From 3de41e7ac47f2f9916c8e147a10fd4aa71efc25d Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Fri, 31 Jul 2020 18:27:34 +0200
Subject: [PATCH] conllu2horizontal replace every number by 42 and every digit
by 0
---
scripts/conllu2horizontal.py | 23 +++++++++++++++++++----
1 file changed, 19 insertions(+), 4 deletions(-)
diff --git a/scripts/conllu2horizontal.py b/scripts/conllu2horizontal.py
index b0997c0..a6ef9c4 100755
--- a/scripts/conllu2horizontal.py
+++ b/scripts/conllu2horizontal.py
@@ -3,6 +3,15 @@
import sys
from readMCD import readMCD
+def isNumber(s) :
+ hasDigit = False
+ for c in s :
+ if c.isdigit() :
+ hasDigit = True
+ if c.isalpha() :
+ return False
+ return hasDigit
+
def printUsageAndExit() :
print("USAGE : %s file.conllu (columnName | LETTERS)"%sys.argv[0], file=sys.stderr)
sys.exit(1)
@@ -36,7 +45,10 @@ if __name__ == "__main__" :
print("ERROR : column %s not found in line '%s'"%(index, line.strip()))
exit(1)
- print(splited[index].replace(" ", "◌"), end=" ")
+ value = splited[index].replace(" ", "◌")
+ if isNumber(splited[index].replace(" ", "").strip()) :
+ value = 42
+ print(value, end=" ")
else :
for line in open(sys.argv[1], "r") :
if line.startswith("#") :
@@ -45,6 +57,9 @@ if __name__ == "__main__" :
col2index, index2col = readMCD(splited[-1].strip())
splited = line.split("text =")
if len(splited) > 1 :
- text = splited[-1].replace("\n", " ").replace(" ", "◌")
- print(" ".join(list(text)))
-
+ text = list(splited[-1].replace("\n", " ").replace(" ", "◌"))
+ for elem in text :
+ if isNumber(str(elem)) :
+ elem = '0'
+ print(elem, end=" ")
+ print("")
--
GitLab