Release v0.6

.all-contributorsrc

@@ -78,6 +78,15 @@
         "review",
         "ideas"
       ]
+    },
+    {
+      "login": "rettinghaus",
+      "name": "Klaus Rettinghaus",
+      "avatar_url": "https://avatars.githubusercontent.com/u/7693447?v=4",
+      "profile": "https://github.com/rettinghaus",
+      "contributions": [
+        "maintenance"
+      ]
     }
   ],
   "contributorsPerLine": 7,

.github/workflows/check.yml

@@ -15,14 +15,14 @@ jobs:
         working-directory: .
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
       - name: Set up Python 3.12
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: "3.12"
 
       - name: Install uv
-        uses: astral-sh/setup-uv@v5
+        uses: astral-sh/setup-uv@v6
         with:
           enable-cache: true
           cache-dependency-glob: "pyproject.toml"

.github/workflows/python-publish.yml

@@ -21,9 +21,9 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v5
     - name: Set up Python
-      uses: actions/setup-python@v3
+      uses: actions/setup-python@v6
       with:
         python-version: '3.x'
     - name: Install dependencies

.github/workflows/unit_tests.yml

@@ -30,11 +30,11 @@ jobs:
         working-directory: .
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
 
       # use github python action instead of uv to take advantage of caching
       - name: Set up Python ${{ matrix.python }}
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: ${{ matrix.python }}
           cache: 'pip'
@@ -54,7 +54,7 @@ jobs:
         if: ${{ matrix.python == env.COV_PYTHON_VERSION }}
 
       - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@v4
+        uses: codecov/codecov-action@v5
         env:
           CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
         if: ${{ matrix.python == env.COV_PYTHON_VERSION }}

CHANGELOG.md

@@ -1,5 +1,12 @@
 # Change Log
 
+## 0.6
+
+- Experimental omnibus date converter + parser (EDTF, Hebrew, Hijri)
+  - Reorganized Lark grammar files to a common location
+- Improved documentation for converter class parsing and serialization
+- Update action versions in GitHub Workflows (contributed by @rettinghaus)
+
 ## 0.5.2
 - repr for Undate, UndateInterval, and UnDelta now produce fully-qualified, constructor-style strings with deterministic field ordering.
 - Added Undate.unknown_year property and is_unknown() method to check for fully-unknown year/month/day.

CONTRIBUTORS.md

@@ -20,6 +20,7 @@ We use [All Contributors](https://allcontributors.org/) because we recognize tha
       <td align="center" valign="top" width="14.28%"><a href="https://github.com/jdamerow"><img src="https://avatars.githubusercontent.com/u/8881141?v=4?s=100" width="100px;" alt="Julia Damerow"/><br /><sub><b>Julia Damerow</b></sub></a><br /><a href="https://github.com/dh-tech/undate-python/commits?author=jdamerow" title="Code">💻</a> <a href="https://github.com/dh-tech/undate-python/pulls?q=is%3Apr+reviewed-by%3Ajdamerow" title="Reviewed Pull Requests">👀</a> <a href="https://github.com/dh-tech/undate-python/commits?author=jdamerow" title="Tests">⚠️</a> <a href="#eventOrganizing-jdamerow" title="Event Organizing">📋</a> <a href="#ideas-jdamerow" title="Ideas, Planning, & Feedback">🤔</a></td>
       <td align="center" valign="top" width="14.28%"><a href="https://github.com/maltevogl"><img src="https://avatars.githubusercontent.com/u/20907912?v=4?s=100" width="100px;" alt="Malte Vogl"/><br /><sub><b>Malte Vogl</b></sub></a><br /><a href="https://github.com/dh-tech/undate-python/commits?author=maltevogl" title="Code">💻</a> <a href="https://github.com/dh-tech/undate-python/pulls?q=is%3Apr+reviewed-by%3Amaltevogl" title="Reviewed Pull Requests">👀</a> <a href="https://github.com/dh-tech/undate-python/commits?author=maltevogl" title="Tests">⚠️</a> <a href="https://github.com/dh-tech/undate-python/commits?author=maltevogl" title="Documentation">📖</a></td>
       <td align="center" valign="top" width="14.28%"><a href="http://taylorarnold.org"><img src="https://avatars.githubusercontent.com/u/5752184?v=4?s=100" width="100px;" alt="Taylor Arnold"/><br /><sub><b>Taylor Arnold</b></sub></a><br /><a href="https://github.com/dh-tech/undate-python/pulls?q=is%3Apr+reviewed-by%3Ataylor-arnold" title="Reviewed Pull Requests">👀</a> <a href="#ideas-taylor-arnold" title="Ideas, Planning, & Feedback">🤔</a></td>
+      <td align="center" valign="top" width="14.28%"><a href="https://github.com/rettinghaus"><img src="https://avatars.githubusercontent.com/u/7693447?v=4?s=100" width="100px;" alt="Klaus Rettinghaus"/><br /><sub><b>Klaus Rettinghaus</b></sub></a><br /><a href="#maintenance-rettinghaus" title="Maintenance">🚧</a></td>
     </tr>
   </tbody>
   <tfoot>

