Pinning newer Docling for EasyOCR deprecation and adjusted CI caching

Author: jamesbraza

.github/workflows/tests.yml

@@ -102,23 +102,15 @@ jobs:
           enable-cache: true
           python-version: ${{ matrix.python-version }}
       - run: uv sync
-      - run: uv run pytest -n auto packages --ignore=packages/paper-qa-docling
-        env:
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-          GEMINI_API_KEY: ${{ secrets.GEMINI_API_KEY }}
-          SEMANTIC_SCHOLAR_API_KEY: ${{ secrets.SEMANTIC_SCHOLAR_API_KEY }}
-          CROSSREF_API_KEY: ${{ secrets.CROSSREF_API_KEY }}
-      - name: Cache Docling and EasyOCR models
+      - name: Cache Docling models
         id: cache-models
         uses: actions/cache@v4
         with:
           path: |
             ~/.cache/docling
-            ~/.EasyOCR
-          key: ${{ runner.os }}-docling-easyocr-${{ hashFiles('uv.lock') }}
-          restore-keys: ${{ runner.os }}-docling-easyocr-
-      - run: uv run pytest packages/paper-qa-docling # Don't use `-n auto` to avoid race condition with EasyOCR model downloads
+          key: ${{ runner.os }}-docling-${{ hashFiles('uv.lock') }}
+          restore-keys: ${{ runner.os }}-docling-
+      - run: uv run pytest -n auto packages
         env:
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
           ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}

packages/paper-qa-docling/pyproject.toml

@@ -36,6 +36,7 @@ requires-python = ">=3.11"
 
 [project.optional-dependencies]
 dev = [
+    "docling>=2.56",  # Pin for EasyOCR deprecation
     "fhlmi>=0.39",  # Pin for bytes_to_string
     "paper-qa>=5.23",  # Pin for PDFParserFn
     "pytest-asyncio",