ssc_codegen 0.8.4__tar.gz → 0.8.5__tar.gz

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ssc_codegen
-Version: 0.8.4
+Version: 0.8.5
 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
@@ -26,6 +26,7 @@ Requires-Dist: ichrome>=4.0.4
 Requires-Dist: jinja2>=3.1.6
 Requires-Dist: lxml>=5.3.0
 Requires-Dist: parsel>=1.10.0
+Requires-Dist: soupsieve>=2.6
 Requires-Dist: typer>=0.15.1
 Requires-Dist: typing-extensions; python_version < '3.11'
 Description-Content-Type: text/markdown

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ssc_codegen"
-version = "0.8.4"
+version = "0.8.5"
 description = "Python-dsl code converter to html parser for web scraping "
 readme = "README.md"
 requires-python = ">=3.10"
@@ -12,6 +12,7 @@ dependencies = [
     "jinja2>=3.1.6",
     "lxml>=5.3.0",
     "parsel>=1.10.0",
+    "soupsieve>=2.6",
     "typer>=0.15.1",
     "typing_extensions;python_version<'3.11'",
 ]
@@ -106,4 +107,4 @@ omit = [
     "*/.venv/*",
     "*/scripts/*",
     "*/tests/*"
-]
+]

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/ssc_codegen/__init__.py

@@ -7,7 +7,7 @@ from ssc_codegen.schema import ItemSchema, DictSchema, ListSchema, FlatListSchem
 
 setup_logger()
 
-VERSION = "0.8.4"
+VERSION = "0.8.5"
 
 
 class __MISSING(object):

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/ssc_codegen/ast_build/utils.py

@@ -95,9 +95,13 @@ def exec_module_code(path: str | Path, add_sys_path: bool = True) -> ModuleType:
     module = ModuleType("_")
     abs_path = path.resolve()
     # required for correct imports (eg: constants)
+    # TODO: calculate configs dirs depth
     if add_sys_path and str(abs_path.parent) not in sys.path:
         sys.path.append(str(abs_path.parent))
 
+    if add_sys_path and str(abs_path.parent.parent) not in sys.path:
+        sys.path.append(str(abs_path.parent.parent))
+
     code = Path(abs_path).read_text()
     exec(code, module.__dict__)
     return module

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/ssc_codegen/cli/main.py

@@ -100,6 +100,13 @@ def generate_code(
         LOGGER.info("TOGGLE debug generated tokens")
         converter.debug_instructions = True
     for file_cfg in ssc_files:
+        # check ssc-gen files
+        if (
+            file_cfg.suffix != ".py"
+            and "ssc_codegen" not in file_cfg.read_text()
+        ):
+            LOGGER.info("Skip %s (not ssc-gen config)", file_cfg.name)
+            continue
         name = file_cfg.name.split(".")[0]
         out_file = f"{prefix}{name}{suffix}"
         LOGGER.info("Make AST %s...", file_cfg.name)

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/ssc_codegen/converters/js_pure.py

@@ -146,6 +146,11 @@ def pre_part_doc(_node: StructPartDocMethod) -> str:
     return "_splitDoc(v) " + BRACKET_START
 
 
+@CONVERTER.post(StructPartDocMethod.kind)
+def post_part_doc(_node: StructPartDocMethod) -> str:
+    return BRACKET_END
+
+
 @CONVERTER(StructParser.kind)
 def pre_struct_parser(node: StructParser) -> str:
     name = node.kwargs["name"]

{ssc_codegen-0.8.4 → ssc_codegen-0.8.5}/ssc_codegen/selector_utils.py

@@ -5,6 +5,12 @@ from cssselect import HTMLTranslator, SelectorSyntaxError
 from lxml import etree
 from lxml.etree import XPathSyntaxError
 
+# cssselect not support CSS4 standard syntax
+# used for particularly provide CSS 4 API
+# https://facelessuser.github.io/soupsieve/selectors/
+from soupsieve import compile as css_compile
+from soupsieve.util import SelectorSyntaxError as SoupSieveSelectorSyntaxError
+
 
 def css_to_xpath(query: str, prefix: str = "descendant-or-self::") -> str:
     """convert css to XPATH selector"""
@@ -24,6 +30,10 @@ def xpath_to_css(query: str) -> str:
     EG FAIL:
         //p[contains(@class, "star-rating")]
 
+
+    NOTE:
+        cssselect not fully support CSS4 standard
+
     """
     # https://stackoverflow.com/a/18421383
     css = re.sub(
@@ -41,8 +51,8 @@ def xpath_to_css(query: str) -> str:
 
 def validate_css_query(query: str) -> None:
     try:
-        HTMLTranslator().css_to_xpath(query.strip('"'))
-    except SelectorSyntaxError:
+        css_compile(query.strip('"'))
+    except SoupSieveSelectorSyntaxError:
         # maybe is XPATH?
         with suppress(XPathSyntaxError):
             etree.XPath(query)  # type: ignore
@@ -57,10 +67,9 @@ def validate_xpath_query(query: str) -> None:
         etree.XPath(query.strip('"'))
         # etree.XPath accept CSS-like queries without throw exception, check it!
         is_css = False
-        with suppress(SelectorSyntaxError):
-            HTMLTranslator().css_to_xpath(query.strip('"'))
+        with suppress(SoupSieveSelectorSyntaxError):
+            css_compile(query.strip('"'))
             is_css = True
-
         if is_css:
             msg = f"`{query}` looks like CSS query, not XPATH"
             raise SelectorSyntaxError(msg)