ssc_codegen 0.12.0__tar.gz → 0.12.1__tar.gz

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ssc_codegen
-Version: 0.12.0
+Version: 0.12.1
 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
@@ -26,7 +26,6 @@ Requires-Dist: colorama>=0.4.6; sys_platform == 'win32'
 Requires-Dist: cssselect>=1.2.0
 Requires-Dist: httpx>=0.28.1
 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
@@ -66,7 +65,7 @@ Current support converters
 | ...                                      | parsel                                 | Y     | Y    | N    | ...                                   | ...                  |
 | ...                                      | selectolax (lexbor)                    | N     | Y    | N    | ...                                   | ...                  |
 | ...                                      | lxml                                   | Y     | Y    | N    | ...                                   | ...                  |
-| js (ES6)`2`                              | pure (firefox/chrome extension/nodejs) | Y     | Y    | Y    | Array, Map`3`                         | prettier             |
+| js (ES6)`2`                              | pure (firefox/chrome extension/nodejs) | Y     | Y    | Y    | JSDoc                                 | prettier             |
 | go (1.10+) **(UNSTABLE)**                | goquery, gjson (`4`)                   | N     | Y    | N    | struct(+json anchors), array, map     | gofmt                |
 | lua (5.2+), luajit(2+) **(UNSTABLE)**`5` | lua-htmlparser, lrexlib(opt), dkjson   | N     | Y    | N    | EmmyLua                               | LuaFormatter         |
 

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/README.md

@@ -29,7 +29,7 @@ Current support converters
 | ...                                      | parsel                                 | Y     | Y    | N    | ...                                   | ...                  |
 | ...                                      | selectolax (lexbor)                    | N     | Y    | N    | ...                                   | ...                  |
 | ...                                      | lxml                                   | Y     | Y    | N    | ...                                   | ...                  |
-| js (ES6)`2`                              | pure (firefox/chrome extension/nodejs) | Y     | Y    | Y    | Array, Map`3`                         | prettier             |
+| js (ES6)`2`                              | pure (firefox/chrome extension/nodejs) | Y     | Y    | Y    | JSDoc                                 | prettier             |
 | go (1.10+) **(UNSTABLE)**                | goquery, gjson (`4`)                   | N     | Y    | N    | struct(+json anchors), array, map     | gofmt                |
 | lua (5.2+), luajit(2+) **(UNSTABLE)**`5` | lua-htmlparser, lrexlib(opt), dkjson   | N     | Y    | N    | EmmyLua                               | LuaFormatter         |
 

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ssc_codegen"
-version = "0.12.0"
+version = "0.12.1"
 description = "Python-dsl code converter to html parser for web scraping "
 readme = "README.md"
 requires-python = ">=3.10"
@@ -10,7 +10,6 @@ dependencies = [
     "cssselect>=1.2.0",
     "httpx>=0.28.1",
     "ichrome>=4.0.4",
-    "jinja2>=3.1.6",
     "lxml>=5.3.0",
     "parsel>=1.10.0",
     "soupsieve>=2.6",
@@ -61,8 +60,10 @@ dev = [
     "coverage>=7.6.12",
     "httpx>=0.28.1",
     "mypy>=1.14.1",
+    "parsel>=1.10.0",
     "pytest>=8.3.4",
     "ruff>=0.9.3",
+    "selectolax>=0.3.27",
 ]
 # python generated code tests
 tests = [

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/ssc_codegen/__init__.py

@@ -7,7 +7,7 @@ from ssc_codegen.schema import ItemSchema, DictSchema, ListSchema, FlatListSchem
 
 setup_logger()
 
-VERSION = "0.12.0"
+VERSION = "0.12.1"
 
 
 class __MISSING(object):

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/ssc_codegen/cli/main.py

@@ -478,7 +478,7 @@ def parse_from_url(
     ] = True,
 ) -> None:
     from ssc_codegen.compiler import Compiler
-    from ssc_codegen.converters.py_parsel import CONVERTER
+    from ssc_codegen.converters.py_bs4 import CONVERTER
 
     cls_target, schema_config = _validate_parser_target(cls_target)
 

