diff options
Diffstat (limited to 'contrib/snowball/algorithms/greek.sbl')
| -rw-r--r-- | contrib/snowball/algorithms/greek.sbl | 706 |
1 files changed, 706 insertions, 0 deletions
diff --git a/contrib/snowball/algorithms/greek.sbl b/contrib/snowball/algorithms/greek.sbl new file mode 100644 index 0000000..02df6c3 --- /dev/null +++ b/contrib/snowball/algorithms/greek.sbl @@ -0,0 +1,706 @@ +// A stemmer for Modern Greek language, based on: +// +// Ntais, Georgios. Development of a Stemmer for the Greek +// Language. Diss. Royal Institute of Technology, 2006. +// https://sais.se/mthprize/2007/ntais2007.pdf +// +// Saroukos, Spyridon. Enhancing a Greek language stemmer. +// University of Tampere, 2008. +// https://tampub.uta.fi/bitstream/handle/10024/80480/gradu03463.pdf + +stringescapes {} + +stringdef a '{U+03B1}' // alpha +stringdef v '{U+03B2}' // beta +stringdef g '{U+03B3}' // gamma +stringdef d '{U+03B4}' // delta +stringdef e '{U+03B5}' // epsilon +stringdef z '{U+03B6}' // zeta +stringdef i '{U+03B7}' // eta +stringdef th '{U+03B8}' // theta +stringdef y '{U+03B9}' // iota +stringdef k '{U+03BA}' // kappa +stringdef l '{U+03BB}' // lamda +stringdef m '{U+03BC}' // mu +stringdef n '{U+03BD}' // nu +stringdef x '{U+03BE}' // xi +stringdef o '{U+03BF}' // omicron +stringdef p '{U+03C0}' // pi +stringdef r '{U+03C1}' // rho +stringdef ss '{U+03C2}' // sigma final +stringdef s '{U+03C3}' // sigma +stringdef t '{U+03C4}' // tau +stringdef u '{U+03C5}' // upsilon +stringdef f '{U+03C6}' // phi +stringdef ch '{U+03C7}' // chi +stringdef ps '{U+03C8}' // psi +stringdef oo '{U+03C9}' // omega + +stringdef A '{U+0391}' // Alpha +stringdef V '{U+0392}' // Beta +stringdef G '{U+0393}' // Gamma +stringdef D '{U+0394}' // Delta +stringdef E '{U+0395}' // Epsilon +stringdef Z '{U+0396}' // Zeta +stringdef I '{U+0397}' // Eta +stringdef Th '{U+0398}' // Theta +stringdef Y '{U+0399}' // Iota +stringdef K '{U+039A}' // Kappa +stringdef L '{U+039B}' // Lamda +stringdef M '{U+039C}' // Mu +stringdef N '{U+039D}' // Nu +stringdef X '{U+039E}' // Xi +stringdef O '{U+039F}' // Omicron +stringdef P '{U+03A0}' // Pi +stringdef R '{U+03A1}' // Rho +stringdef S '{U+03A3}' // Sigma +stringdef T '{U+03A4}' // Tau +stringdef U '{U+03A5}' // Upsilon +stringdef F '{U+03A6}' // Phi +stringdef Ch '{U+03A7}' // Chi +stringdef Ps '{U+03A8}' // Psi +stringdef Oo '{U+03A9}' // Omega + +stringdef Y: '{U+03AA}' // Iota with dialytika +stringdef U: '{U+03AB}' // Upsilon with dialytika + +stringdef a' '{U+03AC}' // alpha with tonos +stringdef e' '{U+03AD}' // epsilon with tonos +stringdef i' '{U+03AE}' // eta with tonos +stringdef y' '{U+03AF}' // iota with tonos +stringdef o' '{U+03CC}' // omicron with tonos +stringdef u' '{U+03CD}' // upsilon with tonos +stringdef oo' '{U+03CE}' // omega with tonos + +stringdef i:' '{U+0390}' // iota with dialytika and tonos +stringdef u:' '{U+03B0}' // upsilon with dialytika and tonos + +stringdef i: '{U+03CA}' // iota with dialytika +stringdef u: '{U+03CB}' // upsilon with dialytika + +stringdef A' '{U+0386}' // Alpha with tonos +stringdef