omlish 0.0.0.dev253__tar.gz → 0.0.0.dev255__tar.gz

{omlish-0.0.0.dev253/omlish.egg-info → omlish-0.0.0.dev255}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: omlish
-Version: 0.0.0.dev253
+Version: 0.0.0.dev255
 Summary: omlish
 Author: wrmsr
 License: BSD-3-Clause

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/__about__.py

@@ -1,5 +1,5 @@
-__version__ = '0.0.0.dev253'
-__revision__ = '87d5615719183def315c54d7b281f939c270c396'
+__version__ = '0.0.0.dev255'
+__revision__ = '68bcce87dfef9de1ab636cae14e4cbf5d71889cd'
 
 
 #

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/dispatch/impls.py

@@ -5,7 +5,6 @@ TODO:
  - multidispatch? never solved..
   - just generic on tuple[A0, A1, ...]
 """
-import contextlib
 import typing as ta
 import weakref
 
@@ -24,8 +23,10 @@ _IMPL_FUNC_CLS_SET_CACHE: ta.MutableMapping[ta.Callable, frozenset[type]] = weak
 
 
 def get_impl_func_cls_set(func: ta.Callable) -> frozenset[type]:
-    with contextlib.suppress(KeyError):
+    try:
         return _IMPL_FUNC_CLS_SET_CACHE[func]
+    except KeyError:
+        pass
 
     ann = getattr(func, '__annotations__', {})
     if not ann:

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/lang/objects.py

@@ -11,11 +11,24 @@ T = ta.TypeVar('T')
 ##
 
 
-def attr_repr(obj: ta.Any, *attrs: str) -> str:
-    return f'{type(obj).__name__}({", ".join(f"{attr}={getattr(obj, attr)!r}" for attr in attrs)})'
+def attr_repr(
+        obj: ta.Any,
+        *attrs: str,
+        with_module: bool = False,
+        use_qualname: bool = False,
+        with_id: bool = False,
+) -> str:
+    return (
+        f'{obj.__class__.__module__ + "." if with_module else ""}'
+        f'{obj.__class__.__qualname__ if use_qualname else obj.__class__.__name__}'
+        f'{("@" + hex(id(obj))[2:]) if with_id else ""}'
+        f'('
+        f'{", ".join(f"{attr}={getattr(obj, attr)!r}" for attr in attrs)}'
+        f')'
+    )
 
 
-def arg_repr(*args, **kwargs) -> str:
+def arg_repr(*args: ta.Any, **kwargs: ta.Any) -> str:
     return ', '.join(*(
         list(map(repr, args)) +
         [f'{k}={v!r}' for k, v in kwargs.items()]
@@ -57,7 +70,7 @@ def new_type(
         name: str,
         bases: ta.Sequence[ta.Any],
         namespace: ta.Mapping[str, ta.Any],
-        **kwargs,
+        **kwargs: ta.Any,
 ) -> type:
     return types.new_class(
         name,
@@ -73,7 +86,7 @@ def super_meta(
         name: str,
         bases: ta.Sequence[ta.Any],
         namespace: ta.MutableMapping[str, ta.Any],
-        **kwargs,
+        **kwargs: ta.Any,
 ) -> type:
     """Per types.new_class"""
     resolved_bases = types.resolve_bases(bases)
@@ -87,7 +100,11 @@ def super_meta(
 ##
 
 
-def deep_subclasses(cls: type[T], *, concrete_only: bool = False) -> ta.Iterator[type[T]]:
+def deep_subclasses(
+        cls: type[T],
+        *,
+        concrete_only: bool = False,
+) -> ta.Iterator[type[T]]:
     seen = set()
     todo = list(reversed(cls.__subclasses__()))
     while todo:

