daplapath 2.0.1__tar.gz → 2.0.3__tar.gz

{daplapath-2.0.1 → daplapath-2.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: daplapath
-Version: 2.0.1
+Version: 2.0.3
 Summary: A pathlib.Path class for dapla
 License: MIT
 Author: ort

{daplapath-2.0.1 → daplapath-2.0.3}/daplapath/path.py

@@ -29,9 +29,8 @@ try:
 except ImportError:
     pass
 
-
 # regex with the prefix '_v' followed by an integer of any length
-VERSION_PATTERN = r"_v(\d+)"
+VERSION_PATTERN = r"_v(\d+)\."
 VERSION_PREFIX = "_v"
 
 # regex with the prefix '_p' followed by four length integer (year) and OPTIONALLY month and date, separated by '-'
@@ -51,12 +50,11 @@ class Config:
 class LocalFileSystem:
     """Mimicks GCS's FileSystem but using standard library (os, glob, shutil)."""
 
-    @classmethod
+    @staticmethod
     def glob(
-        cls,
         path: str,
         recursive: bool = True,
-        detail: bool = True,
+        detail: bool = False,
         include_hidden: bool = False,
         **kwargs,
     ) -> list[dict] | list[str]:
@@ -65,7 +63,7 @@ class LocalFileSystem:
         )
 
         if not detail:
-            return relevant_paths
+            return list(relevant_paths)
         with ThreadPoolExecutor() as executor:
             return list(executor.map(get_file_info, relevant_paths))
 
@@ -270,11 +268,16 @@ class Path(str, _PathBase):
         """Returns a PathSeries of all versions of the file."""
         files_in_folder: Iterable[Path] = self.parent.glob("**", recursive=False)
 
+        if self.version_number:
+            start, _, end = re.split(self._version_pattern, self)
+        else:
+            start, end = self.stem, self.suffix
+
         # create boolean mask. With numpy to make it work with both pandas and list
         arr = np.array(files_in_folder)
-        is_version_of_this_file = (
-            np_str_contains(arr, self.versionless_stem)
-        ) & np_str_endswith(arr, self.suffix)
+        is_version_of_this_file = (np_str_contains(arr, start)) & (
+            np_str_endswith(arr, end)
+        )
         if not include_versionless:
             is_version_of_this_file &= np_str_matches(arr, self._version_pattern)
 
@@ -428,6 +431,8 @@ class Path(str, _PathBase):
         """
         if not isinstance(period, (str, int)):
             raise TypeError(f"'period' should be string or int. Got {type(period)}")
+        if str(period) == self.period:
+            return self
         return self.with_periods(period)
 
     def with_periods(self, from_period: str, to_period: str | None = None) -> "Path":
@@ -464,7 +469,9 @@ class Path(str, _PathBase):
         parent = f"{self.parent}/" if self.parent != "." else ""
 
         return self.__class__(
-            f"{parent}{stem}{period_string}{version_string}{self.suffix}"
+            f"{parent}{stem}{period_string}{version_string}{self.suffix}".replace(
+                "".join(self.periods), period_string.strip(self._period_prefix)
+            )
         )
 
     @property
@@ -502,7 +509,7 @@ class Path(str, _PathBase):
     @property
     def versionless_stem(self) -> str:
         """Return the file stem before the version pattern."""
-        return str(re.sub(self._version_pattern, "", self._path.stem))
+        return self.__class__(re.split(self._version_pattern, self._path.name)[0]).stem
 
     @property
     def parent(self) -> "Path":
@@ -553,6 +560,12 @@ class Path(str, _PathBase):
             except KeyError:
                 return read_nrows(file, 1).columns
 
+    @property
+    def schema(self) -> pyarrow.Schema:
+        """Date types of the file's columns."""
+        with self.open("rb") as file:
+            return pq.read_schema(file)
+
     @property
     def dtypes(self) -> pd.Series:
         """Date types of the file's columns."""
@@ -628,6 +641,10 @@ class Path(str, _PathBase):
         """File size in terrabytes."""
         return self.kb / 1_000_000_000
 
+    @property
+    def partition_root(self) -> "Path":
+        return self.split(".parquet")[0] + ".parquet"
+
     def is_dir(self) -> bool:
         try:
             return self.file_system.isdir(self)
@@ -842,6 +859,10 @@ class PathSeries(pd.Series, _PathBase):
         """Select only the files in the Series."""
         return self[self.is_file()]
 
+    @property
+    def partition_root(self) -> "PathSeries":
+        return self.files.apply(lambda x: x.partition_root).drop_duplicates()
+
     @property
     def dirs(self) -> "PathSeries":
         """Select only the directories in the Series."""
@@ -1469,6 +1490,10 @@ def as_str(obj) -> str:
     raise TypeError(type(obj))
 
 
+def paths_are_equal(path1: Path | str, path2: Path | str) -> bool:
+    return Path(path1).parts == Path(path2).parts
+
+
 def sort_by_period(paths: Iterable[str]) -> Iterable[str]:
     try:
         periods = [pd.Timestamp(path.period) for path in paths]

{daplapath-2.0.1 → daplapath-2.0.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "daplapath"
-version = "2.0.1"
+version = "2.0.3"
 description = "A pathlib.Path class for dapla"
 authors = ["ort <ort@ssb.no>"]
 license = "MIT"