import sys import os import re import collections sys.path.append( os.path.join(os.path.dirname(__file__), '..', '..', 'resource-gcg', 'scripts')) import tree import gcgtree for line in sys.stdin: print(gcgtree.GCGTree(line))
import sys import os import re import collections sys.path.append( os.path.join(os.path.dirname(__file__), '..', '..', 'resource-gcg', 'scripts')) import tree import gcgtree for line in sys.stdin: print(gcgtree.GCGTree( line.strip())) #strip to prevent adding additional blank lines
t.c += xy else: t.c += '-x%' + morph.getLemma(t.c,t.ch[0].c) + '|%Bot' for st in t.ch: ctr = annotY( st, Ymap, ctr ) return ctr ################################################################################ ################################################################################ ## read in corpus and convert to traversal trees... lt = [ ] llpttrav = [ ] for line in sys.stdin: Marked = { } lt.append( gcgtree.GCGTree(line) ) # print( semcuegraph.StoreStateCueGraph(lt[-1]) ) numberTerminals( lt[-1] ) lptFactored = factorConj( lt[-1] ) llpttrav.append( [ (p,t,makeTraversal(semcuegraph.SimpleCueGraph(semcuegraph.SemCueGraph(t)),Marked)) for p,t in lptFactored ] ) # for p,t in lptFactored: # print( '==FULL==>', str( semcuegraph.SemCueGraph(t) ) ) # print( '--SIMPLE->', str( semcuegraph.SimpleCueGraph(semcuegraph.SemCueGraph(t)) ) ) ## set size params... K,L = 0,0 for lpttrav in llpttrav: for p,t,trav in lpttrav: K,l = setKL( trav, KINTS ) L = max(L,l) sys.stderr.write( str(K) + ' predicate constants, ' + str(L) + ' dependency labels.\n' )
## Define discourse graph... G = semcuegraph.SemCueGraph( ) sentctr = 0 ## For each sentence... for line in sys.stdin: if '!ARTICLE' in line: break if sentctr == 0: discctr += 1 sys.stderr.write( 'Discourse ' + str(discctr) + ': ' + line ) ## Initialize new tree with or without tree-lengthening... if RELABEL: t = gcgtree.GCGTree( line ) else: t = tree.Tree( ) t.read( line ) ## Add tree to discourse graph... G.add( t, ('0' if sentctr<10 else '') + str(sentctr) ) sentctr += 1 else: finished = True ## If discourse contained any sentences, finalize and print graph... if sentctr>0: G.finalize() print( str(G) )