feat: added pdf meta data endpoint (#31)

* fix: move metrics to appinfo tag

* feat: added pdf meta data endpoint #28 

* feat: added pdf meta data api tests #28

requirements.in

@@ -13,6 +13,7 @@ ghostscript
 # needed by camelot-py
 opencv-python
 PyPDF2
+pikepdf
 pytest
 pytest-cov
 locust

requirements.txt

@@ -46,6 +46,8 @@ coverage[toml]==7.5.3
     # via pytest-cov
 cryptography==42.0.7
     # via pdfminer-six
+deprecated==1.2.14
+    # via pikepdf
 dnspython==2.6.1
     # via email-validator
 email-validator==2.1.1
@@ -103,6 +105,8 @@ jinja2==3.1.4
     #   flask
 locust==2.28.0
     # via -r requirements.in
+lxml==5.2.2
+    # via pikepdf
 markdown-it-py==3.0.0
     # via rich
 markupsafe==2.1.5
@@ -130,6 +134,7 @@ packaging==24.0
     # via
     #   black
     #   gunicorn
+    #   pikepdf
     #   pytesseract
     #   pytest
 pandas==2.2.2
@@ -140,9 +145,12 @@ pdf2image==1.17.0
     # via -r requirements.in
 pdfminer-six==20231228
     # via camelot-py
+pikepdf==9.0.0
+    # via -r requirements.in
 pillow==10.3.0
     # via
     #   pdf2image
+    #   pikepdf
     #   pytesseract
 platformdirs==4.2.2
     # via black
@@ -241,6 +249,8 @@ werkzeug==3.0.3
     #   flask
     #   flask-login
     #   locust
+wrapt==1.16.0
+    # via deprecated
 zope-event==5.0
     # via gevent
 zope-interface==6.4.post2

teal/api.py

@@ -23,8 +23,9 @@
     LibreOfficePdfProfile,
     HealthCheck,
     ValidatePdfProfile,
+    PdfMetaDataReport,
 )
-from teal.pdf import PdfDataExtractor
+from teal.pdf import PdfDataExtractor, PdfMetaDataExtractor
 from teal.pdfa import PdfAValidator, PdfAConverter
 
 app = FastAPI()
@@ -142,6 +143,23 @@ async def extract_table_from_pdf(
         )
 
 
+if is_feature_enabled("TEAL_FEATURE_PDF_META"):
+    logger.info("feature PDF meta data is enabled")
+
+    @app.post(
+        "/pdf/meta",
+        summary="Extract meata data from a PDF",
+        response_model=PdfMetaDataReport,
+        tags=["pdf"],
+    )
+    async def extract_text_from_pdf(
+        file: UploadFile,
+    ) -> Any:
+        logger.debug(f"extract meta data from pdf file='{file.filename}'")
+        pdf = PdfMetaDataExtractor()
+        return pdf.extract_meta_data(data=await file.read(), filename=file.filename)
+
+
 if is_feature_enabled("TEAL_FEATURE_PDFA_CONVERT"):
     logger.info("feature PDF/A convert is enabled")
 
@@ -274,6 +292,6 @@ def custom_openapi():
         excluded_handlers=["/app/*", "/docs/*", "/openapi.json"]
     )
     instrumentator.instrument(app).expose(
-        app, endpoint="/app/metrics", include_in_schema=True
+        app, endpoint="/app/metrics", include_in_schema=True, tags=["appinfo"]
     )
     instrumentator.add(metrics.requests())

teal/model.py

@@ -32,6 +32,16 @@ class PdfAReport(BaseModel):
     details: dict = {}
 
 
+class PdfMetaDataReport(BaseModel):
+    fileName: str
+    fileSize: int
+    pdfVersion: str
+    pdfaClaim: str | None
+    pages: int
+    docInfo: dict = {}
+    xmp: dict = {}
+
+
 class OcrPdfAProfile(str, Enum):
     PDFA_1B = "pdfa-1b"
     PDFA_2B = "pdfa-2b"

teal/pdf.py

@@ -1,9 +1,11 @@
+import io
 import json
 import logging
 import os
 import tempfile
 
 import camelot.io as camelot
+import pikepdf
 import pypdfium2 as pdfium
 import pytesseract
 from pdf2image import convert_from_bytes
@@ -14,7 +16,7 @@
     make_tesseract_lang_param,
     parse_page_ranges,
 )
-from teal.model import TextExtract, TableExtract
+from teal.model import TextExtract, TableExtract, PdfMetaDataReport
 