docs/undate/converters.rst

@@ -4,10 +4,20 @@ Converters
 Overview
 --------
 
+
+.. automodule:: undate.converters
+
+-----
+
 .. automodule:: undate.converters.base
    :members:
    :undoc-members:
 
+
+.. autoclass:: undate.converters.combined.OmnibusDateConverter
+   :members:
+
+
 Formats
 --------
 
@@ -33,6 +43,8 @@ Extended Date-Time Format (EDTF)
 Calendars
 ---------
 
+.. automodule:: undate.converters.calendars
+
 Gregorian
 ^^^^^^^^^
 

src/undate/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.5.2"
+__version__ = "0.6.0"
 
 from undate.date import DatePrecision, UnDelta
 from undate.undate import Undate, Calendar

src/undate/converters/__init__.py

@@ -1 +1,29 @@
-from undate.converters.base import BaseDateConverter as BaseDateConverter
+"""
+Converter classes add support for parsing and serializing dates
+in a variety of formats. A subset of these are calendar converters
+(:mod:`undate.converters.calendar`), which means they support both parsing
+and conversion from an alternate calendar to a common Gregorian
+for comparison across dates.
+
+To parse a date with a supported converter, use the ``Undate`` class method
+:meth:`~undate.undate.Undate.parse` and specify the date as a string
+with the desired format or calendar, e.g.
+
+.. code-block::
+
+  Undate.parse("2001-05", "EDTF")
+  Undate.parse("7 Heshvan 5425", "Hebrew")
+
+For converters that support it, you can also serialize a date in a specified
+format with ``Undate`` class method :meth:`~undate.undate.Undate.format`:
+
+.. code-block::
+
+  Undate.parse("Rabīʿ ath-Thānī 343", "Islamic").format("EDTF")
+
+
+"""
+
+from undate.converters.base import BaseDateConverter, GRAMMAR_FILE_PATH
+
+__all__ = ["BaseDateConverter", "GRAMMAR_FILE_PATH"]

src/undate/converters/base.py

@@ -44,6 +44,7 @@
 
 import importlib
 import logging
+import pathlib
 import pkgutil
 from functools import cache
 from typing import Dict, Type
@@ -53,6 +54,10 @@
 logger = logging.getLogger(__name__)
 
 
+#: Path to parser grammar files
+GRAMMAR_FILE_PATH = pathlib.Path(__file__).parent / "grammars"
+
+
 class BaseDateConverter:
     """Base class for parsing, formatting, and converting dates to handle
     specific formats and different calendars."""

