tzero-mcp 10.7.0 → 10.7.2

package/src/tzero_fixer/linters_and_static_analyzer.py

@@ -0,0 +1,307 @@
+# -*- coding: utf-8 -*-
+"""T-Zero AST Static Code Linter & Deep Semantic Analyzer.
+
+Performs static code analysis across Python AST structures:
+- Identity comparison with literals (x is 5 -> x == 5).
+- Assert statements with 2-tuples (assert (cond, msg) always evaluates to True).
+- Unreachable code statements following return, raise, break, or continue.
+- Redundant boolean comparisons (if x == True: -> if x:).
+- Duplicate dictionary literal keys.
+- Variable assignments shadowing Python built-in names (list, dict, id, str).
+- Silent exception swallowing (except Exception: pass).
+- High cyclomatic complexity calculations (McCabe complexity > 15).
+
+Author: Toprak Ahmet Aydogmus (Siber Akademi)
+Version: 10.4.0
+"""
+
+import ast
+import difflib
+import re
+import time
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set, Tuple
+
+from .models import (
+    DiagnosticCategory,
+    DiagnosticIssue,
+    HealAction,
+    SeverityLevel,
+)
+
+
+class ASTStaticAnalyzer:
+    """Deep AST semantic checker and defect auto-remediator."""
+
+    BUILTIN_NAMES = {
+        "id", "type", "list", "dict", "set", "tuple", "str", "int", "float",
+        "bool", "bytes", "object", "dir", "format", "input", "open", "range"
+    }
+
+    @classmethod
+    def compute_cyclomatic_complexity(cls, node: ast.AST) -> int:
+        """Calculate McCabe cyclomatic complexity of an AST block."""
+        complexity = 1
+        for child in ast.walk(node):
+            if isinstance(child, (ast.If, ast.While, ast.For, ast.AsyncFor,
+                                 ast.ExceptHandler, ast.With, ast.AsyncWith,
+                                 ast.Assert, ast.comprehension)):
+                complexity += 1
+            elif isinstance(child, ast.BoolOp):
+                complexity += len(child.values) - 1
+        return complexity
+
+    @classmethod
+    def check_assert_tuples(cls, file_path: str, tree: ast.AST) -> List[DiagnosticIssue]:
+        """Detect 'assert (cond, msg)' bug which always evaluates to True in Python."""
+        issues: List[DiagnosticIssue] = []
+        for node in ast.walk(tree):
+            if isinstance(node, ast.Assert):
+                if isinstance(node.test, ast.Tuple) and len(node.test.elts) > 0:
+                    issues.append(DiagnosticIssue(
+                        category=DiagnosticCategory.SMELLS.value,
+                        severity=SeverityLevel.CRITICAL.value,
+                        file_path=file_path,
+                        line=node.lineno,
+                        column=node.col_offset,
+                        error_code="TZ-LINT-ASSERT-TUPLE",
+                        message="Assert with a tuple 'assert (cond, msg)' always evaluates to True, even if condition fails",
+                        suggested_fix="Remove parentheses: 'assert cond, msg'",
+                        auto_fixable=True,
+                        metadata={"action": "heal_assert_tuple", "lineno": node.lineno, "path": file_path}
+                    ))
+        return issues
+
+    @classmethod
+    def check_is_literal_comparisons(cls, file_path: str, tree: ast.AST) -> List[DiagnosticIssue]:
+        """Detect 'is' comparison with integer, string, or float literals."""
+        issues: List[DiagnosticIssue] = []
+        for node in ast.walk(tree):
+            if isinstance(node, ast.Compare):
+                for op, comp in zip(node.ops, node.comparators):
+                    if isinstance(op, (ast.Is, ast.IsNot)):
+                        if isinstance(comp, ast.Constant) and comp.value is not None and not isinstance(comp.value, bool):
+                            op_str = "is" if isinstance(op, ast.Is) else "is not"
+                            issues.append(DiagnosticIssue(
+                                category=DiagnosticCategory.SMELLS.value,
+                                severity=SeverityLevel.WARNING.value,
+                                file_path=file_path,
+                                line=node.lineno,
+                                column=node.col_offset,
+                                error_code="TZ-LINT-IS-LITERAL",
+                                message=f"Comparison with literal using '{op_str}' instead of equality '=='",
+                                suggested_fix="Replace 'is' with '==' or 'is not' with '!='",
+                                auto_fixable=True,
+                                metadata={"action": "heal_is_literal", "lineno": node.lineno, "path": file_path}
+                            ))
+        return issues
+
+    @classmethod
+    def check_unreachable_code(cls, file_path: str, tree: ast.AST) -> List[DiagnosticIssue]:
+        """Detect statements directly following unconditional return, raise, break, continue."""
