Module:sa-verb/data
Jump to navigation
Jump to search
- This module lacks a documentation subpage. Please create it.
- Useful links: root page • root page’s subpages • links • transclusions • testcases • sandbox
-- This is a data module for [[Module:sa-verb]].
local conj_data = {}
local sa_utils = require("Module:sa-utilities")
local to_IAST = require("Module:sa-utilities/translit/SLP1-to-IAST").tr
local sub = mw.ustring.sub
local gsub = mw.ustring.gsub
local gasub = string.gsub -- For when byte by byte comparison is good enough.
local match = mw.ustring.match
local split = mw.text.split
-- Returns { form, stem } if the stem is explicity provided in the form (i.e. form = "rinakti<rinac>" or form = "rincanti<rinc>")
local function split_xs(form)
if form == nil then
return nil
end
local s = split(form, "<")
if s[2] ~= nil then
local sanitized_s2 = sub(s[2], 1, -2)
return {s[1], sanitized_s2}
else
return {form, nil}
end
end
-- Splits "stem1,stem2" into { stem1, stem2 }
local function split_stems(form)
-- TODO: in the older system, we'd have entries where multiple stems could be supplied and show up side-by-side in the same table.
-- I.e. "áricat" and "áraikṣīt" as two aorist variants of "riṇakti". The current system is to make them both separate entries altogether
-- so their conjugation systems don't overlap and cause confusion.
-- Older entries using the older version of this module need to be updated—right now, any extra stems that are supplied are just ignored
-- Note: this doesn't affect manually provided forms, like irregular third-person singular active perfects having long/short-vowel forms
return split(form, ",")[1]
end
local function apply_ending(args, data, tag, stem, es, is_part)
if type(es) == "string" then
es = {es}
end
local forms = data.forms[tag] or {}
local i = #forms + 1
for _, e in ipairs(es) do
local note
if type(e) == "table" then
note = e.note
e = e[1]
end
-- reduce Vedic forms to zero in case of novedic parameter
if args.novedic == true and note and match(note, "[vV]edic$") then
stem = ""; e = ""; note = ""
else
args.has_accent = match(stem, sa_utils.accent)
local set_stem = stem
-- if 'set', add 'i' between consonants, except before 'y' (optative)
if args.set == true and match(stem, sa_utils.consonant .. "$") and match(e, "^[tTDmrvshQz]") then
set_stem = sa_utils.internal_sandhi({
stem = stem,
ending = "i",
has_accent = args.has_accent,
non_final = true,
mono = args.mono,
accent_override = args.accent_override,
no_retroflex_root_s = args.no_retroflex_root_s
})
end
if args.auto_sandhi == true then
forms[i] = sa_utils.internal_sandhi({
stem = set_stem,
ending = e,
has_accent = args.has_accent,
non_final = is_part,
mono = args.mono,
j_to_z = args.j_to_z,
h_to_g = args.h_to_g,
diaspirate = args.diaspirate,
no_syncope = args.no_syncope,
accent_override = args.accent_override,
no_retroflex_root_s = args.no_retroflex_root_s -- for roots 'pis'/'niṃs'/'hiṃs' and 'tresus' (from 'tras')
})
else
forms[i] = set_stem .. e
end
-- Apply special tagging, currently just Ⓛ.
forms[i] = gasub(forms[i],
"a("..sa_utils.accent.."?)("..sa_utils.consonant.."a)Ⓛ", "A%1%2")
forms[i] = gasub(forms[i], "Ⓛ", "") -- Clean up.
if note then
forms["note" .. i] = note
end
i = i + 1
end
end
data.forms[tag] = forms
end
local function make_forms(args, data, stem, forms)
for mood, mood_forms in pairs(forms) do
if args.tense == "nonf" then
local tag = mood
apply_ending(args, data, tag, stem, mood_forms, false)
else
for voice, voice_forms in pairs(mood_forms) do
if mood == "part" then
local tag = mood .. "_" .. voice
apply_ending(args, data, tag, stem, voice_forms, true)
else
for person_number, es in pairs(voice_forms) do
local tag = mood .. "_" .. voice .. "_" .. person_number
apply_ending(args, data, tag, stem, es, false)
end
end
end
end
end
end
local function validate(stem, lemma)
if stem == nil then
error("could not detect stem from " .. to_IAST(lemma) .. "; set args.o to fill in values manually")
end
end
local function detect(t, lemma, match_pattern, oxy_match_pattern, strong)
local prefix = "weak_"
if strong then prefix = "strong_" end
if not t[prefix .. "did_lemma"] then
lemma = split_stems(lemma)
local splitted = split_xs(lemma)
if splitted == nil then return nil end
t[prefix .. "3s"] = splitted[1]
t[prefix .. "stem"] = splitted[2]
t[prefix .. "oxy"] = ""
t[prefix .. "did_lemma"] = true
end
local detected_stem = match(t[prefix .. "3s"], match_pattern)
if detected_stem ~= nil then
if oxy_match_pattern ~= nil then
t[prefix .. "oxy"] = match(t[prefix .. "3s"], oxy_match_pattern)
end
if t[prefix .. "stem"] == nil then
t[prefix .. "stem"] = detected_stem
end
return true
end
return false
end
local function detect_strong(t, lemma, match_pattern, oxy_match_pattern)
return detect(t, lemma, match_pattern, oxy_match_pattern, true)
end
local function detect_weak(t, lemma, match_pattern, oxy_match_pattern)
return detect(t, lemma, match_pattern, oxy_match_pattern, false)
end
local function use_strong_for_weak(t, weak_lemma)
if weak_lemma ~= nil then return false end
t["weak_stem"] = t["strong_stem"]
t["weak_oxy"] = t["strong_oxy"]
return true
end
-- only for class 3 verbs (with 3p -ati), override with 'class=' parameter // does not detect 'iyāy' and 'alar' (Whitney §1002e)
local function detect_intensive(args, t)
-- anchoring detection pattern at the end to allow for prefixes
if match(t.strong_stem, sa_utils.consonant.."a/?[nrlv][iI]"..sa_utils.consonant.."+"..sa_utils.vowel.."[MH]?"..sa_utils.consonant.."*$") then
-- dissyllabic reduplication
return true
elseif match(t.strong_stem, sa_utils.consonant.."[Aeo]/?"..sa_utils.consonant.."+"..sa_utils.vowel.."[MH]?"..sa_utils.consonant.."*$")
or match(t.strong_stem, sa_utils.consonant.."a/?[rlNYRnmMd]"..sa_utils.consonant.."+"..sa_utils.vowel.."[MH]?"..sa_utils.consonant.."?$")
or args.class == "int" then
t["intensive"] = "1"
return true
end
end
conj_data["pres"] = {}
setmetatable(conj_data["pres"], {
__call = function (self, args, data)
local make_thematic_forms = function (oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "A" .. oxy .. "mi",
["2_s"] = "a" .. oxy .. "si",
["3_s"] = "a" .. oxy .. "ti",
["1_d"] = "A" .. oxy .. "vas",
["2_d"] = "a" .. oxy .. "Tas",
["3_d"] = "a" .. oxy .. "tas",
["1_p"] = "A" .. oxy .. "mas",
["2_p"] = "a" .. oxy .. "Ta",
["3_p"] = "a" .. oxy .. "nti"
},
["mv"] = {
["1_s"] = "e" .. oxy .. "",
["2_s"] = "a" .. oxy .. "se",
["3_s"] = "a" .. oxy .. "te",
["1_d"] = "A" .. oxy .. "vahe",
["2_d"] = "e" .. oxy .. "Te",
["3_d"] = "e" .. oxy .. "te",
["1_p"] = "A" .. oxy .. "mahe",
["2_p"] = "a" .. oxy .. "Dve",
["3_p"] = "a" .. oxy .. "nte"
}
},
["imper"] = {
["av"] = {
["1_s"] = "A" .. oxy .. "ni",
-- ["2_s"] = {"a" .. oxy, "a" .. oxy .. "tAt"},
["2_s"] = "a" .. oxy,
["3_s"] = "a" .. oxy .. "tu",
["1_d"] = "A" .. oxy .. "va",
["2_d"] = "a" .. oxy .. "tam",
["3_d"] = "a" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "ma",
["2_p"] = "a" .. oxy .. "ta",
["3_p"] = "a" .. oxy .. "ntu"
},
["mv"] = {
["1_s"] = "E" .. oxy .. "",
["2_s"] = "a" .. oxy .. "sva",
["3_s"] = "a" .. oxy .. "tAm",