src/undate/converters/calendars/hebrew/parser.py

@@ -1,8 +1,8 @@
-import pathlib
-
 from lark import Lark
 
-grammar_path = pathlib.Path(__file__).parent / "hebrew.lark"
+from undate.converters import GRAMMAR_FILE_PATH
+
+grammar_path = GRAMMAR_FILE_PATH / "hebrew.lark"
 
 with open(grammar_path) as grammar:
     # NOTE: LALR parser is faster but can't be used to ambiguity between years and dates

src/undate/converters/calendars/hebrew/transformer.py

@@ -26,12 +26,13 @@ def hebrew_date(self, items):
 
         # initialize and return an undate with year, month, day and
         # configured calendar (hebrew by default)
+        # NOTE: use self.calendar so Seleucid can extend more easily
         return Undate(**parts, calendar=self.calendar)
 
-    # year translation is not needed since we want a tree with name year
-    # this is equivalent to a no-op
-    # def year(self, items):
-    #     return Tree(data="year", children=[items[0]])
+    def year(self, items):
+        # combine multiple parts into a single string
+        value = "".join([str(i) for i in items])
+        return Tree(data="year", children=[value])
 
     def month(self, items):
         # month has a nested tree for the rule and the value

src/undate/converters/calendars/islamic/parser.py

@@ -1,8 +1,8 @@
-import pathlib
-
 from lark import Lark
 
-grammar_path = pathlib.Path(__file__).parent / "islamic.lark"
+from undate.converters import GRAMMAR_FILE_PATH
+
+grammar_path = GRAMMAR_FILE_PATH / "islamic.lark"
 
 with open(grammar_path) as grammar:
     # NOTE: LALR parser is faster but can't be used due to ambiguity between years and days

src/undate/converters/calendars/islamic/transformer.py

@@ -28,8 +28,17 @@ def islamic_date(self, items):
 
     # year translation is not needed since we want a tree with name year
     # this is equivalent to a no-op
-    # def year(self, items):
-    #     return Tree(data="year", children=[items[0]])
+    def year(self, items):
+        # combine multiple parts into a single string
+        # (for some reason we're getting an anonymous token in combined parser)
+        value = "".join([str(i) for i in items])
+        return Tree(data="year", children=[value])
+
+    def day(self, items):
+        # combine multiple parts into a single string
+        # (for some reason we're getting an anonymous token in combined parser)
+        value = "".join([str(i) for i in items])
+        return Tree(data="day", children=[value])
 
     def month(self, items):
         # month has a nested tree for the rule and the value