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/os/filemodes.py

@@ -24,7 +24,7 @@ class FileMode:
 
     https://en.cppreference.com/w/cpp/io/c/fopen
 
-    "r"   read             Open a file for reading        read from start   return NULL and set error
+    "r"   read             Open a file for reading       read from start   return NULL and set error
     "w"   write            Create a file for writing     destroy contents  create new
     "a"   append           Append to a file              write to end      create new
     "r+"  read extended    Open a file for read/write    read from start   return NULL and set error

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/sql/parsing/Minisql.g4

@@ -100,11 +100,14 @@ sortItem
     ;
 
 relation
-    : relation AS? ident                                                     #aliasedRelation
-    | left=relation ty=joinType? JOIN right=relation (ON cond=booleanExpr)?  #joinRelation
-    | '(' select ')'                                                         #selectRelation
-    | '(' relation ')'                                                       #parenRelation
-    | qualifiedName                                                          #tableRelation
+    : relation AS? ident      #aliasedRelation
+    | left=relation
+      ty=joinType?
+      JOIN right=relation
+      (ON cond=booleanExpr)?  #joinRelation
+    | '(' select ')'          #selectRelation
+    | '(' relation ')'        #parenRelation
+    | qualifiedName           #tableRelation
     ;
 
 groupBy

{omlish-0.0.0.dev253 → omlish-0.0.0.dev255}/omlish/subprocesses/wrap.py

@@ -1,5 +1,10 @@
 # ruff: noqa: UP006 UP007
 # @omlish-lite
+"""
+This bypasses debuggers attaching to spawned subprocess children that look like python processes. See:
+
+  https://github.com/JetBrains/intellij-community/blob/e9d8f126c286acf9df3ff272f440b305bf2ff585/python/helpers/pydev/_pydev_bundle/pydev_monkey.py
+"""
 import shlex
 import typing as ta
 

omlish-0.0.0.dev255/omlish/text/decoding.py

@@ -0,0 +1,73 @@
+"""
+TODO:
+ - cext
+"""
+import array
+import codecs
+import typing as ta
+
+
+##
+
+
+def decode_to_list(
+        raw: bytes,
+        encoding: str = 'utf-8',
+        *,
+        errors: str = 'strict',
+) -> list[str]:
+    dec = codecs.getincrementaldecoder(encoding)(errors)
+    end = len(raw) - 1
+    return [dec.decode(bytes([b]), final=i == end) for i, b in enumerate(raw)]
+
+
+#
+
+
+class DecodedStringIndex(ta.NamedTuple):
+    byte_offsets: ta.Sequence[int]
+
+
+def decode_indexed(
+        raw: bytes,
+        encoding: str = 'utf-8',
+        *,
+        errors: str = 'strict',
+) -> tuple[str, DecodedStringIndex]:
+    dec_lst = decode_to_list(
+        raw,
+        encoding,
+        errors=errors,
+    )
+
+    dec_s = ''.join(dec_lst)
+
+    bo_arr_len = len(dec_s) + 1
+    if bo_arr_len < 2**8:
+        fmt = 'B'
+    elif bo_arr_len < 2**16:
+        fmt = 'H'
+    elif bo_arr_len < 2**32:
+        fmt = 'L'
+    else:
+        fmt = 'Q'
+
+    bo_arr = array.array(fmt, [0]) * bo_arr_len
+    so = 0
+    lbo = 0
+    for bo, s in enumerate(dec_lst):
+        if s:
+            bo_arr[so] = lbo
+            so += 1
+            lbo = bo + 1
+
+    if so != len(dec_s):
+        raise RuntimeError
+
+    bo_arr[len(dec_s)] = len(raw)
+
+    dsi = DecodedStringIndex(
+        bo_arr,
+    )
+
+    return (dec_s, dsi)

