Bump libarchive to 3.7.3, pre-upgrade to tip emscripten until 3.1.58 is out.

This commit is contained in:
msqr1
2024-04-20 00:24:35 -07:00
parent 4bca7c142d
commit 5f5a7a7571
9 changed files with 76 additions and 279 deletions

File diff suppressed because one or more lines are too long

45
devel/Vosklet.d.ts vendored
View File

@@ -1,45 +0,0 @@
// TypeScript bindings for emscripten-generated code. Automatically generated at compile time.
interface WasmModule {
_pthread_self(): number;
_malloc(_0: number): number;
__emscripten_tls_init(): number;
__embind_initialize_bindings(): void;
__emscripten_thread_init(_0: number, _1: number, _2: number, _3: number, _4: number, _5: number): void;
__emscripten_thread_crashed(): void;
__emscripten_thread_exit(_0: number): void;
__ZN6__asan9FakeStack17AddrIsInFakeStackEm(_0: number, _1: number): number;
__ZN6__asan9FakeStack8AllocateEmmm(_0: number, _1: number, _2: number, _3: number): number;
___set_stack_limits(_0: number, _1: number): void;
}
export interface epModeValue<T extends number> {
value: T;
}
export type epMode = epModeValue<0>|epModeValue<1>|epModeValue<2>|epModeValue<3>;
export interface genericModel {
extractAndLoad(_0: number, _1: number): void;
delete(): void;
}
export interface recognizer {
reset(): void;
setEndpointerMode(_0: epMode): void;
setSpkModel(_0: genericModel): void;
setWords(_0: boolean): void;
setPartialWords(_0: boolean): void;
setNLSML(_0: boolean): void;
pushData(_0: number, _1: number): void;
setMaxAlternatives(_0: number): void;
setEndpointerDelays(_0: number, _1: number, _2: number): void;
setGrm(_0: ArrayBuffer|Uint8Array|Uint8ClampedArray|Int8Array|string): void;
delete(): void;
}
interface EmbindModule {
epMode: {ANSWER_DEFAULT: epModeValue<0>, ANSWER_SHORT: epModeValue<1>, ANSWER_LONG: epModeValue<2>, ANSWER_VERY_LONG: epModeValue<3>};
genericModel: {new(_0: number, _1: boolean, _2: ArrayBuffer|Uint8Array|Uint8ClampedArray|Int8Array|string, _3: ArrayBuffer|Uint8Array|Uint8ClampedArray|Int8Array|string): genericModel};
recognizer: {new(_0: number, _1: number, _2: genericModel): recognizer; new(_0: number, _1: number, _2: genericModel, _3: genericModel): recognizer; new(_0: number, _1: number, _2: genericModel, _3: ArrayBuffer|Uint8Array|Uint8ClampedArray|Int8Array|string, _4: number): recognizer};
setLogLevel(_0: number): void;
}
export type MainModule = WasmModule & EmbindModule;

File diff suppressed because one or more lines are too long

View File

@@ -1,7 +1,7 @@
<!DOCTYPE html>
<html>
<head>
<script src="Vosklet.js" async defer></script>
<script src="../Vosklet.js" async defer></script>
<script>
async function start() {
let ctx = new AudioContext({sampleRate : 16000})

