omlish 0.0.0.dev103__tar.gz → 0.0.0.dev105__tar.gz

{omlish-0.0.0.dev103/omlish.egg-info → omlish-0.0.0.dev105}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: omlish
-Version: 0.0.0.dev103
+Version: 0.0.0.dev105
 Summary: omlish
 Author: wrmsr
 License: BSD-3-Clause

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/__about__.py

@@ -1,5 +1,5 @@
-__version__ = '0.0.0.dev103'
-__revision__ = 'e402b674aa19d5ca77d769f4290c26e35eb24c22'
+__version__ = '0.0.0.dev105'
+__revision__ = '933f83fc2d561338d613829ed857297f339dcccd'
 
 
 #

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/concurrent/threadlets.py

@@ -94,3 +94,6 @@ class GreenletThreadlets(Threadlets):
 
     def get_current(self) -> Threadlet:
         return GreenletThreadlet(greenlet.getcurrent())
+
+
+GREENLET_THREADLETS = GreenletThreadlets()

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/fnpairs.py

@@ -26,7 +26,6 @@ from . import lang
 if ta.TYPE_CHECKING:
     import bz2 as _bz2
     import gzip as _gzip
-    import json as _json
     import lzma as _lzma
     import pickle as _pickle
     import struct as _struct
@@ -40,10 +39,11 @@ if ta.TYPE_CHECKING:
     import yaml as _yaml
     import zstandard as _zstandard
 
+    from .formats import json as _json
+
 else:
     _bz2 = lang.proxy_import('bz2')
     _gzip = lang.proxy_import('gzip')
-    _json = lang.proxy_import('json')
     _lzma = lang.proxy_import('lzma')
     _pickle = lang.proxy_import('pickle')
     _struct = lang.proxy_import('struct')
@@ -57,6 +57,8 @@ else:
     _yaml = lang.proxy_import('yaml')
     _zstandard = lang.proxy_import('zstandard')
 
+    _json = lang.proxy_import('.formats.json', __package__)
+
 
 ##
 
@@ -277,7 +279,7 @@ def _register_extension(*ss):
     def inner(cls):
         for s in ss:
             if s in _EXTENSION_REGISTRY:
-                raise Exception(s)
+                raise KeyError(s)
             _EXTENSION_REGISTRY[s] = cls
         return cls
     return inner
@@ -409,22 +411,30 @@ class Pickle(ObjectBytes_):
         return _pickle.loads(t)
 
 
+class _Json(ObjectStr_, lang.Abstract):  # noqa
+    def backward(self, t: str) -> ta.Any:
+        return _json.loads(t)
+
+
 @_register_extension('json')
-@dc.dataclass(frozen=True)
-class Json(ObjectStr_):
-    indent: int | str | None = dc.field(default=None, kw_only=True)
-    separators: tuple[str, str] | None = dc.field(default=None, kw_only=True)
+class Json(_Json):
+    def forward(self, f: ta.Any) -> str:
+        return _json.dumps(f)
+
 
+class JsonPretty(_Json):
     def forward(self, f: ta.Any) -> str:
-        return _json.dumps(f, indent=self.indent, separators=self.separators)
+        return _json.dumps_pretty(f)
 
-    def backward(self, t: str) -> ta.Any:
-        return _json.loads(t)
+
+class JsonCompact(_Json):
+    def forward(self, f: ta.Any) -> str:
+        return _json.dumps_compact(f)
 
 
 JSON = Json()
-PRETTY_JSON = Json(indent=2)
-COMPACT_JSON = Json(separators=(',', ':'))
+PRETTY_JSON = JsonPretty()
+COMPACT_JSON = JsonCompact()
 
 
 @_register_extension('jsonl')

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/fnpipes.py

@@ -6,6 +6,9 @@ T = ta.TypeVar('T')
 U = ta.TypeVar('U')
 
 
+##
+
+
 class Fn(abc.ABC, ta.Generic[T]):
     @abc.abstractmethod
     def __call__(self, *args: ta.Any, **kwargs: ta.Any) -> T:
@@ -24,6 +27,9 @@ class Fn(abc.ABC, ta.Generic[T]):
         return self.apply(fn)
 
 
+##
+
+
 class Bind(Fn[T]):
     def __init__(self, fn: ta.Callable[..., T], *args: ta.Any, **kwargs: ta.Any) -> None:
         super().__init__()
@@ -54,6 +60,12 @@ class Bind(Fn[T]):
         return self._fn(*fa, **fkw)
 