omlish-0.0.0.dev255/omlish/text/filecache.py

@@ -0,0 +1,235 @@
+"""
+Features:
+ - get (line, offset)
+ - invalidate
+  - filewatcher
+ - linecache interop
+  - tokenize.open - detect encoding - only for .py
+ - directory / path_parts tree?
+
+TODO:
+ - read raw, decode (detect)
+  - ! index from byteofs -> charofs
+   - charofs -> lineno
+ - 1-based?
+ - maximums?
+  - lru? max_size/max_entries?
+   - collections.cache?
+
+Notes:
+ - linecache is 1-based
+ - linecache appends final newline if missing:
+    if lines and not lines[-1].endswith('\n'):
+        lines[-1] += '\n'
+"""
+import dataclasses as dc
+import io
+import os.path
+import stat as stat_
+import tokenize
+import typing as ta
+
+from .. import cached
+from .. import lang
+from ..os.paths import abs_real_path
+
+
+##
+
+
+class TextFileCache:
+    class FileStat(ta.NamedTuple):
+        size: int
+        mtime: float
+
+    class Entry:
+        def __init__(
+                self,
+                cache: 'TextFileCache',
+                path: str,
+                stat: 'TextFileCache.FileStat',
+                *,
+                encoding: str | None = None,
+        ) -> None:
+            super().__init__()
+
+            self._cache = cache
+            self._path = path
+            self._stat = stat
+            self._given_encoding = encoding
+
+        def __repr__(self) -> str:
+            return lang.attr_repr(self, 'path', 'stat')
+
+        @property
+        def path(self) -> str:
+            return self._path
+
+        @property
+        def stat(self) -> 'TextFileCache.FileStat':
+            return self._stat
+
+        @cached.function
+        def raw(self) -> bytes:
+            return self._cache._read_file(self._path)  # noqa
+
+        @cached.function
+        def encoding(self) -> str:
+            if (ge := self._given_encoding) is not None:
+                return ge
+            return self._cache._determine_encoding(self._path, self.raw())  # noqa
+
+        @cached.function
+        def text(self) -> str:
+            return self.raw().decode(self.encoding())
+
+        @cached.function
+        def lines(self) -> ta.Sequence[str]:
+            return self.text().splitlines(keepends=True)
+
+        def safe_line(self, n: int, default: str = '') -> str:
+            lines = self.lines()
+            if 0 <= n < len(lines):
+                return lines[n]
+            return default
+
+    #
+
+    @dc.dataclass(frozen=True)
+    class Config:
+        max_file_size: int | None = 5 * 1024 * 1024
+
+        default_encoding: str = 'utf-8'
+
+    def __init__(
+            self,
+            config: Config = Config(),
+            *,
+            locking: lang.DefaultLockable = None,
+    ) -> None:
+        super().__init__()
+
+        self._config = config
+
+        self._dct: dict[str, TextFileCache.Entry] = {}
+        self._lock = lang.default_lock(locking)()
+
+    #
+
+    class Error(Exception):
+        pass
+
+    @dc.dataclass()
+    class PathError(Error):
+        path: str
+
+    class PathTypeError(PathError):
+        pass
+
+    class FileTooBigError(PathError):
+        pass
+
+    #
+
+    def _normalize_path(self, path: str) -> str:
+        return abs_real_path(path)
+
+    #
+
+    def _try_stat_file(self, path: str) -> FileStat | Exception:
+        try:
+            st = os.stat(path)
+        except FileNotFoundError as e:
+            return e
+
+        if not stat_.S_ISREG(st.st_mode):
+            return TextFileCache.PathTypeError(path)
+
+        return TextFileCache.FileStat(
+            st.st_size,
+            st.st_mtime,
+        )
+
+    def _read_file(self, path: str) -> bytes:
+        with open(path, 'rb') as f:
+            return f.read()
+
+    def _determine_encoding(self, path: str, raw: bytes) -> str:
+        if path.endswith('.py'):
+            buf = io.BytesIO(raw)
+            encoding, _ = tokenize.detect_encoding(buf.readline)
+            return encoding
+
+        return self._config.default_encoding
+
+    #
+
+    def _get_entry(
+            self,
+            path: str,
+            *,
+            check_stat: bool = False,
+            or_raise: bool = False,
+    ) -> Entry:
+        st_: TextFileCache.FileStat | Exception | None = None
+        try:
+            e = self._dct[path]
+
+        except KeyError:
+            if or_raise:
+                raise
+
+        else:
+            if (
+                    not check_stat or
+                    (st_ := self._try_stat_file(path)) == e.stat
+            ):
+                return e
+
+            del self._dct[path]
+
+        st: TextFileCache.FileStat | Exception
+        if st_ is not None:
+            st = st_
+        else:
+            st = self._try_stat_file(path)
+
+        if isinstance(st, Exception):
+            raise st
+
+        if (mfs := self._config.max_file_size) is not None and st.size > mfs:
+            raise TextFileCache.FileTooBigError(path)
+
+        e = TextFileCache.Entry(
+            self,
+            path,
+            st,
+        )
+
+        self._dct[path] = e
+
+        return e
+
+    def get_entry(
+            self,
+            path: str,
+            *,
+            check_stat: bool = False,
+            or_raise: bool = False,
+    ) -> Entry:
+        path = self._normalize_path(path)
+
+        with self._lock:
+            return self._get_entry(
+                path,
+                check_stat=check_stat,
+                or_raise=or_raise,
+            )
+
+    #
+
+    def invalidate(self, path: str) -> bool:
+        path = self._normalize_path(path)
+
+        with self._lock:
+            return self._dct.pop(path, None) is not None