["1_d"] = "A" .. oxy .. "vahE",
["2_d"] = "e" .. oxy .. "TAm",
["3_d"] = "e" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "mahE",
["2_p"] = "a" .. oxy .. "Dvam",
["3_p"] = "a" .. oxy .. "ntAm"
},
},
["optat"] = {
["av"] = {
["1_s"] = "e" .. oxy .. "yam",
["2_s"] = "e" .. oxy .. "s",
["3_s"] = "e" .. oxy .. "t",
["1_d"] = "e" .. oxy .. "va",
["2_d"] = "e" .. oxy .. "tam",
["3_d"] = "e" .. oxy .. "tAm",
["1_p"] = "e" .. oxy .. "ma",
["2_p"] = "e" .. oxy .. "ta",
["3_p"] = "e" .. oxy .. "yus"
},
["mv"] = {
["1_s"] = "e" .. oxy .. "ya",
["2_s"] = "e" .. oxy .. "TAs",
["3_s"] = "e" .. oxy .. "ta",
["1_d"] = "e" .. oxy .. "vahi",
["2_d"] = "e" .. oxy .. "yATAm",
["3_d"] = "e" .. oxy .. "yAtAm",
["1_p"] = "e" .. oxy .. "mahi",
["2_p"] = "e" .. oxy .. "Dvam",
["3_p"] = "e" .. oxy .. "ran"
},
},
["part"] = {
["av"] = "a" .. oxy .. "t",
["mv"] = "a" .. oxy .. "mAna"
}
}
end
local make_athematic_strong_forms_cons = function ()
return {
["indic"] = {
["av"] = {
["1_s"] = "mi",
["2_s"] = "si",
["3_s"] = "ti",
},
},
["imper"] = {
["av"] = {
["3_s"] = "tu",
}
}
}
end
local make_athematic_strong_forms_vow = function ()
return {
["imper"] = {
["av"] = {
["1_s"] = "Ani",
["1_d"] = "Ava",
["1_p"] = "Ama",
},
["mv"] = {
["1_s"] = "E",
["1_d"] = "AvahE",
["1_p"] = "AmahE",
}
}
}
end
local make_athematic_weak_forms_cons = function (oxy) -- endings starting with consonant not m/v/y
return {
["indic"] = {
["av"] = {
["2_d"] = "Ta" .. oxy .. "s",
["3_d"] = "ta" .. oxy .. "s",
["2_p"] = "Ta" .. oxy,
},
["mv"] = {
["2_s"] = "se" .. oxy,
["3_s"] = "te" .. oxy,
["2_p"] = "Dve" .. oxy,
}
},
["imper"] = {
["av"] = {
-- ["2_s"] = {"Di" .. oxy, "tA" .. oxy .. "t"},
["2_d"] = "ta" .. oxy .. "m",
["3_d"] = "tA" .. oxy .. "m",
["2_p"] = "ta" .. oxy .. "",
},
["mv"] = {
["2_s"] = "sva" .. oxy,
["3_s"] = "tA" .. oxy .. "m",
["2_p"] = "Dva" .. oxy .. "m",
}
}
}
end
local make_athematic_weak_forms_mvy = function (oxy) -- endings starting with m/v/y
return {
["indic"] = {
["av"] = {
["1_d"] = "va" .. oxy .. "s",
["1_p"] = "ma" .. oxy .. "s",
},
["mv"] = {
["1_d"] = "va" .. oxy .. "he",
["1_p"] = "ma" .. oxy .. "he",
}
},
["optat"] = {
["av"] = {
["1_s"] = "yA" .. oxy .. "m",
["2_s"] = "yA" .. oxy .. "s",
["3_s"] = "yA" .. oxy .. "t",
["1_d"] = "yA" .. oxy .. "va",
["2_d"] = "yA" .. oxy .. "tam",
["3_d"] = "yA" .. oxy .. "tAm",
["1_p"] = "yA" .. oxy .. "ma",
["2_p"] = "yA" .. oxy .. "ta",
["3_p"] = "yu" .. oxy .. "s"
}
}
}
end
local make_athematic_weak_forms_vow = function (oxy) -- endings starting with vowels
return {
["indic"] = {
["av"] = {
["3_p"] = "a" .. oxy .. "nti"
},
["mv"] = {
["1_s"] = "e" .. oxy,
["2_d"] = "A" .. oxy .. "Te",
["3_d"] = "A" .. oxy .. "te",
["3_p"] = "a" .. oxy .. "te"
}
},
["imper"] = {
["av"] = {
["3_p"] = "a" .. oxy .. "ntu"
},
["mv"] = {
["2_d"] = "A" .. oxy .. "TAm",
["3_d"] = "A" .. oxy .. "tAm",
["3_p"] = "a" .. oxy .. "tAm"
}
},
["optat"] = {
["mv"] = {
["1_s"] = "Iya" .. oxy,
["2_s"] = "ITA" .. oxy .. "s",
["3_s"] = "Ita" .. oxy,
["1_d"] = "Iva" .. oxy .. "hi",
["2_d"] = "IyA" .. oxy .. "TAm",
["3_d"] = "IyA" .. oxy .. "tAm",
["1_p"] = "Ima" .. oxy .. "hi",
["2_p"] = "IDva" .. oxy .. "m",
["3_p"] = "Ira" .. oxy .. "n"
}
},
["part"] = {
["av"] = "a" .. oxy .. "t",
["mv"] = "Ana" .. oxy,
}
}
end
local make_athematic_class_3_forms_vow = function() -- endings starting with vowel for class 3 (no oxy)
return {
["indic"] = {
["av"] = {
["3_p"] = "ati"
},
["mv"] = {
["1_s"] = "e",
["2_d"] = "ATe",
["3_d"] = "Ate",
["3_p"] = "ate"
}
},
["imper"] = {
["av"] = {
["3_p"] = "atu"
},
["mv"] = {
["2_d"] = "ATAm",
["3_d"] = "AtAm",
["3_p"] = "atAm"
}
},
["optat"] = {
["mv"] = {
["1_s"] = "Iya",
["2_s"] = "ITAs",
["3_s"] = "Ita",
["1_d"] = "Ivahi",
["2_d"] = "IyATAm",
["3_d"] = "IyAtAm",
["1_p"] = "Imahi",
["2_p"] = "IDvam",
["3_p"] = "Iran"
}
},
["part"] = {
["av"] = "at",
["mv"] = "Ana",
}
}
end
local make_extra_class_5_8_forms = function(oxy)
return {
["indic"] = {
["av"] = {
["1_d"] = "va" .. oxy .. "s",
["1_p"] = "ma" .. oxy .. "s",
},
["mv"] = {
["1_d"] = "va" .. oxy .. "he",
["1_p"] = "ma" .. oxy .. "he",
}
},
["imper"] = {
["av"] = {
["2_s"] = {"u" .. oxy, {"uhi" .. oxy, note = "Vedic"}}
}
}
}
end
local make_intensive_weak_forms_extra = function() -- not in case of dissyllabic reduplication (Whitney §1007a)
return {
["indic"] = {
["av"] = {
["1_s"] = "Imi",
["2_s"] = "Izi",
["3_s"] = "Iti",
}
},
["imper"] = {
["av"] = {
["3_s"] = "Itu",
}
}
}
end
local make_athematic_imper_2s_Di = function(oxy)
return {
["imper"] = {
["av"] = {
["2_s"] = "Di" .. oxy
}
}
}
end
local make_athematic_imper_2s_hi = function(oxy)
return {
["imper"] = {
["av"] = {
["2_s"] = "hi" .. oxy
}
}
}
end
local make_class_9_cons_imper_2s = function(oxy)
return {
["imper"] = {
["av"] = {
["2_s"] = "Ana" .. oxy
}
}
}
end
local t = {}
local is_thematic = args.weak_lemma == nil
if is_thematic then
if not detect_strong(t, args.strong_lemma, "(.+)a" .. sa_utils.accent .. "?ti$", "a(" .. sa_utils.accent .. "?)ti$") then
if not detect_strong(t, args.strong_lemma, "(.+)a" .. sa_utils.accent .. "?te$", "a(" .. sa_utils.accent .. "?)te$") then
validate(t["strong_stem"], args.strong_lemma)
else
args.n = "m" -- deponent
end
end
make_forms(args, data, t["strong_stem"], make_thematic_forms(t["strong_oxy"]))
-- class 10 (Whitney §1043f)
if (args.class == "10" or not args.class) and match(t.strong_stem, sa_utils.vowel
.."M?"..sa_utils.consonant.."+a" .. sa_utils.accent .. "?y$") then
args.accent_override = true
make_forms(args, data, t["strong_stem"], {
["part"] = {
["mv"] = {{ "Ana", note = "Later Sanskrit" }},
}
})
end
-- active athematic verbs
elseif detect_strong(t, args.strong_lemma, "(.+)[tDwQ]i$") or match(args.strong_lemma, "^<.+>$") then
if not detect_weak(t, args.weak_lemma, "(.+)a" .. sa_utils.accent .. "?nti$", "a(" .. sa_utils.accent .. "?)nti$") then
if not detect_weak(t, args.weak_lemma, "(.+A)" .. sa_utils.accent .. "?nti$", "A(" .. sa_utils.accent .. "?)nti$") then
if not detect_weak(t, args.weak_lemma, "(.+)ati$", sa_utils.vowel .. "(" .. sa_utils.accent .. "?).+ati$") then
validate(t["weak_stem"], args.weak_lemma)
else
t.ati_3p = true
end
end
end
-- if no strong stem supplied, check weak stem in case of possible sandhi
if match(args.strong_lemma, "[ktp]ti$") then
t.strong_stem = gasub(t.strong_stem, ".$", gasub(t.weak_stem, ".+(.)$", "%1"))
-- if 3s ends on -dhi/-ṭi
elseif match(args.strong_lemma, "[Dw]i$") then
if match(t.weak_stem, "kz$") then
t.strong_stem = gasub(t.strong_stem, "z$", "kz")
else
t.strong_stem = gasub(t.strong_stem, ".$", gasub(t.weak_stem, ".+(.)$", "%1"))
if match(t.strong_stem, "h$") then
args.h_to_g = true
elseif match(t.strong_stem, "j$") then
args.j_to_z = true
end
end
-- if 3s ends on -ḍhi
elseif match(args.strong_lemma, "Qi$") then
t.strong_stem = gasub(t.strong_stem, "^(.*)$", "%1h")
-- error for roots on -i/ī/u/ū when no weak stem provided
elseif match(args.weak_lemma, sa_utils.consonant.."i?ya/?n?ti$")
or ( match(args.weak_lemma, "uva/?n?ti$") and not match(t.strong_stem, "[nR]o/?")) then
error("Please add weak stem for roots on -i/-ī and -u/-ū.")
