songleash

commit 73f0f51928dfbccd6857ca0fa2d82c05decd6264
parent 63838ada2f96be79473763307350400fe455ab4c
Author: Antoine Amarilli <a3nm@a3nm.net>
Date:   Wed, 26 Aug 2015 23:00:10 +0200

lexique actually provides all that we need

Diffstat:
common.py
 | 
2
+-
only3.py
 | 
22
++++++++++------------
script.sh
 | 
3
++-

3 files changed, 13 insertions(+), 14 deletions(-)
diff --git a/common.py b/common.py
@@ -2,7 +2,7 @@
 
 import unicodedata
 
-vowels = 'io9@2EeaOy#$u()'
+vowels = 'io9@2EeaOy#$u()§°51'
 vowels_script = "aeiouy"
 semivowels = 'j8w'
 
diff --git a/only3.py b/only3.py
@@ -7,22 +7,20 @@ import sys
 for l in sys.stdin.readlines():
     f = l.split('\t')
     vowels_regexp = re.compile('([' + vowels + '])')
-    f[1] = f[1].strip()
-    parse = re.split(vowels_regexp, f[1])
-    parse = [x for x in parse if len(x) > 0]
-    # count number of vowel sounds
-    s = sum([1 for x in parse if x[0] in vowels])
-    if s != 3:
-        continue
+    f[-1] = f[-1].strip()
+    if len(f) <= 3:
+        f.append("")
+    chunks = f[1].split('-')
     # words cannot start with a vowel
     # as last consonant before last vowel will be kept
-    if parse[0][0] in vowels:
+    if chunks[0][0] in vowels:
         continue
     # sound preceding the last vowel sound must be a consonant
-    if parse[-1][0] in vowels and parse[-2][0] in vowels:
+    if chunks[-1][-1] in vowels and (len(chunks[-1]) == 1 or chunks[-1][-2] in
+            vowels):
         continue
-    if (parse[-1][0] not in vowels and parse[-2][0] in vowels
-            and parse[-3][0] in vowels):
+    if (chunks[-1][-1] not in vowels and chunks[-1][-2] in vowels
+            and (len(chunks[-1]) == 2 or chunks[-1][-3] in vowels)):
         continue
-    print("%s\t%s" % (f[0], ' '.join(parse)))
+    print('\t'.join(f))
 
diff --git a/script.sh b/script.sh
@@ -5,7 +5,8 @@
 
 #cat Lexique371/Bases+Scripts/Lexique3.txt |
 #  ./lexique_fix.sh| cut -f1,2,4 | grep NOM | uniq > lexique.txt
-pv lexique.txt| ./only3.py > lexique3
+cat ~/documents/lexique/lexique | cut -f1,2,4,23,24,28 | grep NOM | grep '\s3\s' | cut -f1,4,6 | rev | uniq -f 2 | rev > lexique_full
+pv lexique_full | ./only3.py > lexique3
 cat lexique3 | ./graph.py > graph
 cat graph | ./cycle.py