mirror of
https://github.com/moses-smt/mosesdecoder.git
synced 2024-12-26 05:14:36 +03:00
346 lines
11 KiB
Plaintext
346 lines
11 KiB
Plaintext
#BUILDING MOSES
|
|
|
|
#PACKAGES
|
|
#Language models (optional):
|
|
#--with-irstlm=/path/to/irstlm
|
|
#--with-srilm=/path/to/srilm See moses/LM/Jamfile for more options.
|
|
#--with-maxent-srilm=true (requires a maxent-enabled version of SRILM to be specified via --with-srilm)
|
|
#--with-nplm=/path/to/nplm
|
|
#--with-randlm=/path/to/randlm
|
|
#KenLM is always compiled.
|
|
#
|
|
#--with-boost=/path/to/boost
|
|
#If Boost is in a non-standard location, specify it here. This directory is
|
|
#expected to contain include and lib or lib64.
|
|
#
|
|
#--with-xmlrpc-c=/path/to/xmlrpc-c for libxmlrpc-c (used by server)
|
|
#Note that, like language models, this is the --prefix where the library was
|
|
#installed, not some executable within the library.
|
|
#
|
|
#--no-xmlrpc-c
|
|
# Don't use xmlrpc-c library, even if it exists. Don't build moses server
|
|
#
|
|
#Compact phrase table and compact lexical reordering table
|
|
#--with-cmph=/path/to/cmph
|
|
#
|
|
#Thread-caching malloc (if present, used for multi-threaded builds by default)
|
|
#--without-tcmalloc does not compile with tcmalloc even if present
|
|
#--full-tcmalloc links against the full version (useful for memory profiling)
|
|
#
|
|
#REGRESSION TESTING
|
|
#--with-regtest=/path/to/moses-reg-test-data
|
|
#
|
|
#INSTALLATION
|
|
#--prefix=/path/to/prefix sets the install prefix [default is source root].
|
|
#--bindir=/path/to/prefix/bin sets the bin directory [PREFIX/bin]
|
|
#--libdir=/path/to/prefix/lib sets the lib directory [PREFIX/lib]
|
|
#--includedir=/path/to/prefix/include installs headers.
|
|
# Does not install if missing. No argument defaults to PREFIX/include .
|
|
#--install-scripts=/path/to/scripts copies scripts into a directory.
|
|
# Does not install if missing. No argument defaults to PREFIX/scripts .
|
|
#--git appends the git revision to the prefix directory.
|
|
#
|
|
#
|
|
#BUILD OPTIONS
|
|
# By default, the build is multi-threaded, optimized, and statically linked.
|
|
# Pass these to change the build:
|
|
#
|
|
# threading=single|multi controls threading (default multi)
|
|
#
|
|
# variant=release|debug|profile builds optimized (default), for debug, or for
|
|
# profiling
|
|
#
|
|
# link=static|shared controls preferred linking (default static)
|
|
# --static forces static linking (the default will fall
|
|
# back to shared)
|
|
#
|
|
# debug-symbols=on|off include or exclude (default) debugging
|
|
# information also known as -g
|
|
# --notrace compiles without TRACE macros
|
|
#
|
|
# --enable-boost-pool uses Boost pools for the memory SCFG tabgle
|
|
#
|
|
# --enable-mpi switch on mpi
|
|
# --without-libsegfault does not link with libSegFault
|
|
#
|
|
# --max-kenlm-order maximum ngram order that kenlm can process (default 6)
|
|
#
|
|
# --max-factors maximum number of factors (default 4)
|
|
#
|
|