{ssc_codegen-0.12.0 → ssc_codegen-0.12.1}/ssc_codegen/converters/js_pure.py

@@ -149,19 +149,7 @@ from ssc_codegen.converters.helpers import (
     is_first_node_cond,
     is_prev_node_atomic_cond,
 )
-from ssc_codegen.converters.templates.js_pure import (
-    HELPER_FUNCTIONS,
-    J2_STRUCT_INIT,
-    J2_PRE_LIST_STR_TRIM,
-    J2_PRE_STR_LEFT_TRIM,
-    J2_PRE_LIST_STR_LEFT_TRIM,
-    J2_PRE_STR_RIGHT_TRIM,
-    J2_PRE_LIST_STR_RIGHT_TRIM,
-    J2_PRE_XPATH,
-    J2_PRE_XPATH_ALL,
-    J2_PRE_STR_TRIM,
-    J2_IS_XPATH,
-)
+from ssc_codegen.converters.templates.js_pure import HELPER_FUNCTIONS
 from ssc_codegen.str_utils import (
     wrap_backtick,
     to_upper_camel_case,
@@ -299,7 +287,17 @@ def pre_parse_field(node: StructFieldMethod) -> str:
 
 @CONVERTER(StructInitMethod.kind)
 def pre_struct_init(_node: StructInitMethod) -> str:
-    return J2_STRUCT_INIT
+    code = [
+        "constructor(doc){",
+        "if (typeof doc === 'string'){",
+        "this._doc = new DOMParser().parseFromString(doc, 'text/html');",
+        "} else if (doc instanceof Document || doc instanceof Element){",
+        "this._doc = doc.cloneNode(true);",
+        "} else {",
+        'throw new Error("Invalid input: Expected a Document, Element, or string");}',
+        "}",
+    ]
+    return "\n".join(code)
 
 
 JSDOC_TYPES = {
@@ -576,42 +574,88 @@ def pre_list_str_fmt(node: ExprListStringFormat) -> str:
 def pre_str_trim(node: ExprStringTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_STR_TRIM.render(nxt=nxt, prv=prv, substr=substr)
+    code = [
+        f"let {nxt} = (function (str, chars) {{",
+        "return str.replace(new RegExp(`^[${chars}]+|[${chars}]+$`, 'g'), '');",
+        f"}})({prv}, {substr});",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprListStringTrim.kind)
 def pre_list_str_trim(node: ExprListStringTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_LIST_STR_TRIM.render(prv=prv, nxt=nxt, substr=substr)
+    code = [
+        f"let {nxt} = {prv}.map(e =>",
+        "(function (str, chars) {",
+        "return str.replace(new RegExp(`^[${chars}]+|[${chars}]+$`, 'g'), '');",
+        f"}})(e, {substr})",
+        ");",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprStringLeftTrim.kind)
 def pre_str_left_trim(node: ExprStringLeftTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_STR_LEFT_TRIM.render(prv=prv, nxt=nxt, substr=substr)
+    """
+    let {{ nxt }} = (function (str, chars) {
+        return str.replace(new RegExp(`^[${chars}]+`, 'g'), '');
+    })({{ prv }}, {{ substr }});
+    """
+    code = [
+        f"let {nxt} = (function (str, chars) {{",
+        "return str.replace(new RegExp(`^[${chars}]+`, 'g'), '');",
+        f"}})({prv}, {substr});",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprListStringLeftTrim.kind)
 def pre_list_str_left_trim(node: ExprListStringLeftTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_LIST_STR_LEFT_TRIM.render(prv=prv, nxt=nxt, substr=substr)
+
+    code = [
+        f"let {nxt} = {prv}.map(e =>",
+        "(function (str, chars) {",
+        "return str.replace(new RegExp(`^[${chars}]+`, 'g'), '');",
+        f"}})(e, {substr})",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprStringRightTrim.kind)
 def pre_str_right_trim(node: ExprStringRightTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_STR_RIGHT_TRIM.render(prv=prv, nxt=nxt, substr=substr)
+    """
+    let {{ nxt }} = (function (str, chars) {
+        return str.replace(new RegExp(`[${chars}]+$`, 'g'), '');
+    })({{ prv }}, {{ substr }});
+    """
+    code = [
+        f"let {nxt} = (function (str, chars) {{",
+        "return str.replace(new RegExp(`[${chars}]+$`, 'g'), '');",
+        f"}})({prv}, {substr});",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprListStringRightTrim.kind)
 def pre_list_str_right_trim(node: ExprListStringRightTrim) -> str:
     prv, nxt = prev_next_var(node)
     substr = js_get_classvar_hook_or_value(node, "substr")
-    return J2_PRE_LIST_STR_RIGHT_TRIM.render(prv=prv, nxt=nxt, substr=substr)
+    code = [
+        f"let {nxt} = {prv}.map(e =>",
+        "(function (str, chars) {",
+        "return str.replace(new RegExp(`[${chars}]+$`, 'g'), '');",
+        f"}})(e, {substr})",
+        ");",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprStringSplit.kind)
@@ -814,11 +858,14 @@ def pre_is_xpath(node: ExprIsXpath) -> str:
     prv, nxt = prev_next_var(node)
     query = js_get_classvar_hook_or_value(node, "query")
     msg = js_get_classvar_hook_or_value(node, "msg")
-
-    expr = J2_IS_XPATH.render(prv=prv, nxt=nxt, msg=msg, query=query)
-    if is_last_var_no_ret(node):
-        return expr
-    return expr + f"let {nxt} = {prv};"
+    code = [
+        f"if (document.evaluate({query}, {prv}, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null).singleNodeValue === null) {{",
+        f"throw new Error({msg});",
+        "}",
+    ]
+    if not is_last_var_no_ret(node):
+        code.append(f"let {nxt} = {prv};")
+    return "\n".join(code)
 
