From c49d6847f382edbaa1d5dc6ca9ebbb9094b9c0d5 Mon Sep 17 00:00:00 2001
From: "Peter J. Holzer" <hjp@hjp.at>
Date: Tue, 27 Aug 2019 22:38:00 +0200
Subject: [PATCH] Write used evidence to database

---
 judge_message | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

diff --git a/judge_message b/judge_message
index c52dae2..35d6eb3 100755
--- a/judge_message
+++ b/judge_message
@@ -32,9 +32,17 @@ def add_message(msg):
                     msg.get_payload(decode=True)
                         .decode(charset, errors='replace'))
 
-def extract_features():
+def extract_features(msgid):
     db = psycopg2.connect("dbname=bayes")
     csr = db.cursor(cursor_factory=ppe.DictCursor)
+    csr.execute(
+            """
+            insert into messages(id, message_id)
+            values(default, %s)
+            returning id
+            """,
+            (msgid,))
+    msg_pk = csr.fetchone()["id"]
     evidence = []
     for t in components:
         prev = {""}
@@ -82,9 +90,14 @@ def extract_features():
     p2 = 1
     for i in interesting_evidence:
         print("#", i["spam_prob"], i["type"], i["length"], i["feature"], sep="\t")
+        csr.execute(
+            "insert into used_evidence(message, spam_prob, type, length, feature) values(%s, %s, %s, %s, %s)",
+            (msg_pk, i["spam_prob"], i["type"], i["length"], i["feature"]))
         p1 *= i["spam_prob"]
         p2 *= 1 - i["spam_prob"]
     p = p1 / (p1 + p2)
+    csr.execute("update messages set type=%s where id = %s", ("%.6f" % p, msg_pk))
+    db.commit()
     return p
 
 def main():
@@ -98,7 +111,7 @@ def main():
     parser = email.parser.BytesParser(policy=email.policy.default)
     msg = parser.parse(fh)
     add_message(msg)
-    p = extract_features()
+    p = extract_features(msg["Message-Id"])
     print(p, "spam" if p > 0.5 else "ham")
 
 main()