# --unlabelled-source ignore source labels (redundant in hiero or string-to-tree system)
|
|
# for better performance
|
|
#CONTROLLING THE BUILD
|
|
#-a to build from scratch
|
|
#-j$NCPUS to compile in parallel
|
|
#--clean to clean
|
|
#--debug-build to build with Og. Only available with gcc 4.8+
|
|
|
|
import os ;
|
|
import option ;
|
|
import modules ;
|
|
import path ;
|
|
path-constant TOP : . ;
|
|
|
|
include $(TOP)/jam-files/sanity.jam ;
|
|
|
|
home = [ os.environ "HOME" ] ;
|
|
if [ path.exists $(home)/moses-environment.jam ]
|
|
{
|
|
# for those of use who don't like typing in command line bjam options all day long
|
|
include $(home)/moses-environment.jam ;
|
|
}
|
|
include $(TOP)/jam-files/check-environment.jam ; # get resource locations
|
|
# from environment variables
|
|
include $(TOP)/jam-files/xmlrpc-c.jam ; # xmlrpc-c stuff for the server
|
|
# include $(TOP)/jam-files/curlpp.jam ; # curlpp stuff for bias lookup (MMT only)
|
|
|
|
# exit "done" : 0 ;
|
|
|
|
max-order = [ option.get "max-kenlm-order" : 6 : 6 ] ;
|
|
if ! [ option.get "max-kenlm-order" ]
|
|
{
|
|
# some classes in Moses pull in header files from KenLM, so this needs to be
|
|
# defined here, not in moses/lm/Jamfile
|
|
option.set "max-kenlm-order" : 6 ;
|
|
requirements += <define>KENLM_MAX_ORDER=$(max-order) ;
|
|
}
|
|
# exit "all done" : 0 ;
|
|
|
|
boost 104400 ;
|
|
external-lib z ;
|
|
|
|
#lib dl : : <runtime-link>static:<link>static <runtime-link>shared:<link>shared ;
|
|
#requirements += <library>dl ;
|
|
requirements += <cxxflags>-std=c++0x ;
|
|
|
|
# Allow moses to report the git commit hash of the version used for compilation
|
|
moses_githash = [ _shell "git describe --dirty" ] ;
|
|
requirements += <define>MOSES_VERSION_ID=\\\"$(moses_githash)\\\" ;
|
|
|
|
if ! [ option.get "without-tcmalloc" : : "yes" ] && [ test_library "tcmalloc_minimal" ] {
|
|
if [ option.get "full-tcmalloc" : : "yes" ] {
|
|
external-lib unwind ;
|
|
external-lib tcmalloc_and_profiler : : unwind ;
|
|
requirements += <library>tcmalloc_and_profiler <library>unwind <cflags>-fno-omit-frame-pointer <cxxflags>-fno-omit-frame-pointer ;
|
|
} else {
|
|
external-lib tcmalloc_minimal ;
|
|
requirements += <threading>multi:<library>tcmalloc_minimal ;
|
|
}
|
|
} else {
|
|
echo "Tip: install tcmalloc for faster threading. See BUILD-INSTRUCTIONS.txt for more information." ;
|
|
}
|
|
|
|
if [ option.get "filter-warnings" : : "yes" ] {
|
|
# given the low coding standards in Moses, we may want to filter out
|
|
# warnings about poor coding practice that no-one is ever going to fix
|
|
# anyway ...
|
|
requirements += <cxxflags>-Wno-deprecated ;
|
|
requirements += <cxxflags>-Wno-reorder ;
|
|
requirements += <cxxflags>-Wno-sign-compare ;
|
|
requirements += <cxxflags>-Wno-unused-but-set-variable ;
|
|
requirements += <cxxflags>-Wno-unused-result ;
|
|
requirements += <cxxflags>-Wno-unused-variable ;
|
|
requirements += <cxxflags>-Wno-comment ;
|
|
requirements += <cxxflags>-Wno-strict-aliasing ;
|
|
requirements += <cxxflags>-Wno-overloaded-virtual ;