src/undate/converters/combined.py

@@ -0,0 +1,85 @@
+"""
+**Experimental** combined parser. Supports EDTF, Hebrew, and Hijri
+where dates are unambiguous. (Year-only dates are parsed as EDTF in
+Gregorian calendar.)
+"""
+
+from typing import Union
+
+from lark import Lark
+from lark.exceptions import UnexpectedCharacters
+from lark.visitors import Transformer, merge_transformers
+
+from undate import Undate, UndateInterval
+from undate.converters import BaseDateConverter, GRAMMAR_FILE_PATH
+from undate.converters.edtf.transformer import EDTFTransformer
+from undate.converters.calendars.hebrew.transformer import HebrewDateTransformer
+from undate.converters.calendars.islamic.transformer import IslamicDateTransformer
+
+
+class CombinedDateTransformer(Transformer):
+    def start(self, children):
+        # trigger the transformer for the appropriate part of the grammar
+        return children
+
+
+# NOTE: currently year-only dates in combined parser are interpreted as
+# EDTF and use Gregorian calendar.
+# In future, we could refine by adding calendar names & abbreviations
+# to the parser in order to recognize years from other calendars.
+
+combined_transformer = merge_transformers(
+    CombinedDateTransformer(),
+    edtf=EDTFTransformer(),
+    hebrew=HebrewDateTransformer(),
+    islamic=IslamicDateTransformer(),
+)
+
+
+# open based on filename so we can specify relative import path based on grammar file
+parser = Lark.open(
+    str(GRAMMAR_FILE_PATH / "combined.lark"), rel_to=__file__, strict=True
+)
+
+
+class OmnibusDateConverter(BaseDateConverter):
+    """
+    Combination parser that aggregates existing parser grammars.
+    Currently supports EDTF, Hebrew, and Hijri  where dates are unambiguous.
+    (Year-only dates are parsed as EDTF in Gregorian calendar.)
+
+    Does not support serialization.
+
+    Example usage::
+
+        Undate.parse("Tammuz 4816", "omnibus")
+
+    """
+
+    #: converter name: omnibus
+    name: str = "omnibus"
+
+    def __init__(self):
+        self.transformer = combined_transformer
+
+    def parse(self, value: str) -> Union[Undate, UndateInterval]:
+        """
+        Parse a string in a supported format and return an :class:`~undate.undate.Undate`
+        or :class:`~undate.undate.UndateInterval`.
+        """
+        if not value:
+            raise ValueError("Parsing empty/unset string is not supported")
+
+        # parse the input string, then transform to undate object
+        try:
+            parsetree = parser.parse(value)
+            # transform returns a list; we want the first item in the list
+            return self.transformer.transform(parsetree)[0]
+        except UnexpectedCharacters:
+            raise ValueError(
+                "Parsing failed: '%s' is not in a recognized date format" % value
+            )
+
+    def to_string(self, undate: Union[Undate, UndateInterval]) -> str:
+        "Not supported by this converter. Will raise :class:`ValueError`"
+        raise ValueError("Omnibus converter does not support serialization")

src/undate/converters/edtf/parser.py

@@ -1,8 +1,8 @@
-import pathlib
-
 from lark import Lark
 
-grammar_path = pathlib.Path(__file__).parent / "edtf.lark"
+from undate.converters import GRAMMAR_FILE_PATH
+
+grammar_path = GRAMMAR_FILE_PATH / "edtf.lark"
 
 with open(grammar_path) as grammar:
     edtf_parser = Lark(grammar.read(), start="edtf")

src/undate/converters/edtf/transformer.py

@@ -66,7 +66,10 @@ def day_unspecified(self, items):
     def date_level1(self, items):
         return self.date(items)
 
-    # year (including negative years) use default transformation
+    def year(self, items):
+        # combine parts (numeric & unknown) into a single string
+        value = "".join(self.get_values(items))
+        return Tree(data="year", children=[value])
 
     def year_fivedigitsplus(self, items):
         # strip off the leading Y and convert to integer

src/undate/converters/grammars/combined.lark

@@ -0,0 +1,32 @@
+%import common.WS
+%ignore WS
+
+start: (edtf__start | hebrew__hebrew_date  | islamic__islamic_date )
+
+// Renaming of the import variables is required, as they receive the namespace of this file.
+// See: https://github.com/lark-parser/lark/pull/973#issuecomment-907287565
+
+// All grammars are in the same file, so we can use relative imports
+
+// relative import from edtf.lark
+%import .edtf.edtf -> edtf__start
+
+// relative import from hebrew.lark
+%import .hebrew.hebrew_date -> hebrew__hebrew_date
+%import .hebrew.day  -> hebrew__day
+%import .hebrew.month -> hebrew__month
+%import .hebrew.year -> hebrew__year
+
+// relative import from islamic.lark
+%import .islamic.islamic_date -> islamic__islamic_date
+%import .islamic.day  -> islamic__day
+%import .islamic.month -> islamic__month
+%import .islamic.year -> islamic__year
+
+
+// override hebrew date to omit year-only, since year without calendar is ambiguous
+// NOTE: potentially support year with calendar label
+%override hebrew__hebrew_date: hebrew__day hebrew__month hebrew__year | hebrew__month hebrew__year 
+
+// same for islamic date, year alone is ambiguous
+%override islamic__islamic_date: islamic__day islamic__month islamic__year | islamic__month islamic__year