E' '{U+0388}' // Epsilon with tonos +stringdef I' '{U+0389}' // Eta with tonos +stringdef Y' '{U+038A}' // Iota with tonos +stringdef O' '{U+038C}' // Omicron with tonos +stringdef U' '{U+038E}' // Upsilon with tonos +stringdef OO' '{U+038F}' // Omega with tonos + +externals ( stem ) + +booleans ( test1 ) + +groupings ( v v2 ) + +routines ( tolower has_min_length + steps1 steps2 steps3 steps4 steps5 steps6 steps7 + steps8 steps9 steps10 + step1 step2a step2b step2c step2d step3 step4 + step5a step5b step5c step5d step5e step5f + step5g step5h step5i + step5j step5k step5l step5m + step6 step7 ) + +define v '{a}{e}{i}{y}{o}{u}{oo}' +define v2 '{a}{e}{i}{y}{o}{oo}' + +backwardmode ( + define has_min_length as ( + $(len >= 3) + ) + + define tolower as ( + repeat ( + [substring] among ( + '{A}' (<- '{a}') + '{V}' (<- '{v}') + '{G}' (<- '{g}') + '{D}' (<- '{d}') + '{E}' (<- '{e}') + '{Z}' (<- '{z}') + '{I}' (<- '{i}') + '{Th}' (<- '{th}') + '{Y}' (<- '{y}') + '{K}' (<- '{k}') + '{L}' (<- '{l}') + '{M}' (<- '{m}') + '{N}' (<- '{n}') + '{X}' (<- '{x}') + '{O}' (<- '{o}') + '{P}' (<- '{p}') + '{R}' (<- '{r}') + '{S}' (<- '{s}') + '{T}' (<- '{t}') + '{U}' (<- '{u}') + '{F}' (<- '{f}') + '{Ch}' (<- '{ch}') + '{Ps}' (<- '{ps}') + '{Oo}' (<- '{oo}') + '{Y:}' (<- '{y}') + '{U:}' (<- '{u}') + '{a'}' (<- '{a}') + '{e'}' (<- '{e}') + '{i'}' (<- '{i}') + '{y'}' (<- '{y}') + '{o'}' (<- '{o}') + '{u'}' (<- '{u}') + '{oo'}' (<- '{oo}') + '{i:'}' (<- '{i}') + '{u:'}' (<- '{u}') + '{i:}' (<- '{i}') + '{u:}' (<- '{u}') + '{A'}' (<- '{a}') + '{E'}' (<- '{e}') + '{I'}' (<- '{i}') + '{Y'}' (<- '{y}') + '{O'}' (<- '{o}') + '{U'}' (<- '{u}') + '{OO'}' (<- '{oo}') + '{ss}' (<- '{s}') + '' (next) + ) + ) + ) + + define step1 as ( + [substring] among ( + '{f}{a}{g}{y}{a}' '{f}{a}{g}{y}{o}{u}' '{f}{a}{g}{y}{oo}{n}' (<- '{f}{a}') + '{s}{k}{a}{g}{y}{a}' '{s}{k}{a}{g}{y}{o}{u}' '{s}{k}{a}{g}{y}{oo}{n}' (<- '{s}{k}{a}') + '{o}{l}{o}{g}{y}{o}{u}' '{o}{l}{o}{g}{y}{a}' '{o}{l}{o}{g}{y}{oo}{n}' (<- '{o}{l}{o}') + '{s}{o}{g}{y}{o}{u}' '{s}{o}{g}{y}{a}' '{s}{o}{g}{y}{oo}{n}' (<- '{s}{o}') + '{t}{a}{t}{o}{g}{y}{a}' '{t}{a}{t}{o}{g}{y}{o}{u}' '{t}{a}{t}{o}{g}{y}{oo}{n}' (<- '{t}{a}{t}{o}') + '{k}{r}{e}{a}{s}' '{k}{r}{e}{a}{t}{o}{s}' '{k}{r}{e}{a}{t}{a}' '{k}{r}{e}{a}{t}{oo}{n}' (<- '{k}{r}{e}') + '{p}{e}{r}{a}{s}' '{p}{e}{r}{a}{t}{o}{s}' '{p}{e}{r}{a}{t}{i}' '{p}{e}{r}{a}{t}{a}' '{p}{e}{r}{a}{t}{oo}{n}' (<- '{p}{e}{r}') + '{t}{e}{r}{a}{s}' '{t}{e}{r}{a}{t}{o}{s}' '{t}{e}{r}{a}{t}{a}' '{t}{e}{r}{a}{t}{oo}{n}' (<- '{t}{e}{r}') + '{f}{oo}{s}' '{f}{oo}{t}{o}{s}' '{f}{oo}{t}{a}' '{f}{oo}{t}{oo}{n}' (<- '{f}{oo}') + '{k}{a}{th}{e}{s}{t}{oo}{s}' '{k}{a}{th}{e}{s}{t}{oo}{t}{o}{s}' '{k}{a}{th}{e}{s}{t}{oo}{t}{a}' '{k}{a}{th}{e}{s}{t}{oo}{t}{oo}{n}' (<- '{k}{a}{th}{e}{s}{t}') + '{g}{e}{g}{o}{n}{o}{s}' '{g}{e}{g}{o}{n}{o}{t}{o}{s}' '{g}{e}{g}{o}{n}{o}{t}{a}' '{g}{e}{g}{o}{n}{o}{t}{oo}{n}' (<- '{g}{e}{g}{o}{n}') + ) + unset test1 + ) + + define steps1 as ( + [substring] among ( + '{y}{z}{a}' '{y}{z}{e}{s}' '{y}{z}{e}' '{y}{z}{a}{m}{e}' '{y}{z}{a}{t}{e}' '{y}{z}{a}{n}' '{y}{z}{a}{n}{e}' '{y}{z}{oo}' '{y}{z}{e}{y}{s}' '{y}{z}{e}{y}' + '{y}{z}{o}{u}{m}{e}' '{y}{z}{e}{t}{e}' '{y}{z}{o}{u}{n}' '{y}{z}{o}{u}{n}{e}' ( + delete + unset test1 + ([] substring atlimit among ( + '{a}{n}{a}{m}{p}{a}' '{e}{m}{p}{a}' '{e}{p}{a}' '{x}{a}{n}{a}{p}{a}' '{p}{a}' '{p}{e}{r}{y}{p}{a}' '{a}{th}{r}{o}' '{s}{u}{n}{a}{th}{r}{o}' '{d}{a}{n}{e}' + (<- '{y}') + )) or + ([] substring atlimit among ( + '{m}{a}{r}{k}' '{k}{o}{r}{n}' '{a}{m}{p}{a}{r}' '{a}{r}{r}' '{v}{a}{th}{u}{r}{y}' '{v}{a}{r}{k}' '{v}' '{v}{o}{l}{v}{o}{r}' '{g}{k}{r}' + '{g}{l}{u}{k}{o}{r}' '{g}{l}{u}{k}{u}{r}' '{y}{m}{p}' '{l}' '{l}{o}{u}' '{m}{a}{r}' '{m}' '{p}{r}' '{m}{p}{r}' '{p}{o}{l}{u}{r}' '{p}' + '{r}' '{p}{y}{p}{e}{r}{o}{r}' + (<- '{y}{z}') + )) + ) + ) + ) + + define steps2 as ( + [substring] among ( + '{oo}{th}{i}{k}{a}' '{oo}{th}{i}{k}{e}{s}' '{oo}{th}{i}{k}{e}' '{oo}{th}{i}{k}{a}{m}{e}' '{oo}{th}{i}{k}{a}{t}{e}' '{oo}{th}{i}{k}{a}{n}' '{oo}{th}{i}{k}{a}{n}{e}' ( + delete + unset test1 + [] substring atlimit among ( + '{a}{l}' '{v}{y}' '{e}{n}' '{u}{ps}' '{l}{y}' '{z}{oo}' '{s}' '{ch}' (<- '{oo}{n}') + ) + ) + ) + ) + + define steps3 as ( + [substring] among ( + '{y}{s}{a}' '{y}{s}{e}{s}' '{y}{s}{e}' '{y}{s}{a}{m}{e}' '{y}{s}{a}{t}{e}' '{y}{s}{a}{n}' '{y}{s}{a}{n}{e}' ( + delete + unset test1 + ('{y}{s}{a}' atlimit <- '{y}{s}') or + ([] substring atlimit among ( + '{a}{n}{a}{m}{p}{a}' '{a}{th}{r}{o}' '{e}{m}{p}{a}' '{e}{s}{e}' '{e}{s}{oo}{k}{l}{e}' '{e}{p}{a}' '{x}{a}{n}{a}{p}{a}' '{e}{p}{e}' '{p}{e}{r}{y}{p}{a}' + '{s}{u}{n}{a}{th}{r}{o}' '{d}{a}{n}{e}' '{k}{l}{e}' '{ch}{a}{r}{t}{o}{p}{a}' '{e}{x}{a}{r}{ch}{a}' '{m}{e}{t}{e}{p}{e}' '{a}{p}{o}{k}{l}{e}' + '{a}{p}{e}{k}{l}{e}' '{e}{k}{l}{e}' '{p}{e}' + (<- '{y}') + )) or + ([] substring atlimit among ( + '{a}{n}' '{a}{f}' '{g}{e}' '{g}{y}{g}{a}{n}{t}{o}{a}{f}' '{g}{k}{e}' '{d}{i}{m}{o}{k}{r}{a}{t}' '{k}{o}{m}' '{g}{k}' '{m}' '{p}' + '{p}{o}{u}{k}{a}{m}' '{o}{l}{o}' '{l}{a}{r}' + (<- '{y}{s}') + )) + ) + ) + ) + + define steps4 as ( + [substring] among ( + '{y}{s}{oo}' '{y}{s}{e}{y}{s}' '{y}{s}{e}{y}' '{y}{s}{o}{u}{m}{e}' '{y}{s}{e}{t}{e}' '{y}{s}{o}{u}{n}' '{y}{s}{o}{u}{n}{e}' ( + delete + unset test1 + [] substring atlimit among ( + '{a}{n}{a}{m}{p}{a}' '{e}{m}{p}{a}' '{e}{s}{e}' '{e}{s}{oo}{k}{l}{e}' '{e}{p}{a}' '{x}{a}{n}{a}{p}{a}' '{e}{p}{e}' '{p}{e}{r}{y}{p}{a}' '{a}{th}{r}{o}' + '{s}{u}{n}{a}{th}{r}{o}' '{d}{a}{n}{e}' '{k}{l}{e}' '{ch}{a}{r}{t}{o}{p}{a}' '{e}{x}{a}{r}{ch}{a}' '{m}{e}{t}{e}{p}{e}' '{a}{p}{o}{k}{l}{e}' '{a}{p}{e}{k}{l}{e}' + '{e}{k}{l}{e}' '{p}{e}' + (<- '{y}') + ) + ) + ) + ) + + define steps5 as ( + [substring] among ( + '{y}{s}{t}{o}{s}' '{y}{s}{t}{o}{u}' '{y}{s}{t}{o}' '{y}{s}{t}{e}' '{y}{s}{t}{o}{y}' '{y}{s}{t}{oo}{n}' '{y}{s}{t}{o}{u}{s}' '{y}{s}{t}{i}' '{y}{s}{t}{i}{s}' + '{y}{s}{t}{a}' '{y}{s}{t}{e}{s}' ( + delete + unset test1 + ([] substring atlimit among ( + '{d}{a}{n}{e}' '{s}{u}{n}{a}{th}{r}{o}' '{k}{l}{e}' '{s}{e}' '{e}{s}{oo}{k}{l}{e}' '{a}{s}{e}' '{p}{l}{e}' + (<- '{y}') + )) or + ([] substring atlimit among ( + '{m}' '{p}' '{a}{p}' '{a}{r}' '{i}{d}' '{k}{t}' '{s}{k}' '{s}{ch}' '{u}{ps}' '{f}{a}' '{ch}{r}' '{ch}{t}' '{a}{k}{t}' + '{a}{o}{r}' '{a}{s}{ch}' '{a}{t}{a}' '{a}{ch}{n}' '{a}{ch}{t}' '{g}{e}{m}' '{g}{u}{r}' '{e}{m}{p}' '{e}{u}{p}' '{e}{ch}{th}' '{i}{f}{a}' + '{k}{a}{th}' '{k}{a}{k}' '{k}{u}{l}' '{l}{u}{g}' '{m}{a}{k}' '{m}{e}{g}' '{t}{a}{ch}' '{f}{y}{l}' '{ch}{oo}{r}' + (<- '{y}{s}{t}') + )) + ) + ) + ) + + define steps6 as ( + [substring] among ( + '{y}{s}{m}{o}' '{y}{s}{m}{o}{y}' '{y}{s}{m}{o}{s}' '{y}{s}{m}{o}{u}' '{y}{s}{m}{o}{u}{s}' '{y}{s}{m}{oo}{n}' ( + delete + unset test1 + ([] substring atlimit among ( + '{s}{e}' '{m}{e}{t}{a}{s}{e}' '{m}{y}{k}{r}{o}{s}{e}' '{e}{g}{k}{l}{e}' '{a}{p}{o}{k}{l}{e}' + (<- '{y}{s}{m}') + )) or + ([] substring atlimit among ( + '{d}{a}{n}{e}' '{a}{n}{t}{y}{d}{a}{n}{e}' + (<- '{y}') + )) or + ([substring] among ( + '{a}{g}{n}{oo}{s}{t}{y}{k}' (<- '{a}{g}{n}{oo}{s}{t}') + '{a}{t}{o}{m}{y}{k}' (<- '{a}{t}{o}{m}') + '{g}{n}{oo}{s}{t}{y}{k}' (<- '{g}{n}{oo}{s}{t}') + '{e}{th}{n}{y}{k}' (<- '{e}{th}{n}') + '{e}{k}{l}{e}{k}{t}{y}{k}' (<- '{e}{k}{l}{e}{k}{t}') + '{s}{k}{e}{p}{t}{y}{k}' (<- '{s}{k}{e}{p}{t}') + '{t}{o}{p}{y}{k}' (<- '{t}{o}{p}') + '{a}{l}{e}{x}{a}{n}{d}{r}{y}{n}' (<- '{a}{l}{e}{x}{a}{n}{d}{r}') + '{v}{u}{z}{a}{n}{t}{y}{n}' (<- '{v}{u}{z}{a}{n}{t}') + '{th}{e}{a}{t}{r}{y}{n}' (<- '{th}{e}{a}{t}{r}') + )) + ) + ) + ) + + define steps7 as ( + [substring] among ( + '{a}{r}{a}{k}{y}' '{a}{r}{a}{k}{y}{a}' '{o}{u}{d}{a}{k}{y}' '{o}{u}{d}{a}{k}{y}{a}' ( + delete + unset test1 + [] substring atlimit among ( + '{s}' '{ch}' + (<- '{a}{r}{a}{k}') + ) + ) + ) + ) + + define steps8 as ( + [substring] among ( + '{a}{k}{y}' '{a}{k}{y}{a}' '{y}{t}{s}{a}' '{y}{t}{s}{a}{s}' '{y}{t}{s}{e}{s}' '{y}{t}{s}{oo}{n}' '{a}{r}{a}{k}{y}' '{a}{r}{a}{k}{y}{a}' ( + delete + unset test1 + ([] substring atlimit among ( + '{v}{a}{m}{v}' '{v}{r}' '{k}{a}{y}{m}' '{k}{o}{n}' '{k}{o}{r}' '{l}{a}{v}{r}' '{l}{o}{u}{l}' '{m}{e}{r}' '{m}{o}{u}{s}{t}' + '{n}{a}{g}{k}{a}{s}' '{p}{l}' '{r}' '{r}{u}' '{s}' '{s}{k}' '{s}{o}{k}' '{s}{p}{a}{n}' '{t}{z}' '{f}{a}{r}{m}' '{ch}' '{k}{a}{p}{a}{k}' + '{a}{l}{y}{s}{f}' '{a}{m}{v}{r}' '{a}{n}{th}{r}' '{k}' '{f}{u}{l}' '{k}{a}{t}{r}{a}{p}' '{k}{l}{y}{m}' '{m}{a}{l}' '{s}{l}{o}{v}' '{f}' + '{s}{f}' '{t}{s}{e}{ch}{o}{s}{l}{o}{v}' + (<- '{a}{k}') + )) or + ([] substring atlimit among ( + '{v}' '{v}{a}{l}' '{g}{y}{a}{n}' '{g}{l}' '{z}' '{i}{g}{o}{u}{m}{e}{n}' '{k}{a}{r}{d}' '{k}{o}{n}' '{m}{a}{k}{r}{u}{n}' '{n}{u}{f}' + '{p}{a}{t}{e}{r}' '{p}' '{s}{k}' '{t}{o}{s}' '{t}{r}{y}{p}{o}{l}' + (<- '{y}{t}{s}') + )) or + ([] '{k}{o}{r}' <- '{y}{t}{s}') + ) + ) + ) + + define steps9 as ( + [substring] among ( + '{y}{d}{y}{o}' '{y}{d}{y}{a}' '{y}{d}{y}{oo}{n}' ( + delete + unset test1 + ([] substring atlimit among ( + '{a}{y}{f}{n}' '{y}{r}' '{o}{l}{o}' '{ps}{a}{l}' (<- '{y}{d}') + )) or + ([] substring among ( + '{e}' '{p}{a}{y}{ch}{n}' (<- '{y}{d}') + )) + ) + ) + ) + + define steps10 as ( + [substring] among ( + '{y}{s}{k}{o}{s}' '{y}{s}{k}{o}{u}' '{y}{s}{k}{o}' '{y}{s}{k}{e}' ( + delete + unset test1 + [] substring atlimit among ( + '{d}' '{y}{v}' '{m}{i}{n}' '{r}' '{f}{r}{a}{g}{k}' '{l}{u}{k}' '{o}{v}{e}{l}' + (<- '{y}{s}{k}') + ) + ) + ) + ) + + define