View File

@@ -1,25 +1,50 @@
# Build the js file into this directory
cd .. &&
SRC=$(realpath src) &&
KALDI=$(realpath kaldi) &&
VOSK=$(realpath vosk) &&
OPENFST=$(realpath openfst) &&
LIBARCHIVE=$(realpath libarchive) &&
CLAPACK_WASM=$(realpath clapack-wasm) &&
#!/bin/bash
MAX_MEMORY=${MAX_MEMORY:-300mb}
MAX_THREADS=${MAX_THREADS:-1}
EMSDK=${EMSDK:-$(realpath ../emsdk)}
. $EMSDK/emsdk_env.sh &&
EMSDK=${EMSDK:-../emsdk}
JOBS=${JOBS:-$(nproc)}
cd $SRC &&
MODE=1 && # 0: Ultra debug info, 1: Extremely optimized release, else custom
if [ $EMSDK != ../emsdk ] && [ ! -d $EMSDK ]; then
echo "Invalid emsdk path"
exit 1
fi
if [ $MAX_THREADS -lt 1 ]; then
echo "MAX_THREADS must be greater than or equal to 1"
exit 1
fi
if [ $JOBS -lt 1 ]; then
echo "JOBS must be greater than or equal to 1"
exit 1
fi
if ! [[ $MAX_MEMORY =~ ^[0-9]+([kmgt]b)?$ ]]; then
echo "MAX_MEMORY valid suffixes are kb, mb, gb, tb, none (bytes)"
exit 1
fi
if [ $EMSDK = ../emsdk ] && [ ! -d $EMSDK ]; then
echo "Installing emsdk + Emscripten..."
git clone --depth=1 https://github.com/emscripten-core/emsdk.git ../emsdk &&
cd ../emsdk &&
./emsdk install tot &&
./emsdk activate tot
fi
. $(realpath $EMSDK)/emsdk_env.sh &&
export PATH=:$PATH:$(realpath $EMSDK)/upstream/bin
cd ..
KALDI=$(realpath kaldi)
VOSK=$(realpath vosk)
OPENFST=$(realpath openfst)
LIBARCHIVE=$(realpath libarchive)
CLAPACK_WASM=$(realpath clapack-wasm)
cd $(realpath src)
MODE=0 && # 0: Ultra debug info, 1: Extremely optimized release, else custom
echo "Mode = $MODE" &&
if [ $MODE = 0 ]; then
em++ -O0 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sRUNTIME_DEBUG -sSTACK_OVERFLOW_CHECK=2 -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sASSERTIONS=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sDISABLE_EXCEPTION_CATCHING=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sPOLYFILL=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 --embind-emit-tsd Vosklet.d.ts -fsanitize=undefined -fsanitize=address -fsanitize=leak -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -g3 --pre-js pre.js -o ../devel/Vosklet.js
em++ -O0 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sRUNTIME_DEBUG -sSTACK_OVERFLOW_CHECK=2 -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sASSERTIONS=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sDISABLE_EXCEPTION_CATCHING=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sPOLYFILL=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 --emit-tsd Vosklet.d.ts -fsanitize=undefined -fsanitize=address -fsanitize=leak -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -g3 --pre-js pre.js -o ../Vosklet.js
elif [ $MODE = 1 ]; then
em++ -O3 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext --pre-js pre.js -o ../devel/Vosklet.js
em++ -O3 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals --pre-js pre.js -o ../Vosklet.js
else
em++ -O0 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext --pre-js pre.js -o ../devel/Vosklet.js
em++ -O0 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals --pre-js pre.js -o ../Vosklet.js
fi
rm -f ../devel/Vosklet.worker.js
rm -f Vosklet.worker.js

View File

@@ -7,7 +7,7 @@ index bf7eda6..5fedaf6 100644
#-----------------------------------------------------------------------------
-CFLAGS = -DADD_ -O3
+CFLAGS = -DADD_ -O3 -flto -msimd128
+CFLAGS = -DADD_ -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals
#-----------------------------------------------------------------------------
# Archive programs and flags
@@ -20,7 +20,7 @@ index 80037d0..a964b2d 100644
# if no wrapping of the blas library is needed, uncomment next line
CC = emcc # -DNO_BLAS_WRAP
-CFLAGS = -I$(TOPDIR)/INCLUDE -I$(TOPDIR)/../libf2c -O3
+CFLAGS = -I$(TOPDIR)/INCLUDE -I$(TOPDIR)/../libf2c -O3 -flto -msimd128
+CFLAGS = -I$(TOPDIR)/INCLUDE -I$(TOPDIR)/../libf2c -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals
LOADER = $(CC)
LOADOPTS =
NOOPT = -O0 -I$(TOPDIR)/INCLUDE -I$(TOPDIR)/../libf2c
@@ -33,7 +33,7 @@ index e071614..4647c2b 100644
#
CC = emcc
-CFLAGS = -I../libf2c -O3
+CFLAGS = -I../libf2c -O3 -flto -msimd128
+CFLAGS = -I../libf2c -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals
DRVOPTS = $(OPTS)
NOOPT =
LOADER = emcc
@@ -46,7 +46,7 @@ index 6221401..d93b87f 100644
CC = emcc
SHELL = /bin/sh
-CFLAGS = -DNON_UNIX_STDIO -O3
+CFLAGS = -DNON_UNIX_STDIO -O3 -flto -msimd128
+CFLAGS = -DNON_UNIX_STDIO -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals
LD = wasm-ld
RANLIB = emranlib