 _logger = logging.getLogger("teal.pdf")
 
@@ -124,3 +126,42 @@ def extract_table(
                         f.close()
 
         return extracts
+
+
+class PdfMetaDataExtractor:
+    def __init__(self):
+        self.supported_file_extensions = [".pdf"]
+
+    def extract_meta_data(
+        self,
+        data: bytes,
+        filename: str,
+    ) -> PdfMetaDataReport | JSONResponse:
+        file_ext = os.path.splitext(filename)[1]
+        if file_ext not in self.supported_file_extensions:
+            return create_json_err_response(
+                400, f"file extension '{file_ext}' is not supported ({filename})."
+            )
+        meta_data = {}
+        pdf = pikepdf.open(io.BytesIO(data))
+        meta = pdf.open_metadata()
+        for m in meta:
+            meta_data[m] = meta.get(m)
+
+        doc_info = {}
+        for key, value in pdf.docinfo.items():
+            doc_info[key] = str(value)
+
+        return PdfMetaDataReport.model_validate(
+            {
+                "fileName": filename,
+                "fileSize": len(data),
+                "pdfVersion": pdf.pdf_version,
+                "pdfaClaim": (
+                    None if meta.pdfa_status == "" else str(meta.pdfa_status)
+                ),
+                "pages": len(pdf.pages),
+                "docInfo": doc_info,
+                "xmp": meta_data,
+            }
+        )

tests/test_api_libreoffice_convert.py

@@ -19,6 +19,51 @@ def test_libreoffice_convert_docx_default():
                 assert response.status_code == 200
                 assert len(response.json()) == 3
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] is None
+
+
+def test_libreoffice_convert_docx_pdf15():
+    client = TestClient(api.app, raise_server_exceptions=False)
+    with open(get_path("data/doc/word_document.docx"), "rb") as f:
+        response = client.post(
+            url="/libreoffice/convert?profile=pdf-1.5", files={"file": f}
+        )
+        assert response.status_code == 200
+        with tempfile.NamedTemporaryFile(suffix=".pdf") as tmp:
+            tmp.write(response.content)
+            with open(tmp.name, "rb") as pdf_file:
+                response = client.post(url="/pdf/text", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert len(response.json()) == 3
+
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.5"
+                assert response.json()["pdfaClaim"] is None
+
+
+def test_libreoffice_convert_docx_pdf16():
+    client = TestClient(api.app, raise_server_exceptions=False)
+    with open(get_path("data/doc/word_document.docx"), "rb") as f:
+        response = client.post(
+            url="/libreoffice/convert?profile=pdf-1.6", files={"file": f}
+        )
+        assert response.status_code == 200
+        with tempfile.NamedTemporaryFile(suffix=".pdf") as tmp:
+            tmp.write(response.content)
+            with open(tmp.name, "rb") as pdf_file:
+                response = client.post(url="/pdf/text", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert len(response.json()) == 3
+
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] is None
+
 
 def test_libreoffice_convert_txt_default():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -32,6 +77,11 @@ def test_libreoffice_convert_txt_default():
                 assert response.status_code == 200
                 assert len(response.json()) == 1
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] is None
+
 
 def test_libreoffice_convert_pdf_default():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -45,6 +95,11 @@ def test_libreoffice_convert_pdf_default():
                 assert response.status_code == 200
                 assert len(response.json()) == 2
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] is None
+
 
