Adds test data and code to regenerate it

tests/testdata/Makefile

@@ -0,0 +1,4 @@
+SUBDIRS = copy delete ice_g2p identity reverse tur_inflection upper
+
+all:
+	for dir in $(SUBDIRS); do $(MAKE) -C $$dir; done

tests/testdata/README.md

@@ -0,0 +1,2 @@
+This directory contains test data for tests involving actual
+sequence-to-sequence training. To regenerate the data, run `make -j`.

tests/testdata/copy/Makefile

@@ -0,0 +1,8 @@
+SHELL := /bin/bash
+
+all: train.tsv dev.tsv test.tsv
+
+%.tsv:
+	.././generate.py --vocabulary 16 --length 5 --seed 1951 --size 100 > /tmp/copy_$@
+	paste /tmp/copy_$@ <(paste -d'\0' /tmp/copy_$@ /tmp/copy_$@) > $@
+	rm /tmp/copy_$@

tests/testdata/copy/dev.tsv

@@ -0,0 +1,100 @@
+eplnjcdd	eplnjcddeplnjcdd
+p	pp
+ibejfod	ibejfodibejfod
+pab	pabpab
+dlfcgbk	dlfcgbkdlfcgbk
+olfop	olfopolfop
+gcfohaobgcf	gcfohaobgcfgcfohaobgcf
+ejnpc	ejnpcejnpc
+mo	momo
+pifg	pifgpifg
+cjefakb	cjefakbcjefakb
+o	oo
+pjcjlek	pjcjlekpjcjlek
+hehdgc	hehdgchehdgc
+mmhf	mmhfmmhf
+nih	nihnih
+l	ll
+hgahdmme	hgahdmmehgahdmme
+llhekbhgg	llhekbhggllhekbhgg
+fjmcif	fjmciffjmcif
+kpkji	kpkjikpkji
+dlef	dlefdlef
+naljdol	naljdolnaljdol
+dapf	dapfdapf
+ae	aeae
+pikn	piknpikn
+kdnkma	kdnkmakdnkma
+jmoe	jmoejmoe
+hkcd	hkcdhkcd
+bae	baebae
+npdchffilp	npdchffilpnpdchffilp
+feci	fecifeci
+cfhm	cfhmcfhm
+eb	ebeb
+ej	ejej
+kha	khakha
+fhhen	fhhenfhhen
+clobjl	clobjlclobjl
+bdcnnfk	bdcnnfkbdcnnfk
+apjnmed	apjnmedapjnmed
+epopgd	epopgdepopgd
+eoa	eoaeoa
+mak	makmak
+ilg	ilgilg
+odnfeap	odnfeapodnfeap
+jebn	jebnjebn
+mo	momo
+ijiele	ijieleijiele
+efolbe	efolbeefolbe
+hjfmp	hjfmphjfmp
+g	gg
+ohog	ohogohog
+lmggl	lmggllmggl
+ncbkj	ncbkjncbkj
+bk	bkbk
+odkcf	odkcfodkcf
+ojnoi	ojnoiojnoi
+eeaf	eeafeeaf
+iljalco	iljalcoiljalco
+cbod	cbodcbod
+fjh	fjhfjh
+cfkmlfe	cfkmlfecfkmlfe
+bljo	bljobljo
+gdejbbjh	gdejbbjhgdejbbjh
+opofbihg	opofbihgopofbihg
+koj	kojkoj
+mn	mnmn
+ne	nene
+hkkjk	hkkjkhkkjk
+bjlifgeo	bjlifgeobjlifgeo
+blb	blbblb
+neeb	neebneeb
+dogb	dogbdogb
+llenp	llenpllenp
+acia	aciaacia
+cceg	ccegcceg
+ipbfghh	ipbfghhipbfghh
+icbmm	icbmmicbmm
+jiakjo	jiakjojiakjo
+ndaek	ndaekndaek
+minelhf	minelhfminelhf
+fmcbh	fmcbhfmcbh
+icmp	icmpicmp
+oigkcacd	oigkcacdoigkcacd
+efn	efnefn
+jednkp	jednkpjednkp
+eajcb	eajcbeajcb
+calemkd	calemkdcalemkd
+hokdk	hokdkhokdk
+ahb	ahbahb
+biikpho	biikphobiikpho
+lbjdide	lbjdidelbjdide
+kjlob	kjlobkjlob
+jhhd	jhhdjhhd
+aiffmono	aiffmonoaiffmono
+gfnpe	gfnpegfnpe
+bjbbkj	bjbbkjbjbbkj
+jabie	jabiejabie
+jmlokf	jmlokfjmlokf
+hmjb	hmjbhmjb