end
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_cons())
if t.ati_3p == true and match (t.strong_stem, "I$") then -- assuming 3s on -īti is intensive (as a normal class 3 root on -ī would have -eti)
args.n = "a"
make_forms(args, data, t["weak_stem"], make_athematic_strong_forms_vow()) -- using weak stem with (normally) strong endings
elseif t.ati_3p == true and (args.class == "int" or not args.class) and detect_intensive(args, t) then -- intensive verbs
args.n = "a" -- no middle (maybe add middle participle?)
make_forms(args, data, t["weak_stem"], make_athematic_strong_forms_vow())
if t.intensive == "1" then -- no dissyllabic reduplication
if match(t.strong_stem, sa_utils.vowel_with_accent.."$") then -- Whitney §1004a
make_forms(args, data, t["strong_stem"], make_intensive_weak_forms_extra())
else
make_forms(args, data, t["weak_stem"], make_intensive_weak_forms_extra())
end
end
else
if match(t.strong_stem, "O/?$") and match(t.weak_stem, "u$") then -- Whitney §626
t.strong_stem = gasub(t.strong_stem, "O(/?)$", "a%1v")
elseif match(t.strong_stem, "a/?vI$") and match(t.weak_stem, "[uU]v?$") then -- e.g. 'brū', see Whitney §632-3
t.strong_stem = gasub(t.strong_stem, "I$", "")
elseif match(t.strong_stem, "i$") and match(t.weak_stem, "[^iy]$") then -- for 'set' verbs (Whitney §631)
t.strong_stem = gasub(t.strong_stem, "i$", "")
end
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_vow())
end
if match(args.weak_lemma, "ati$") then -- if 3p on -ati and no weak stem provided
args.accent_override = true
-- for class 5/8 verbs on vowel + -noti
elseif match(t.strong_stem, sa_utils.vowel .. "[nR]o/?$") then
t.temp_stem = gasub(t.strong_stem, "o/?$", "")
make_forms(args, data, t["temp_stem"], make_extra_class_5_8_forms(t["weak_oxy"])) -- includes active imper. 2s
t.imper_2s_created = true
-- for verbs like āpnoti
elseif match(t.strong_stem, sa_utils.consonant .. "[nR]o/?$") then
args.no_syncope = true
t.weak_stem = gasub(t.weak_stem, "uv$", "u")
-- class 9
elseif match(t.strong_stem, "[nR]A/?$") and match(t.weak_stem, "[nR]$") then
t.weak_stem = gasub(t.weak_stem, "^.*$", "%1I")
end
if t.imper_2s_created == true then
-- do nothing
elseif match(t.weak_stem, sa_utils.consonant .."[nR]I$") then -- Whitney §722
t.temp_stem = gasub(t.weak_stem, "[nR]I$", "")
make_forms(args, data, t["temp_stem"], make_class_9_cons_imper_2s(t["weak_oxy"]))
elseif match(t.weak_stem, sa_utils.vowel .. "$") or match(t.weak_stem, sa_utils.consonant.."[yrv]$") or match(t.weak_stem, "^y$") then
make_forms(args, data, t["weak_stem"], make_athematic_imper_2s_hi(t["weak_oxy"]))
else
make_forms(args, data, t["weak_stem"], make_athematic_imper_2s_Di(t["weak_oxy"]))
end
make_forms(args, data, t["weak_stem"], make_athematic_weak_forms_cons(t["weak_oxy"]))
if args.extra_1p_stem then -- optional extra stem for endings starting with m/v/y
t.weak_stem = args.extra_1p_stem
end
make_forms(args, data, t["weak_stem"], make_athematic_weak_forms_mvy(t["weak_oxy"]))
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "a?" .. sa_utils.accent .. "?n?ti<.*>$", "")
elseif args.extra_1p_stem or match(t.weak_stem, "[nR]I$") then
t.weak_stem = gasub(args.weak_lemma, "a?" .. sa_utils.accent .. "?nti$", "")
end
if t.ati_3p == true then
args.accent_override = false
make_forms(args, data, t["weak_stem"], make_athematic_class_3_forms_vow())
else
make_forms(args, data, t["weak_stem"], make_athematic_weak_forms_vow(t["weak_oxy"]))
end
-- deponent verbs
elseif detect_strong(t, args.strong_lemma, "(.+)[tDwQ]e/?$", "[tDwQ]e(/?)$") then
args.n = "m" -- deponent
if not detect_weak(t, args.weak_lemma, "(.+"..sa_utils.accent..".+)ate$", sa_utils.vowel.."("..sa_utils.accent..").+ate$") then
if not detect_weak(t, args.weak_lemma, "(.+)a"..sa_utils.accent.."?te$", "a("..sa_utils.accent.."?)te$") then
if not detect_weak(t, args.weak_lemma, "(.+A/?)te$") then
validate(t["weak_stem"], args.weak_lemma)
end
end
elseif t.strong_oxy ~= "" then
t.accented_class_3 = true
end
-- for class 5/8 verbs on vowel + -nute (+ tarute)
-- adding '-no' for detection as this might be the strong stem supplied to form 1st person imperative
if match(t.strong_stem, sa_utils.vowel .. "[nRr][uo]/?$")
and (not args.class or match(args.class, "^[58]$")) and not args.extra_1p_stem then
t.temp_stem = gasub(t.strong_stem, "[uo]/?$", "")
make_forms(args, data, t["temp_stem"], make_extra_class_5_8_forms(t["weak_oxy"]))
-- for verbs like āpnoti, but middle (the regex also includes hnute, but this is ok)
elseif match(t.strong_stem, sa_utils.consonant .. "[nR][uo]/?$") then
args.no_syncope = true
end
-- produce 1st person imperative if strong stem is supplied with 3sg
if match(args.strong_lemma, "<.*>$") then
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_vow())
-- or if root has 'a/ā' (when guṇa would be the same),
-- and not accented class 3 (accent on strong stem unpredictable) or potentially class 7
elseif match(t.strong_stem, "[aA]"..sa_utils.accent.."?"..sa_utils.consonant.."*$")
and not t.accented_class_3 == true and not match(t.strong_stem, "[NYRnm]"..sa_utils.consonant.."$") then
-- look at 3p to avoid possible sandhi before 3s -te
t.strong_stem = gasub(t.weak_stem, "([aA])("..sa_utils.consonant.."*)$", "%1"..t.strong_oxy.."%2")
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_vow())
-- or if stem ends on -[nṇ]u (including hnu/snu is ok)
elseif match(t.strong_stem, "[nR]u$") then
t.strong_stem = gasub(t.strong_stem, "u$", "o"..t.strong_oxy)
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_vow())
-- or class 9
elseif match(t.strong_stem, "[nR]I$") and match(t.weak_stem, "[nR]$") then
t.strong_stem = gasub(t.strong_stem, "I$", "A"..t.strong_oxy)
make_forms(args, data, t["strong_stem"], make_athematic_strong_forms_vow())
end
if match(args.strong_lemma, sa_utils.vowel_with_accent .. "te/?$") then
t.weak_stem_before_cons = gasub(args.strong_lemma, "te/?$", "")
elseif match(args.strong_lemma, sa_utils.vowel_with_accent .. "te/?<.*>$") then
t.weak_stem_before_cons = gasub(args.strong_lemma, "te/?<.*>$", "")
else
t.weak_stem_before_cons = t.weak_stem
end
if t.accented_class_3 == true then
args.accent_override = true -- no accent on class 3 stem if ending starts with consonant
end
make_forms(args, data, t["weak_stem_before_cons"], make_athematic_weak_forms_cons(t["strong_oxy"]))
if args.extra_1p_stem then
t.weak_stem_before_cons = args.extra_1p_stem
end
make_forms(args, data, t["weak_stem_before_cons"], make_athematic_weak_forms_mvy(t["strong_oxy"]))
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "a?" .. sa_utils.accent .. "?te<.*>$", "")
end
if t.accented_class_3 == true then
args.accent_override = false
make_forms(args, data, t["weak_stem"], make_athematic_class_3_forms_vow())
else
make_forms(args, data, t["weak_stem"], make_athematic_weak_forms_vow(t["weak_oxy"]))
end
else
validate(t["strong_stem"], args.strong_lemma)
end
end
})
conj_data["impf"] = {}
setmetatable(conj_data["impf"], {
__call = function (self, args, data)
-- accent for augmentless forms
