Merge pull request #264 from gerlero/files

Update serialization

Author: gerlero

foamlib/_files/_parsing.py

@@ -253,24 +253,24 @@ def _unpack_binary_field(
     )
 )
 _TOKEN = QuotedString('"', unquote_results=False) | _IDENTIFIER
-_DATA = Forward()
-_KEYWORD = _TOKEN | _list_of(_IDENTIFIER).set_parse_action(
+DATA = Forward()
+KEYWORD = _TOKEN | _list_of(_IDENTIFIER).set_parse_action(
     lambda tks: "(" + " ".join(tks[0]) + ")"
 )
-_KEYWORD_ENTRY = Dict(Group(_keyword_entry_of(_KEYWORD, _DATA)), asdict=True)
+_KEYWORD_ENTRY = Dict(Group(_keyword_entry_of(KEYWORD, DATA)), asdict=True)
 _DATA_ENTRY = Forward()
 _LIST_ENTRY = _KEYWORD_ENTRY | _DATA_ENTRY
 _LIST = _list_of(_LIST_ENTRY)
 _NUMBER = common.signed_integer ^ common.ieee_float
 _DATA_ENTRY <<= _FIELD | _LIST | _DIMENSIONED | _DIMENSIONS | _NUMBER | _SWITCH | _TOKEN
 
-_DATA <<= _DATA_ENTRY[1, ...].set_parse_action(
+DATA <<= _DATA_ENTRY[1, ...].set_parse_action(
     lambda tks: tuple(tks) if len(tks) > 1 else [tks[0]]
 )
 
 _FILE = (
     Dict(
-        Group(_keyword_entry_of(_KEYWORD, Opt(_DATA, default=""), located=True))[...]
+        Group(_keyword_entry_of(KEYWORD, Opt(DATA, default=""), located=True))[...]
         + Opt(
             Group(
                 Located(
@@ -280,7 +280,7 @@ def _unpack_binary_field(
                 )
             )
         )
-        + Group(_keyword_entry_of(_KEYWORD, Opt(_DATA, default=""), located=True))[...]
+        + Group(_keyword_entry_of(KEYWORD, Opt(DATA, default=""), located=True))[...]
     )
     .ignore(cpp_style_comment)
     .ignore(Literal("#include") + ... + LineEnd())  # type: ignore [no-untyped-call]

foamlib/_files/_serialization.py

@@ -1,9 +1,7 @@
 from __future__ import annotations
 
 import array
-import contextlib
 import itertools
-import re
 import sys
 from enum import Enum, auto
 from typing import cast, overload
@@ -13,6 +11,7 @@
 else:
     from typing import Mapping, Sequence
 
+from ._parsing import DATA, KEYWORD
 from ._types import Data, DataEntry, Dimensioned, DimensionSet
 from ._util import is_sequence
 
@@ -34,9 +33,6 @@ class Kind(Enum):
     DIMENSIONS = auto()
 
 
-_TOKENS = re.compile(r'(?:[^\s"]|"(?:[^"])*")+')
-
-
 @overload
 def normalize(data: DataEntry, *, kind: Kind = Kind.DEFAULT) -> DataEntry: ...
 
@@ -69,46 +65,19 @@ def normalize(data: Data, *, kind: Kind = Kind.DEFAULT) -> Data:
 
         return [normalize(d, kind=Kind.SINGLE_ENTRY) for d in data]
 
-    if isinstance(data, str):
-        data = data.strip()
-
-        if data.startswith("(") and data.endswith(")"):
-            data = data[1:-1].split()
-            if kind == Kind.KEYWORD:
-                return "(" + " ".join(data) + ")"
-            return [normalize(d, kind=Kind.SINGLE_ENTRY) for d in data]
-
-        if data.startswith("[") and data.endswith("]"):
-            data = data[1:-1].split()
-            return normalize(data, kind=Kind.DIMENSIONS)
-
-        with contextlib.suppress(ValueError):
-            return int(data)
-
-        with contextlib.suppress(ValueError):
-            return float(data)
-
-        if kind != Kind.KEYWORD:
-            if data in ("yes", "true", "on", "y", "t"):
-                return True
-            if data in ("no", "false", "off", "n", "f"):
-                return False
-
-        tokens: list[str] = re.findall(_TOKENS, data)
-
-        if len(tokens) == 1:
-            return tokens[0]
-
-        if kind == Kind.KEYWORD:
-            return " ".join(tokens)
-
-        return tuple(tokens)
-
     if isinstance(data, Dimensioned):
         value = normalize(data.value, kind=Kind.SINGLE_ENTRY)
         assert isinstance(value, (int, float, list))
         return Dimensioned(value, data.dimensions, data.name)
 
+    if isinstance(data, str):
+        if kind == Kind.KEYWORD:
+            data = KEYWORD.parse_string(data)[0]
+            assert isinstance(data, str)
+            return data
+
+        return cast(DataEntry, DATA.parse_string(data)[0])
+
     if isinstance(
         data,
         (int, float, bool, tuple, DimensionSet),