From c21954ebba4a0da953511c80066861f7a9a5e8f9 Mon Sep 17 00:00:00 2001
From: James Turk <dev@jamesturk.net>
Date: Thu, 13 Jul 2023 16:54:13 -0500
Subject: [PATCH] avoid duplicate processing

---
 examples/fruits.py     | 10 +++++++---
 src/beakers/beakers.py |  3 +++
 src/beakers/recipe.py  |  7 ++++++-
 3 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/examples/fruits.py b/examples/fruits.py
index eff576a..c9615ed 100644
--- a/examples/fruits.py
+++ b/examples/fruits.py
@@ -43,15 +43,19 @@ recipe.add_conditional(
     lambda cw: cw.is_fruit,
     "fruits",
 )
-recipe.add_transform("fruits", "sentence", lambda x: f"I love a fresh {x['word']}")
+recipe.add_transform(
+    "fruits",
+    "sentence",
+    lambda x: Sentence(sentence=f"I love a fresh {x.normalized_word}".split()),
+)
 
 recipe.add_seed(
     "word",
     [
         Word(word="apple"),
-        Word(word="banana"),
+        Word(word="bAnAnA"),
         Word(word="hammer"),
         Word(word="orange"),
-        Word(word="egg"),
+        Word(word="EGG"),
     ],
 )
diff --git a/src/beakers/beakers.py b/src/beakers/beakers.py
index 07ae9e2..8908362 100644
--- a/src/beakers/beakers.py
+++ b/src/beakers/beakers.py
@@ -40,6 +40,9 @@ class Beaker(abc.ABC):
         for item in items:
             self.add_item(item)
 
+    def id_set(self) -> set[str]:
+        return set(id for id, _ in self.items())
+
 
 class TempBeaker(Beaker):
     def __init__(self, name: str, model: PydanticModel, recipe: "Recipe"):
diff --git a/src/beakers/recipe.py b/src/beakers/recipe.py
index cef4232..9ff1c1b 100644
--- a/src/beakers/recipe.py
+++ b/src/beakers/recipe.py
@@ -272,11 +272,14 @@ class Recipe:
 
                 from_beaker = self.beakers[from_b]
                 to_beaker = self.beakers[to_b]
+                already_processed = from_beaker.id_set() & to_beaker.id_set()
+
                 log.info(
                     "transform",
                     from_b=from_b,
                     to_b=to_b,
-                    items=len(from_beaker),
+                    to_process=len(from_beaker) - len(already_processed),
+                    already_processed=len(already_processed),
                     transform=edge["transform"].name,
                 )
 
@@ -289,6 +292,8 @@ class Recipe:
                     t_func = transform.transform_func
 
                 for id, item in from_beaker.items():
+                    if id in already_processed:
+                        continue
                     try:
                         transformed = t_func(item)
                         if transformed: