ssc_codegen 0.29.5__tar.gz → 0.30.0__tar.gz

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ssc_codegen
-Version: 0.29.5
+Version: 0.30.0
 Summary: Python-dsl code converter to html parser for web scraping 
 Project-URL: Documentation, https://github.com/vypivshiy/selector_schema_codegen#readme
 Project-URL: Issues, https://github.com/vypivshiy/selector_schema_codegen/issues
@@ -20,13 +20,13 @@ Classifier: Topic :: Text Processing :: Markup :: HTML
 Classifier: Topic :: Utilities
 Requires-Python: >=3.10
 Requires-Dist: bs4>=0.0.2
-Requires-Dist: click<8.2.0
 Requires-Dist: colorama>=0.4.6; sys_platform == 'win32'
 Requires-Dist: cssselect>=1.2.0
+Requires-Dist: kdlquery>=1.1.2
 Requires-Dist: lxml>=5.3.0
 Requires-Dist: pyyaml>=6.0
 Requires-Dist: soupsieve>=2.6
-Requires-Dist: typer>=0.15.1
+Requires-Dist: typer>0.26
 Requires-Dist: typing-extensions; python_version < '3.11'
 Description-Content-Type: text/markdown
 

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ssc_codegen"
-version = "0.29.5"
+version = "0.30.0"
 description = "Python-dsl code converter to html parser for web scraping "
 readme = "README.md"
 requires-python = ">=3.10"
@@ -10,12 +10,10 @@ dependencies = [
     "cssselect>=1.2.0",
     "lxml>=5.3.0",
     "soupsieve>=2.6",
-    "typer>=0.15.1",
+    "typer>0.26",
     "typing_extensions;python_version<'3.11'",
-    # https://github.com/fastapi/typer/discussions/1215
-    # https://github.com/fastapi/typer/pull/1145
-    "click<8.2.0",
     "pyyaml>=6.0",
+    "kdlquery>=1.1.2",
 ]
 
 classifiers = [

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/converters/js_pure.py

@@ -588,6 +588,7 @@ def pre_struct_check_method(node: a.CheckMethod, ctx: ConverterContext):
     method_name = to_camel_case(node.name)
     return [
         f"{ctx.indent}{method_name}() " + "{",
+        f"{ctx.deeper().indent}let {ctx.var_name} = this._doc;",
     ]
 
 

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/converters/py_bs4.py

@@ -307,7 +307,7 @@ def pre_utilities(node: a.Utilities, ctx: ConverterContext):
     runtime = ctx.meta.get("runtime_module")
     if runtime:
         names = runtime_export_names(node)
-        return [f"from {runtime} import " + ", ".join(names), ""]
+        return [f"from .{runtime} import " + ", ".join(names), ""]
 
     if module_is_rest_only(node):
         lines = []
@@ -808,16 +808,10 @@ def pre_struct_check_method(node: a.CheckMethod, ctx: ConverterContext):
     from ssc_codegen.converters.helpers import to_snake_case
 
     method_name = to_snake_case(node.name)
-    return [f"    def {method_name}(self) -> bool:"]
-
-
-# @PY_BASE_CONVERTER.post(a.CheckMethod)
-# def post_struct_check_method(node: a.CheckMethod, ctx: ConverterContext):
-#     return [
-#         "            return True",
-#         "        except Exception:",
-#         "            return False",
-#     ]
+    return [
+        f"    def {method_name}(self) -> bool:",
+        f"        {ctx.var_name} = self._doc",
+    ]
 
 
 @PY_BASE_CONVERTER(a.SplitDoc)

