Integrate ctcdecode into build system

mozilla · Oct 25, 2018 · 0002d0f · 0002d0f
1 parent c34fc5b
commit 0002d0f
Showing 1 changed file with 29 additions and 24 deletions.
diff --git a/native_client/BUILD b/native_client/BUILD
@@ -10,6 +10,27 @@ genrule(
     tools = [":ds_git_version.sh"]
 )
 
+KENLM_SOURCES = glob(["kenlm/lm/*.cc", "kenlm/util/*.cc", "kenlm/util/double-conversion/*.cc",
+                      "kenlm/lm/*.hh", "kenlm/util/*.hh", "kenlm/util/double-conversion/*.h"],
+                     exclude = ["kenlm/*/*test.cc", "kenlm/*/*main.cc"]) + glob(["boost_locale/**/*.hpp"])
+
+KENLM_INCLUDES = [
+    "kenlm",
+    "boost_locale"
+]
+
+DECODER_SOURCES = glob([
+    "ctcdecode/*.h",
+    "ctcdecode/*.cpp",
+    "ctcdecode/third_party/openfst-1.6.7/src/lib/*.cc"
+]) + KENLM_SOURCES
+
+DECODER_INCLUDES = [
+    ".",
+    "ctcdecode/third_party/openfst-1.6.7/src/include",
+    "ctcdecode/third_party/ThreadPool"
+] + KENLM_INCLUDES
+
 tf_cc_shared_object(
     name = "libdeepspeech.so",
     srcs = ["deepspeech.cc",
@@ -26,10 +47,7 @@ tf_cc_shared_object(
             "kiss_fft130/_kiss_fft_guts.h",
             "kiss_fft130/tools/kiss_fftr.h",
             "ds_version.h"] +
-           glob(["kenlm/lm/*.cc", "kenlm/util/*.cc", "kenlm/util/double-conversion/*.cc",
-                 "kenlm/lm/*.hh", "kenlm/util/*.hh", "kenlm/util/double-conversion/*.h"],
-                exclude = ["kenlm/*/*test.cc", "kenlm/*/*main.cc"]) +
-           glob(["boost_locale/**/*.hpp"]),
+           DECODER_SOURCES,
     # -Wno-sign-compare to silent a lot of warnings from tensorflow itself,
     # which makes it harder to see our own warnings
     copts = ["-Wno-sign-compare", "-fvisibility=hidden"],
@@ -76,7 +94,7 @@ tf_cc_shared_object(
     ] + if_cuda([
         "//tensorflow/core:core",
     ]),
-    includes = ["kenlm", "boost_locale", "c_speech_features", "kiss_fft130"],
+    includes = ["c_speech_features", "kiss_fft130"] + DECODER_INCLUDES,
     defines = ["KENLM_MAX_ORDER=6"],
 )
 
@@ -88,11 +106,8 @@ tf_cc_shared_object(
             "alphabet.h",
             "trie_node.h"
            ] +
-           glob(["kenlm/lm/*.cc", "kenlm/util/*.cc", "kenlm/util/double-conversion/*.cc",
-                 "kenlm/lm/*.hh", "kenlm/util/*.hh", "kenlm/util/double-conversion/*.h"],
-                exclude = ["kenlm/*/*test.cc", "kenlm/*/*main.cc"]) +
-           glob(["boost_locale/**/*.hpp"]),
-    includes = ["kenlm", "boost_locale"],
+           KENLM_SOURCES,
+    includes = KENLM_INCLUDES,
     copts = ["-std=c++11"],
     defines = ["KENLM_MAX_ORDER=6"],
     deps = ["//tensorflow/core:framework_headers_lib",
@@ -105,14 +120,9 @@ cc_binary(
     name = "generate_trie",
     srcs = [
             "generate_trie.cpp",
-            "trie_node.h",
             "alphabet.h",
-           ] +
-           glob(["kenlm/lm/*.cc", "kenlm/util/*.cc", "kenlm/util/double-conversion/*.cc",
-                 "kenlm/lm/*.hh", "kenlm/util/*.hh", "kenlm/util/double-conversion/*.h"],
-                exclude = ["kenlm/*/*test.cc", "kenlm/*/*main.cc"]) +
-           glob(["boost_locale/**/*.hpp"]),
-    includes = ["kenlm", "boost_locale"],
+           ] + DECODER_SOURCES,
+    includes = DECODER_INCLUDES,
     copts = ["-std=c++11"],
     linkopts = ["-lm"],
     defines = ["KENLM_MAX_ORDER=6"],
@@ -122,14 +132,9 @@ cc_binary(
     name = "trie_load",
     srcs = [
             "trie_load.cc",
-            "trie_node.h",
             "alphabet.h",
-           ] +
-           glob(["kenlm/lm/*.cc", "kenlm/util/*.cc", "kenlm/util/double-conversion/*.cc",
-                 "kenlm/lm/*.hh", "kenlm/util/*.hh", "kenlm/util/double-conversion/*.h"],
-                exclude = ["kenlm/*/*test.cc", "kenlm/*/*main.cc"]) +
-           glob(["boost_locale/**/*.hpp"]),
-    includes = ["kenlm", "boost_locale"],
+           ] + DECODER_SOURCES,
+    includes = DECODER_INCLUDES,
     copts = ["-std=c++11"],
     linkopts = ["-lm"],
     defines = ["KENLM_MAX_ORDER=6"],