chore: Reformatted the order-processing sample by ensuring black, isort, and flake8 pass

python/agents/personalized-shopping/deployment/deploy.py

@@ -12,13 +12,13 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import vertexai
-from vertexai.preview.reasoning_engines import AdkApp
-from vertexai import agent_engines
-from dotenv import load_dotenv
 import os
 
+import vertexai
+from dotenv import load_dotenv
 from personalized_shopping.agent import root_agent
+from vertexai import agent_engines
+from vertexai.preview.reasoning_engines import AdkApp
 
 load_dotenv()
 

python/agents/personalized-shopping/personalized_shopping/__init__.py

@@ -21,17 +21,17 @@
 os.environ["GOOGLE_CLOUD_LOCATION"] = "global"
 os.environ.setdefault("GOOGLE_GENAI_USE_VERTEXAI", "True")
 
-import torch
+import torch  # noqa: E402
 
 # Workaround to Resolve the PyTorch-Streamlit Incompatibility Issue
 torch.classes.__path__ = []
 
 # Initialize webshop environment (requires Java)
 # If Java is not available (e.g., in CI), set webshop_env to None
 try:
-    from .shared_libraries.init_env import init_env, webshop_env
+    from .shared_libraries.init_env import init_env, webshop_env  # noqa: E402
 except Exception:
     webshop_env = None
     init_env = None
 
-from . import agent
+from . import agent  # noqa: F401, E402

python/agents/personalized-shopping/personalized_shopping/agent.py

@@ -15,10 +15,9 @@
 from google.adk.agents import Agent
 from google.adk.tools import FunctionTool
 
-from .tools.search import search
-from .tools.click import click
-
 from .prompt import personalized_shopping_agent_instruction
+from .tools.click import click
+from .tools.search import search
 
 root_agent = Agent(
     model="gemini-2.5-flash",

python/agents/personalized-shopping/personalized_shopping/shared_libraries/init_env.py

@@ -41,5 +41,7 @@ def get_webshop_env():
     if _webshop_env is None:
         _webshop_env = init_env(num_product_items)
         _webshop_env.reset()
-        print(f"Finished initializing WebshopEnv with {num_product_items} items.")
+        print(
+            f"Finished initializing WebshopEnv with {num_product_items} items."
+        )
     return _webshop_env

python/agents/personalized-shopping/personalized_shopping/shared_libraries/search_engine/convert_product_file_format.py

@@ -14,11 +14,12 @@
 
 import json
 import sys
+
 from tqdm import tqdm
 
 sys.path.insert(0, "../")
 
-from web_agent_site.engine.engine import load_products
+from web_agent_site.engine.engine import load_products  # noqa: E402
 
 all_products, *_ = load_products(filepath="../data/items_shuffle.json")
 

python/agents/personalized-shopping/personalized_shopping/shared_libraries/web_agent_site/__init__.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from .envs.web_agent_text_env import WebAgentTextEnv
+from .envs.web_agent_text_env import WebAgentTextEnv  # noqa: F401

python/agents/personalized-shopping/personalized_shopping/shared_libraries/web_agent_site/engine/engine.py

@@ -14,13 +14,13 @@
 
 """ """
 
-from ast import literal_eval
-from collections import defaultdict
-from decimal import Decimal
 import json
 import os
 import random
 import re
+from ast import literal_eval
+from collections import defaultdict
+from decimal import Decimal
 
 from flask import render_template_string
 from pyserini.search.lucene import LuceneSearcher
@@ -179,7 +179,9 @@ def get_top_n_product_from_keywords(
         docs = [search_engine.doc(hit.docid) for hit in hits]
         top_n_asins = [json.loads(doc.raw())["id"] for doc in docs]
         top_n_products = [
-            product_item_dict[asin] for asin in top_n_asins if asin in product_item_dict
+            product_item_dict[asin]
+            for asin in top_n_asins
+            if asin in product_item_dict
         ]
     return top_n_products
 
@@ -334,7 +336,10 @@ def load_products(filepath, num_products=None, human_goals=True):
                 option_values = []
                 for option_content in option_contents:
                     option_value = (
-                        option_content["value"].strip().replace("/", " | ").lower()
+                        option_content["value"]
+                        .strip()
+                        .replace("/", " | ")
+                        .lower()
                     )
                     option_image = option_content.get("image", None)
 
@@ -364,7 +369,9 @@ def load_products(filepath, num_products=None, human_goals=True):
             if asin in human_attributes:
                 products[i]["instructions"] = human_attributes[asin]
         else:
-            products[i]["instruction_text"] = attributes[asin].get("instruction", None)
+            products[i]["instruction_text"] = attributes[asin].get(
+                "instruction", None
+            )
 
             products[i]["instruction_attributes"] = attributes[asin].get(
                 "instruction_attributes", None

python/agents/personalized-shopping/personalized_shopping/shared_libraries/web_agent_site/engine/goal.py

@@ -14,12 +14,14 @@
 
 """Functions for specifying goals and reward calculations."""
 