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/converters/py_lxml.py

@@ -87,7 +87,7 @@ def pre_utilities(node: a.Utilities, ctx: ConverterContext):
     if runtime:
         names = py_bs4.runtime_export_names(node)
         names.append(_FALLBACK_HTML_EXPORT)
-        return [f"from {runtime} import " + ", ".join(names), ""]
+        return [f"from .{runtime} import " + ", ".join(names), ""]
 
     if py_bs4.module_is_rest_only(node):
         lines = []
@@ -352,7 +352,7 @@ def pre_expr_attr(node: a.Attr, ctx: ConverterContext):
 @PY_LXML_CONVERTER(a.PredCss)
 def pre_expr_pred_css(node: a.PredCss, ctx: ConverterContext):
     query = repr(node.query)
-    cond = f"bool(i.cssselect({query}))"
+    cond = f"i.cssselect({query}) is not None"
     if ctx.index == 0:
         return ctx.indent + cond
     return ctx.indent + f"and {cond}"
@@ -361,7 +361,7 @@ def pre_expr_pred_css(node: a.PredCss, ctx: ConverterContext):
 @PY_LXML_CONVERTER(a.PredXpath)
 def pre_expr_pred_xpath(node: a.PredXpath, ctx: ConverterContext):
     query = repr(node.query)
-    cond = f"bool(i.xpath({query}))"
+    cond = f"i.xpath({query}) is not None"
     if ctx.index == 0:
         return ctx.indent + cond
     return ctx.indent + f"and {cond}"

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/core/__init__.py

@@ -1,10 +1,10 @@
 """ssc_codegen.core — unified KDL -> Module AST reader with integrated linting."""
-from ssc_codegen.core.reader import parse_module, SscReader, ReadDiagnostic
+from ssc_codegen.core.reader import parse_module
 from ssc_codegen.core.format import format_diagnostics, format_diagnostic
+from kdlquery import ReadDiagnostic
 __all__ = [
     "parse_module",
-    "SscReader",
     "ReadDiagnostic",
     "format_diagnostics",
     "format_diagnostic"
-]
+]

