From 2c88ebe6972a6f2eeb84463363bf0442d45e533f Mon Sep 17 00:00:00 2001
From: Tyler Goodlet <jgbt@protonmail.com>
Date: Wed, 13 Sep 2023 11:59:51 -0400
Subject: [PATCH] binance: implement `Client.search_symbols()` using
 `rapidfuzz`

Change the deats inside the method and have the `brokerd` search task
just call it as needed since we already do internal mem caching on the
lookup table.

APIs changed so we need to make some tweaks as per:
- https://github.com/maxbachmann/RapidFuzz/blob/main/api_differences.md
- https://github.com/maxbachmann/RapidFuzz/blob/main/api_differences.md#differences-in-processor-functions

The main motivation is to get better wheel pkging support (for nixos),
better impl in C++, and a more simply licensed dep.
---
 piker/brokers/binance/api.py  | 26 ++++++++++++++++++++------
 piker/brokers/binance/feed.py | 19 ++++++++++---------
 2 files changed, 30 insertions(+), 15 deletions(-)

diff --git a/piker/brokers/binance/api.py b/piker/brokers/binance/api.py
index 9f5476a0..f9ecc8fd 100644
--- a/piker/brokers/binance/api.py
+++ b/piker/brokers/binance/api.py
@@ -31,6 +31,8 @@ from pprint import pformat
 from typing import (
     Any,
     Callable,
+    Hashable,
+    Sequence,
     Type,
 )
 import hmac
@@ -549,7 +551,7 @@ class Client:
         if sym:
             return pair_table[sym]
         else:
-            self._pairs
+            return self._pairs
 
     async def get_assets(
         self,
@@ -596,14 +598,26 @@ class Client:
 
         fq_pairs: dict = await self.exch_info()
 
-        matches = fuzzy.extractBests(
-            pattern,
-            fq_pairs,
+        # TODO: cache this list like we were in
+        # `open_symbol_search()`?
+        keys: list[str] = list(fq_pairs)
+
+        matches: list[tuple[
+            Sequence[Hashable],  # matching input key
+            Any,  # scores
+            Any,
+        ]] = fuzzy.extract(
+            query=pattern.upper(),  # since all keys are uppercase
+            choices=keys,
             score_cutoff=50,
         )
         # repack in dict form
-        return {item[0]['symbol']: item[0]
-                for item in matches}
+        matched_pairs: dict[str, Pair] = {}
+        for item in matches:
+            pair_key: str = item[0]
+            matched_pairs[pair_key] = self._pairs[pair_key]
+
+        return matched_pairs
 
     async def bars(
         self,
diff --git a/piker/brokers/binance/feed.py b/piker/brokers/binance/feed.py
index 713f76c6..1416d6a7 100644
--- a/piker/brokers/binance/feed.py
+++ b/piker/brokers/binance/feed.py
@@ -42,7 +42,7 @@ from trio_typing import TaskStatus
 from pendulum import (
     from_timestamp,
 )
-from fuzzywuzzy import process as fuzzy
+from rapidfuzz import process as fuzzy
 import numpy as np
 import tractor
 
@@ -533,14 +533,15 @@ async def open_symbol_search(
 
             pattern: str
             async for pattern in stream:
-                matches = fuzzy.extractBests(
+                # NOTE: pattern fuzzy-matching is done within
+                # the methd impl.
+                pairs: dict[str, Pair] = await client.search_symbols(
                     pattern,
-                    client._pairs,
-                    score_cutoff=50,
                 )
 
-                # repack in dict form
-                await stream.send({
-                    item[0].bs_fqme: item[0]
-                    for item in matches
-                })
+                # repack in fqme-keyed table
+                byfqme: dict[start, Pair] = {}
+                for pair in pairs.values():
+                    byfqme[pair.bs_fqme] = pair
+
+                await stream.send(byfqme)