kpriv/pkgsrc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
pkgsrc/audio/festvox-us3/Makefile

38 lines
1 KiB
Makefile
Raw Normal View History

Update festival to 1.4.1, and fix dependent packages up while I'm here (including fixing lots of PLISTs). Changes since 1.4.0, from the NEWS file: . SSFF (for emulabel) track support . Java fixes . various minor bug fixes . WFST with proper quoting . Wagon sample counts . gcc-2.95.1 support Also use newer distfile for mbrola. Addresses pkg/10671.
2001-07-18 17:19:35 +02:00
# $NetBSD: Makefile,v 1.9 2001/07/18 15:19:45 wiz Exp $
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
DISTNAME= festvox_us3
PKGNAME= festvox-us3-1.4.0
WRKSRC= ${WRKDIR}/festival
CATEGORIES= audio
Update festival to 1.4.1, and fix dependent packages up while I'm here (including fixing lots of PLISTs). Changes since 1.4.0, from the NEWS file: . SSFF (for emulabel) track support . Java fixes . various minor bug fixes . WFST with proper quoting . Wagon sample counts . gcc-2.95.1 support Also use newer distfile for mbrola. Addresses pkg/10671.
2001-07-18 17:19:35 +02:00
MASTER_SITES= ftp://ftp.cstr.ed.ac.uk/pub/festival/1.4.1/ \
http://www.speech.cs.cmu.edu/festival/cstr/festival/1.4.1/
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
MAINTAINER= proff@iq.org
HOMEPAGE= http://www.cstr.ed.ac.uk/projects/festival.html
Update to new COMMENT style: COMMENT var in Makefile instead of pkg/COMMENT.
2001-02-16 14:30:32 +01:00
COMMENT= MBROLA based American English male voice for festival
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
Update festival to 1.4.1, and fix dependent packages up while I'm here (including fixing lots of PLISTs). Changes since 1.4.0, from the NEWS file: . SSFF (for emulabel) track support . Java fixes . various minor bug fixes . WFST with proper quoting . Wagon sample counts . gcc-2.95.1 support Also use newer distfile for mbrola. Addresses pkg/10671.
2001-07-18 17:19:35 +02:00
DEPENDS+= festival>=1.4.0:../festival
DEPENDS+= festlex-poslex>=1.4.0:../festlex-poslex
DEPENDS+= festlex-cmu>=1.4.0:../festlex-cmu
When multi-line dependencies occur, use a "DEPENDS+= package" format for each of the continuation lines, rather than using backslashes to continue a single, long definition. This makes it much easier to spot pre-requisite packages and other dependencies.
2000-02-15 10:26:12 +01:00
DEPENDS+= mbrola-301h:../mbrola
DEPENDS+= mbrolavox-us3-990208:../mbrolavox-us3
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
LICENSE= no-commercial-use
NO_BUILD= yes
FHOME= ${PREFIX}/share/festival
MBROLA= ${FHOME}/lib/voices/english/us3_mbrola
post-extract:
pkglint silencing: o use DISTNAME and PKGNAME instead of DISTFILES o replace rm by ${RM}, chmod by ${CHMOD}, pax by ${PAX}, echo by ${ECHO} o reorder lines, remove unnecessary blank o modify pkg/COMMENT and pkg/DESCR o remove lines from files/patch-sum for files that aren't there
1999-12-25 23:38:54 +01:00
@ ${CHMOD} -R a+r ${WRKSRC}
Don't use find ... -print0 | xargs -0 ... as -0 isn't an acceptable option on Solaris' xargs.
2001-07-14 14:37:45 +02:00
@ ${FIND} ${WRKSRC} -type d -print | ${XARGS} ${CHMOD} 755
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
do-install:
pkglint silencing: o use DISTNAME and PKGNAME instead of DISTFILES o replace rm by ${RM}, chmod by ${CHMOD}, pax by ${PAX}, echo by ${ECHO} o reorder lines, remove unnecessary blank o modify pkg/COMMENT and pkg/DESCR o remove lines from files/patch-sum for files that aren't there
1999-12-25 23:38:54 +01:00
cd ${WRKSRC} && ${PAX} -r -w * ${FHOME}
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
#ugh
replaced some more commands by their ${COMMAND} counterparts
1999-12-28 05:19:57 +01:00
${LN} -fs ../../../../../mbrola/voices/us3 ${MBROLA}
${LN} -fs ../../../../../mbrola/voices/us3/us3mrpa ${MBROLA}
MBROLA based American English male voice for the festival speech synthesis system. This voice provides a American English male voice using the MBROLA synthesis method. It uses a modified CMU lexicon for pronunciations. Prosodic phrasing is provided by a statistically trained model using part of speech and local distribution of breaks. Intonation is provided by a CART tree predicting ToBI accents and an F0 contour generated from a model trained from natural speech. The duration model is also trained from data using a CART tree. The quality of this voice is not as high as us1 and us2 This voice can be activated via (voice_us3_mbrola) -Julian Assange <proff@iq.org>
1999-10-11 18:51:41 +02:00
.include "../../mk/bsd.pkg.mk"
Reference in a new issue Copy permalink