Merge branch 'release/0.4.0'

.github/workflows/ci.yml

@@ -151,8 +151,8 @@ jobs:
       -
         name: Display release info
         run: |
-          echo "$Version: ${{ steps.query-release-info.outputs.version }}"
-          echo "$Date: ${{ steps.query-release-info.outputs.release-date }}"
+          echo "Version: ${{ steps.query-release-info.outputs.version }}"
+          echo "Date: ${{ steps.query-release-info.outputs.release-date }}"
           echo "${{ steps.query-release-info.outputs.release-notes }}"
       -
         uses: ncipollo/release-action@v1
@@ -177,4 +177,4 @@ jobs:
           path: dist
       -
         name: Publish build to PyPI
-        uses: pypa/[email protected].7
+        uses: pypa/[email protected].8

.pre-commit-config.yaml

@@ -37,7 +37,7 @@ repos:
         exclude: "(^Pipfile\\.lock$)"
   # Python hooks
   - repo: https://github.com/charliermarsh/ruff-pre-commit
-    rev: 'v0.0.278'
+    rev: 'v0.0.280'
     hooks:
       - id: ruff
   - repo: https://github.com/psf/black

CHANGELOG.md

@@ -5,6 +5,18 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.4.0] - 2023-07-27
+
+### Added
+
+- More extensive testing of date and time strings in various formats, including
+  [RFC-3339](https://www.ietf.org/rfc/rfc3339.txt), ISO-8061 and things in between
+
+### Changed
+
+- Date parsing is now does not assume a timezone if none is provided (the parsed datetime will be naive)
+- `pypa/gh-action-pypi-publish` updated to v1.8.8
+
 ## [0.3.0] - 2023-07-19
 
 ### Added

pyproject.toml

@@ -168,7 +168,8 @@ ban-relative-imports = "all"
                 "S101",
                 "TID252",
                 # Allow more complex pytest.raises
-                "PT012"
+                "PT012",
+                "DTZ001"
                 ]
 
 [tool.coverage.run]

src/tika_client/__about__.py

@@ -1 +1 @@
-__version__ = "0.3.0"
+__version__ = "0.4.0"

src/tika_client/client.py

@@ -1,5 +1,8 @@
 import logging
+from types import TracebackType
 from typing import Dict
+from typing import Optional
+from typing import Type
 
 from httpx import Client
 
@@ -35,7 +38,7 @@ def __init__(
         self.tika = Tika(self._client, compress=compress)
         self.rmeta = Recursive(self._client, compress=compress)
 
-    def add_headers(self, header: Dict[str, str]):  # pragma: no cover
+    def add_headers(self, header: Dict[str, str]) -> None:  # pragma: no cover
         """
         Updates the httpx Client headers with the given values
         """
@@ -44,5 +47,10 @@ def add_headers(self, header: Dict[str, str]):  # pragma: no cover
     def __enter__(self) -> "TikaClient":
         return self
 
-    def __exit__(self, exc_type, exc_val, exc_tb) -> None:
+    def __exit__(
+        self,
+        exc_type: Optional[Type[BaseException]],
+        exc_val: Optional[BaseException],
+        exc_tb: Optional[TracebackType],
+    ) -> None:
         self._client.close()

src/tika_client/data_models.py

@@ -2,6 +2,7 @@
 import re
 from datetime import datetime
 from datetime import timedelta
+from datetime import timezone
 from enum import Enum
 from typing import Dict
 from typing import List
@@ -11,7 +12,17 @@
 # Based on https://cwiki.apache.org/confluence/display/TIKA/Metadata+Overview
 
 logger = logging.getLogger("tika-client.data")
-_FRACTION_REGEX = re.compile("(.*)([\\.,][0-9]+)(.*)")
+_TIME_RE = re.compile(
+    r"(?P<year>\d{4})-"
+    r"(?P<month>\d{2})-"
+    r"(?P<day>\d{2})"
+    r"[ tT]"
+    r"(?P<hour>\d{2}):"
+    r"(?P<minute>\d{2}):"
+    r"(?P<second>\d{2})"
+    r"(?P<fractional_seconds>\.\d+)?"
+    r"(?P<timezone>[zZ]|[+-]\d{2}:\d{2})?",
+)
 
 
 class TikaKey(str, Enum):
@@ -105,31 +116,34 @@ def get_optional_datetime(self, key: Union[TikaKey, DublinCoreKey, XmpKey, str])
 
         date_str: str = self.data[key]
 