 def test_libreoffice_convert_with_selection():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -87,6 +142,11 @@ def test_libreoffice_convert_pdfa1():
                 assert response.json()["compliant"] is True
                 assert response.json()["profile"] == "PDF/A-1A"
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.4"
+                assert response.json()["pdfaClaim"] == "1A"
+
 
 def test_libreoffice_convert_pdfa2():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -103,6 +163,11 @@ def test_libreoffice_convert_pdfa2():
                 assert response.json()["compliant"] is True
                 assert response.json()["profile"] == "PDF/A-2B"
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] == "2B"
+
 
 def test_libreoffice_convert_pdfa3():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -119,6 +184,11 @@ def test_libreoffice_convert_pdfa3():
                 assert response.json()["compliant"] is True
                 assert response.json()["profile"] == "PDF/A-3B"
 
+                response = client.post(url="/pdf/meta", files={"file": pdf_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] == "3B"
+
 
 def test_libreoffice_convert_wrong_file_type():
     client = TestClient(api.app, raise_server_exceptions=False)

tests/test_api_pdf_meta.py

@@ -0,0 +1,28 @@
+from starlette.testclient import TestClient
+
+from teal import api
+from tests import get_path
+
+
+def test_pdf_extract_meta():
+    client = TestClient(api.app, raise_server_exceptions=False)
+    with open(get_path("data/digital_pdf/document_one_page.pdf"), "rb") as f:
+        response = client.post(url="/pdf/meta", files={"file": f})
+        assert response.status_code == 200
+        assert response.json()["fileName"] == "document_one_page.pdf"
+        assert response.json()["fileSize"] > 100
+        assert response.json()["pdfVersion"] == "1.3"
+        assert response.json()["pages"] == 1
+        assert response.json()["pdfaClaim"] is None
+        assert response.json()["docInfo"] is not None
+        assert response.json()["xmp"] is not None
+
+
+def test_pdf_extract_meta_with_wrong_file_ending():
+    client = TestClient(api.app, raise_server_exceptions=False)
+    with open(get_path("data/doc/word_document.docx"), "rb") as f:
+        response = client.post(url="/pdf/text", files={"file": f})
+        assert response.status_code == 400
+        assert response.json() == {
+            "message": "file extension '.docx' is not supported (word_document.docx)."
+        }

tests/test_api_pdfa_convert.py

@@ -23,6 +23,11 @@ def test_pdfa_convert_digital_pdf():
                 assert response.status_code == 200
                 assert len(response.json()) == 2
 
+                response = client.post(url="/pdf/meta", files={"file": pdfa_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.4"
+                assert response.json()["pdfaClaim"] == "1B"
+
 
 def test_pdfa_convert_scanned_document_with_ocr():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -41,6 +46,11 @@ def test_pdfa_convert_scanned_document_with_ocr():
                 assert response.status_code == 200
                 assert len(response.json()) == 10
 
+                response = client.post(url="/pdf/meta", files={"file": pdfa_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.4"
+                assert response.json()["pdfaClaim"] == "1B"
+
 
 def test_pdfa_convert_digital_pdf_with_lang():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -159,6 +169,11 @@ def test_pdfa_convert_to_pdfa2():
                 assert response.json()["compliant"] is True
                 assert response.json()["profile"] == "PDF/A-2B"
 
+                response = client.post(url="/pdf/meta", files={"file": pdfa_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.7"
+                assert response.json()["pdfaClaim"] == "2B"
+
 
 def test_pdfa_convert_to_pdfa3():
     client = TestClient(api.app, raise_server_exceptions=False)
@@ -173,6 +188,11 @@ def test_pdfa_convert_to_pdfa3():
                 assert response.json()["compliant"] is True
                 assert response.json()["profile"] == "PDF/A-3B"
 
+                response = client.post(url="/pdf/meta", files={"file": pdfa_file})
+                assert response.status_code == 200
+                assert response.json()["pdfVersion"] == "1.6"
+                assert response.json()["pdfaClaim"] == "3B"
+
 
 def test_pdfa_convert_wrong_file_type():
     client = TestClient(api.app, raise_server_exceptions=False)