drime

commit 4da377ba38cfb840d3c45f3e3bc50e711e214823
parent 160ed739006ce8f2eac1ef9cce887f0ccd2f5e67
Author: Antoine Amarilli <a3nm@a3nm.net>
Date:   Thu, 29 Sep 2011 15:18:33 +0200

continuing cleanup

Diffstat:
README
 | 
4
++++
make_db.py
 | 
37
++++++++++++++++---------------------
make_db.sh
 | 
7
++++---

3 files changed, 24 insertions(+), 24 deletions(-)
diff --git a/README b/README
@@ -11,6 +11,10 @@ drime is a French rhyme dictionary engine with advanced features, most
 notably rime selection based on phonetic or visual similarity,
 frequency, syllable count, and rhyme genre.
 
+== 2. Requirements ==
+
+drime requires the haspirater module <http://gitorious.org/haspirater/>.
+
 == 2. Generating the DB ==
 
 The program database isn't shipped, but scripts are provided to build it
diff --git a/make_db.py b/make_db.py
@@ -7,26 +7,26 @@
 
 import haspirater
 import metric
+from common import is_vowels, is_consonants, sure_end_fem
 import sys
 
 seen = {}
 
-vowels = "aàâãeéèêëiîïoôöuùûüy"
-consonants = "bcçdfghjklmnpqrstvwxz"
-
+# phonetic vowel sounds
 phon_vowels = "()$#289aeEioOuy@"
-
-sure_end_fem = ['e', 'es', 'ent']
+# not a feminine ending, independently of spelling
 phon_non_end_fem = ['#', ')']
 
 class Word:
   @property
   def elidable(self):
-    return self.word[0] in vowels or (self.word[0] == 'h' and
+    """Can this word cause elision in the previous word?"""
+    return is_vowels(self.word[0]) or (self.word[0] == 'h' and
         not haspirater.lookup(self.word))
 
   @property
   def phon_ending(self):
+    """Compute minimal phonetic rhyme"""
     l = []
     w = list(self.phon)
     w.reverse()
@@ -39,14 +39,15 @@ class Word:
       
   @property
   def ending(self):
+    """Compute minimal visual rhyme"""
     l = []
     w = list(self.word)
     count = 0
     w.reverse()
     for x in w:
-      if x in vowels or x in consonants:
+      if is_vowels(x) or is_consonants(x):
         l.append(x)
-      if x in vowels and count >= 1:
+      if is_vowels(x) and count >= 1:
         break
       count += 1
     l.reverse()
@@ -54,11 +55,14 @@ class Word:
 
   @property
   def feminine(self):
+    """Would this word be a feminine rhyme?"""
     for end in sure_end_fem:
       if self.word.endswith(end):
         return True
     if not self.word.endswith('ent'):
       return False
+    # word ends in -ent, it's hard to tell from writing, so look at phon
+    # example: "tient" vs. "lient"
     for end in phon_non_end_fem:
       if self.phon.endswith(end):
         return False
@@ -75,7 +79,6 @@ class Word:
         + str(self.nsyl[1]) + ', "'
         + self.ending + '", "'
         + self.phon_ending + '", '
-        + str(int(self.mult)) + ', '
         + str(int(self.elidable)) + ', '
         + str(int(self.feminine)) + ');')
 
@@ -86,14 +89,13 @@ class Word:
         return True
     return False
 
-  def __init__(self, word, phon, base, kind, freq, nsyl, mult):
+  def __init__(self, word, phon, base, kind, freq, nsyl):
     self.word = word
     self.phon = phon
     self.base = base
     self.kind = kind
     self.freq = freq
     self.nsyl = [nsyl, nsyl]
-    self.mult = mult
     self.redundant = False
     self.do_extends()
 
@@ -131,7 +133,7 @@ def derives(a, b):
 
 print ("""CREATE TABLE words(word varchar(100), phon varchar(100), base
 varchar(100), kind varchar(10), freq float, min_nsyl int, max_nsyl int,
-word_end varchar(10), phon_end varchar(10), multiple bool, elidable
+word_end varchar(10), phon_end varchar(10), elidable
 bool, feminine bool);""")
 
 while True:
@@ -143,21 +145,14 @@ while True:
   phon = l.pop(0)
   base = l.pop(0)
   kind = l.pop(0)
-  freq = ((float(l[0]) + float(l[1]))/2 +
-        100*(float(l[2]) + float(l[3]))/2)
-  l.pop(0)
-  l.pop(0)
-  l.pop(0)
-  l.pop(0)
+  freq = float(l.pop(0))
   #print ("DBG for %s: %d and %d" % (word, int(l[0]), 1+len([x for x in l[1]
     #if x == ' ' or x == '-'])))
   nsyl = max(int(l[0]), 1+len([x for x in l[1] if x == ' ' or x == '-']))
   l.pop(0)
   l.pop(0)
-  mult = ',' in l[0]
-  l.pop(0)
   assert(len(l) == 0)
-  w = Word(word, phon, base, kind, freq, nsyl, mult)
+  w = Word(word, phon, base, kind, freq, nsyl)
   if word not in bases.keys():
     bases[word] = []
   bases[word].append(base)
diff --git a/make_db.sh b/make_db.sh
@@ -2,6 +2,7 @@
 
 cd "$( dirname "$0" )"
 
-cat - additions | cut -f 1,2,3,4,7,8,9,10,24,28,29 |
-  sort -k1,1 |
-  ./make_db.py
+cat - additions | # add custom exceptions
+  cut -f 1,2,3,4,7,8,9,10,24,28 | # select relevant fields
+  awk '{FS="	"; OFS="	"; print $1, $2, $3, $4, ($5+$6)/2 + 100*($7+$8)/2, $9, $10}' #| # aggregate frequencies
+  #./make_db.py