 
+bind = Bind
+
+
+##
+
+
 class Pipe(Fn[T]):
     def __init__(self, lfns: ta.Sequence[ta.Callable], rfn: ta.Callable[..., T]) -> None:
         super().__init__()
@@ -66,6 +78,14 @@ class Pipe(Fn[T]):
         return o
 
 
+def pipe(*fns: ta.Callable) -> Pipe:
+    *lfns, rfn = fns
+    return Pipe(lfns, rfn)
+
+
+##
+
+
 class Apply(Fn[T]):
     def __init__(self, *fns: ta.Callable[[T], ta.Any]) -> None:
         super().__init__()
@@ -77,6 +97,4 @@ class Apply(Fn[T]):
         return o
 
 
-bind = Bind
-pipe = Pipe
 apply = Apply

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/formats/json/__init__.py

@@ -3,6 +3,8 @@ from .consts import (  # noqa
     COMPACT_SEPARATORS,
     PRETTY_INDENT,
     PRETTY_KWARGS,
+    PRETTY_SEPARATORS,
+    Separators,
 )
 
 from .encoding import (  # noqa

omlish-0.0.0.dev105/omlish/formats/json/cli/cli.py

@@ -0,0 +1,299 @@
+"""
+TODO:
+ - read from http
+ - jmespath output flat, unquoted strs like jq '.[]'
+
+==
+
+jq Command options)
+  -n, --null-input          use `null` as the single input value;
+  -R, --raw-input           read each line as string instead of JSON;
+  -s, --slurp               read all inputs into an array and use it as the single input value;
+  -c, --compact-output      compact instead of pretty-printed output;
+  -r, --raw-output          output strings without escapes and quotes;
+      --raw-output0         implies -r and output NUL after each output;
+  -j, --join-output         implies -r and output without newline after each output;
+  -a, --ascii-output        output strings by only ASCII characters using escape sequences;
+  -S, --sort-keys           sort keys of each object on output;
+  -C, --color-output        colorize JSON output;
+  -M, --monochrome-output   disable colored output;
+      --tab                 use tabs for indentation;
+      --indent n            use n spaces for indentation (max 7 spaces);
+      --unbuffered          flush output stream after each output;
+      --stream              parse the input value in streaming fashion;
+      --stream-errors       implies --stream and report parse error as an array;
+      --seq                 parse input/output as application/json-seq;
+  -f, --from-file file      load filter from the file;
+  -L directory              search modules from the directory;
+      --arg name value      set $name to the string value;
+      --argjson name value  set $name to the JSON value;
+      --slurpfile name file set $name to an array of JSON values read from the file;
+      --rawfile name file   set $name to string contents of file;
+      --args                consume remaining arguments as positional string values;
+      --jsonargs            consume remaining arguments as positional JSON values;
+  -e, --exit-status         set exit status code based on the output;
+  -V, --version             show the version;
+  --build-configuration     show jq's build configuration;
+  -h, --help                show the help;
+  --                        terminates argument processing;
+"""
+import argparse
+import contextlib
+import dataclasses as dc
+import io
+import os
+import subprocess
+import sys
+import typing as ta
+
+from .... import check
+from .... import fnpipes as fp
+from .... import lang
+from .formats import FORMATS_BY_NAME
+from .formats import Format
+from .formats import Formats
+from .parsing import DelimitingParser
+from .parsing import EagerParser
+from .parsing import StreamBuilder
+from .parsing import StreamParser
+from .processing import ProcessingOptions
+from .processing import Processor
+from .rendering import EagerRenderer
+from .rendering import RenderingOptions
+from .rendering import StreamRenderer
+
+
+T = ta.TypeVar('T')
+U = ta.TypeVar('U')
+
+
+def _build_args_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument('file', nargs='?')
+
+    parser.add_argument('--stream', action='store_true')
+    parser.add_argument('--stream-build', action='store_true')
+
+    parser.add_argument('-l', '--lines', action='store_true')
+
+    parser.add_argument('--read-buffer-size', type=int, default=0x4000)
+
+    parser.add_argument('-f', '--format')
+
+    parser.add_argument('-x', '--jmespath-expr')
+    parser.add_argument('-F', '--flat', action='store_true')
+
+    parser.add_argument('-z', '--compact', action='store_true')
+    parser.add_argument('-p', '--pretty', action='store_true')
+    parser.add_argument('-i', '--indent')
+    parser.add_argument('-s', '--sort-keys', action='store_true')
+    parser.add_argument('-R', '--raw', action='store_true')
+    parser.add_argument('-U', '--unicode', action='store_true')
+    parser.add_argument('-c', '--color', action='store_true')
+
+    parser.add_argument('-L', '--less', action='store_true')
+
+    return parser
+
+
+def _parse_args(args: ta.Any = None) -> ta.Any:
+    return _build_args_parser().parse_args(args)
+
+
+@dc.dataclass(frozen=True, kw_only=True)
+class RunConfiguration:
+    format: Format
+    processing: ProcessingOptions
+    rendering: RenderingOptions
+
+
+def _process_args(args: ta.Any) -> RunConfiguration:
+    fmt_name = args.format
+    if fmt_name is None:
+        if args.file is not None:
+            ext = args.file.rpartition('.')[2]
+            if ext in FORMATS_BY_NAME:
+                fmt_name = ext
+    if fmt_name is None:
+        fmt_name = 'json'
+    format = FORMATS_BY_NAME[fmt_name]  # noqa
+
+    if args.stream:
+        check.arg(format is Formats.JSON.value)
+
+    #
+
+    processing = ProcessingOptions(
+        jmespath_expr=args.jmespath_expr,
+        flat=args.flat,
+    )
+
+    #
+
+    separators = None
+    if args.compact:
+        separators = (',', ':')
+
+    indent = None
+    if args.pretty:
+        indent = 2
+    if args.indent:
+        try:
+            indent = int(args.indent)
+        except ValueError:
+            indent = args.indent
+
+    rendering = RenderingOptions(
+        indent=indent,
+        separators=separators,
+        sort_keys=args.sort_keys,
+        raw=args.raw,
+        unicode=args.unicode,
+        color=args.color,
+    )
+
+    #
+
+    return RunConfiguration(
+        format=format,
+        processing=processing,
+        rendering=rendering,
+    )
+
+
+def _main() -> None:
+    args = _parse_args()
+
+    #
+
+    cfg = _process_args(args)
+
+    #
+
+    with contextlib.ExitStack() as es:
+        if args.file is None:
+            in_file = sys.stdin.buffer
+
+        else:
+            in_file = es.enter_context(open(args.file, 'rb'))
+
+        def yield_input() -> ta.Generator[bytes, None, None]:
+            fd = check.isinstance(in_file.fileno(), int)
+
+            while True:
+                buf = os.read(fd, args.read_buffer_size)
+
+                yield buf
+
+                if not buf:
+                    break
+
+        #
+
+        if args.less:
+            less = subprocess.Popen(
+                [
+                    'less',
+                    *(['-R'] if cfg.rendering.color else []),
+                ],
+                stdin=subprocess.PIPE,
+                encoding='utf-8',
+            )
+            out = check.not_none(less.stdin)
+
+            def close_less() -> None:
+                out.close()
+                less.wait()
+
+            es.enter_context(lang.defer(close_less))  # noqa
+
+        else:
+            out = sys.stdout
+
+        #
+
+        parser: ta.Any
+        renderer: ta.Any
+
+        if args.stream:
+            with contextlib.ExitStack() as es2:
+                parser = es2.enter_context(StreamParser())
+
+                def flush_output(
+                        fn: ta.Callable[[T], ta.Iterable[U]],
+                        i: T,
+                ) -> ta.Generator[U, None, None]:
+                    n = 0
+                    for o in fn(i):
+                        yield o
+                        n += 1
+                    if n:
+                        out.flush()
+
+                pipeline: ta.Any
+
+                if args.stream_build:
+                    builder: StreamBuilder = es2.enter_context(StreamBuilder())
+                    processor = Processor(cfg.processing)
+                    renderer = EagerRenderer(cfg.rendering)
+                    trailing_newline = False
+
+                    def append_newlines(
+                            fn: ta.Callable[[T], ta.Iterable[str]],
+                            i: T,
+                    ) -> ta.Generator[str, None, None]:
+                        yield from fn(i)
+                        yield '\n'
+
+                    pipeline = lambda v: (renderer.render(v),)  # Any -> [str]  # noqa
+                    pipeline = fp.bind(append_newlines, pipeline)  # Any -> [str]
+                    pipeline = fp.bind(lang.flatmap, pipeline)  # [Any] -> [str]
+                    pipeline = fp.pipe(fp.bind(lang.flatmap, processor.process), pipeline)  # [Any] -> [str]
+                    pipeline = fp.pipe(fp.bind(lang.flatmap, builder.build), pipeline)  # [JsonStreamParserEvent] -> [str]  # noqa
+                    pipeline = fp.pipe(parser.parse, pipeline)  # bytes -> [str]
+
+                else:
+                    renderer = StreamRenderer(cfg.rendering)
+                    trailing_newline = True
+
+                    pipeline = renderer.render  # JsonStreamParserEvent -> [str]
+                    pipeline = fp.bind(lang.flatmap, pipeline)  # [JsonStreamParserEvent] -> [str]
+                    pipeline = fp.pipe(parser.parse, pipeline)  # bytes -> [str]
+
+                pipeline = fp.bind(flush_output, pipeline)  # bytes -> [str]
+
+                for buf in yield_input():
+                    for s in pipeline(buf):
+                        print(s, file=out, end='')
+
+                if trailing_newline:
+                    print(file=out)
+
+        elif args.lines:
+            parser = DelimitingParser(cfg.format)
+            processor = Processor(cfg.processing)
+            renderer = EagerRenderer(cfg.rendering)
+
+            for buf in yield_input():
+                if buf:
+                    for v in parser.parse(buf):
+                        for e in processor.process(v):
+                            s = renderer.render(e)
+                            print(s, file=out)
+
+        else:
+            parser = EagerParser(cfg.format)
+            processor = Processor(cfg.processing)
+            renderer = EagerRenderer(cfg.rendering)
+
+            with io.TextIOWrapper(in_file) as tf:
+                v = parser.parse(tf)
+
+            for e in processor.process(v):
+                s = renderer.render(e)
+                print(s, file=out)
+
+
+if __name__ == '__main__':
+    _main()