tests/testdata/copy/test.tsv

@@ -0,0 +1,100 @@
+eplnjcdd	eplnjcddeplnjcdd
+p	pp
+ibejfod	ibejfodibejfod
+pab	pabpab
+dlfcgbk	dlfcgbkdlfcgbk
+olfop	olfopolfop
+gcfohaobgcf	gcfohaobgcfgcfohaobgcf
+ejnpc	ejnpcejnpc
+mo	momo
+pifg	pifgpifg
+cjefakb	cjefakbcjefakb
+o	oo
+pjcjlek	pjcjlekpjcjlek
+hehdgc	hehdgchehdgc
+mmhf	mmhfmmhf
+nih	nihnih
+l	ll
+hgahdmme	hgahdmmehgahdmme
+llhekbhgg	llhekbhggllhekbhgg
+fjmcif	fjmciffjmcif
+kpkji	kpkjikpkji
+dlef	dlefdlef
+naljdol	naljdolnaljdol
+dapf	dapfdapf
+ae	aeae
+pikn	piknpikn
+kdnkma	kdnkmakdnkma
+jmoe	jmoejmoe
+hkcd	hkcdhkcd
+bae	baebae
+npdchffilp	npdchffilpnpdchffilp
+feci	fecifeci
+cfhm	cfhmcfhm
+eb	ebeb
+ej	ejej
+kha	khakha
+fhhen	fhhenfhhen
+clobjl	clobjlclobjl
+bdcnnfk	bdcnnfkbdcnnfk
+apjnmed	apjnmedapjnmed
+epopgd	epopgdepopgd
+eoa	eoaeoa
+mak	makmak
+ilg	ilgilg
+odnfeap	odnfeapodnfeap
+jebn	jebnjebn
+mo	momo
+ijiele	ijieleijiele
+efolbe	efolbeefolbe
+hjfmp	hjfmphjfmp
+g	gg
+ohog	ohogohog
+lmggl	lmggllmggl
+ncbkj	ncbkjncbkj
+bk	bkbk
+odkcf	odkcfodkcf
+ojnoi	ojnoiojnoi
+eeaf	eeafeeaf
+iljalco	iljalcoiljalco
+cbod	cbodcbod
+fjh	fjhfjh
+cfkmlfe	cfkmlfecfkmlfe
+bljo	bljobljo
+gdejbbjh	gdejbbjhgdejbbjh
+opofbihg	opofbihgopofbihg
+koj	kojkoj
+mn	mnmn
+ne	nene
+hkkjk	hkkjkhkkjk
+bjlifgeo	bjlifgeobjlifgeo
+blb	blbblb
+neeb	neebneeb
+dogb	dogbdogb
+llenp	llenpllenp
+acia	aciaacia
+cceg	ccegcceg
+ipbfghh	ipbfghhipbfghh
+icbmm	icbmmicbmm
+jiakjo	jiakjojiakjo
+ndaek	ndaekndaek
+minelhf	minelhfminelhf
+fmcbh	fmcbhfmcbh
+icmp	icmpicmp
+oigkcacd	oigkcacdoigkcacd
+efn	efnefn
+jednkp	jednkpjednkp
+eajcb	eajcbeajcb
+calemkd	calemkdcalemkd
+hokdk	hokdkhokdk
+ahb	ahbahb
+biikpho	biikphobiikpho
+lbjdide	lbjdidelbjdide
+kjlob	kjlobkjlob
+jhhd	jhhdjhhd
+aiffmono	aiffmonoaiffmono
+gfnpe	gfnpegfnpe
+bjbbkj	bjbbkjbjbbkj
+jabie	jabiejabie
+jmlokf	jmlokfjmlokf
+hmjb	hmjbhmjb

