From d9248687b933960e4889c13657d09f1f54330910 Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Mon, 15 Jun 2020 17:03:51 +0200
Subject: [PATCH] updated pretrainEmbeddings script
---
UD_any/data/pretrainEmbeddings.py | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/UD_any/data/pretrainEmbeddings.py b/UD_any/data/pretrainEmbeddings.py
index 3817567..7e1d9f1 100755
--- a/UD_any/data/pretrainEmbeddings.py
+++ b/UD_any/data/pretrainEmbeddings.py
@@ -18,10 +18,15 @@ if __name__ == "__main__" :
splited = os.path.splitext(pathToFile)
target = sys.argv[3]
- if which("word2vec") is None or which("udpipe") is None :
+ if which("word2vec") is None :
+ print("word2vec not installed")
exit(0)
- p = subprocess.Popen("udpipe --output=horizontal none %s > in.text && rm in.text"%pathToFile, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
+ if which("udpipe") is None :
+ print("udpipe not installed")
+ exit(0)
+
+ p = subprocess.Popen("udpipe --output=horizontal none %s > in.text"%pathToFile, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
p.wait()
args = [
@@ -38,6 +43,7 @@ if __name__ == "__main__" :
"-min-count 2",
"-train in.text",
"-output %s"%target,
+ "&& rm in.text"
]
p = subprocess.Popen(" ".join(args), stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
--
GitLab