From 16aeaa50db79e3c17194128a93d22370bf5daa33 Mon Sep 17 00:00:00 2001
From: Arjun Satarkar <me@arjunsatarkar.net>
Date: Tue, 19 Mar 2024 23:07:54 -0400
Subject: markov: avoid float division in uint_to_bytes

It seems as though the float division could cause an issue where the
actual value should be an integer but the float representation is
slightly higher than that, causing the addition of an extra unnecessary
byte. Although this works fine as long as the same function is used,
it could cause problems if seemingly inconsequential details change.

I switch to using divmod and checking for the presence of a remainder.
This should ensure use of the exact minimum required number of bytes in
the blob.
---
 markov/markov.py | 31 +++++++++++++++++--------------
 1 file changed, 17 insertions(+), 14 deletions(-)

(limited to 'markov/markov.py')

diff --git a/markov/markov.py b/markov/markov.py
index e206800..73ac08e 100644
--- a/markov/markov.py
+++ b/markov/markov.py
@@ -151,13 +151,29 @@ class Markov(commands.Cog):
     def uint_to_bytes(self, x: int):
         if x < 0:
             raise ValueError(f"x must be non-negative (got {x})")
-        return x.to_bytes(math.ceil(x.bit_length() / 8), byteorder="big", signed=False)
+        byte_length, remainder = divmod(x.bit_length(), 8)
+        if remainder:
+            byte_length += 1
+        return x.to_bytes(byte_length, byteorder="big", signed=False)
 
     def get_base_channel(self, channel_or_thread):
         if isinstance(channel_or_thread, discord.Thread):
             return channel_or_thread.parent
         return channel_or_thread
 
+    def append_token(self, text, token):
+        # NOTE: if changing PUNCTUATION, also change the regex in process_message() with the corresponding note
+        PUNCTUATION = r".,!?/;()"
+        if token == "/":
+            text = text[:-1] + token
+        elif token == "(":
+            text += token
+        elif token in PUNCTUATION:
+            text = text[:-1] + token + " "
+        else:
+            text += token + " "
+        return text
+
     @commands.group()
     async def markov(self, _ctx):
         """
@@ -385,19 +401,6 @@ class Markov(commands.Cog):
             await db.commit()
         await ctx.reply("All markov data for this guild has been deleted.")
 
-    def append_token(self, text, token):
-        # NOTE: if changing PUNCTUATION, also change the regex in process_message() with the corresponding note
-        PUNCTUATION = r".,!?/;()"
-        if token == "/":
-            text = text[:-1] + token
-        elif token == "(":
-            text += token
-        elif token in PUNCTUATION:
-            text = text[:-1] + token + " "
-        else:
-            text += token + " "
-        return text
-
     @markov.command()
     async def generate(self, ctx, member: discord.Member | None):
         if not await self.config.guild(ctx.guild).use_messages():
-- 
cgit v1.2.3-57-g22cb