local make_thematic_impf_forms = function(oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "a" .. oxy .. "m",
["2_s"] = "a" .. oxy .. "s",
["3_s"] = "a" .. oxy .. "t",
["1_d"] = "A" .. oxy .. "va",
["2_d"] = "a" .. oxy .. "tam",
["3_d"] = "a" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "ma",
["2_p"] = "a" .. oxy .. "ta",
["3_p"] = "a" .. oxy .. "nt"
},
["mv"] = {
["1_s"] = "e" .. oxy,
["2_s"] = "a" .. oxy .. "TAs",
["3_s"] = "a" .. oxy .. "ta",
["1_d"] = "A" .. oxy .. "vahi",
["2_d"] = "e" .. oxy .. "TAm",
["3_d"] = "e" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "mahi",
["2_p"] = "a" .. oxy .. "Dvam",
["3_p"] = "a" .. oxy .. "nta"
},
}
}
end
local athematic_impf_strong_endings_cons = {
["indic"] = {
["av"] = {
["2_s"] = "s",
["3_s"] = "t",
},
},
}
local athematic_impf_strong_endings_vow = {
["indic"] = {
["av"] = {
["1_s"] = "am",
},
},
}
-- accent for augmentless forms
local make_athematic_impf_weak_forms_cons = function(oxy) -- endings starting with consonants not m/v/y
return {
["indic"] = {
["av"] = {
["2_d"] = "ta" .. oxy .. "m",
["3_d"] = "tA" .. oxy .. "m",
["2_p"] = "ta" .. oxy,
},
["mv"] = {
["2_s"] = "TA" .. oxy .. "s",
["3_s"] = "ta" .. oxy,
["2_p"] = "Dva" .. oxy .. "m",
}
}
}
end
local make_athematic_impf_weak_forms_mvy = function(oxy) -- endings starting with m/v/y
return {
["indic"] = {
["av"] = {
["1_d"] = "va" .. oxy,
["1_p"] = "ma" .. oxy,
},
["mv"] = {
["1_d"] = "va" .. oxy .. "hi",
["1_p"] = "ma" .. oxy .. "hi",
}
}
}
end
local make_athematic_impf_weak_forms_vow = function(oxy) -- endings starting with vowels (without active 3p)
return {
["indic"] = {
["av"] = {
-- ["3_p"] = "a" .. oxy .. "n",
},
["mv"] = {
["1_s"] = "i" .. oxy,
["2_d"] = "A" .. oxy .. "TAm",
["3_d"] = "A" .. oxy .. "tAm",
["3_p"] = "a" .. oxy .. "ta"
}
}
}
end
local make_athematic_impf_3p_an = function(oxy)
return {
["indic"] = {
["av"] = {
["3_p"] = "a" .. oxy .. "n"
}
}
}
end
local make_athematic_impf_3p_us = function(oxy)
return {
["indic"] = {
["av"] = {
["3_p"] = "u" .. oxy .. "s"
}
}
}
end
local intensive_impf_weak_endings_extra = {
["indic"] = {
["av"] = {
["2_s"] = "Is",
["3_s"] = "It",
},
},
}
local t = {}
local is_thematic = args.weak_lemma == nil
if is_thematic then
if not detect_strong(t, args.strong_lemma, "(.+)a/?t$", "a(/?)t$") then
if not detect_strong(t, args.strong_lemma, "(.+)a/?ta$", "a(/?)ta$") then
validate(t["strong_stem"], args.strong_lemma)
else
args.n = "m" -- deponent
end
end
make_forms(args, data, t["strong_stem"], make_thematic_impf_forms(t["strong_oxy"]))
-- active verbs
elseif detect_strong(t, args.strong_lemma, "(.*" .. sa_utils.vowel_with_accent .. ")t$")
or detect_strong(t, args.strong_lemma, "(.+[kNwRtnprlsH])$") or match(args.strong_lemma, "^<.+>$") then
if not detect_weak(t, args.weak_lemma, "(.+)a/?n$", "a(/?)n$") then
if not detect_weak(t, args.weak_lemma, "(.+A)/?n$", "A(/?)n$") then
if not detect_weak(t, args.weak_lemma, "(.+)u[Hs]$") then
validate(t["weak_stem"], args.weak_lemma)
else
t.us_3p = true
end
end
end
-- look at weak stem if 3s '-t' disappeared and no strong stem is supplied
if match(args.strong_lemma, "[kwpH]$") then
if args.diaspirate == true then
t.strong_stem = gasub(t.strong_stem, "([GDB])(.+)$", function(cons, post) return sa_utils.deaspirate[cons] .. post end)
end
if match(args.strong_lemma, "k$") then
t.strong_stem = gasub(t.strong_stem, ".$", gasub(t.weak_stem, ".+(.)$", "%1"))
if match(t.strong_stem, "h$") then
args.h_to_g = true
end
elseif match(t.weak_stem, "kz$") then
t.strong_stem = gasub(t.strong_stem, ".$", "kz")
else
t.strong_stem = gasub(t.strong_stem, ".$", gasub(t.weak_stem, ".+(.)$", "%1"))
if match(t.strong_stem, "j$") then args.j_to_z = true
elseif match(t.strong_stem, "f$") then t.strong_stem = gasub(t.strong_stem, ".$", "r")
end
end
elseif match(t.weak_stem, "[tTdD]$")
and match(args.strong_lemma, "[^tTdD]" .. sa_utils.vowel_with_accent.."t$") then
t.strong_stem = gasub(t.strong_stem, "^(.+)$", "%1" .. gasub(t.weak_stem, ".+(.)$", "%1"))
-- error for roots on -i/ī/u/ū when no weak stem provided
elseif match(args.weak_lemma, sa_utils.consonant.."i?ya/?n$")
or ( match(args.weak_lemma, "uva/?n$") and not match(t.strong_stem, "[nR]o/?")) then
error("Please add weak stem for roots on -i/-ī and -u/-ū.")
end
-- intensive verbs
if t.us_3p == true and match(t.strong_stem, "I$") then -- assuming 3s on -īt is intensive (as a normal class 3 root on -ī would have -et)
args.n = "a"
make_forms(args, data, t["weak_stem"], athematic_impf_strong_endings_vow) -- using weak stem with (normally) strong ending
elseif t.us_3p == true and (args.class == "int" or not args.class) and detect_intensive(args, t) then
args.n = "a" -- no middle
make_forms(args, data, t["weak_stem"], athematic_impf_strong_endings_vow)
else -- non-intensive
if match(args.strong_lemma, "O/?t$") and match(t.weak_stem, "u$") then -- Whitney §626
t.strong_stem_before_vow = gasub(t.strong_stem, "O(/?)$", "a%1v")
-- verbs with inserted 'ī' in 2/3 sg., see Whitney §631-3 (roots on -ī should have -e in strong stem)
elseif match(args.strong_lemma, "It$") then
t.strong_stem_before_vow = gasub(t.strong_stem, "I$", "")
else
t.strong_stem_before_vow = t.strong_stem
end
make_forms(args, data, t["strong_stem_before_vow"], athematic_impf_strong_endings_vow)
end
-- use 3s form for 2s and 3s (if appropriate)
local syllable = sa_utils.vowel_with_accent..sa_utils.consonant.."+"
if match(args.strong_lemma, "^"..syllable..sa_utils.vowel_with_accent
.."t<"..syllable..sa_utils.vowel_with_accent.."[tTdD]>$")
or match(args.strong_lemma, "^"..syllable..syllable..sa_utils.vowel_with_accent.."t<"
..syllable..syllable..sa_utils.vowel_with_accent.."[tTdD]>$") then
-- do nothing
elseif match(args.strong_lemma, sa_utils.vowel_with_accent.."t<.+>$") then
t.strong_stem = gasub(args.strong_lemma, "t<.+>$", "")
end
make_forms(args, data, t["strong_stem"], athematic_impf_strong_endings_cons)
-- if intensive without dissyllabic reduplication
if t.intensive == "1" then
if match(t.strong_stem, sa_utils.vowel_with_accent.."$") then -- Whitney §1004a
make_forms(args, data, t["strong_stem"], intensive_impf_weak_endings_extra)
else
make_forms(args, data, t["weak_stem"], intensive_impf_weak_endings_extra)
end
end
-- e.g. āpnot
if match(t.strong_stem, sa_utils.consonant .. "[nR]o/?$") then
args.no_syncope = true
t.weak_stem = gasub(t.weak_stem, "uv$", "u")
-- extra forms for verbs on vowel + -noti
elseif match(t.strong_stem, sa_utils.vowel .. "[nR]o/?$") then
t.temp_stem = gasub(t.strong_stem, "o/?$", "")
make_forms(args, data, t["temp_stem"], make_athematic_impf_weak_forms_mvy(t["weak_oxy"]))
-- class 9
elseif match(t.strong_stem, "[nR]A/?$") and match(t.weak_stem, "[nR]$") then
t.weak_stem = gasub(t.weak_stem, "^.*$", "%1I")
-- class 3 roots on -ṛ
elseif match(args.weak_lemma, ".aru[Hs]$") and not match(t.strong_stem, "A/?$") then
t.weak_stem = gasub(t.weak_stem, "ar$", "f")
-- class 3 roots on -i/ī/u/ū with no weak stem provided
elseif match(args.weak_lemma, ".a[vy]u[Hs]$") and not match(t.strong_stem, "A/?$") then
error("Please add weak stem for class 3 roots on -i/-ī and -u/-ū.")
