Unattested
]
(* Usage, MW *)
| "gha.t.t" -> if gana = 1 then
if upasarga = "" then raise Unattested
else Atma (* only "vi" | "sam", NOT "" *)
else (* gana = 10 *) Para
| "i.s#1" when gana = 1 -> match upasarga with
[ "" -> raise Unattested
| _ -> Ubha
]
| root -> voices_of_gana gana root
]
;
(*i end; i*)