-from collections import defaultdict
 import itertools
 import random
-from rich import print
+from collections import defaultdict
+
 import spacy
+from rich import print
 from thefuzz import fuzz
+
 from .normalize import normalize_color
 
 nlp = spacy.load("en_core_web_sm")
@@ -53,7 +55,9 @@ def get_human_goals(all_products, product_prices):
                 price_range = [p for p in PRICE_RANGE if p > price][:4]
                 if len(price_range) >= 2:
                     _, price_upper = sorted(random.sample(price_range, 2))
-                    price_text = f", and price lower than {price_upper:.2f} dollars"
+                    price_text = (
+                        f", and price lower than {price_upper:.2f} dollars"
+                    )
                 else:
                     price_upper = 1000000
                     price_text = ""
@@ -67,7 +71,8 @@ def get_human_goals(all_products, product_prices):
                     "query": item["query"],
                     "name": item["name"],
                     "product_category": item["product_category"],
-                    "instruction_text": product["instruction"].strip(".") + price_text,
+                    "instruction_text": product["instruction"].strip(".")
+                    + price_text,
                     "attributes": attributes,
                     "price_upper": price_upper,
                     "goal_options": product["instruction_options"],
@@ -86,7 +91,10 @@ def get_synthetic_goals(all_products, product_prices):
     goals = []
     cnt_atts = defaultdict(int)
     for product in all_products:
-        if "instruction_text" not in product or product["instruction_text"] is None:
+        if (
+            "instruction_text" not in product
+            or product["instruction_text"] is None
+        ):
             continue
         product_goals = []
         asin = product["asin"]
@@ -111,14 +119,18 @@ def get_synthetic_goals(all_products, product_prices):
         options = product["options"]
         option_names = sorted(options)
         combinations = list(
-            itertools.product(*(options[option_name] for option_name in option_names))
+            itertools.product(
+                *(options[option_name] for option_name in option_names)
+            )
         )
         for combination in combinations:
             goal_options = dict()
             for i, o in enumerate(combination):
                 #                option_text.append(f'{option_names[i]}: {o}')
                 goal_options[option_names[i]] = o
-            option_text = ", and ".join([f"{k}: {v}" for k, v in goal_options.items()])
+            option_text = ", and ".join(
+                [f"{k}: {v}" for k, v in goal_options.items()]
+            )
             option_text = " with " + option_text if option_text else ""
             product_goals.append(
                 {
@@ -138,9 +150,9 @@ def get_synthetic_goals(all_products, product_prices):
                 cnt_atts[att] += 1
         goals += product_goals
     for goal in goals:
-        goal["weight"] = sum(1.0 / cnt_atts[att] for att in goal["attributes"]) / len(
-            goal["attributes"]
-        )
+        goal["weight"] = sum(
+            1.0 / cnt_atts[att] for att in goal["attributes"]
+        ) / len(goal["attributes"])
     return goals
 
 
@@ -152,7 +164,9 @@ def get_type_reward(purchased_product, goal):
     purchased_product_category = [
         x.strip() for x in purchased_product["product_category"].split("›")
     ]
-    goal_product_category = [x.strip() for x in goal["product_category"].split("›")]
+    goal_product_category = [
+        x.strip() for x in goal["product_category"].split("›")
+    ]
     category_match = (
         len(set(purchased_product_category) & set(goal_product_category)) >= 2
     )
@@ -245,15 +259,21 @@ def get_option_reward(purchased_options, goal_options):
                 break
 
     # Calculate option reward as fraction of goal options hit
-    r_option = num_option_matches / len(goal_options) if len(goal_options) > 0 else None
+    r_option = (
+        num_option_matches / len(goal_options)
+        if len(goal_options) > 0
+        else None
+    )
     return r_option, num_option_matches
 
 
 def get_reward(purchased_product, goal, price, options, **kwargs):
     """Get cumulative reward score for purchased product and goal"""
     r_type_dict = get_type_reward(purchased_product, goal)
 
-    r_price = (price <= goal["price_upper"]) if goal["price_upper"] > 0 else None
+    r_price = (
+        (price <= goal["price_upper"]) if goal["price_upper"] > 0 else None
+    )
 
     r_att, num_attr_matches = get_attribute_reward(purchased_product, goal)
 

python/agents/personalized-shopping/personalized_shopping/shared_libraries/web_agent_site/envs/web_agent_text_env.py