View File

@@ -7,7 +7,7 @@ JOBS=${JOBS:-$(nproc)}
if [ $EMSDK != ../emsdk ] && [ ! -d $EMSDK ]; then
echo "Invalid emsdk path"
exit 1
fi
fi
if [ $MAX_THREADS -lt 1 ]; then
echo "MAX_THREADS must be greater than or equal to 1"
exit 1
@@ -17,20 +17,20 @@ if [ $JOBS -lt 1 ]; then
exit 1
fi
if ! [[ $MAX_MEMORY =~ ^[0-9]+([kmgt]b)?$ ]]; then
echo "MAX_MEMORY valid suffixes are kb, mb, gb, tb, none (bytes"
echo "MAX_MEMORY valid suffixes are kb, mb, gb, tb, none (bytes)"
exit 1
fi
if [ $EMSDK = ../emsdk ]; then
if [ $EMSDK = ../emsdk ] && [ ! -d $EMSDK ]; then
echo "Installing emsdk + Emscripten..."
git clone --depth=1 https://github.com/emscripten-core/emsdk.git ../emsdk &&
cd ../emsdk &&
./emsdk install 3.1.56 &&
./emsdk activate 3.1.56
./emsdk install tot &&
./emsdk activate tot
fi
. $(realpath $EMSDK)/emsdk_env.sh &&
export PATH=:$PATH:$(realpath $EMSDK)/upstream/bin
export PATH=:$PATH:$(realpath $EMSDK)/upstream/bin
cd ..
cd ..
SRC=$(realpath src)
KALDI=$(realpath kaldi)
VOSK=$(realpath vosk)
@@ -40,11 +40,12 @@ CLAPACK_WASM=$(realpath clapack-wasm)
if [ ! -d $LIBARCHIVE ]; then
rm -rf /tmp/libarchive &&
git clone -b v3.7.2 --depth=1 https://github.com/libarchive/libarchive /tmp/libarchive &&
git clone -b v3.7.3 --depth=1 https://github.com/libarchive/libarchive /tmp/libarchive &&
cd /tmp/libarchive &&
build/autogen.sh &&
CFLAGS="-O3 -flto -msimd128" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix=$LIBARCHIVE --without-lz4 --without-lzma --without-zlib --without-bz2lib --without-xml2 --without-expat --without-cng --without-openssl --without-libb2 --without-zstd --disable-bsdunzip --disable-xattr --disable-acl --disable-bsdcpio --disable-bsdcat --disable-rpath --disable-maintainer-mode --disable-dependency-tracking --enable-static --disable-shared &&
emmake make -j$JOBS install
CFLAGS="-O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix=$LIBARCHIVE --without-lz4 --without-lzma --without-zlib --without-bz2lib --without-xml2 --without-expat --without-cng --without-openssl --without-libb2 --without-zstd --disable-bsdunzip --disable-xattr --disable-acl --disable-bsdcpio --disable-bsdcat --disable-rpath --disable-maintainer-mode --disable-dependency-tracking --enable-static --disable-shared &&
emmake make -j$JOBS install &&
rm -rf /tmp/libarchive
fi
if [ ! -d $OPENFST ]; then
@@ -52,9 +53,10 @@ if [ ! -d $OPENFST ]; then
git clone --depth=1 https://github.com/alphacep/openfst /tmp/openfst &&
cd /tmp/openfst &&
autoreconf -is &&
CXXFLAGS="-r -O3 -flto -msimd128" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix=$OPENFST --enable-static --disable-shared --enable-lookahead-fsts --enable-ngram-fsts --disable-bin &&
CXXFLAGS="-r -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix=$OPENFST --enable-static --disable-shared --enable-lookahead-fsts --enable-ngram-fsts --disable-bin &&
emmake make -j$JOBS install &&
echo "PACKAGE_VERSION = 1.8.0" >> $OPENFST/Makefile
echo "PACKAGE_VERSION = 1.8.0" >> $OPENFST/Makefile &&
rm -rf /tmp/openfst
fi
if [ ! -d $CLAPACK_WASM ]; then
@@ -68,7 +70,7 @@ if [ ! -d $KALDI ]; then
git clone -b vosk --depth=1 https://github.com/alphacep/kaldi $KALDI &&
cd $KALDI/src &&
git apply $SRC/kaldi.patch &&
CXXFLAGS="-O3 -UHAVE_EXECINFO_H -flto -msimd128 -Wno-unused-variable -Wno-unused-but-set-variable -g0" LDFLAGS="-O3 -lembind -flto -g0" emconfigure ./configure --use-cuda=no --with-cudadecoder=no --static --static-math=yes --static-fst=yes --debug-level=0 --fst-root=$OPENFST --clapack-root=$CLAPACK_WASM --host=WASM &&
CXXFLAGS="-O3 -UHAVE_EXECINFO_H -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-unused-variable -Wno-unused-but-set-variable -g0" LDFLAGS="-O3 -lembind -flto -g0" emconfigure ./configure --use-cuda=no --with-cudadecoder=no --static --static-math=yes --static-fst=yes --debug-level=0 --fst-root=$OPENFST --clapack-root=$CLAPACK_WASM --host=WASM &&
emmake make -j$JOBS online2 rnnlm
fi
@@ -77,12 +79,10 @@ if [ ! -d $VOSK ]; then
cd $VOSK/src &&
git apply $SRC/vosk.patch &&
VOSK_FILES="recognizer.cc language_model.cc model.cc spk_model.cc vosk_api.cc" &&
em++ -O3 -flto -msimd128 -Wno-deprecated -I. -I$KALDI/src -I$OPENFST/include $VOSK_FILES -c &&
em++ -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-deprecated -I. -I$KALDI/src -I$OPENFST/include $VOSK_FILES -c &&
emar -rcs vosk.a ${VOSK_FILES//.cc/.o}
fi
cd $SRC &&
em++ -O3 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext --pre-js pre.js -o ../Vosklet.js &&
rm -f ../Vosklet.worker.js &&
rm -rf /tmp/libarchive &&
rm -rf /tmp/openfst
em++ -O3 link.cc genericModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals --pre-js pre.js -o ../Vosklet.js &&
rm -f Vosklet.worker.js