 
 @CONVERTER(ExprToInt.kind)
@@ -884,7 +931,13 @@ def pre_css_all(node: ExprCssAll) -> str:
 def pre_xpath(node: ExprXpath) -> str:
     prv, nxt = prev_next_var(node)
     query = js_get_classvar_hook_or_value(node, "query")
-    return J2_PRE_XPATH.render(prv=prv, nxt=nxt, query=query)
+
+    code = [
+        f"let {nxt} = document.evaluate(",
+        f"{query}, {prv}, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null",
+        ").singleNodeValue;",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprXpathAll.kind)
@@ -892,9 +945,15 @@ def pre_xpath_all(node: ExprXpathAll) -> str:
     prv, nxt = prev_next_var(node)
     query = js_get_classvar_hook_or_value(node, "query")
     snapshot_var = f"s{nxt}"
-    return J2_PRE_XPATH_ALL.render(
-        prv=prv, nxt=nxt, query=query, snapshot_var=snapshot_var
-    )
+    code = [
+        f"let {snapshot_var} = {prv}.evaluate(",
+        f"{query}, document, null, XPathResult.ORDERED_NODE_SNAPSHOT_TYPE, null",
+        ");",
+        f"let {nxt} = Array.from({{ length: {snapshot_var}.snapshotLength }}, (_, i) => ",
+        f"{snapshot_var}.snapshotItem(i)",
+        ");",
+    ]
+    return "\n".join(code)
 
 
 @CONVERTER(ExprGetHtmlAttr.kind)

ssc_codegen-0.12.1/ssc_codegen/converters/templates/js_pure.py

@@ -0,0 +1,38 @@
+# used functions instead consts arrow functions for allow rewrite it wout errors
+HELPER_FUNCTIONS = r"""
+function sscUnescape(v) {
+  return v
+    .replace(/&/g, '&')
+    .replace(/&lt;/g, '<')
+    .replace(/&gt;/g, '>')
+    .replace(/&quot;/g, '"')
+    .replace(/&#039;/g, "'")
+    .replace(/&#x2F;/g, '/')
+    .replace(/&nbsp;/g, ' ')
+    .replace(/&#x([0-9a-fA-F]+);/g, function(_, hex) {
+      return String.fromCharCode(parseInt(hex, 16));
+    })
+    .replace(/\\u([0-9a-fA-F]{4})/g, function(_, hex) {
+      return String.fromCharCode(parseInt(hex, 16));
+    })
+    .replace(/\\x([0-9a-fA-F]{2})/g, function(_, hex) {
+      return String.fromCharCode(parseInt(hex, 16));
+    })
+    .replace(/\\([bfnrt])/g, function(_, ch) {
+      return { b: '\b', f: '\f', n: '\n', r: '\r', t: '\t' }[ch];
+    });
+}
+
+function sscRmPrefix(v, p) {
+  return v.startsWith(p) ? v.slice(p.length) : v;
+}
+
+function sscRmSuffix(v, s) {
+  return v.endsWith(s) ? v.slice(0, -s.length) : v;
+}
+
+function sscRmPrefixSuffix(v, p, s) {
+  return sscRmSuffix(sscRmPrefix(v, p), s);
+}
+ 
+"""