-        # Handle fractional seconds
-        frac = _FRACTION_REGEX.match(date_str)
-        if frac is not None:
-            logger.info("Located fractional seconds")
-            delta = timedelta(seconds=float(frac.group(2)))
-            date_str = frac.group(1)
-            # Attempt to include the timezone info still
-            if frac.group(3) is not None:
-                date_str += frac.group(3)
-        else:
-            delta = timedelta()
-
-        # Handle Zulu time as UTC
-        if "Z" in date_str:
-            date_str = date_str.replace("Z", "+00:00")
-
-        # Assume UTC if it is not set
-        if "+" not in date_str:
-            date_str += "+00:00"
-
-        try:
-            return datetime.fromisoformat(date_str) + delta
-        except ValueError as e:
-            logger.error(f"{e} during datetime parsing")
-        return None
+        m = _TIME_RE.match(date_str)
+        if not m:
+            return None
+
+        (year, month, day, hour, minute, second, frac_sec, timezone_str) = m.groups()
+
+        microseconds = int(float(frac_sec) * 1000000.0) if frac_sec is not None else 0
+        tzinfo = None
+        if timezone_str is not None:
+            if timezone_str.lower() == "z":
+                tzinfo = timezone.utc
+            else:
+                multi = -1 if timezone_str[0:1] == "-" else 1
+                hours = int(timezone_str[1:3])
+                minutes = int(timezone_str[4:])
+                delta = timedelta(hours=hours, minutes=minutes) * multi
+                tzinfo = timezone(delta)
+
+        return datetime(
+            year=int(year),
+            month=int(month),
+            day=int(day),
+            hour=int(hour),
+            minute=int(minute),
+            second=int(second),
+            microsecond=microseconds,
+            tzinfo=tzinfo,
+        )
 
     def get_optional_string(self, key: Union[TikaKey, DublinCoreKey, XmpKey, str]) -> Optional[str]:
         if key not in self.data:

tests/test_datetime_formats.py

@@ -0,0 +1,131 @@
+from datetime import datetime
+from datetime import timedelta
+from datetime import timezone
+
+import magic
+from pytest_httpx import HTTPXMock
+
+from tests.conftest import SAMPLE_DIR
+from tika_client.client import TikaClient
+from tika_client.data_models import DublinCoreKey
+from tika_client.data_models import TikaKey
+
+
+class TestDateTimeFormat:
+    def test_parse_offset_date_format_utc(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time with a UTC timezone in the +xx:yy format
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: "2023-05-17T16:30:44+00:00"},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created == datetime(
+            year=2023,
+            month=5,
+            day=17,
+            hour=16,
+            minute=30,
+            second=44,
+            tzinfo=timezone.utc,
+        )
+
+    def test_parse_offset_date_format_zulu(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time with a UTC timezone in the Z format
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: "2023-01-17T16:35:44Z"},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created == datetime(
+            year=2023,
+            month=1,
+            day=17,
+            hour=16,
+            minute=35,
+            second=44,
+            tzinfo=timezone.utc,
+        )
+
+    def test_parse_offset_date_format_positive(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time with a timezone in the +xx:yy format offset from UTC
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: "2023-06-17T16:30:44+08:00"},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created == datetime(
+            year=2023,
+            month=6,
+            day=17,
+            hour=16,
+            minute=30,
+            second=44,
+            tzinfo=timezone(timedelta(hours=8)),
+        )
+
+    def test_parse_offset_date_format_negative(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time with a timezone in the -xx:yy format offset from UTC
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: "2023-06-17T16:30:44-08:00"},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created == datetime(
+            year=2023,
+            month=6,
+            day=17,
+            hour=16,
+            minute=30,
+            second=44,
+            tzinfo=timezone(timedelta(hours=-8)),
+        )
+
+    def test_parse_offset_date_format_python_isoformat(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time with a timezone in the ISO 8061 format (as done by Python)
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        expected = datetime.now(tz=timezone.utc)
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: expected.isoformat()},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created == expected
+
+    def test_parse_offset_date_no_match(self, tika_client: TikaClient, httpx_mock: HTTPXMock):
+        """
+        Test the datetime parsing properly handles a time string which doesn't match the correct formats
+        """
+        test_file = SAMPLE_DIR / "sample-libre-office.odt"
+
+        httpx_mock.add_response(
+            json={TikaKey.ContentType: "test", TikaKey.Parsers: [], DublinCoreKey.Created: "202-06-17T16:30:44-0"},
+        )
+
+        resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
+
+        assert resp.created is None

tests/test_file_formats.py

@@ -1,5 +1,4 @@
 from datetime import datetime
-from datetime import timezone
 
 import magic
 
@@ -9,6 +8,9 @@
 
 class TestLibreOfficeFormats:
     def test_parse_libre_office_writer_document(self, tika_client: TikaClient):
+        """
+        Test handling of a ODT document produced by LibreOffice
+        """
         test_file = SAMPLE_DIR / "sample-libre-office.odt"
         resp = tika_client.tika.as_html.from_file(test_file, magic.from_file(str(test_file), mime=True))
 
@@ -27,5 +29,5 @@ def test_parse_libre_office_writer_document(self, tika_client: TikaClient):
             minute=30,
             second=44,
             microsecond=719000,
-            tzinfo=timezone.utc,
+            tzinfo=None,
         )

tests/test_image_files.py

@@ -6,12 +6,18 @@
 
 class TestParseImageMetadata:
     def test_image_jpeg(self, tika_client: TikaClient):
+        """
+        Test the handling of a JPEG file metadata retrieval
+        """
         test_file = SAMPLE_DIR / "sample.jpg"
         resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))
 
         assert resp.type == "image/jpeg"
 
     def test_image_png(self, tika_client: TikaClient):
+        """
+        Test the handling of a PNG file metadata retrieval
+        """
         test_file = SAMPLE_DIR / "sample.png"
         resp = tika_client.metadata.from_file(test_file, magic.from_file(str(test_file), mime=True))