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_weak_forms_cons(t["weak_oxy"]))
if args.extra_1p_stem then
t.weak_stem = args.extra_1p_stem
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_weak_forms_mvy(t["weak_oxy"]))
if match(args.weak_lemma, "n<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "a?/?n<.*>$", "")
elseif args.extra_1p_stem or match(t.weak_stem, "[nR]I$") then
t.weak_stem = gasub(args.weak_lemma, "a/?n$", "")
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_weak_forms_vow(t["weak_oxy"]))
if t.us_3p == true then
-- active 3p for class 3 verbs is often not from the weak stem
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, 'u[Hs]<.*>$', '')
else
t.weak_stem = gasub(t.weak_stem, "f$", "ar")
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_3p_us(t["weak_oxy"]))
else
if match(t.strong_stem, "A$") and match(t.weak_stem, "A$") then -- Whitney §621a
t.temp_stem = gasub(t.weak_stem, "A$", "")
make_forms(args, data, t["temp_stem"], make_athematic_impf_3p_us(t["weak_oxy"]))
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_3p_an(t["weak_oxy"]))
end
-- deponent verbs
elseif detect_strong(t, args.strong_lemma, "(.+)[tDwQ]a/?$") then
if detect_weak(t, args.weak_lemma, "(.+)a/?ta$", "a(/?)ta$") or detect_weak(t, args.weak_lemma, "(.+A/?)ta$") then
args.n = "m" -- deponent
-- for verbs on vowel + -nuta (+ impf. of tarute)
if match(t.strong_stem, sa_utils.vowel .. "[nRr]u$")
and (not args.class or match(args.class, "^[58]$")) and not args.extra_1p_stem then
t.temp_stem = gasub(t.strong_stem, "u$", "")
make_forms(args, data, t["temp_stem"], make_athematic_impf_weak_forms_mvy(t["weak_oxy"]))
-- e.g. āpnot, but middle
elseif match(t.strong_stem, sa_utils.consonant .. "[nR]u?$") then
args.no_syncope = true
end
if match(args.strong_lemma, sa_utils.vowel_with_accent .. "ta$") then
t.weak_stem_before_cons = gasub(args.strong_lemma, "ta$", "")
else
t.weak_stem_before_cons = t.weak_stem
end
make_forms(args, data, t["weak_stem_before_cons"], make_athematic_impf_weak_forms_cons(t["weak_oxy"]))
if args.extra_1p_stem then
t.weak_stem_before_cons = args.extra_1p_stem
end
make_forms(args, data, t["weak_stem_before_cons"], make_athematic_impf_weak_forms_mvy(t["weak_oxy"]))
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "a?/?ta<.*>$", "")
end
make_forms(args, data, t["weak_stem"], make_athematic_impf_weak_forms_vow(t["weak_oxy"]))
else
validate(t["weak_stem"], args.weak_lemma)
end
else
validate(t["strong_stem"], args.strong_lemma)
end
end
})
conj_data["fut"] = {}
setmetatable(conj_data["fut"], {
__call = function (self, args, data)
local make_strong_forms = function (oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "syA" .. oxy .. "mi",
["2_s"] = "sya" .. oxy .. "si",
["3_s"] = "sya" .. oxy .. "ti",
["1_d"] = "syA" .. oxy .. "vas",
["2_d"] = "sya" .. oxy .. "Tas",
["3_d"] = "sya" .. oxy .. "tas",
["1_p"] = "syA" .. oxy .. "mas",
["2_p"] = "sya" .. oxy .. "Ta",
["3_p"] = "sya" .. oxy .. "nti"
},
["mv"] = {
["1_s"] = "sye" .. oxy,
["2_s"] = "sya" .. oxy .. "se",
["3_s"] = "sya" .. oxy .. "te",
["1_d"] = "syA" .. oxy .. "vahe",
["2_d"] = "sye" .. oxy .. "Te",
["3_d"] = "sye" .. oxy .. "te",
["1_p"] = "syA" .. oxy .. "mahe",
["2_p"] = "sya" .. oxy .. "Dve",
["3_p"] = "sya" .. oxy .. "nte"
}
},
["part"] = {
["av"] = "sya" .. oxy .. "t",
["mv"] = "sya" .. oxy .. "mAna"
}
}
end
local t = {}
if not detect_strong(t, args.strong_lemma, "(.+).ya" .. sa_utils.accent .. "?ti$", "(" .. sa_utils.accent .. "?)ti$") then
if not detect_strong(t, args.strong_lemma, "(.+).ya" .. sa_utils.accent .. "?te$", "(" .. sa_utils.accent .. "?)te$") then
validate(t["strong_stem"], args.strong_lemma)
else
args.n = "m" -- deponent
end
end
make_forms(args, data, t["strong_stem"], make_strong_forms(t["strong_oxy"]))
table.insert(data.categories, "Sanskrit verbs with s-future")
end
})
conj_data["pfut"] = {}
setmetatable(conj_data["pfut"], {
__call = function (self, args, data)
local make_perph_forms = function (oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "tA" .. oxy .. "smi",
["2_s"] = "tA" .. oxy .. "si",
["3_s"] = "tA" .. oxy,
["1_d"] = "tA" .. oxy .. "svas",
["2_d"] = "tA" .. oxy .. "sTas",
["3_d"] = {"tA" .. oxy .. "rO", {"tA" .. oxy .. "rA", note = "Vedic"}}, -- 'gantārā' occurs in Rigveda 8.13.10
["1_p"] = "tA" .. oxy .. "smas",
["2_p"] = "tA" .. oxy .. "sTa",
["3_p"] = "tA" .. oxy .. "ras"
},
["mv"] = {
["1_s"] = "tA" .. oxy .. "he",
["2_s"] = "tA" .. oxy .. "se",
["3_s"] = "tA" .. oxy,
["1_d"] = "tA" .. oxy .. "svahe",
["2_d"] = "tA" .. oxy .. "sATe",
["3_d"] = "tA" .. oxy .. "rO",
["1_p"] = "tA" .. oxy .. "smahe",
["2_p"] = "tA" .. oxy .. "Dve",
["3_p"] = "tA" .. oxy .. "ras"
}
},
}
end
local t = {}
if not detect_strong(t, args.strong_lemma, "(.+)tA" .. sa_utils.accent .. "?$", "(" .. sa_utils.accent .. "?)$") then
validate(t["strong_stem"], args.strong_lemma)
-- for the periphrastic future, there is no way to tell if the verb is deponent from the lemma (third-person singular form)
-- atmanepada verbs require the "n=m" argument to avoid showing the active forms
end
make_forms(args, data, t["strong_stem"], make_perph_forms(t["strong_oxy"]))
table.insert(data.categories, "Sanskrit verbs with periphrastic future")
end
})
conj_data["cond"] = {}
setmetatable(conj_data["cond"], {
__call = function (self, args, data)
local t = {}
if not detect_strong(t, args.strong_lemma, "(.+).yat$") then
if not detect_strong(t, args.strong_lemma, "(.+).yata$") then
validate(t["strong_stem"], args.strong_lemma)
else
args.n = "m" -- deponent
end
end
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["1_s"] = "syam",
["2_s"] = "syas",
["3_s"] = "syat",
["1_d"] = "syAva",
["2_d"] = "syatam",
["3_d"] = "syatAm",
["1_p"] = "syAma",
["2_p"] = "syata",
["3_p"] = "syant"
},
["mv"] = {
["1_s"] = "sye",
["2_s"] = "syaTAs",
["3_s"] = "syata",
["1_d"] = "syAvahi",
["2_d"] = "syeTAm",
["3_d"] = "syetAm",
["1_p"] = "syAmahi",
["2_p"] = "syaDvam",
["3_p"] = "syanta"
}
}
})
end
})
conj_data["aor"] = {}
setmetatable(conj_data["aor"], {
__call = function (self, args, data)
-- there is some added complexity here compared to the other tenses due to the number of aorist types in Sanskrit
local strong_is_endings = {
["indic"] = {
["av"] = {
["1_s"] = "izam",
["2_s"] = "Is",
["3_s"] = "It",
["1_d"] = "izva",
["2_d"] = "izwam",
["3_d"] = "izwAm",
["1_p"] = "izma",
["2_p"] = "izwa",
["3_p"] = "izus"
},
}
}
local weak_is_endings = {