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/core/contexts.py

@@ -1,235 +1,197 @@
-"""Parse/lint contexts, error codes, and supporting data classes."""
-
-from __future__ import annotations
-
-from dataclasses import dataclass, field
-from enum import Enum, auto
-from pathlib import Path
-
-from ssc_codegen.ast import VariableType
-from ssc_codegen.kdl import KdlNode, ReadDiagnostic, Severity
-from ssc_codegen.kdl.parser import Span
-
-
-# ── Walk context enum ──────────────────────────────────────────────────────────
-
-
-class WalkCtx(Enum):
-    MODULE = auto()
-    STRUCT_BODY = auto()
-    INIT_BLOCK = auto()
-    PIPELINE = auto()
-    JSON_TYPEDEF = auto()
-    SPECIAL_FIELD = auto()
-
-
-# ── Error codes ────────────────────────────────────────────────────────────────
-
-
-class ErrorCode(Enum):
-    INVALID_SYNTAX = "E000"
-    MISSING_ARGUMENT = "E001"
-    INVALID_ARGUMENT = "E002"
-    EMPTY_BLOCK = "E003"
-    UNEXPECTED_CHILDREN = "E004"
-    TYPE_MISMATCH = "E100"
-    INCOMPATIBLE_OPERATION = "E101"
-    UNKNOWN_OPERATION = "E200"
-    UNKNOWN_FIELD = "E201"
-    MISSING_REQUIRED_FIELD = "E202"
-    INVALID_FIELD_FOR_TYPE = "E203"
-    UNDEFINED_REFERENCE = "E300"
-    INIT_FIELD_NOT_FOUND = "E301"
-    DEFINE_NOT_FOUND = "E302"
-    INVALID_STRUCT_TYPE = "E400"
-    MISSING_SPECIAL_FIELD = "E401"
-    DEPRECATED_SYNTAX = "W001"
-    UNUSED_FIELD = "W002"
-
-
-class DefineKind(Enum):
-    SCALAR = auto()
-    BLOCK = auto()
-
-
-@dataclass
-class RawArg:
-    value: str
-    is_identifier: bool
-    span: Span
-
-
-@dataclass
-class DefineInfo:
-    name: str
-    kind: DefineKind
-    value: str | None
-    node: KdlNode
-
-
-@dataclass
-class TransformInfo:
-    name: str
-    accept: str
-    ret: str
-    langs: list[str]
-    node: KdlNode
-
-
-# ── Parse context ───────────────────────────────────────────────────────────────
-
-
-@dataclass
-class ParseContext:
-    property_defines: dict[str, str | int | float | bool] = field(
-        default_factory=dict
-    )
-    children_defines: dict[str, list[KdlNode]] = field(default_factory=dict)
-    transforms: dict[str, TransformDef] = field(default_factory=dict)
-    structs: dict[str, Struct] = field(default_factory=dict)
-    json_defs: dict[str, JsonDef] = field(default_factory=dict)
-    source_path: Path | None = None
-
-    def all_names(self) -> set[str]:
-        return (
-            set(self.property_defines)
-            | set(self.children_defines)
-            | set(self.transforms)
-            | set(self.structs)
-            | set(self.json_defs)
-        )
-
-
-# ── Lint context ───────────────────────────────────────────────────────────────
-
-
-@dataclass
-class LintContext:
-    """Lint state: defines, transforms, init_fields, walk context, path, diagnostics."""
-
-    defines: dict[str, DefineInfo] = field(default_factory=dict)
-    transforms: dict[str, TransformInfo] = field(default_factory=dict)
-    init_fields: set[str] = field(default_factory=set)
-    walk_context: WalkCtx = WalkCtx.MODULE
-    _path_segments: list[str] = field(default_factory=list)
-    diagnostics: list[ReadDiagnostic] = field(default_factory=list)
-    inferred_define_types: dict[str, tuple[VariableType, VariableType]] = field(
-        default_factory=dict
-    )
-    dsl_names: set[str] = field(default_factory=set)
-    json_kdl_nodes: dict[str, "KdlNode"] = field(default_factory=dict)
-    rest_response_refs: list[tuple["KdlNode", str]] = field(
-        default_factory=list
-    )
-    rest_error_refs: list[tuple["KdlNode", str]] = field(default_factory=list)
-    _predicate_depth: int = field(default=0)
-    _predicate_context: str = ""
-
-    def error(
-        self,
-        node: KdlNode,
-        *,
-        message: str,
-        code: str = "",
-        hint: str = "",
-        label: str | None = None,
-        notes: list[str] | None = None,
-    ) -> None:
-        self.diagnostics.append(
-            ReadDiagnostic(
-                message=message,
-                severity=Severity.ERROR,
-                span=node.span,
-                path=self.path,
-                hint=hint,
-                code=code,
-                label=label,
-                notes=tuple(notes) if notes else (),
-            )
-        )
-
-    def warning(
-        self,
-        node: KdlNode,
-        *,
-        message: str,
-        code: str = "",
-        hint: str = "",
-        label: str | None = None,
-        notes: list[str] | None = None,
-    ) -> None:
-        self.diagnostics.append(
-            ReadDiagnostic(
-                message=message,
-                severity=Severity.WARNING,
-                span=node.span,
-                path=self.path,
-                hint=hint,
-                code=code,
-                label=label,
-                notes=tuple(notes) if notes else (),
-            )
-        )
-
-    @property
-    def path(self) -> str:
-        return "/".join(self._path_segments)
-
-    def push(self, segment: str) -> None:
-        self._path_segments.append(segment)
-
-    def pop(self) -> None:
-        self._path_segments.pop()
-
-    def node_name(self, node: KdlNode) -> str:
-        return node.name
-
-    def get_args(self, node: KdlNode) -> list[str]:
-        return [str(a.value) for a in node.args]
-
-    def get_raw_args(self, node: KdlNode) -> list[RawArg]:
-        return [
-            RawArg(
-                value=str(a.value), is_identifier=a.is_identifier, span=a.span
-            )
-            for a in node.args
-        ]
-
-    def get_arg(self, node: KdlNode, index: int) -> str | None:
-        args = self.get_args(node)
-        return args[index] if index < len(args) else None
-
-    def get_prop(self, node: KdlNode, key: str) -> str | None:
-        entry = node.properties.get(key)
-        return str(entry.value) if entry else None
-
-    def get_children_nodes(self, node: KdlNode) -> list[KdlNode]:
-        return list(node.children)
-
