From 1494d4dfedb73981905bcfc73b41f945a7a1c7eb Mon Sep 17 00:00:00 2001 From: voje Date: Mon, 15 Apr 2019 00:26:08 +0200 Subject: [PATCH] cjvt-corpusparser parallel upgrade --- Makefile | 10 ++++++---- src/pkg/cjvt-corpusparser | 2 +- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/Makefile b/Makefile index d7373d7..750f1a4 100644 --- a/Makefile +++ b/Makefile @@ -12,8 +12,9 @@ SSJ_FILE = "$(MAKE_ROOT)/data/samples/ssj_xml/ssj500k-sl.body.sample.xml" KRES_FOLDER = "$(MAKE_ROOT)/data/samples/kres_xml" KRES_SRL_FOLDER = "$(MAKE_ROOT)/data/samples/kres_srl_json" -OUTPUT = "db" -OUTDIR = "/home/voje/workdir/test_out" +# OUTPUT = "db" +OUTPUT = "file" +OUTDIR = "/tmp/three" # if you're running this in docker, make sure to mount the volume DBADDR = "0.0.0.0:27017" # don't use localhost DB_ADM_USER = valadmin @@ -56,12 +57,13 @@ data/samples: cd data; tar xzvf samples.tar.gz # from inside python-env container: +# you can set OUTPUT = "file" and a valid OUTDIR to test writing to json files instead of DB fill-database: data/samples python3 src/pkg/cjvt-corpusparser/corpusparser/main.py --kres-folder $(KRES_FOLDER) \ --ssj-file $(SSJ_FILE) --kres-srl-folder $(KRES_SRL_FOLDER) \ --output $(OUTPUT) --outdir $(OUTDIR) --dbaddr $(DBADDR) \ - --dbuser $(DB_USR_USER) --dbpass $(DB_USR_PASS) - + --dbuser $(DB_USR_USER) --dbpass $(DB_USR_PASS) \ + --cores 2 ## Frontend ## Run from host diff --git a/src/pkg/cjvt-corpusparser b/src/pkg/cjvt-corpusparser index c17361f..86e5676 160000 --- a/src/pkg/cjvt-corpusparser +++ b/src/pkg/cjvt-corpusparser @@ -1 +1 @@ -Subproject commit c17361fbda194d845cbac674065e40490ebc04cc +Subproject commit 86e56767ddb72b83adcb144c32373b3e92e215dc