["indic"] = {
["mv"] = {
["1_s"] = "izi",
["2_s"] = "izWAs",
["3_s"] = "izwa",
["1_d"] = "izvahi",
["2_d"] = "izATAm",
["3_d"] = "izAtAm",
["1_p"] = "izmahi",
["2_p"] = "iQvam",
["3_p"] = "izata"
},
}
}
local strong_s_endings = {
["indic"] = {
["av"] = {
["1_s"] = "sam",
["2_s"] = {"sIs", {"s", note = "Vedic"}},
["3_s"] = {"sIt", {"s", note = "Vedic"}},
["1_d"] = "sva",
["2_d"] = "stam",
["3_d"] = "stAm",
["1_p"] = "sma",
["2_p"] = "sta",
["3_p"] = "sus"
},
}
}
local weak_s_endings = {
["indic"] = {
["mv"] = {
["1_s"] = "si",
["2_s"] = "sTAs",
["3_s"] = "sta",
["1_d"] = "svahi",
["2_d"] = "sATAm",
["3_d"] = "sAtAm",
["1_p"] = "smahi",
["2_p"] = "sDvam", -- further treated by sandhi module
["3_p"] = "sata"
},
}
}
-- accent for augmentless forms
local make_strong_sa_forms = function(oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "sa" .. oxy .. "m",
["2_s"] = "sa" .. oxy .. "s",
["3_s"] = "sa" .. oxy .. "t",
["1_d"] = "sA" .. oxy .. "va",
["2_d"] = "sa" .. oxy .. "tam",
["3_d"] = "sa" .. oxy .. "tAm",
["1_p"] = "sA" .. oxy .. "ma",
["2_p"] = "sa" .. oxy .. "ta",
["3_p"] = "sa" .. oxy .. "n"
},
}
}
end
local make_weak_sa_forms = function (oxy)
return {
["indic"] = {
["mv"] = {
["1_s"] = "si" .. oxy,
["2_s"] = "sa" .. oxy .. "TAs",
["3_s"] = "sa" .. oxy .. "ta",
["1_d"] = "sA" .. oxy .. "vahi",
["2_d"] = "sA" .. oxy .. "TAm",
["3_d"] = "sA" .. oxy .. "tAm",
["1_p"] = "sA" .. oxy .. "mahi",
["2_p"] = "sa" .. oxy .. "Dvam",
["3_p"] = "sa" .. oxy .. "nta"
},
}
}
end
-- accent for augmentless forms
local make_strong_a_forms = function(oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "a" .. oxy .. "m",
["2_s"] = "a" .. oxy .. "s",
["3_s"] = "a" .. oxy .. "t",
["1_d"] = "A" .. oxy .. "va",
["2_d"] = "a" .. oxy .. "tam",
["3_d"] = "a" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "ma",
["2_p"] = "a" .. oxy .. "ta",
["3_p"] = "a" .. oxy .. "nt"
},
}
}
end
local make_weak_a_forms = function(oxy)
return {
["indic"] = {
["mv"] = {
["1_s"] = "e" .. oxy,
["2_s"] = "a" .. oxy .. "TAs",
["3_s"] = "a" .. oxy .. "ta",
["1_d"] = "A" .. oxy .. "vahi",
["2_d"] = "e" .. oxy .. "TAm",
["3_d"] = "e" .. oxy .. "tAm",
["1_p"] = "A" .. oxy .. "mahi",
["2_p"] = "a" .. oxy .. "Dvam",
["3_p"] = "a" .. oxy .. "nta"
},
}
}
end
local strong_root_endings = { -- without 3p
["indic"] = {
["av"] = {
["1_s"] = "am",
["2_s"] = "s",
["3_s"] = "t",
["1_d"] = "va",
["2_d"] = "tam",
["3_d"] = "tAm",
["1_p"] = "ma",
["2_p"] = "ta",
-- ["3_p"] = "us" -- for bhū: -an (abhūvan)
}
}
}
-- only for some Vedic forms, the root-aorist can only be in active voice in Classical Sanskrit
-- to some extent speculative
local weak_root_endings = {
["indic"] = {
["mv"] = {
["1_s"] = "i",
["2_s"] = "TAs",
["3_s"] = "ta",
["1_d"] = "vahi",
["2_d"] = "ATAm",
["3_d"] = "AtAm", -- 'akrātām' (Whitney §834a)
["1_p"] = "mahi",
["2_p"] = "Dvam",
["3_p"] = "ata" -- but '-ran' frequently occurs, see Whitney §834b
},
}
}
local t = {}
local weak_lemma = args.weak_lemma or args.passive_lemma
-- detection of oxytone accent in case of injunctive
if (not args.aor or args.aor == "sa") and detect_strong(t, args.strong_lemma, "(.+k)za/?t$", "kza(/?)t$") then
table.insert(data.categories, "Sanskrit verbs with sa-aorist")
make_forms(args, data, t["strong_stem"], make_strong_sa_forms(t["strong_oxy"]))
if use_strong_for_weak(t, weak_lemma) then
make_forms(args, data, t["weak_stem"], make_weak_sa_forms(t["weak_oxy"]))
end
elseif (not args.aor or args.aor == "s" or args.aor == "ṣ") and detect_strong(t, args.strong_lemma, "(.+)[sz]It$") then
table.insert(data.categories, "Sanskrit verbs with s-aorist")
make_forms(args, data, t["strong_stem"], strong_s_endings)
if use_strong_for_weak(t, weak_lemma) then
make_forms(args, data, t["weak_stem"], weak_s_endings)
end
elseif (not args.aor or args.aor == "iṣ" or args.aor == "siṣ") and detect_strong(t, args.strong_lemma, "(.+)It$") then
if args.aor == "siṣ" then
table.insert(data.categories, "Sanskrit verbs with siṣ-aorist")
make_forms(args, data, t["strong_stem"], strong_is_endings)
-- middle forms for siṣ-aorist are not allowed by the grammarians (but might have existed: Whitney §915)
else
table.insert(data.categories, "Sanskrit verbs with iṣ-aorist")
make_forms(args, data, t["strong_stem"], strong_is_endings)
if use_strong_for_weak(t, weak_lemma) then
make_forms(args, data, t["weak_stem"], weak_is_endings)
end
end
elseif (not args.aor or args.aor == "a") and detect_strong(t, args.strong_lemma, "(.+)a/?t$", "a(/?)t$") then
table.insert(data.categories, "Sanskrit verbs with a-aorist")
make_forms(args, data, t["strong_stem"], make_strong_a_forms(t["strong_oxy"]))
if use_strong_for_weak(t, weak_lemma) then
make_forms(args, data, t["weak_stem"], make_weak_a_forms(t["weak_oxy"]))
end
-- the 't' of 'avart' should be detected as part of the stem
elseif (not args.aor or args.aor == "root") and (detect_strong(t, args.strong_lemma, "(.+" .. sa_utils.vowel .. ")t$")
or detect_strong(t, args.strong_lemma, "(.+[kNwRtnprlsH])$") ) then
table.insert(data.categories, "Sanskrit verbs with root aorist")
make_forms(args, data, t["strong_stem"], strong_root_endings)
-- for active 3p, remove final ā from stem
if match(t.strong_stem, "A$") then
t.strong_stem = gasub(t.strong_stem, "A$", "")
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["3_p"] = "us"
}
}
})
else
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["3_p"] = "an"
}
}
})
end
elseif args.strong_lemma ~= "-" then
error("Could not detect aorist type from " .. to_IAST(args.strong_lemma)) -- failed to recognize an aorist type
end
if weak_lemma then
-- if ends on -gdha, -ddha or -bdha, then the stem necessarily ends on resp. -h/dh/bh
local aspirate = { ['g'] = 'h', ['d'] = 'D', ['b'] = 'B' }
weak_lemma = gasub(weak_lemma, '(.+)([gdb])(Da)$', function(a,b,c) return a .. aspirate[b].. c end)
if match(weak_lemma, "hDa$") then args.h_to_g = true end
if (not args.aor or args.aor == "iṣ") and detect_weak(t, weak_lemma, "(.+)izwa$") then
table.insert(data.categories, "Sanskrit verbs with iṣ-aorist")
make_forms(args, data, t["weak_stem"], weak_is_endings)
elseif (not args.aor or args.aor == "sa") and detect_weak(t, weak_lemma, "(.+k)za/?ta$", "kza(/?)ta$") then
table.insert(data.categories, "Sanskrit verbs with sa-aorist")
make_forms(args, data, t["weak_stem"], make_weak_sa_forms(t["weak_oxy"]))
elseif (not args.aor or args.aor == "[sṣ]") and (detect_weak(t, weak_lemma, "(.+)sta$") or detect_weak(t, weak_lemma, "(.+)zwa$")
