Issue #1000

issue1000/step1.py

@@ -0,0 +1,43 @@
+import sqlite3
+import sys
+
+STRUCTURE_ID = '1'
+
+con = sqlite3.connect(sys.argv[1])
+cur = con.cursor()
+
+data_out = {}
+
+cur.execute("""SELECT Matches.match_id, Matches.word_id, Colocations.colocation_id
+FROM Matches, Colocations, ColocationMatches
+WHERE Matches.match_id = ColocationMatches.mid_match_id 
+AND Colocations.colocation_id = ColocationMatches.mid_colocation_id
+AND Colocations.structure_id = ?""", (STRUCTURE_ID, ))
+
+prev_mid = None
+idx = 0
+
+while True:
+    row = cur.fetchone()
+    if row is None:
+        break
+
+    mid, wid, cid = row
+    if mid == prev_mid:
+        continue
+    elif cid not in data_out:
+        data_out[cid] = []
+
+    wid_int = int(wid[2:9])
+    data_out[cid].append(wid_int) 
+
+    prev_mid = mid
+    idx += 1
+    if(idx % 10000 == 0):
+        print("\r{}".format(idx), end="", flush=True, file=sys.stderr)
+
+print("", file=sys.stderr)
+for mid, wids in data_out.items():
+    print(mid, *wids)
+
+con.close()