{omlish-0.0.0.dev103 → omlish-0.0.0.dev105}/omlish/formats/json/cli/formats.py

@@ -44,12 +44,19 @@ class Format:
 
 class Formats(enum.Enum):
     JSON = Format(['json'], json.load)
+
     YAML = Format(['yaml', 'yml'], lambda f: yaml.safe_load(f))
+
     TOML = Format(['toml'], lambda f: tomllib.loads(f.read()))
+
     ENV = Format(['env', 'dotenv'], lambda f: dotenv.dotenv_values(stream=f))
+
     PROPS = Format(['properties', 'props'], lambda f: dict(props.Properties().load(f.read())))
+
     PY = Format(['py', 'python', 'repr'], lambda f: ast.literal_eval(f.read()))
+
     XML = Format(['xml'], lambda f: xml.build_simple_element(xml.parse_tree(f.read()).getroot()).as_dict())
+
     CSV = Format(['csv'], lambda f: list(csv.DictReader(f)))
     TSV = Format(['tsv'], lambda f: list(csv.DictReader(f, delimiter='\t')))
     FLAT_CSV = Format(['fcsv'], lambda f: list(csv.reader(f)))

omlish-0.0.0.dev105/omlish/formats/json/cli/io.py

@@ -0,0 +1,74 @@
+# """
+# TODO:
+#  - -I/-Ogz, lz4, etc
+#   - fnpairs? or not yet just do it
+# """
+# import abc
+# import contextlib
+# import dataclasses as dc
+# import gzip
+# import os
+# import typing as ta
+#
+# from .... import lang
+#
+#
+# if ta.TYPE_CHECKING:
+#     import bz2 as _bz2
+#     import gzip as _gzip
+#     import lzma as _lzma
+# else:
+#     _bz2 = lang.proxy_import('bz2')
+#     _gzip = lang.proxy_import('gzip')
+#     _lzma = lang.proxy_import('lzma')
+#
+#
+# ##
+#
+#
+# class BaseIo(lang.Abstract):
+#     def close(self) -> None:
+#         raise NotImplementedError
+#
+#     def fileno(self) -> int | None:
+#         return None
+#
+#
+# class Input(BaseIo):
+#     @abc.abstractmethod
+#     def read(self, sz: int | None = None) -> bytes:
+#         raise NotImplementedError
+#
+#
+# class Output(BaseIo):
+#     @abc.abstractmethod
+#     def write(self, data: bytes) -> int:
+#         raise NotImplementedError
+#
+#
+# #
+#
+#
+# DEFAULT_READ_SZ = 0x4000
+#
+#
+# @dc.dataclass(frozen=True)
+# class FdIo(Input, Output):
+#     fd: int
+#
+#     default_read_sz: int = DEFAULT_READ_SZ
+#
+#     def read(self, sz: int | None = None) -> bytes:
+#         return os.read(self.fd, sz or self.default_read_sz)
+#
+#     def write(self, data: bytes) -> int:
+#         return os.write(self.fd, data)
+#
+#
+# ##
+#
+#
+# @contextlib.contextmanager
+# def gzip_io_codec(f: ta.IO, mode: str) -> ta.ContextManager[ta.IO]:
+#     with gzip.open(f, mode) as o:
+#         yield o

