Upgrade to Python 3.12 base layer

.github/workflows/production-deploy.yml

@@ -16,7 +16,7 @@ on:
         description: 'Branch to release from'
         required: true
         type: string
-        default: 'main'
+        default: 'master'
       seed:
         description: 'Random seed for reproducible sampling (float in [-1.0, 1.0]). Leave blank to auto-generate.'
         required: false
@@ -34,15 +34,15 @@ on:
         default: '5'
 jobs:
   deploy:
-    uses: lambda-feedback/evaluation-function-workflows/.github/workflows/deploy.yml@deploy-request
+    uses: lambda-feedback/evaluation-function-workflows/.github/workflows/deploy.yml@main
     with:
       template-repository-name: 'lambda-feedback/evaluation-function-boilerplate-python'
       build-file: "app/Dockerfile"
       build-context: "./app"
       environment: "production"
       version-bump: ${{ inputs.version-bump }}
       branch: ${{ inputs.branch }}
-      run-database-tests: false
+      run-database-tests: true
       seed: ${{ inputs.seed }}
       request_delay: ${{ inputs.request_delay }}
       max_concurrency: ${{ inputs.max_concurrency }}

.github/workflows/staging-deploy.yml

@@ -3,7 +3,7 @@ name: Deploy to Staging
 on:
   push:
     branches:
-      - main
+      - master
   workflow_dispatch:
 
 jobs:
@@ -18,7 +18,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.8"]
+        python-version: ["3.12"]
     steps:
       - name: Checkout
         uses: actions/checkout@v4
@@ -59,7 +59,7 @@ jobs:
           if-no-files-found: warn
   deploy:
     needs: test
-    uses: lambda-feedback/evaluation-function-workflows/.github/workflows/deploy.yml@deploy-request
+    uses: lambda-feedback/evaluation-function-workflows/.github/workflows/deploy.yml@main
     with:
       template-repository-name: "lambda-feedback/evaluation-function-boilerplate-python"
       build-file: "app/Dockerfile"

app/Dockerfile

@@ -1,15 +1,21 @@
-# Base image that bundles AWS Lambda Python 3.8 image with some middleware functions
-# FROM base-eval-tmp
-FROM ghcr.io/lambda-feedback/baseevalutionfunctionlayer:main-3.8
+# Base image that bundles the AWS Lambda Python image with the lambda-feedback middleware (app.handler)
+FROM ghcr.io/lambda-feedback/baseevalutionfunctionlayer:main-3.12
+
+# The 3.12 image is Amazon Linux 2023, which doesn't ship `find` (used below).
+# git is needed to pip install latex2sympy from GitHub.
+RUN dnf install -y git findutils && dnf clean all
 
 WORKDIR /app
 
-RUN yum install -y git
+ENV PYTHONPATH=/app
 
 # Copy and install any packages/modules needed for your evaluation script.
 COPY requirements.txt .
 RUN pip3 install -r requirements.txt
 
+# The healthcheck command runs the pytest suite
+RUN pip3 install pytest
+
 # Copy the evaluation and testing scripts
 COPY evaluation.py ./app/
 COPY evaluation_test.py ./app/
@@ -34,5 +40,6 @@ COPY docs/user.md ./app/docs/user.md
 RUN chmod 644 $(find . -type f)
 RUN chmod 755 $(find . -type d)
 
-# The entrypoint for AWS is to invoke the handler function within the app package
-CMD [ "/app/app.handler" ]
+# The entrypoint for AWS is to invoke the handler function within the app package.
+# Python 3.12's Lambda runtime imports this as a module name, so it can't be a path.
+CMD [ "app.handler" ]

app/expression_utilities.py

@@ -13,7 +13,9 @@
 from sympy.parsing.sympy_parser import T as parser_transformations
 from sympy.printing.latex import LatexPrinter
 from sympy import Basic, Symbol
+import io
 import re
+import tokenize
 from typing import Dict, List, TypedDict
 
 class ModifiedLatexPrinter(LatexPrinter):
@@ -573,6 +575,24 @@ def create_sympy_parsing_params(params, unsplittable_symbols=tuple(), symbol_ass
     return parsing_params
 
 
+def reject_invalid_identifiers(expr):
+    '''
+    Raise ValueError if expr contains a name token that is not a valid Python
+    identifier (e.g. `∞`).
+
+    Python <= 3.11 tokenized such characters as ERRORTOKEN, which made
+    parse_expr fail. From Python 3.12 (PEP 701) they are emitted as NAME
+    tokens, and SymPy's auto_symbol silently turns them into symbols.
+    '''
+    try:
+        tokens = list(tokenize.generate_tokens(io.StringIO(expr).readline))
+    except (tokenize.TokenError, SyntaxError):
+        return  # Leave genuine tokenizer errors for parse_expr to report
+    for token in tokens:
+        if token.type == tokenize.NAME and not token.string.isidentifier():
+            raise ValueError(f"Invalid symbol `{token.string}` in expression `{expr}`")
+
+
 def parse_expression(expr, parsing_params):
     '''
     Input:
@@ -610,6 +630,7 @@ def parse_expression(expr, parsing_params):
         transformations = parser_transformations[0:5, 6]+extra_transformations+(split_symbols_custom(can_split),)+parser_transformations[8]
     if parsing_params.get("rationalise", False):
         transformations += parser_transformations[11]
+    reject_invalid_identifiers(expr)
     if parsing_params.get("simplify", False):
         parsed_expr = parse_expr(expr, transformations=transformations, local_dict=symbol_dict)
         parsed_expr = parsed_expr.simplify()

app/requirements.txt

@@ -1,5 +1,5 @@
 typing_extensions
 mpmath==1.2.1
-sympy==1.10.1
+sympy==1.12.1
 antlr4-python3-runtime==4.7.2
 git+https://github.com/lambda-feedback/latex2sympy.git@master#egg=latex2sympy2

pytest.ini

@@ -0,0 +1,4 @@
+[pytest]
+# evaluation_tests.py is the name the base layer's healthcheck imports, so
+# include it alongside pytest's default patterns.
+python_files = test_*.py *_test.py *_tests.py

readme.md

@@ -52,7 +52,7 @@ Once the code passes all these tests, it will then be uploaded to AWS and will b
 ## Pre-requisites
 Although all programming can be done through the GitHub interface, it is recommended you do this locally on your machine. To do this, you must have installed:
 
-- Python 3.8 or higher.
+- Python 3.12 or higher.
 
 - GitHub Desktop or the `git` CLI.