songleash

commit 50379b128bc817f4d7d18eb03556f3bf3c03a506
parent 1672857fa9897ecd847992501ba142c11f33a7bf
Author: Antoine Amarilli <a3nm@a3nm.net>
Date:   Tue,  4 Aug 2015 18:32:38 +0200

TODOs

Diffstat:
common.py
 | 
1
+
cycle.py
 | 
4
++++
graph.py
 | 
18
+++++++-----------
script.sh
 | 
3
+++

4 files changed, 15 insertions(+), 11 deletions(-)
diff --git a/common.py b/common.py
@@ -6,6 +6,7 @@ vowels = 'io9@2EeaOy#$u()'
 vowels_script = "aeiouy"
 semivowels = 'j8w'
 
+# TODO replace by startswith
 def is_pref(u, v):
     if len(v) < len(u):
         return False
diff --git a/cycle.py b/cycle.py
@@ -3,6 +3,8 @@
 import sys
 from common import fem, vowels_script, rmacc, semivowels
 
+# TODO use defaultdict
+
 g = {}
 START = "sEl"
 bestlen = 0
@@ -55,6 +57,8 @@ def print_list(l):
     print("------------------")
     print("")
 
+# TODO: do something more clever
+
 def dfs(l):
     global g
     global bestlen
diff --git a/graph.py b/graph.py
@@ -3,6 +3,8 @@
 import sys
 from common import vowels, is_pref, semivowels, fem
 
+# TODO use defaultdict
+
 g = {}
 
 def nedd(myg):
@@ -55,6 +57,8 @@ for l in sys.stdin.readlines():
     p = p.split(' ')
     for i in range(len(p[2])+1):
         f = p[0] + p[1] + p[2][:i]
+        # TODO: this cut is non-optimal: should be "radio/gramme", "géo/graphe",
+        # but how to tell?
         t = ((p[-2][-2:] if p[-2][-1] in semivowels else p[-2][-1]) if p[-1][0] in vowels else ((p[-3][-2:] if p[-3][-1] in
             semivowels else p[-3][-1]) + p[-2])) + p[-1]
         # print ("%s : %s -> %s" % (w, f, t))
@@ -77,6 +81,9 @@ for l in sys.stdin.readlines():
             continue
         g[f][t].add(w)
 
+
+# TODO: replace this by a true SCC
+
 ned = nedd(g)
 print(ned, file=sys.stderr)
 
@@ -96,15 +103,4 @@ for f in g.keys():
         for w in g[f][t]:
             print ("%s %s %s" % (f, t, w))
 
-#
-#for j in range(100):
-#    print("-----------------")
-#    f = list(g.keys())[j]
-#    print(f)
-#    for i in range(100):
-#        t = list(g[f])[0]
-#        if t[0] not in g.keys():
-#            continue
-#        print("%s -[%s]-> %s" % (f, t[1], t[0]))
-#        f = t[0]
 
diff --git a/script.sh b/script.sh
@@ -1,5 +1,8 @@
 #!/bin/bash
 
+# TODO: add n-grams from wikipedia article titles with unambiguous
+# pronunciation, known words, and suitable POS
+
 #cat Lexique371/Bases+Scripts/Lexique3.txt |
 #  ./lexique_fix.sh| cut -f1,2,4 | uniq > lexique.txt
 pv lexique.txt| ./only3.py > lexique3