ssc_codegen-0.12.0/ssc_codegen/converters/templates/js_pure.py

@@ -1,135 +0,0 @@
-from jinja2 import Template
-
-# used functions instead consts arrow functions for allow rewrite it wout errors
-HELPER_FUNCTIONS = r"""
-function sscUnescape(v) {
-  return v
-    .replace(/&/g, '&')
-    .replace(/&lt;/g, '<')
-    .replace(/&gt;/g, '>')
-    .replace(/&quot;/g, '"')
-    .replace(/&#039;/g, "'")
-    .replace(/&#x2F;/g, '/')
-    .replace(/&nbsp;/g, ' ')
-    .replace(/&#x([0-9a-fA-F]+);/g, function(_, hex) {
-      return String.fromCharCode(parseInt(hex, 16));
-    })
-    .replace(/\\u([0-9a-fA-F]{4})/g, function(_, hex) {
-      return String.fromCharCode(parseInt(hex, 16));
-    })
-    .replace(/\\x([0-9a-fA-F]{2})/g, function(_, hex) {
-      return String.fromCharCode(parseInt(hex, 16));
-    })
-    .replace(/\\([bfnrt])/g, function(_, ch) {
-      return { b: '\b', f: '\f', n: '\n', r: '\r', t: '\t' }[ch];
-    });
-}
-
-function sscRmPrefix(v, p) {
-  return v.startsWith(p) ? v.slice(p.length) : v;
-}
-
-function sscRmSuffix(v, s) {
-  return v.endsWith(s) ? v.slice(0, -s.length) : v;
-}
-
-function sscRmPrefixSuffix(v, p, s) {
-  return sscRmSuffix(sscRmPrefix(v, p), s);
-}
- 
-"""
-
-J2_STRUCT_INIT = """
-constructor(doc){
-if (typeof doc === 'string'){
-this._doc = new DOMParser().parseFromString(doc, 'text/html');
-} else if (doc instanceof Document || doc instanceof Element){
-this._doc = doc.cloneNode(true);
-} else {
-throw new Error("Invalid input: Expected a Document, Element, or string");}
-}
-"""
-
-J2_PRE_STR_TRIM = Template(
-    """
-    let {{ nxt }} = (function (str, chars) {
-        return str.replace(new RegExp(`^[${chars}]+|[${chars}]+$`, 'g'), '');
-    })({{ prv }}, {{ substr }});
-    """
-)
-
-
-J2_PRE_LIST_STR_TRIM = Template(
-    """
-    let {{ nxt }} = {{ prv }}.map(e =>
-        (function (str, chars) {
-            return str.replace(new RegExp(`^[${chars}]+|[${chars}]+$`, 'g'), '');
-        })(e, {{ substr }})
-    ); 
-    """
-)
-
-J2_PRE_STR_LEFT_TRIM = Template(
-    """
-    let {{ nxt }} = (function (str, chars) {
-        return str.replace(new RegExp(`^[${chars}]+`, 'g'), '');
-    })({{ prv }}, {{ substr }});
-    """
-)
-
-J2_PRE_LIST_STR_LEFT_TRIM = Template(
-    """
-    let {{ nxt }} = {{ prv }}.map(e =>
-        (function (str, chars) {
-            return str.replace(new RegExp(`^[${chars}]+`, 'g'), '');
-        })(e, {{ substr }})
-    );
-    """
-)
-
-J2_PRE_STR_RIGHT_TRIM = Template(
-    """
-    let {{ nxt }} = (function (str, chars) {
-        return str.replace(new RegExp(`[${chars}]+$`, 'g'), '');
-    })({{ prv }}, {{ substr }});
-    """
-)
-
-J2_PRE_LIST_STR_RIGHT_TRIM = Template(
-    """
-    let {{ nxt }} = {{ prv }}.map(e =>
-        (function (str, chars) {
-            return str.replace(new RegExp(`[${chars}]+$`, 'g'), '');
-        })(e, {{ substr }})
-    );
-    """
-)
-
-J2_PRE_XPATH = Template("""
-let {{ nxt }} = document.evaluate(
-    "{{ query }}", {{ prv }}, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null
-).singleNodeValue;
-""")
-
-J2_PRE_XPATH_ALL = Template("""
-let {{ snapshot_var }} = {{ prv }}.evaluate(
-    "{{ query }}", document, null, XPathResult.ORDERED_NODE_SNAPSHOT_TYPE, null
-); 
-let {{ nxt }} = Array.from({ length: {{ snapshot_var }}.snapshotLength }, (_, i) => 
-    {{ snapshot_var }}.snapshotItem(i)
-);
-""")
-
-
-J2_IS_XPATH = Template("""
-if (document.evaluate("{{ query }}", {{ prv }}, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null).singleNodeValue === null) throw new Error({{ msg }});
-""")
-# TODO: use in covverter
-JS_HAS_ATTR = Template("""
-if (!{{prv}}?.hasAttribute({{key}}) throw new Error({{ msg }});
-""")
-
-
-JS_HAS_ATTR_ALL = Template("""
-if (!{{prv}}.every(e => e?.hasAttribute({{key}}));}) throw new Error({{ msg }});
-""")