-    def has_empty_block(self, node: KdlNode) -> bool:
-        return len(node.children) == 0
-
-    def is_define_ref(self, arg: str) -> bool:
-        return arg in self.defines
-
-    def resolve_scalar_arg(self, arg: str) -> str | None:
-        info = self.defines.get(arg)
-        if info is not None and info.kind == DefineKind.SCALAR:
-            return info.value
-        return None
-
-    @property
-    def in_predicate(self) -> bool:
-        return self._predicate_depth > 0
-
-    @property
-    def in_assert(self) -> bool:
-        return self._predicate_depth > 0 and self._predicate_context == "assert"
-
-
-# lazy imports for forward refs
-from ssc_codegen.ast import (  # noqa: E402
-    JsonDef,
-    Struct,
-    TransformDef,
-)
+"""Parse/lint contexts, error codes, and supporting data classes."""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from enum import Enum, auto
+from pathlib import Path
+
+from ssc_codegen.ast import VariableType
+from kdlquery import KdlNode, ReadDiagnostic, Severity
+from kdlquery.types import Span
+
+
+# ── Walk context enum ──────────────────────────────────────────────────────────
+
+
+class WalkCtx(Enum):
+    MODULE = auto()
+    STRUCT_BODY = auto()
+    INIT_BLOCK = auto()
+    PIPELINE = auto()
+    JSON_TYPEDEF = auto()
+    SPECIAL_FIELD = auto()
+
+
+# ── Error codes ────────────────────────────────────────────────────────────────
+
+
+class ErrorCode(Enum):
+    INVALID_SYNTAX = "E000"
+    MISSING_ARGUMENT = "E001"
+    INVALID_ARGUMENT = "E002"
+    EMPTY_BLOCK = "E003"
+    UNEXPECTED_CHILDREN = "E004"
+    TYPE_MISMATCH = "E100"
+    INCOMPATIBLE_OPERATION = "E101"
+    UNKNOWN_OPERATION = "E200"
+    UNKNOWN_FIELD = "E201"
+    MISSING_REQUIRED_FIELD = "E202"
+    INVALID_FIELD_FOR_TYPE = "E203"
+    UNDEFINED_REFERENCE = "E300"
+    INIT_FIELD_NOT_FOUND = "E301"
+    DEFINE_NOT_FOUND = "E302"
+    INVALID_STRUCT_TYPE = "E400"
+    MISSING_SPECIAL_FIELD = "E401"
+    DEPRECATED_SYNTAX = "W001"
+    UNUSED_FIELD = "W002"
+
+
+class DefineKind(Enum):
+    SCALAR = auto()
+    BLOCK = auto()
+
+
+@dataclass
+class RawArg:
+    value: str
+    span: Span
+
+
+@dataclass
+class DefineInfo:
+    name: str
+    kind: DefineKind
+    value: str | None
+    node: KdlNode
+
+
+@dataclass
+class TransformInfo:
+    name: str
+    accept: str
+    ret: str
+    langs: list[str]
+    node: KdlNode
+
+
+# ── Parse context ───────────────────────────────────────────────────────────────
+
+
+@dataclass
+class ParseContext:
+    property_defines: dict[str, str | int | float | bool] = field(
+        default_factory=dict
+    )
+    children_defines: dict[str, list[KdlNode]] = field(default_factory=dict)
+    transforms: dict[str, TransformDef] = field(default_factory=dict)
+    structs: dict[str, Struct] = field(default_factory=dict)
+    json_defs: dict[str, JsonDef] = field(default_factory=dict)
+    source_path: Path | None = None
+
+    def all_names(self) -> set[str]:
+        return (
+            set(self.property_defines)
+            | set(self.children_defines)
+            | set(self.transforms)
+            | set(self.structs)
+            | set(self.json_defs)
+        )
+
+
+# ── Lint context ───────────────────────────────────────────────────────────────
+
+
+@dataclass
+class LintContext:
+    """Lint state for reader-side validation (type inference, per-op checks)."""
+
+    defines: dict[str, DefineInfo] = field(default_factory=dict)
+    transforms: dict[str, TransformInfo] = field(default_factory=dict)
+    init_fields: set[str] = field(default_factory=set)
+    walk_context: WalkCtx = WalkCtx.MODULE
+    _path_segments: list[str] = field(default_factory=list)
+    diagnostics: list[ReadDiagnostic] = field(default_factory=list)
+    inferred_define_types: dict[str, tuple[VariableType, VariableType]] = field(
+        default_factory=dict
+    )
+    _predicate_depth: int = field(default=0)
+    _predicate_context: str = ""
+
+    def error(
+        self,
+        node: KdlNode,
+        *,
+        message: str,
+        code: str = "",
+        hint: str = "",
+        label: str | None = None,
+        notes: list[str] | None = None,
+    ) -> None:
+        self.diagnostics.append(
+            ReadDiagnostic(
+                message=message,
+                severity=Severity.ERROR,
+                span=node.span,
+                path=self.path,
+                hint=hint,
+                code=code,
+                label=label,
+                notes=tuple(notes) if notes else (),
+            )
+        )
+
+    def warning(
+        self,
+        node: KdlNode,
+        *,
+        message: str,
+        code: str = "",
+        hint: str = "",
+        label: str | None = None,
+        notes: list[str] | None = None,
+    ) -> None:
+        self.diagnostics.append(
+            ReadDiagnostic(
+                message=message,
+                severity=Severity.WARNING,
+                span=node.span,
+                path=self.path,
+                hint=hint,
+                code=code,
+                label=label,
+                notes=tuple(notes) if notes else (),
+            )
+        )
+
+    @property
+    def path(self) -> str:
+        return "/".join(self._path_segments)
+
+    def push(self, segment: str) -> None:
+        self._path_segments.append(segment)
+
+    def pop(self) -> None:
+        self._path_segments.pop()
+
+    def resolve_scalar_arg(self, arg: str) -> str | None:
+        info = self.defines.get(arg)
+        if info is not None and info.kind == DefineKind.SCALAR:
+            return info.value
+        return None
+
+    @property
+    def in_predicate(self) -> bool:
+        return self._predicate_depth > 0
+
+    @property
+    def in_assert(self) -> bool:
+        return self._predicate_depth > 0 and self._predicate_context == "assert"
+
+
+# lazy imports for forward refs
+from ssc_codegen.ast import (  # noqa: E402
+    JsonDef,
+    Struct,
+    TransformDef,
+)

