From 8c20295adf7bd83798922a376c278ffd0c40987b Mon Sep 17 00:00:00 2001
From: Ozbolt Menegatti <ozbolt.menegatti@gmail.com>
Date: Thu, 27 Jun 2019 22:04:33 +0200
Subject: [PATCH] Adding dispersions to sqlite, finished moving to it.

---
 src/match.py       |  4 ++--
 src/match_store.py | 20 +++++++++-----------
 2 files changed, 11 insertions(+), 13 deletions(-)

diff --git a/src/match.py b/src/match.py
index 5e241d7..294029d 100644
--- a/src/match.py
+++ b/src/match.py
@@ -2,7 +2,7 @@ from word import Word
 
 class StructureMatch:
     def __init__(self, match_id, structure):
-        self.match_id = match_id
+        self.match_id = str(match_id)
         self.structure = structure
 
         self.matches = []
@@ -10,7 +10,7 @@ class StructureMatch:
     
     @staticmethod
     def from_db(db, colocation_id, structure):
-        result = StructureMatch(str(colocation_id), structure)
+        result = StructureMatch(colocation_id, structure)
         for match_id in db.execute("SELECT mid_match_id FROM ColocationMatches WHERE mid_colocation_id=?", (colocation_id,)):
             to_add = {}
 
diff --git a/src/match_store.py b/src/match_store.py
index e305847..2a712f2 100644
--- a/src/match_store.py
+++ b/src/match_store.py
@@ -1,4 +1,5 @@
 from collections import defaultdict
+from ast import literal_eval
 
 from match import StructureMatch
 from representation_assigner import RepresentationAssigner
@@ -7,14 +8,13 @@ from progress_bar import progress
 class MatchStore:
     def __init__(self, args, db):
         self.db = db
-        self.data = {}
         self.min_frequency = args.min_freq
         self.dispersions = {}
         self.match_num = 0
 
         self.db.init("""CREATE TABLE Colocations (
             colocation_id INTEGER PRIMARY KEY,
-            structure_id INTEGER,
+            structure_id varchar(8),
             key varchar(256))
             """)
         self.db.init("""CREATE TABLE Matches (
@@ -73,24 +73,20 @@ class MatchStore:
         
         self.match_num += 1
 
-        if key not in self.data:
-            self.data[key] = StructureMatch(str(len(self.data) + 1), structure)
-        self.data[key].append(match)
-
     def add_matches(self, matches):
-        for structure, nms in matches.items():
+        for structure, nms in progress(matches.items(), 'adding-matches'):
             for nm in nms:
                 self._add_match(nm[1], structure, nm[0])
 
     def get_matches_for(self, structure):
-        print(structure.id)
         for cid in self.db.execute("SELECT colocation_id FROM Colocations WHERE structure_id=?",
                                    (structure.id,)):
             yield StructureMatch.from_db(self.db, cid[0], structure)
 
     def set_representations(self, word_renderer, structures):
+        structures_dict = {s.id: s for s in structures}
         for cid, sid in progress(self.db.execute("SELECT colocation_id, structure_id FROM Colocations"), "representations"):
-            structure = structures[sid - 1]
+            structure = structures_dict[sid]
             match = StructureMatch.from_db(self.db, cid, structure)
             RepresentationAssigner.set_representations(match, word_renderer)
             for component_id, text in match.representations.items():
@@ -101,7 +97,9 @@ class MatchStore:
 
     def determine_colocation_dispersions(self):
         dispersions = defaultdict(int)
-        for (structure_id, *word_tups) in self.data.keys():
+        for structure_id, word_tups_str in self.db.execute("SELECT structure_id, key FROM Colocations"):
+            word_tups = literal_eval(word_tups_str)
             for component_id, lemma in word_tups:
-                dispersions[(structure_id, component_id, lemma)] += 1
+                dispersions[(str(structure_id), component_id, lemma)] += 1
+            
         self.dispersions = dict(dispersions)