View File

@@ -19,173 +19,8 @@ let processorURL = URL.createObjectURL(new Blob(['(', (() => {
}
})
}).toString(), ')()'], { type : "text/javascript" }))
let pthreadURL = URL.createObjectURL(new Blob(['(', (() => {
/**
* @license
* Copyright 2015 The Emscripten Authors
* SPDX-License-Identifier: MIT
*/
// Pthread Web Worker startup routine:
// This is the entry point file that is loaded first by each Web Worker
// that executes pthreads on the Emscripten application.
'use strict';
var Module = {};
// Thread-local guard variable for one-time init of the JS state
var initializedJS = false;
function assert(condition, text) {
if (!condition) abort('Assertion failed: ' + text);
}
function threadPrintErr(...args) {
var text = args.join(' ');
console.error(text);
}
function threadAlert(...args) {
var text = args.join(' ');
postMessage({cmd: 'alert', text, threadId: Module['_pthread_self']()});
}
// We don't need out() for now, but may need to add it if we want to use it
// here. Or, if this code all moves into the main JS, that problem will go
// away. (For now, adding it here increases code size for no benefit.)
var out = () => { throw 'out() is not defined in worker.js.'; }
var err = threadPrintErr;
self.alert = threadAlert;
var dbg = threadPrintErr;
Module['instantiateWasm'] = (info, receiveInstance) => {
// Instantiate from the module posted from the main thread.
// We can just use sync instantiation in the worker.
var module = Module['wasmModule'];
// We don't need the module anymore; new threads will be spawned from the main thread.
Module['wasmModule'] = null;
var instance = new WebAssembly.Instance(module, info);
// TODO: Due to Closure regression https://github.com/google/closure-compiler/issues/3193,
// the above line no longer optimizes out down to the following line.
// When the regression is fixed, we can remove this if/else.
return receiveInstance(instance);
}
// Turn unhandled rejected promises into errors so that the main thread will be
// notified about them.
self.onunhandledrejection = (e) => {
throw e.reason || e;
};
function handleMessage(e) {
try {
if (e.data.cmd === 'load') { // Preload command that is called once per worker to parse and load the Emscripten code.
// Until we initialize the runtime, queue up any further incoming messages.
let messageQueue = [];
self.onmessage = (e) => messageQueue.push(e);
// And add a callback for when the runtime is initialized.
self.startWorker = (instance) => {
Module = instance;
// Notify the main thread that this thread has loaded.
postMessage({ 'cmd': 'loaded' });
// Process any messages that were queued before the thread was ready.
for (let msg of messageQueue) {
handleMessage(msg);
}
// Restore the real message handler.
self.onmessage = handleMessage;
};
// Module and memory were sent from main thread
Module['wasmModule'] = e.data.wasmModule;
// Use `const` here to ensure that the variable is scoped only to
// that iteration, allowing safe reference from a closure.
for (const handler of e.data.handlers) {
Module[handler] = (...args) => {
postMessage({ cmd: 'callHandler', handler, args: args });
}
}
Module['wasmMemory'] = e.data.wasmMemory;
Module['buffer'] = Module['wasmMemory'].buffer;
Module['workerID'] = e.data.workerID;
Module['ENVIRONMENT_IS_PTHREAD'] = true;
if (typeof e.data.urlOrBlob == 'string') {
importScripts(e.data.urlOrBlob);
} else {
var objectUrl = URL.createObjectURL(e.data.urlOrBlob);