tests/testdata/copy/train.tsv

@@ -0,0 +1,100 @@
+eplnjcdd	eplnjcddeplnjcdd
+p	pp
+ibejfod	ibejfodibejfod
+pab	pabpab
+dlfcgbk	dlfcgbkdlfcgbk
+olfop	olfopolfop
+gcfohaobgcf	gcfohaobgcfgcfohaobgcf
+ejnpc	ejnpcejnpc
+mo	momo
+pifg	pifgpifg
+cjefakb	cjefakbcjefakb
+o	oo
+pjcjlek	pjcjlekpjcjlek
+hehdgc	hehdgchehdgc
+mmhf	mmhfmmhf
+nih	nihnih
+l	ll
+hgahdmme	hgahdmmehgahdmme
+llhekbhgg	llhekbhggllhekbhgg
+fjmcif	fjmciffjmcif
+kpkji	kpkjikpkji
+dlef	dlefdlef
+naljdol	naljdolnaljdol
+dapf	dapfdapf
+ae	aeae
+pikn	piknpikn
+kdnkma	kdnkmakdnkma
+jmoe	jmoejmoe
+hkcd	hkcdhkcd
+bae	baebae
+npdchffilp	npdchffilpnpdchffilp
+feci	fecifeci
+cfhm	cfhmcfhm
+eb	ebeb
+ej	ejej
+kha	khakha
+fhhen	fhhenfhhen
+clobjl	clobjlclobjl
+bdcnnfk	bdcnnfkbdcnnfk
+apjnmed	apjnmedapjnmed
+epopgd	epopgdepopgd
+eoa	eoaeoa
+mak	makmak
+ilg	ilgilg
+odnfeap	odnfeapodnfeap
+jebn	jebnjebn
+mo	momo
+ijiele	ijieleijiele
+efolbe	efolbeefolbe
+hjfmp	hjfmphjfmp
+g	gg
+ohog	ohogohog
+lmggl	lmggllmggl
+ncbkj	ncbkjncbkj
+bk	bkbk
+odkcf	odkcfodkcf
+ojnoi	ojnoiojnoi
+eeaf	eeafeeaf
+iljalco	iljalcoiljalco
+cbod	cbodcbod
+fjh	fjhfjh
+cfkmlfe	cfkmlfecfkmlfe
+bljo	bljobljo
+gdejbbjh	gdejbbjhgdejbbjh
+opofbihg	opofbihgopofbihg
+koj	kojkoj
+mn	mnmn
+ne	nene
+hkkjk	hkkjkhkkjk
+bjlifgeo	bjlifgeobjlifgeo
+blb	blbblb
+neeb	neebneeb
+dogb	dogbdogb
+llenp	llenpllenp
+acia	aciaacia
+cceg	ccegcceg
+ipbfghh	ipbfghhipbfghh
+icbmm	icbmmicbmm
+jiakjo	jiakjojiakjo
+ndaek	ndaekndaek
+minelhf	minelhfminelhf
+fmcbh	fmcbhfmcbh
+icmp	icmpicmp
+oigkcacd	oigkcacdoigkcacd
+efn	efnefn
+jednkp	jednkpjednkp
+eajcb	eajcbeajcb
+calemkd	calemkdcalemkd
+hokdk	hokdkhokdk
+ahb	ahbahb
+biikpho	biikphobiikpho
+lbjdide	lbjdidelbjdide
+kjlob	kjlobkjlob
+jhhd	jhhdjhhd
+aiffmono	aiffmonoaiffmono
+gfnpe	gfnpegfnpe
+bjbbkj	bjbbkjbjbbkj
+jabie	jabiejabie
+jmlokf	jmlokfjmlokf
+hmjb	hmjbhmjb

tests/testdata/delete/Makefile

@@ -0,0 +1,9 @@
+SHELL := /bin/bash
+
+all: train.tsv dev.tsv test.tsv
+
+%.tsv:
+	paste \
+		<(.././generate.py --vocabulary 16 --length 5 --seed 1953 --size 100) \
+		/dev/null \
+		> $@