@@ -12,18 +12,20 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from collections import defaultdict
 import json
 import random
 import string
 import time
+from collections import defaultdict
+
+import gym
+import numpy as np
+import torch
 from bs4 import BeautifulSoup
 from bs4.element import Comment
 from flask import Flask
-import gym
 from gym.envs.registration import register
-import numpy as np
-import torch
+
 from ..engine.engine import (
     ACTION_TO_TEMPLATE,
     BACK_TO_SEARCH,
@@ -45,7 +47,6 @@
     random_idx,
 )
 
-
 app = Flask(__name__)
 
 
@@ -124,7 +125,11 @@ def step(self, action):
         action_name, action_arg = parse_action(action)
         if action_arg is not None:
             action_arg = action_arg.lower()
-        if action_name == "search" and action_arg is not None and action_arg != "":
+        if (
+            action_name == "search"
+            and action_arg is not None
+            and action_arg != ""
+        ):
             status = self.browser.search(action_arg)
         elif (
             action_name == "click"
@@ -214,7 +219,9 @@ def observation(self):
         elif self.observation_mode == "url":
             return self.state["url"]
         else:
-            raise ValueError(f"Observation mode {self.observation_mode} not supported.")
+            raise ValueError(
+                f"Observation mode {self.observation_mode} not supported."
+            )
 
     @property
     def state(self):
@@ -246,13 +253,20 @@ def convert_html_to_text(self, html, simple=False):
                     processed_t = f"[button] {t} [button_]"
                 elif t.parent.name == "label":  # options
                     if f'"{t}"' in self.state["url"]:
-                        processed_t = f"  [clicked button] {t} [clicked button_]"
+                        processed_t = (
+                            f"  [clicked button] {t} [clicked button_]"
+                        )
                         observation = f"You have clicked {t}.\n" + observation
                     else:
                         processed_t = f"  [button] {t} [button_]"
                 elif t.parent.get("class") == ["product-link"]:  # product asins
-                    if f"{t}" in self.server.user_sessions[self.session]["asins"]:
-                        processed_t = f"\n[clicked button] {t} [clicked button_]"
+                    if (
+                        f"{t}"
+                        in self.server.user_sessions[self.session]["asins"]
+                    ):
+                        processed_t = (
+                            f"\n[clicked button] {t} [clicked button_]"
+                        )
                     else:
                         processed_t = f"\n[button] {t} [button_]"
                 else:  # regular, unclickable text
@@ -273,7 +287,9 @@ def reset(self, session=None, instruction_text=None):
             self.session = self.session_prefix + self.session
 
         init_url = f"{self.base_url}/{self.session}"
-        self.browser.get(init_url, session_id=self.session, session_int=session_int)
+        self.browser.get(
+            init_url, session_id=self.session, session_int=session_int
+        )
 
         self.text_to_clickable = None
         self.instruction_text = (
@@ -295,7 +311,9 @@ def close(self):
 
 def tag_visible(element):
     ignore = {"style", "script", "head", "title", "meta", "[document]"}
-    return element.parent.name not in ignore and not isinstance(element, Comment)
+    return element.parent.name not in ignore and not isinstance(
+        element, Comment
+    )
 
 
 class SimServer:
@@ -332,7 +350,9 @@ def __init__(
             )
         )
         self.search_engine = init_search_engine(num_products=num_products)
-        self.goals = get_goals(self.all_products, self.product_prices, human_goals)
+        self.goals = get_goals(
+            self.all_products, self.product_prices, human_goals
+        )
         self.show_attrs = show_attrs
 
         # Fix outcome for random shuffling of goals
@@ -342,7 +362,9 @@ def __init__(
         # Apply `filter_goals` parameter if exists to select speific goal(s)
         if filter_goals is not None:
             self.goals = [
-                goal for (i, goal) in enumerate(self.goals) if filter_goals(i, goal)
+                goal
+                for (i, goal) in enumerate(self.goals)
+                if filter_goals(i, goal)
             ]
 
         # Imposes `limit` on goals via random selection
@@ -561,7 +583,9 @@ def receive(self, session_id, current_url, session_int=None, **kwargs):
             if session_id not in self.user_sessions:
                 idx = (
                     session_int
-                    if (session_int is not None and isinstance(session_int, int))
+                    if (
+                        session_int is not None and isinstance(session_int, int)
+                    )
                     else random_idx(self.cum_weights)
                 )
                 goal = self.goals[idx]
@@ -679,7 +703,9 @@ def __init__(self, server):
 
     def get(self, url, session_id=None, session_int=None):
         """Set browser variables to corresponding link, page HTML for URL"""
-        self.session_id = url.split("/")[-1] if session_id is None else session_id
+        self.session_id = (
+            url.split("/")[-1] if session_id is None else session_id
+        )
         self.page_source, _, _ = self.server.receive(
             self.session_id, self.current_url, session_int=session_int
         )