or detect_weak(t, weak_lemma, "(.+[ktp])ta$") or detect_weak(t, weak_lemma, "(.+[ghdDbB])Da$")) then
table.insert(data.categories, "Sanskrit verbs with s-aorist")
make_forms(args, data, t["weak_stem"], weak_s_endings)
elseif (not args.aor or args.aor == "a") and detect_weak(t, weak_lemma, "(.+)a/?ta$", "a(/?)ta$") then
table.insert(data.categories, "Sanskrit verbs with a-aorist")
make_forms(args, data, t["weak_stem"], make_weak_a_forms(t["weak_oxy"]))
elseif (not args.aor or args.aor == "root") and (detect_weak(t, weak_lemma, "(.+)ta$")
or detect_weak(t, weak_lemma, "(.+[ghdDbB])Da$")) then
table.insert(data.categories, "Sanskrit verbs with root-aorist")
make_forms(args, data, t["weak_stem"], weak_root_endings)
else
error("Could not detect aorist type from " .. to_IAST(weak_lemma)) -- failed to recognize an aorist type
end
end
end
})
conj_data["bene"] = {}
setmetatable(conj_data["bene"], {
__call = function (self, args, data)
local make_strong_forms = function (oxy)
return {
["optat"] = {
["av"] = {
["1_s"] = "yA" .. oxy .. "sam",
["2_s"] = "yA" .. oxy .. "s",
["3_s"] = "yA" .. oxy .. "t",
["1_d"] = "yA" .. oxy .. "sva",
["2_d"] = "yA" .. oxy .. "stam",
["3_d"] = "yA" .. oxy .. "stAm",
["1_p"] = "yA" .. oxy .. "sma",
["2_p"] = "yA" .. oxy .. "sta",
["3_p"] = "yA" .. oxy .. "sus"
},
}
}
end
local make_weak_forms = function (oxy)
return {
["optat"] = {
["mv"] = {
["1_s"] = "sIya" .. oxy,
["2_s"] = "sIzWA" .. oxy .. "s",
["3_s"] = "sIzwa" .. oxy,
["1_d"] = "sIva" .. oxy .. "hi",
["2_d"] = {{"sIyA" .. oxy .. "sTAm", note = "Uncertain"}}, -- Whitney §924a: "of very questionable value"
["3_d"] = {{"sIyA" .. oxy .. "stAm", note = "Uncertain"}},
["1_p"] = "sIma" .. oxy .. "hi",
["2_p"] = "sIQva" .. oxy .. "m",
["3_p"] = "sIra" .. oxy .. "nt"
},
}
}
end
local t = {}
if detect_strong(t, args.strong_lemma, "(.+)yA" .. sa_utils.accent .. "?t$", "yA(" .. sa_utils.accent .. "?)t$") then
make_forms(args, data, t["strong_stem"], make_strong_forms(t["strong_oxy"]))
elseif args.strong_lemma ~= "-" then
validate(t["strong_stem"], args.strong_lemma)
end
local weak_lemma = args.weak_lemma or args.passive_lemma
if use_strong_for_weak(t, weak_lemma) then
make_forms(args, data, t["weak_stem"], make_weak_forms(t["weak_oxy"]))
elseif detect_weak(t, weak_lemma, "(.+)[sz]Izwa" .. sa_utils.accent .. "?$", "[sz]Izwa(" .. sa_utils.accent .. "?)$") then
make_forms(args, data, t["weak_stem"], make_weak_forms(t["weak_oxy"]))
else
validate(t["weak_stem"], weak_lemma)
end
end
})
conj_data["perf"] = {}
setmetatable(conj_data["perf"], {
__call = function (self, args, data)
local make_weak_forms_cons = function (oxy) -- endings starting with consonant
return {
["indic"] = {
["av"] = {
["1_d"] = "va" .. oxy,
["1_p"] = "ma" .. oxy,
},
["mv"] = {
["2_s"] = "se" .. oxy,
["1_d"] = "va" .. oxy .. "he",
["1_p"] = "ma" .. oxy .. "he",
["2_p"] = "Dve" .. oxy,
}
},
["part"] = {
-- ["av"] = "vA" .. oxy .. "Ms",
}
}
end
local make_weak_forms_vow = function (oxy) -- endings starting with vowel
return {
["indic"] = {
["av"] = {
["2_d"] = "a" .. oxy .. "Tus",
["3_d"] = "a" .. oxy .. "tus",
["2_p"] = "a" .. oxy,
["3_p"] = "u" .. oxy .. "s"
},
["mv"] = {
["1_s"] = "e" .. oxy,
["3_s"] = "e" .. oxy,
["2_d"] = "A" .. oxy .. "Te",
["3_d"] = "A" .. oxy .. "te",
["3_p"] = "ire" .. oxy
}
},
["part"] = {
["mv"] = "Ana" .. oxy
}
}
end
local make_vedic_anit_forms = function (oxy)
return {
["indic"] = {
["av"] = {
["1_d"] = {{"va" .. oxy, note = "Vedic"}},
["1_p"] = {{"ma" .. oxy, note = "Vedic"}},
},
["mv"] = {
["2_s"] = {{"se" .. oxy, note = "Vedic"}},
["1_d"] = {{"va" .. oxy .. "he", note = "Vedic"}},
["1_p"] = {{"ma" .. oxy .. "he", note = "Vedic"}},
["2_p"] = {{"Dve" .. oxy, note = "Vedic"}},
["3_p"] = {{"re" .. oxy, note = "Vedic"}},
}
}
}
end
local make_anit_perf_participle = function(oxy)
return { ["part"] = { ["av"] = "vA" .. oxy .. "Ms" } }
end
local make_set_perf_participle = function(oxy)
return { ["part"] = { ["av"] = "ivA" .. oxy .. "Ms" } }
end
local make_strong_au_forms = function(oxy)
return {
["indic"] = {
["av"] = {
["1_s"] = "O" .. oxy,
["2_s"] = {"A" .. oxy .. "Ta", {"iTa" .. oxy, note = "Later Sanskrit"}},
["3_s"] = "O" .. oxy,
},
}
}
end
local anit_perf_2s_ending = {
["indic"] = {
["av"] = {
["2_s"] = "Ta"
}
}
}
local set_perf_2s_ending = {
["indic"] = {
["av"] = {
["2_s"] = "iTa"
}
}
}
local make_weak_perf_2s = function(oxy)
return {
["indic"] = {
["av"] = {
["2_s"] = {{"iTa" .. oxy, note = "Later Sanskrit"}}
}
}
}
end
local t = {}
if detect_strong(t, args.strong_lemma, "(.+)O/?$", "O(/?)$") then
make_forms(args, data, t["strong_stem"], make_strong_au_forms(t["strong_oxy"]))
args.set = true
make_forms(args, data, t["strong_stem"], make_weak_forms_cons(t["strong_oxy"]))
make_forms(args, data, t["strong_stem"], make_set_perf_participle(t["strong_oxy"]))
make_forms(args, data, t["strong_stem"], make_weak_forms_vow(t["strong_oxy"]))
else
if not detect_strong(t, args.strong_lemma, "(.+)a$") then
if not detect_strong(t, args.strong_lemma, "(.+)e" .. sa_utils.accent .. "?$", "e(" .. sa_utils.accent .. "?)$") then
validate(t["strong_stem"], args.strong_lemma)
else
args.n = "m" -- deponent
end
end
local weak_lemma = args.weak_lemma or args.passive_lemma
if not use_strong_for_weak(t, weak_lemma) then
if not detect_weak(t, weak_lemma, "(.+)u" .. sa_utils.accent .. "?[Hs]$", "u(" .. sa_utils.accent .. "?)[Hs]$") then
validate(t["weak_stem"], weak_lemma)
end
end
-- if no strong stem supplied in case of e.g. cakāra, look at weak stem
if match(args.strong_lemma, "A"..sa_utils.accent.."?"..sa_utils.consonant.."a$")
and not match(t.weak_stem, "A"..sa_utils.accent.."?"..sa_utils.consonant.."$") then
t.strong_stem = gasub(t.strong_stem, "A("..sa_utils.accent.."?.)$", "a%1")
end
-- if Brugmann's law is applicable
if match(t.strong_stem, "a"..sa_utils.accent.."?"..sa_utils.consonant.."$")
or match(t.strong_stem, "[eo]"..sa_utils.accent.."?$") then
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["1_s"] = {"a", {"aⓁ", note = "Later Sanskrit"}}, -- only 3 verbs with vriddhi-form in Vedic (Whitney §793d)
["3_s"] = "aⓁ",
},
}
})
else
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["1_s"] = "a",
["3_s"] = "a",
},
}
})
end
if args.weak_lemma and ( match(args.weak_lemma, sa_utils.consonant.."i?yu/?[Hs]$") or match(args.weak_lemma, "uvu/?[Hs]$") ) then
error("Please add weak stem for active participle for roots on -i/-ī and -u/-ū.")
elseif match(t.weak_stem, sa_utils.consonant .. 'r$') then
t.weak_stem = gasub(t.weak_stem, 'r$', 'f') -- change -r of weak stem to -ṛ if after consonant
elseif match(t.weak_stem, "[uU]$") or match(t.weak_stem, sa_utils.consonant..sa_utils.consonant.."[iI]$") then
args.no_syncope = true
end
-- anit roots (on -ṛ or -u)
if match(t.weak_stem, "[uf]$") and not args.set == true then
make_forms(args, data, t["strong_stem"], anit_perf_2s_ending)
make_forms(args, data, t["weak_stem"], make_weak_forms_cons(t["weak_oxy"]))
make_forms(args, data, t["weak_stem"], make_anit_perf_participle(t["weak_oxy"]))
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "u/?[Hs]<.*>$", "")
end
make_forms(args, data, t["weak_stem"], make_weak_forms_vow(t["weak_oxy"]))
else
args.set = false -- as this interferes with adding -tha
-- rules for 2s
if (match(t.strong_stem, "uva/?.$") and match(t.weak_stem, "U.$")) -- e.g. uvac-
or (match(t.strong_stem, "iya/?.$") and match(t.weak_stem, "I.$")) -- iyaj-
or (match(t.strong_stem, "[aA]/?[NYRnmM]?[^yrv]$") and (match(t.weak_stem, "e.$") -- e.g. tatan- / ten-
or match(t.weak_stem, sa_utils.consonant..sa_utils.consonant.."$"))) -- e.g. jajan- / jajñ-
or match(t.strong_stem, "U/?v$") -- babhūv-
or match(t.strong_stem, "a/?[yvr]$") or match(t.strong_stem, "[eo]/?$") then -- e.g. ninay-, dadhar-
make_forms(args, data, t["strong_stem"], anit_perf_2s_ending) -- add -tha
t["Ta_added"] = true
end
if match(t.strong_stem, "a/?[^yrlv]$") and match(t.weak_stem, "e.$") then
-- or match(t.weak_stem, sa_utils.consonant..sa_utils.consonant.."$")) then -- this specific line might be wrong
make_forms(args, data, t["weak_stem"], make_weak_perf_2s(t["weak_oxy"]))
else
make_forms(args, data, t["strong_stem"], set_perf_2s_ending) -- add -itha
end
if match(t.strong_stem, "a"..sa_utils.accent.."?"..sa_utils.consonant.."$") and not t.Ta_added then
make_forms(args, data, t["strong_stem"], {
["indic"] = {
["av"] = {
["2_s"] = {{"Ta", note = "Vedic"}}
}
}
})
end
-- use (supplied) weak stem for participle
if sa_utils.is_monosyllabic(t.weak_stem) or args.mono == true then -- not accurate for some verbs (Whitney §803a)
make_forms(args, data, t["weak_stem"], make_set_perf_participle(t["weak_oxy"]))
else
make_forms(args, data, t["weak_stem"], make_anit_perf_participle(t["weak_oxy"]))
end
-- change weak stem to form used in 3p
if args.weak_lemma then
if match(args.weak_lemma, ".<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "u/?[Hs]<.*>$", "")
else
t.weak_stem = gasub(t.weak_stem, "("..sa_utils.consonant..")f$", "%1r")
t.weak_stem = gasub(t.weak_stem, "[uU]$", "uv")
t.weak_stem = gasub(t.weak_stem, "("..sa_utils.vowel..sa_utils.consonant..")[iI]$", "%1y")
t.weak_stem = gasub(t.weak_stem, "("..sa_utils.consonant..sa_utils.consonant..")[iI]$", "%1iy")
end
end
args.set = true
make_forms(args, data, t["weak_stem"], make_weak_forms_cons(t["weak_oxy"]))
make_forms(args, data, t["weak_stem"], make_weak_forms_vow(t["weak_oxy"]))
-- Vedic rules from Whitney §798a (omitting y/r/l/v from the consonants
-- as r+r is not allowed and none of them are among Wh.'s examples)
if match(t.weak_stem, "[aiufx][kKgGNcCjJYwWqQRtTdDnpPbBmLSzsh]$") then
args.set = false
if args.weak_lemma and match(args.weak_lemma, "<.*>$") then
t.weak_stem = gasub(args.weak_lemma, "^.*<(.*)>$", "%1")
end
make_forms(args, data, t["weak_stem"], make_vedic_anit_forms(t["weak_oxy"]))
end
end
end
end
})
local function legacy_split_stems(form)
return split(form, ",")
end
-- Gets stems for COMPLEX_FORM given a pattern in MATCH_RE
local function legacy_get_stem(complex_form, match_re)
local s = split_xs(complex_form)
if s[2] ~= nil then return s[2] end
return match(s[1], match_re)
end
-- Returns { stems, accents }
local function legacy_get_stems_from_lemmas(lemma, match_pattern, oxy_match_pattern)
local strong_lemmas = legacy_split_stems(lemma)
local strong_stems = {}
local oxys = nil
if oxy_match_pattern ~= nil then
oxys = {}
end
for i, strong_lemma in ipairs(strong_lemmas) do
local prov = split_xs(strong_lemma)[1]
if oxy_match_pattern ~= nil then
oxys[i] = match(prov, oxy_match_pattern)
end
strong_stems[i] = legacy_get_stem(strong_lemma, match_pattern)
validate(strong_stems[i], strong_lemma)
end
return { strong_stems, oxys }
end
conj_data["nonf"] = {}
setmetatable(conj_data["nonf"], {
-- TODO: this whole system needs to be removed
-- it's kept just so as to not break verbal entries which use this module, but non-finite forms
-- should now be part of the root derivations and not tied to the verbal conjugation
__call = function (self, args, data)
local make_strong_forms = function (oxy)
return {
["gerundive_mn"] = { "ya", "tavya" .. oxy, "anI" .. oxy .. "ya" },
["gerundive_f"] = { "yA", "tavyA" .. oxy, "anI" .. oxy .. "yA" }
}
end
local make_weak_forms = function (oxy)
return {
["gerund"] = "tvA" .. oxy,
["part_mn"] = "ta" .. oxy,
["part_f"] = "tA" .. oxy
}
end
local strong_stems = legacy_get_stems_from_lemmas(
args.strong_lemma,
"(.+)ya$",
"(" .. sa_utils.accent .. "?)"
)
for i, strong_stem in ipairs(strong_stems[1]) do
make_forms(args, data, strong_stem, {
["inf"] = "tum"
})
strong_stem = gsub(strong_stem, sa_utils.accent, "")
make_forms(args, data, strong_stem, make_strong_forms(strong_stems[2][i]))
local ya_form = #(data.forms["gerundive_mn"]) - 2 -- TODO: This is not a great way of doing this
local prov = split_xs(legacy_split_stems(args.strong_lemma)[i])[1]
data.forms["gerundive_mn"][1] = prov
data.forms["gerundive_f"][1] = match(prov, "(.+)a$") .. "A"
end
local weak_lemma = args.weak_lemma or args.passive_lemma
if weak_lemma == nil then
for i, strong_stem in ipairs(strong_stems[1]) do
make_forms(args, data, gsub(strong_stem, sa_utils.accent, ""), make_weak_forms(strong_stems[2][i]))
end
else
local weak_stems = legacy_get_stems_from_lemmas(
weak_lemma,
"(.+).a" .. sa_utils.accent .. "?$",
"a(" .. sa_utils.accent .. "?)$"
)
for i, weak_stem in ipairs(weak_stems[1]) do
make_forms(args, data, weak_stem, make_weak_forms(weak_stems[2][i]))
end
end
end
})
return conj_data