step2a as ( + [substring] among ( + '{a}{d}{e}{s}' '{a}{d}{oo}{n}' (delete) + ) + not ([substring] among ( + '{o}{k}' '{m}{a}{m}' '{m}{a}{n}' '{m}{p}{a}{m}{p}' '{p}{a}{t}{e}{r}' '{g}{y}{a}{g}{y}' '{n}{t}{a}{n}{t}' '{k}{u}{r}' '{th}{e}{y}' '{p}{e}{th}{e}{r}' + )) + insert '{a}{d}' + ) + + define step2b as ( + [substring] among ( + '{e}{d}{e}{s}' '{e}{d}{oo}{n}' (delete) + ) + [] substring among ( + '{o}{p}' '{y}{p}' '{e}{m}{p}' '{u}{p}' '{g}{i}{p}' '{d}{a}{p}' '{k}{r}{a}{s}{p}' '{m}{y}{l}' (<- '{e}{d}') + ) + ) + + define step2c as ( + [substring] among ( + '{o}{u}{d}{e}{s}' '{o}{u}{d}{oo}{n}' (delete) + ) + [] substring among ( + '{a}{r}{k}' '{k}{a}{l}{y}{a}{k}' '{p}{e}{t}{a}{l}' '{l}{y}{ch}' '{p}{l}{e}{x}' '{s}{k}' '{s}' '{f}{l}' '{f}{r}' '{v}{e}{l}' '{l}{o}{u}{l}' '{ch}{n}' + '{s}{p}' '{t}{r}{a}{g}' '{f}{e}' (<- '{o}{u}{d}') + ) + ) + + define step2d as ( + [substring] among ( + '{e}{oo}{s}' '{e}{oo}{n}' (delete unset test1) + ) + [] substring atlimit among ( + '{th}' '{d}' '{e}{l}' '{g}{a}{l}' '{n}' '{p}' '{y}{d}' '{p}{a}{r}' (<- '{e}') + ) + ) + + define step3 as ( + [substring] among ( + '{y}{a}' '{y}{o}{u}' '{y}{oo}{n}' (delete unset test1) + ) + ([] v <- '{y}') + ) + + define step4 as ( + [substring] among ( + '{y}{k}{a}' '{y}{k}{o}' '{y}{k}{o}{u}' '{y}{k}{oo}{n}' (delete unset test1) + ) + ([] v <- '{y}{k}') or + [] substring atlimit among ( + '{a}{l}' '{a}{d}' '{e}{n}{d}' '{a}{m}{a}{n}' '{a}{m}{m}{o}{ch}{a}{l}' '{i}{th}' '{a}{n}{i}{th}' '{a}{n}{t}{y}{d}' '{f}{u}{s}' '{v}{r}{oo}{m}' '{g}{e}{r}' + '{e}{x}{oo}{d}' '{k}{a}{l}{p}' '{k}{a}{l}{l}{y}{n}' '{k}{a}{t}{a}{d}' '{m}{o}{u}{l}' '{m}{p}{a}{n}' '{m}{p}{a}{g}{y}{a}{t}' '{m}{p}{o}{l}' '{m}{p}{o}{s}' + '{n}{y}{t}' '{x}{y}{k}' '{s}{u}{n}{o}{m}{i}{l}' '{p}{e}{t}{s}' '{p}{y}{t}{s}' '{p}{y}{k}{a}{n}{t}' '{p}{l}{y}{a}{t}{s}' '{p}{o}{s}{t}{e}{l}{n}' '{p}{r}{oo}{t}{o}{d}' + '{s}{e}{r}{t}' '{s}{u}{n}{a}{d}' '{t}{s}{a}{m}' '{u}{p}{o}{d}' '{f}{y}{l}{o}{n}' '{f}{u}{l}{o}{d}' '{ch}{a}{s}' + (<- '{y}{k}') + ) + ) + + define step5a as ( + do ('{a}{g}{a}{m}{e}' atlimit <- '{a}{g}{a}{m}') + do ( + [substring] among ( + '{a}{g}{a}{m}{e}' '{i}{s}{a}{m}{e}' '{o}{u}{s}{a}{m}{e}' '{i}{k}{a}{m}{e}' '{i}{th}{i}{k}{a}{m}{e}' (delete unset test1) + ) + ) + ['{a}{m}{e}'] + delete + unset test1 + [] substring atlimit among ( + '{a}{n}{a}{p}' '{a}{p}{o}{th}' '{a}{p}{o}{k}' '{a}{p}{o}{s}{t}' '{v}{o}{u}{v}' '{x}{e}{th}' '{o}{u}{l}' '{p}{e}{th}' '{p}{y}{k}{r}' '{p}{o}{t}' '{s}{y}{ch}' '{ch}' + (<- '{a}{m}') + ) + ) + + define step5b as ( + do ( + [substring] among ( + '{a}{g}{a}{n}{e}' '{i}{s}{a}{n}{e}' '{o}{u}{s}{a}{n}{e}' '{y}{o}{n}{t}{a}{n}{e}' '{y}{o}{t}{a}{n}{e}' '{y}{o}{u}{n}{t}{a}{n}{e}' '{o}{n}{t}{a}{n}{e}' '{o}{t}{a}{n}{e}' + '{o}{u}{n}{t}{a}{n}{e}' '{i}{k}{a}{n}{e}' '{i}{th}{i}{k}{a}{n}{e}' ( + delete + unset test1 + [] substring atlimit among ( + '{t}{r}' '{t}{s}' (<- '{a}{g}{a}{n}') + ) + ) + ) + ) + ['{a}{n}{e}'] + delete + unset test1 + ([] v2 <- '{a}{n}') or + [] substring atlimit among ( + '{v}{e}{t}{e}{r}' '{v}{o}{u}{l}{k}' '{v}{r}{a}{ch}{m}' '{g}' '{d}{r}{a}{d}{o}{u}{m}' + '{th}' '{k}{a}{l}{p}{o}{u}{z}' '{k}{a}{s}{t}{e}{l}' '{k}{o}{r}{m}{o}{r}' '{l}{a}{o}{p}{l}' '{m}{oo}{a}{m}{e}{th}' + '{m}' '{m}{o}{u}{s}{o}{u}{l}{m}' '{n}' '{o}{u}{l}' '{p}' '{p}{e}{l}{e}{k}' '{p}{l}' '{p}{o}{l}{y}{s}' + '{p}{o}{r}{t}{o}{l}' '{s}{a}{r}{a}{k}{a}{t}{s}' '{s}{o}{u}{l}{t}' '{t}{s}{a}{r}{l}{a}{t}' '{o}{r}{f}' + '{t}{s}{y}{g}{g}' '{t}{s}{o}{p}' '{f}{oo}{t}{o}{s}{t}{e}{f}' '{ch}' '{ps}{u}{ch}{o}{p}{l}' '{a}{g}' + '{g}{a}{l}' '{g}{e}{r}' '{d}{e}{k}' '{d}{y}{p}{l}' '{a}{m}{e}{r}{y}{k}{a}{n}' '{o}{u}{r}' '{p}{y}{th}' + '{p}{o}{u}{r}{y}{t}' '{s}' '{z}{oo}{n}{t}' '{y}{k}' '{k}{a}{s}{t}' '{k}{o}{p}' '{l}{y}{ch}' + '{l}{o}{u}{th}{i}{r}' '{m}{a}{y}{n}{t}' '{m}{e}{l}' '{s}{y}{g}' '{s}{p}' '{s}{t}{e}{g}' '{t}{r}{a}{g}' + '{t}{s}{a}{g}' '{f}' '{e}{r}' '{a}{d}{a}{p}' '{a}{th}{y}{g}{g}' '{a}{m}{i}{ch}' '{a}{n}{y}{k}' + '{a}{n}{o}{r}{g}' '{a}{p}{i}{g}' '{a}{p}{y}{th}' '{a}{t}{s}{y}{g}{g}' '{v}{a}{s}' '{v}{a}{s}{k}' + '{v}{a}{th}{u}{g}{a}{l}' '{v}{y}{o}{m}{i}{ch}' '{v}{r}{a}{ch}{u}{k}' '{d}{y}{a}{t}' '{d}{y}{a}{f}' '{e}{n}{o}{r}{g}' + '{th}{u}{s}' '{k}{a}{p}{n}{o}{v}{y}{o}{m}{i}{ch}' '{k}{a}{t}{a}{g}{a}{l}' '{k}{l}{y}{v}' '{k}{o}{y}{l}{a}{r}{f}' + '{l}{y}{v}' '{m}{e}{g}{l}{o}{v}{y}{o}{m}{i}{ch}' '{m}{y}{k}{r}{o}{v}{y}{o}{m}{i}{ch}' '{n}{t}{a}{v}' + '{x}{i}{r}{o}{k}{l}{y}{v}' '{o}{l}{y}{g}{o}{d}{a}{m}' '{o}{l}{o}{g}{a}{l}' '{p}{e}{n}{t}{a}{r}{f}' '{p}{e}{r}{i}{f}' + '{p}{e}{r}{y}{t}{r}' '{p}{l}{a}{t}' '{p}{o}{l}{u}{d}{a}{p}' '{p}{o}{l}{u}{m}{i}{ch}' '{s}{t}{e}{f}' '{t}{a}{v}' + '{t}{e}{t}' '{u}{p}{e}{r}{i}{f}' '{u}{p}{o}{k}{o}{p}' '{ch}{a}{m}{i}{l}{o}{d}{a}{p}' '{ps}{i}{l}{o}{t}{a}{v}' + (<- '{a}{n}') + ) + ) + + define step5c as ( + do ( + [substring] among ( + '{i}{s}{e}{t}{e}' (delete unset test1) + ) + ) + ['{e}{t}{e}'] + delete + unset test1 + ([] v2 <- '{e}{t}') or + ([] substring among ( + '{o}{d}' '{a}{y}{r}' '{f}{o}{r}' '{t}{a}{th}' '{d}{y}{a}{th}' '{s}{ch}' '{e}{n}{d}' '{e}{u}{r}' '{t}{y}{th}' '{u}{p}{e}{r}{th}' + '{r}{a}{th}' '{e}{n}{th}' '{r}{o}{th}' '{s}{th}' '{p}{u}{r}' '{a}{y}{n}' '{s}{u}{n}{d}' '{s}{u}{n}' '{s}{u}{n}{th}' '{ch}{oo}{r}' + '{p}{o}{n}' '{v}{r}' '{k}{a}{th}' '{e}{u}{th}' '{e}{k}{th}' '{n}{e}{t}' '{r}{o}{n}' '{a}{r}{k}' '{v}{a}{r}' '{v}{o}{l}' '{oo}{f}{e}{l}' + (<- '{e}{t}') + )) or + [] substring atlimit among ( + '{a}{v}{a}{r}' '{v}{e}{n}' '{e}{n}{a}{r}' '{a}{v}{r}' '{a}{d}' '{a}{th}' '{a}{n}' '{a}{p}{l}' '{v}{a}{r}{o}{n}' '{n}{t}{r}' '{s}{k}' '{k}{o}{p}' + '{m}{p}{o}{r}' '{n}{y}{f}' '{p}{a}{g}' '{p}{a}{r}{a}{k}{a}{l}' '{s}{e}{r}{p}' '{s}{k}{e}{l}' '{s}{u}{r}{f}' '{t}{o}{k}' '{u}' '{d}' '{e}{m}' + '{th}{a}{r}{r}' '{th}' + (<- '{e}{t}') + ) + ) + + define step5d as ( + [substring] among ( + '{o}{n}{t}{a}{s}' '{oo}{n}{t}{a}{s}' ( + delete + unset test1 + ([] '{a}{r}{ch}' atlimit <- '{o}{n}{t}') or + ([] '{k}{r}{e}' <- '{oo}{n}{t}') + ) + ) + ) + + define step5e as ( + [substring] among ( + '{o}{m}{a}{s}{t}{e}' '{y}{o}{m}{a}{s}{t}{e}' ( + delete + unset test1 + ([] '{o}{n}' atlimit <- '{o}{m}{a}{s}{t}') + ) + ) + ) + + define step5f as ( + do ( + ['{y}{e}{s}{t}{e}'] + delete + unset test1 + [] substring atlimit among ( + '{p}' '{a}{p}' '{s}{u}{m}{p}' '{a}{s}{u}{m}{p}' '{a}{k}{a}{t}{a}{p}' '{a}{m}{e}{t}{a}{m}{f}' (<- '{y}{e}{s}{t}') + ) + ) + ['{e}{s}{t}{e}'] + delete + unset test1 + [] substring atlimit among ( + '{a}{l}' '{a}{r}' '{e}{k}{t}{e}{l}' '{z}' '{m}' '{x}' '{p}{a}{r}{a}{k}{a}{l}' '{p}{r}{o}' '{n}{y}{s}' + (<- '{y}{e}{s}{t}') + ) + ) + + define step5g as ( + do ( + [substring] among ( + '{i}{th}{i}{k}{a}' '{i}{th}{i}{k}{e}{s}' '{i}{th}{i}{k}{e}' (delete unset test1) + ) + ) + [substring] among ( + '{i}{k}{a}' '{i}{k}{e}{s}' '{i}{k}{e}' ( + delete + unset test1 + ([] substring among ( + '{s}{k}{oo}{l}' '{s}{k}{o}{u}{l}' '{n}{a}{r}{th}' '{s}{f}' '{o}{th}' '{p}{y}{th}' (<- '{i}{k}') + )) or + ([] substring atlimit among ( + '{d}{y}{a}{th}' '{th}' '{p}{a}{r}{a}{k}{a}{t}{a}{th}' '{p}{r}{o}{s}{th}' '{s}{u}{n}{th}' (<- '{i}{k}') + )) + ) + ) + ) + + define step5h as ( + [substring] among ( + '{o}{u}{s}{a}' '{o}{u}{s}{e}{s}' '{o}{u}{s}{e}' ( + delete + unset test1 + ([] substring among ( + '{p}{o}{d}{a}{r}' '{v}{l}{e}{p}' '{p}{a}{n}{t}{a}{ch}' '{f}{r}{u}{d}' '{m}{a}{n}{t}{y}{l}' '{m}{a}{l}{l}' '{k}{u}{m}{a}{t}' '{l}{a}{ch}' '{l}{i}{g}' + '{f}{a}{g}' '{o}{m}' '{p}{r}{oo}{t}' (<- '{o}{u}{s}') + + )) or + ([] substring atlimit among ( + '{f}{a}{r}{m}{a}{k}' '{ch}{a}{d}' '{a}{g}{k}' '{a}{n}{a}{r}{r}' '{v}{r}{o}{m}' '{e}{k}{l}{y}{p}' '{l}{a}{m}{p}{y}{d}' '{l}{e}{ch}' '{m}' '{p}{a}{t}' + '{r}' '{l}' '{m}{e}{d}' '{m}{e}{s}{a}{z}' '{u}{p}{o}{t}{e}{y}{n}' '{a}{m}' '{a}{y}{th}' '{a}{n}{i}{k}' '{d}{e}{s}{p}{o}{z}' + '{e}{n}{d}{y}{a}{f}{e}{r}' '{d}{e}' '{d}{e}{u}{t}{e}{r}{e}{u}' '{k}{a}{th}{a}{r}{e}{u}' '{p}{l}{e}' '{t}{s}{a}' + (<- '{o}{u}{s}') + )) + ) + ) + ) + + define step5i as ( + [substring] among ( + '{a}{g}{a}' '{a}{g}{e}{s}' '{a}{g}{e}' ( + delete + unset test1 + ([] '{k}{o}{l}{l}' <- '{a}{g}') or ( + not ([substring] among ('{ps}{o}{f}' '{n}{a}{u}{l}{o}{ch}')) + ([] substring among ( + '{o}{f}' '{p}{e}{l}' '{ch}{o}{r}{t}' '{l}{l}' '{s}{f}' '{r}{p}' '{f}{r}' '{p}{r}' '{l}{o}{ch}' '{s}{m}{i}{n}' + (<- '{a}{g}') + )) or + ([] substring atlimit among ( + '{a}{v}{a}{s}{t}' '{p}{o}{l}{u}{f}' '{a}{d}{i}{f}' '{p}{a}{m}{f}' '{r}' '{a}{s}{p}' '{a}{f}' '{a}{m}{a}{l}' '{a}{m}{a}{l}{l}{y}' + '{a}{n}{u}{s}{t}' '{a}{p}{e}{r}' '{a}{s}{p}{a}{r}' '{a}{ch}{a}{r}' '{d}{e}{r}{v}{e}{n}' '{d}{r}{o}{s}{o}{p}' '{x}{e}{f}' '{n}{e}{o}{p}' + '{n}{o}{m}{o}{t}' '{o}{l}{o}{p}' '{o}{m}{o}{t}' '{p}{r}{o}{s}{t}' '{p}{r}{o}{s}{oo}{p}{o}{p}' '{s}{u}{m}{p}' '{s}{u}{n}{t}' '{t}' '{u}{p}{o}{t}' + '{ch}{a}{r}' '{a}{e}{y}{p}' '{a}{y}{m}{o}{s}{t}' '{a}{n}{u}{p}' '{a}{p}{o}{t}' '{a}{r}{t}{y}{p}' '{d}{y}{a}{t}' '{e}{n}' '{e}{p}{y}{t}' + '{k}{r}{o}{k}{a}{l}{o}{p}' '{s}{y}{d}{i}{r}{o}{p}' '{l}' '{n}{a}{u}' '{o}{u}{l}{a}{m}' '{o}{u}{r}' '{p}' '{t}{r}' '{m}' + (<- '{a}{g}') + )) + ) + ) + ) + ) + + define step5j as ( + [substring] among ( + '{i}{s}{e}' '{i}{s}{o}{u}' '{i}{s}{a}' (delete unset test1) + ) + [] substring atlimit among ( + '{n}' '{ch}{e}{r}{s}{o}{n}' '{d}{oo}{d}{e}{k}{a}{n}' '{e}{r}{i}{m}{o}{n}' '{m}{e}{g}{a}{l}{o}{n}' '{e}{p}{t}{a}{n}' (<- '{i}{s}') + ) + ) + + define step5k as ( + [substring] among ( + '{i}{s}{t}{e}' (delete unset test1) + ) + [] substring atlimit among ( + '{a}{s}{v}' '{s}{v}' '{a}{ch}{r}' '{ch}{r}' '{a}{p}{l}' '{a}{e}{y}{m}{n}' '{d}{u}{s}{ch}{r}' '{e}{u}{ch}{r}' '{k}{o}{y}{n}{o}{ch}{r}' '{p}{a}{l}{y}{m}{ps}' + (<- '{i}{s}{t}') + ) + ) + + define step5l as ( + [substring] among ( + '{o}{u}{n}{e}' '{i}{s}{o}{u}{n}{e}' '{i}{th}{o}{u}{n}{e}' (delete unset test1) + ) + [] substring atlimit among ( + '{n}' '{r}' '{s}{p}{y}' '{s}{t}{r}{a}{v}{o}{m}{o}{u}{t}{s}' '{k}{a}{k}{o}{m}{o}{u}{t}{s}' '{e}{x}{oo}{n}' (<- '{o}{u}{n}') + ) + ) + + define step5m as ( + [substring] among ( + '{o}{u}{m}{e}' '{i}{s}{o}{u}{m}{e}' '{i}{th}{o}{u}{m}{e}' (delete unset test1) + ) + [] substring atlimit among ( + '{p}{a}{r}{a}{s}{o}{u}{s}' '{f}' '{ch}' '{oo}{r}{y}{o}{p}{l}' '{a}{z}' '{a}{l}{l}{o}{s}{o}{u}{s}' '{a}{s}{o}{u}{s}' + (<- '{o}{u}{m}') + ) + ) + + define step6 as ( + do ( + [substring] among ( + '{m}{a}{t}{a}' '{m}{a}{t}{oo}{n}' '{m}{a}{t}{o}{s}' (<- '{m}{a}') + ) + ) + test1 + [substring] among ( + '{a}' '{a}{g}{a}{t}{e}' '{a}{g}{a}{n}' '{a}{e}{y}' '{a}{m}{a}{y}' '{a}{n}' '{a}{s}' '{a}{s}{a}{y}' '{a}{t}{a}{y}' '{a}{oo}' '{e}' '{e}{y}' + '{e}{y}{s}' '{e}{y}{t}{e}' '{e}{s}{a}{y}' '{e}{s}' '{e}{t}{a}{y}' '{y}' '{y}{e}{m}{a}{y}' '{y}{e}{m}{a}{s}{t}{e}' '{y}{e}{t}{a}{y}' '{y}{e}{s}{a}{y}' + '{y}{e}{s}{a}{s}{t}{e}' '{y}{o}{m}{a}{s}{t}{a}{n}' '{y}{o}{m}{o}{u}{n}' '{y}{o}{m}{o}{u}{n}{a}' '{y}{o}{n}{t}{a}{n}' '{y}{o}{n}{t}{o}{u}{s}{a}{n}' '{y}{o}{s}{a}{s}{t}{a}{n}' + '{y}{o}{s}{a}{s}{t}{e}' '{y}{o}{s}{o}{u}{n}' '{y}{o}{s}{o}{u}{n}{a}' '{y}{o}{t}{a}{n}' '{y}{o}{u}{m}{a}' '{y}{o}{u}{m}{a}{s}{t}{e}' '{y}{o}{u}{n}{t}{a}{y}' + '{y}{o}{u}{n}{t}{a}{n}' '{i}' '{i}{d}{e}{s}' '{i}{d}{oo}{n}' '{i}{th}{e}{y}' '{i}{th}{e}{y}{s}' '{i}{th}{e}{y}{t}{e}' '{i}{th}{i}{k}{a}{t}{e}' '{i}{th}{i}{k}{a}{n}' + '{i}{th}{o}{u}{n}' '{i}{th}{oo}' '{i}{k}{a}{t}{e}' '{i}{k}{a}{n}' '{i}{s}' '{i}{s}{a}{n}' '{i}{s}{a}{t}{e}' '{i}{s}{e}{y}' '{i}{s}{e}{s}' '{i}{s}{o}{u}{n}' + '{i}{s}{oo}' '{o}' '{o}{y}' '{o}{m}{a}{y}' '{o}{m}{a}{s}{t}{a}{n}' '{o}{m}{o}{u}{n}' '{o}{m}{o}{u}{n}{a}' '{o}{n}{t}{a}{y}' '{o}{n}{t}{a}{n}' + '{o}{n}{t}{o}{u}{s}{a}{n}' '{o}{s}' '{o}{s}{a}{s}{t}{a}{n}' '{o}{s}{a}{s}{t}{e}' '{o}{s}{o}{u}{n}' '{o}{s}{o}{u}{n}{a}' '{o}{t}{a}{n}' '{o}{u}' '{o}{u}{m}{a}{y}' + '{o}{u}{m}{a}{s}{t}{e}' '{o}{u}{n}' '{o}{u}{n}{t}{a}{y}' '{o}{u}{n}{t}{a}{n}' '{o}{u}{s}' '{o}{u}{s}{a}{n}' '{o}{u}{s}{a}{t}{e}' '{u}' '{u}{s}' '{oo}' + '{oo}{n}' (delete) + ) + ) + + define step7 as ( + [substring] among ( + '{e}{s}{t}{e}{r}' '{e}{s}{t}{a}{t}' '{o}{t}{e}{r}' '{o}{t}{a}{t}' '{u}{t}{e}{r}' '{u}{t}{a}{t}' '{oo}{t}{e}{r}' '{oo}{t}{a}{t}' (delete) + ) + ) +) + +define stem as ( + backwards ( + do tolower + has_min_length + set test1 + do step1 + do steps1 + do steps2 + do steps3 + do steps4 + do steps5 + do steps6 + do steps7 + do steps8 + do steps9 + do steps10 + do step2a + do step2b + do step2c + do step2d + do step3 + do step4 + do step5a + do step5b + do step5c + do step5d + do step5e + do step5f + do step5g + do step5h + do step5j + do step5i + do step5k + do step5l + do step5m + do step6 + do step7 + ) +) |