importScripts(objectUrl);
URL.revokeObjectURL(objectUrl);
}
loadVosklet(Module);
} else if (e.data.cmd === 'run') {
// Pass the thread address to wasm to store it for fast access.
Module['__emscripten_thread_init'](e.data.pthread_ptr, /*is_main=*/0, /*is_runtime=*/0, /*can_block=*/1);
// Await mailbox notifications with `Atomics.waitAsync` so we can start
// using the fast `Atomics.notify` notification path.
Module['__emscripten_thread_mailbox_await'](e.data.pthread_ptr);
assert(e.data.pthread_ptr);
// Also call inside JS module to set up the stack frame for this pthread in JS module scope
Module['establishStackSpace']();
Module['PThread'].receiveObjectTransfer(e.data);
Module['PThread'].threadInitTLS();
if (!initializedJS) {
// Embind must initialize itself on all threads, as it generates support JS.
// We only do this once per worker since they get reused
Module['__embind_initialize_bindings']();
initializedJS = true;
}
try {
Module['invokeEntryPoint'](e.data.start_routine, e.data.arg);
} catch(ex) {
if (ex != 'unwind') {
// The pthread "crashed". Do not call `_emscripten_thread_exit` (which
// would make this thread joinable). Instead, re-throw the exception
// and let the top level handler propagate it back to the main thread.
throw ex;
}
}
} else if (e.data.cmd === 'cancel') { // Main thread is asking for a pthread_cancel() on this thread.
if (Module['_pthread_self']()) {
Module['__emscripten_thread_exit'](-1);
}
} else if (e.data.target === 'setimmediate') {
// no-op
} else if (e.data.cmd === 'checkMailbox') {
if (initializedJS) {
Module['checkMailbox']();
}
} else if (e.data.cmd) {
// The received message looks like something that should be handled by this message
// handler, (since there is a e.data.cmd field present), but is not one of the
// recognized commands:
err(`worker.js received unknown command ${e.data.cmd}`);
err(e.data);
}
} catch(ex) {
err(`worker.js onmessage() captured an uncaught exception: ${ex}`);
if (ex?.stack) err(ex.stack);
Module['__emscripten_thread_crashed']?.();
throw ex;
}
};
self.onmessage = handleMessage;
}).toString(), ')()'], { type : "text/javascript" }))
Module.cleanUp = () => {
objs.forEach(obj => obj.obj.delete())
URL.revokeObjectURL(pthreadURL)
URL.revokeObjectURL(processorURL)
}
Module.createTransferer = async (ctx, bufferSize) => {
@@ -198,10 +33,6 @@ Module.createTransferer = async (ctx, bufferSize) => {
processorOptions : { maxCount: bufferSize / 128 }
})
}
Module.locateFile = (path, scriptDir) => {
if(path === "Vosklet.worker.js") return pthreadURL
return scriptDir + path
}
async function getFileHandle(path, create = false) {
let components = path.split("/")
let prevDir = await navigator.storage.getDirectory()

View File

@@ -91,17 +91,17 @@ index 7fc09df..608d28a 100644
using namespace kaldi;
using namespace std;
diff --git a/src/recognizer.cc b/src/recognizer.cc
index 1da6e6b..5e0652a 100644
index 1da6e6b..44c4d8c 100644
--- a/src/recognizer.cc
+++ b/src/recognizer.cc
@@ -384,8 +384,10 @@ bool Recognizer::AcceptWaveform(const float *fdata, int len)
@@ -384,8 +384,9 @@ bool Recognizer::AcceptWaveform(const float *fdata, int len)
{
Vector<BaseFloat> wave;
wave.Resize(len, kUndefined);
- for (int i = 0; i < len; i++)
- wave(i) = fdata[i];
+ for (int i = 0; i < len; i++) {
wave(i) = fdata[i];
+ wave(i) *= 32768;
+ wave(i) = fdata[i] * 32768.0f;
+ }
return AcceptWaveform(wave);
}