diff --git a/Makefile b/Makefile index 33d04fc..1a3ec8e 100644 --- a/Makefile +++ b/Makefile @@ -10,10 +10,10 @@ MAKE_ROOT = $(shell pwd) # (for each kres.xml file there is a kres.json file with srl tags) # SSJ_FILE = "$(MAKE_ROOT)/data/samples/ssj_xml/ssj500k-sl.body.sample.xml" SSJ_FILE = "$(MAKE_ROOT)/data/ssj_file_link" -KRES_FOLDER = "$(MAKE_ROOT)/data/samples/kres_xml" -# KRES_FOLDER = "$(MAKE_ROOT)/data/kres_xml_folder_link" -KRES_SRL_FOLDER = "$(MAKE_ROOT)/data/samples/kres_srl_json" -# KRES_SRL_FOLDER = "$(MAKE_ROOT)/data/kres_json_folder_link" +# KRES_FOLDER = "$(MAKE_ROOT)/data/samples/kres_xml" +KRES_FOLDER = "$(MAKE_ROOT)/data/kres_xml_folder_link" +# KRES_SRL_FOLDER = "$(MAKE_ROOT)/data/samples/kres_srl_json" +KRES_SRL_FOLDER = "$(MAKE_ROOT)/data/kres_json_folder_link" OUTPUT = "db" # OUTPUT = "file" @@ -26,7 +26,7 @@ include env.local N_CORES = 5 # insert kres files into database in chunks, for fewer connections -KRES_CHUNK_SIZE = 3 +KRES_CHUNK_SIZE = 30 export .PHONY: python-env fill-database diff --git a/data/kres_json_folder_link b/data/kres_json_folder_link index 9c386ed..965aba4 120000 --- a/data/kres_json_folder_link +++ b/data/kres_json_folder_link @@ -1 +1 @@ -/home/kristjan/workdir/final_json/ \ No newline at end of file +/home/kristjan/kres_data/payload/kres_json/ \ No newline at end of file diff --git a/data/kres_json_folder_link_prod b/data/kres_json_folder_link_prod new file mode 120000 index 0000000..9c386ed --- /dev/null +++ b/data/kres_json_folder_link_prod @@ -0,0 +1 @@ +/home/kristjan/workdir/final_json/ \ No newline at end of file diff --git a/data/kres_xml_folder_link b/data/kres_xml_folder_link index b5b31cd..b3fbfac 120000 --- a/data/kres_xml_folder_link +++ b/data/kres_xml_folder_link @@ -1 +1 @@ -/home/kristjan/kres_mount/kres_parsed/tei/ \ No newline at end of file +/home/kristjan/kres_data/payload/kres_xml/ \ No newline at end of file diff --git a/data/kres_xml_folder_link_proc b/data/kres_xml_folder_link_proc new file mode 120000 index 0000000..b5b31cd --- /dev/null +++ b/data/kres_xml_folder_link_proc @@ -0,0 +1 @@ +/home/kristjan/kres_mount/kres_parsed/tei/ \ No newline at end of file diff --git a/src/pkg/cjvt-corpusparser b/src/pkg/cjvt-corpusparser index 2582314..01adf47 160000 --- a/src/pkg/cjvt-corpusparser +++ b/src/pkg/cjvt-corpusparser @@ -1 +1 @@ -Subproject commit 2582314c4d69aa6f6a7be9fdf898a2c3fa590d5e +Subproject commit 01adf47b9b63b43f86bff52429792b0de2327ddd