{ssc_codegen-0.29.5 → ssc_codegen-0.30.0}/ssc_codegen/core/expressions.py

@@ -76,12 +76,12 @@ from ssc_codegen.ast import (
     XpathSelectAll,
 )
 from ssc_codegen.exceptions import BuildTimeError
-from ssc_codegen.kdl import KdlArg, KdlNode
+from kdlquery import KdlNode
 from ssc_codegen.regex_utils import normalize_regex_pattern
 from typing import cast
 
 from ssc_codegen.core.contexts import LintContext, ParseContext, WalkCtx
-from ssc_codegen.core.linting import (
+from ssc_codegen.core.linter import (
     lint_pipeline_op,
     lint_validate_css,
     lint_validate_xpath,
@@ -866,12 +866,7 @@ def _expr_len(
 def _expr_unique(
     node: KdlNode, parent: FieldLikeNode, ctx: ParseContext, lint: LintContext
 ):
-    keep_order = bool(
-        node.properties.get(
-            "keep-order",
-            KdlArg(value=False, span=node.span, is_identifier=False),
-        ).value
-    )
+    keep_order = bool(node.get_prop("keep-order") or False)
     return Unique(parent=parent, keep_order=keep_order)
 
 
@@ -917,11 +912,7 @@ def _expr_jsonify(
     node: KdlNode, parent: FieldLikeNode, ctx: ParseContext, lint: LintContext
 ):
     schema_name = str(node.args[0].value)
-    path = str(
-        node.properties.get(
-            "path", KdlArg(value="", span=node.span, is_identifier=False)
-        ).value
-    )
+    path = node.get_prop("path") or ""
     json_def = ctx.json_defs.get(schema_name)
     if json_def is None:
         lint.error(