|
|
}
|
|
|
|
if [ option.get "debug-build" : : "yes" ] {
|
|
requirements += <cxxflags>-Og ;
|
|
echo "Building with -Og to enable easier profiling and debugging. Only available on gcc 4.8+." ;
|
|
}
|
|
|
|
if [ option.get "with-address-sanitizer" : : "yes" ] {
|
|
requirements += <cxxflags>-fsanitize=address ;
|
|
requirements += <cxxflags>-fno-omit-frame-pointer ;
|
|
requirements += <linkflags>-fsanitize=address ;
|
|
echo "Building with AddressSanitizer to enable debugging of memory errors. Only available on gcc 4.8+." ;
|
|
}
|
|
|
|
if [ option.get "enable-mpi" : : "yes" ] {
|
|
import mpi ;
|
|
using mpi ;
|
|
external-lib boost_mpi ;
|
|
external-lib boost_serialization ;
|
|
requirements += <define>MPI_ENABLE ;
|
|
requirements += <library>mpi ;
|
|
requirements += <library>boost_mpi ;
|
|
requirements += <library>boost_serialization ;
|
|
}
|
|
|
|
mmt = [ option.get "mmt" ] ;
|
|
if $(mmt) {
|
|
requirements += <define>MMT ;
|
|
requirements += <include>$(mmt) ;
|
|
mmt_githash = [ _shell "cd $(mmt) && git describe --dirty" ] ;
|
|
requirements += <define>MMT_VERSION_ID=\\\"$(mmt_githash)\\\" ;
|
|
}
|
|
|
|
requirements += [ option.get "notrace" : <define>TRACE_ENABLE=1 ] ;
|
|
requirements += [ option.get "enable-boost-pool" : : <define>USE_BOOST_POOL ] ;
|
|
requirements += [ option.get "with-mm" : : <define>PT_UG ] ;
|
|
requirements += [ option.get "with-mm" : : <define>MAX_NUM_FACTORS=4 ] ;
|
|
requirements += [ option.get "unlabelled-source" : : <define>UNLABELLED_SOURCE ] ;
|
|
|
|
if [ option.get "with-oxlm" ] {
|
|
external-lib gomp ;
|
|
requirements += <library>boost_serialization ;
|
|
requirements += <library>gomp ;
|
|
}
|
|
|
|
if [ option.get "with-cmph" : : "yes" ] {
|
|
requirements += <define>HAVE_CMPH ;
|
|
}
|
|
|
|
if [ option.get "with-icu" : : "yes" ]
|
|
{
|
|
external-lib icuuc ;
|
|
external-lib icuio ;
|
|
external-lib icui18n ;
|
|
requirements += <library>icuuc/<link>shared ;
|
|
requirements += <library>icuio/<link>shared ;
|
|
requirements += <library>icui18n/<link>shared ;
|
|
requirements += <cxxflags>-fPIC ;
|
|
requirements += <address-model>64 ;
|
|
# requirements += <runtime-link>shared ;
|
|
}
|
|
|
|
# for probing pt
|
|
external-lib boost_serialization ;
|
|
requirements += <library>boost_serialization/<runtime-link>static ;
|
|
|
|
if [ option.get "with-vw" ] {
|
|
requirements += <define>HAVE_VW ;
|
|
}
|
|
|
|
project : default-build
|
|
<threading>multi
|
|
<warnings>on
|
|
<debug-symbols>off
|
|
<variant>release
|
|
<link>static
|
|
;
|
|
|
|
#Apparently OS X likes to link against iconv for fgetsUTF8.
|
|
lib iconv ;
|
|
requirements += <os>MACOSX:<library>iconv ;
|
|
|
|
project : requirements
|
|
<threading>multi:<define>WITH_THREADS
|
|
<threading>multi:<library>boost_thread
|
|
<library>boost_system
|
|
<library>boost_program_options
|
|
<define>_FILE_OFFSET_BITS=64 <define>_LARGE_FILES
|
|
$(requirements)
|
|
<include>.
|
|
;
|
|
|
|
|
|
#Add directories here if you want their incidental targets too (i.e. tests).