omlish-0.0.0.dev255/omlish/text/linecache.py

@@ -0,0 +1,80 @@
+"""
+TODO:
+ - reserver / generator
+
+==
+
+@dc.dataclass()
+class _ReservedFilenameEntry:
+    unique_id: str
+    seq: int = 0
+
+
+_RESERVED_FILENAME_UUID_THREAD_LOCAL = threading.local()
+
+
+def reserve_filename(prefix: str) -> str:
+    try:
+        e = _RESERVED_FILENAME_UUID_THREAD_LOCAL.unique_id
+    except AttributeError:
+        e = _RESERVED_FILENAME_UUID_THREAD_LOCAL.unique_id = _ReservedFilenameEntry(str(uuid.uuid4()))
+    while True:
+        unique_filename = f'<generated:{prefix}:{e.seq}>'
+        cache_line = (1, None, (e.unique_id,), unique_filename)
+        e.seq += 1
+        if linecache.cache.setdefault(unique_filename, cache_line) == cache_line:
+            return unique_filename
+"""
+import typing as ta
+
+
+##
+
+
+class LinecacheKey(ta.NamedTuple):
+    size: int  # os.stat().st_size
+    mtime: float | None  # os.stat().st_mtime
+    lines: ta.Sequence[str]
+    fullname: str
+
+
+LinecacheLazyLoader: ta.TypeAlias = tuple[ta.Callable[[], str]]
+
+LinecacheEntry: ta.TypeAlias = LinecacheKey | LinecacheLazyLoader
+
+
+class LinecacheProtocol(ta.Protocol):
+    @property
+    def cache(self) -> ta.MutableMapping[str, LinecacheEntry]: ...
+
+    def clearcache(self) -> None: ...
+
+    def getline(
+        self,
+        filename: str,
+        lineno: int,
+        module_globals: ta.Mapping[str, ta.Any] | None = None,
+    ) -> str: ...
+
+    def getlines(
+        self,
+        filename: str,
+        module_globals: ta.Mapping[str, ta.Any] | None = None,
+    ) -> ta.Sequence[str]: ...
+
+    def checkcache(
+        self,
+        filename: str | None = None,
+    ) -> None: ...
+
+    def updatecache(
+        self,
+        filename: str,
+        module_globals: ta.Mapping[str, ta.Any] | None = None,
+    ) -> ta.Sequence[str]: ...
+
+    def lazycache(
+        self,
+        filename: str,
+        module_globals: ta.Mapping[str, ta.Any],
+    ) -> bool: ...