omlish-0.0.0.dev105/omlish/formats/json/cli/parsing.py

@@ -0,0 +1,82 @@
+import codecs
+import io
+import typing as ta
+
+from .... import check
+from .... import lang
+from ....lite.io import DelimitingBuffer
+from ..stream.build import JsonObjectBuilder
+from ..stream.lex import JsonStreamLexer
+from ..stream.parse import JsonStreamParser
+from ..stream.parse import JsonStreamParserEvent
+from .formats import Format
+
+
+##
+
+
+class EagerParser:
+    def __init__(self, fmt: Format) -> None:
+        super().__init__()
+
+        self._fmt = fmt
+
+    def parse(self, f: ta.TextIO) -> ta.Generator[ta.Any, None, None]:
+        return self._fmt.load(f)
+
+
+##
+
+
+class DelimitingParser:
+    def __init__(
+            self,
+            fmt: Format,
+            *,
+            delimiters: ta.Iterable[int] = b'\n',
+    ) -> None:
+        super().__init__()
+
+        self._fmt = fmt
+
+        self._db = DelimitingBuffer(delimiters)
+
+    def parse(self, b: bytes) -> ta.Generator[ta.Any, None, None]:
+        for chunk in self._db.feed(b):
+            s = check.isinstance(chunk, bytes).decode('utf-8')
+            v = self._fmt.load(io.StringIO(s))
+            yield v
+
+
+##
+
+
+class StreamBuilder(lang.ExitStacked):
+    _builder: JsonObjectBuilder | None = None
+
+    def __enter__(self) -> ta.Self:
+        super().__enter__()
+        self._builder = self._enter_context(JsonObjectBuilder())
+        return self
+
+    def build(self, e: JsonStreamParserEvent) -> ta.Generator[ta.Any, None, None]:
+        yield from check.not_none(self._builder)(e)
+
+
+class StreamParser(lang.ExitStacked):
+    _decoder: codecs.IncrementalDecoder
+    _lex: JsonStreamLexer
+    _parse: JsonStreamParser
+
+    def __enter__(self) -> ta.Self:
+        super().__enter__()
+        self._decoder = codecs.getincrementaldecoder('utf-8')()
+        self._lex = self._enter_context(JsonStreamLexer())
+        self._parse = self._enter_context(JsonStreamParser())
+        return self
+
+    def parse(self, b: bytes) -> ta.Generator[JsonStreamParserEvent, None, None]:
+        for s in self._decoder.decode(b, not b):
+            for c in s:
+                for t in self._lex(c):
+                    yield from self._parse(t)

omlish-0.0.0.dev105/omlish/formats/json/cli/processing.py

@@ -0,0 +1,44 @@
+import dataclasses as dc
+import typing as ta
+
+from .... import lang
+
+
+if ta.TYPE_CHECKING:
+    from ....specs import jmespath
+else:
+    jmespath = lang.proxy_import('....specs.jmespath', __package__)
+
+
+##
+
+
+@dc.dataclass(frozen=True, kw_only=True)
+class ProcessingOptions:
+    jmespath_expr: ta.Any | None = None
+    flat: bool = False
+
+
+class Processor:
+    def __init__(self, opts: ProcessingOptions) -> None:
+        super().__init__()
+
+        self._opts = opts
+
+        jmespath_expr = opts.jmespath_expr
+        if isinstance(jmespath_expr, str):
+            jmespath_expr = jmespath.compile(jmespath_expr)
+        self._jmespath_expr: ta.Any | None = jmespath_expr
+
+    def process(self, v: ta.Any) -> ta.Iterable[ta.Any]:
+        if self._jmespath_expr is not None:
+            v = self._jmespath_expr.search(v)
+
+        if self._opts.flat:
+            if isinstance(v, str):
+                raise TypeError(f'Flat output must be arrays, got {type(v)}', v)
+
+            yield from v
+
+        else:
+            yield v