|
|
build-projects lm util phrase-extract phrase-extract/syntax-common search moses moses/LM mert moses-cmd scripts regression-testing ;
|
|
# contrib/mira
|
|
|
|
if [ option.get "with-mm-extras" : : "yes" ]
|
|
{
|
|
alias mm-extras :
|
|
moses/TranslationModel/UG//bitext-find
|
|
moses/TranslationModel/UG//ptable-describe-features
|
|
moses/TranslationModel/UG//count-ptable-features
|
|
moses/TranslationModel/UG//ptable-sigtest-filter
|
|
moses/TranslationModel/UG//ptable-lookup
|
|
moses/TranslationModel/UG//ptable-lookup-corpus
|
|
moses/TranslationModel/UG//check-coverage
|
|
moses/TranslationModel/UG/mm//mtt-demo1
|
|
moses/TranslationModel/UG/mm//mtt-dump
|
|
moses/TranslationModel/UG/mm//mam2symal
|
|
moses/TranslationModel/UG/mm//mam_verify
|
|
moses/TranslationModel/UG/mm//mmlex-lookup
|
|
moses/TranslationModel/UG/mm//mtt-count-words
|
|
moses/TranslationModel/UG/mm//calc-coverage
|
|
moses/TranslationModel/UG//try-align
|
|
;
|
|
}
|
|
else
|
|
{
|
|
alias mm-extras ;
|
|
}
|
|
|
|
if [ option.get "with-mm" : : "yes" ]
|
|
{
|
|
alias mm :
|
|
moses/TranslationModel/UG/mm//mtt-build
|
|
moses/TranslationModel/UG/mm//symal2mam
|
|
moses/TranslationModel/UG/mm//mmlex-build
|
|
;
|
|
}
|
|
else
|
|
{
|
|
alias mm ;
|
|
}
|
|
|
|
if [ option.get "with-rephraser" : : "yes" ]
|
|
{
|
|
alias rephraser :
|
|
contrib/rephraser//paraphrase
|
|
;
|
|
}
|
|
else
|
|
{
|
|
alias rephraser ;
|
|
}
|
|
|
|
alias programs :
|
|
lm//programs
|
|
moses-cmd//programs
|
|
OnDiskPt//CreateOnDiskPt
|
|
OnDiskPt//queryOnDiskPt
|
|
mert//programs
|
|
misc//programs
|
|
symal
|
|
phrase-extract
|
|
phrase-extract//lexical-reordering
|
|
phrase-extract//extract-ghkm
|
|
phrase-extract//pcfg-extract
|
|
phrase-extract//pcfg-score
|
|
phrase-extract//extract-mixed-syntax
|
|
phrase-extract//score-stsg
|
|
phrase-extract//filter-rule-table
|
|
phrase-extract//postprocess-egret-forests
|
|
biconcor
|
|
# contrib/mira//mira
|
|
contrib/server//mosesserver
|
|
mm
|
|
mm-extras
|
|
rephraser
|
|
contrib/c++tokenizer//tokenizer
|
|
contrib/expected-bleu-training//train-expected-bleu
|
|
contrib/expected-bleu-training//prepare-expected-bleu-training
|
|
|
|
probingpt//programs
|
|
moses2//programs
|
|
;
|
|
|
|
|
|
install-bin-libs programs ;
|
|
install-headers headers-base : [ path.glob-tree biconcor contrib lm mert misc moses-cmd OnDiskPt phrase-extract symal util : *.hh *.h ] : . ;
|
|
install-headers headers-moses : moses//headers-to-install : moses ;
|
|
|
|
alias install : prefix-bin prefix-lib headers-base headers-moses ;
|
|
|
|
if ! [ option.get "includedir" : : $(prefix)/include ] {
|
|
explicit install headers-base headers-moses ;
|
|
}
|
|
|
|
if [ path.exists $(TOP)/dist ] && $(prefix) != dist {
|
|
echo "You have a $(TOP)/dist directory, but the build system now places files directly in the root i.e. $(TOP)/bin ." ;
|
|
echo "To disable this message, delete $(TOP)/dist ." ;
|
|
echo ;
|
|
}
|
|
|
|
#local temp = [ _shell "bash source ./s.sh" ] ;
|
|
local temp = [ _shell "mkdir -p $(PREFIX)/bin" ] ;
|
|
local temp = [ _shell "rm -f $(PREFIX)/bin/moses_chart" ] ;
|
|
local temp = [ _shell "cd $(PREFIX)/bin && ln -sf moses moses_chart" ] ;
|
|
local temp = [ _shell "cd $(PREFIX)/bin && ln -sf CreateProbingPT CreateProbingPT2" ] ;
|
|
|