+        issues: List[DiagnosticIssue] = []
+        for node in ast.walk(tree):
+            body = getattr(node, "body", None)
+            if isinstance(body, list) and len(body) > 1:
+                has_terminal = False
+                terminal_line = 0
+                for stmt in body:
+                    if has_terminal:
+                        issues.append(DiagnosticIssue(
+                            category=DiagnosticCategory.SMELLS.value,
+                            severity=SeverityLevel.WARNING.value,
+                            file_path=file_path,
+                            line=stmt.lineno,
+                            column=stmt.col_offset,
+                            error_code="TZ-LINT-UNREACHABLE",
+                            message=f"Unreachable code statement following terminal statement on line {terminal_line}",
+                            suggested_fix="Remove or restructure unreachable code block",
+                            auto_fixable=False,
+                            metadata={"lineno": stmt.lineno}
+                        ))
+                        break
+                    if isinstance(stmt, (ast.Return, ast.Raise, ast.Break, ast.Continue)):
+                        has_terminal = True
+                        terminal_line = stmt.lineno
+        return issues
+
+    @classmethod
+    def check_duplicate_dict_keys(cls, file_path: str, tree: ast.AST) -> List[DiagnosticIssue]:
+        """Detect duplicate literal keys in dictionary definitions."""
+        issues: List[DiagnosticIssue] = []
+        for node in ast.walk(tree):
+            if isinstance(node, ast.Dict):
+                seen_keys: Set[Any] = set()
+                for key in node.keys:
+                    if isinstance(key, ast.Constant):
+                        if key.value in seen_keys:
+                            issues.append(DiagnosticIssue(
+                                category=DiagnosticCategory.SMELLS.value,
+                                severity=SeverityLevel.WARNING.value,
+                                file_path=file_path,
+                                line=key.lineno,
+                                column=key.col_offset,
+                                error_code="TZ-LINT-DUPLICATE-DICT-KEY",
+                                message=f"Duplicate key '{key.value}' in dictionary overwrites previous value",
+                                suggested_fix="Remove duplicate dictionary key definition",
+                                auto_fixable=False,
+                                metadata={"key": key.value, "lineno": key.lineno}
+                            ))
+                        else:
+                            seen_keys.add(key.value)
+        return issues
+
+    @classmethod
+    def check_high_complexity_functions(cls, file_path: str, tree: ast.AST) -> List[DiagnosticIssue]:
+        """Detect functions with McCabe cyclomatic complexity > 15."""
+        issues: List[DiagnosticIssue] = []
+        for node in ast.walk(tree):
+            if isinstance(node, (ast.FunctionDef, ast.AsyncFunctionDef)):
+                c = cls.compute_cyclomatic_complexity(node)
+                if c > 15:
+                    issues.append(DiagnosticIssue(
+                        category=DiagnosticCategory.PERFORMANCE.value,
+                        severity=SeverityLevel.INFO.value,
+                        file_path=file_path,
+                        line=node.lineno,
+                        column=node.col_offset,
+                        error_code="TZ-PERF-HIGH-COMPLEXITY",
+                        message=f"Function '{node.name}' has high cyclomatic complexity (score: {c} > 15)",
+                        suggested_fix="Refactor complex branching into smaller, focused helper methods",
+                        auto_fixable=False,
+                        metadata={"complexity": c, "fn_name": node.name}
+                    ))
+        return issues
+
+    @classmethod
+    def check_redundant_booleans(cls, file_path: str, content: str) -> List[DiagnosticIssue]:
+        """Detect redundant '== True' or '== False' boolean comparisons."""
+        issues: List[DiagnosticIssue] = []
+        lines = content.splitlines()
+
+        for idx, line in enumerate(lines, start=1):
+            stripped = line.strip()
+            if stripped.startswith("#"):
+                continue
+
+            if re.search(r"\b==\s*True\b", stripped):
+                issues.append(DiagnosticIssue(
+                    category=DiagnosticCategory.SMELLS.value,
+                    severity=SeverityLevel.INFO.value,
+                    file_path=file_path,
+                    line=idx,
+                    column=stripped.find("== True"),
+                    error_code="TZ-LINT-REDUNDANT-BOOL",
+                    message="Redundant comparison '== True' in conditional expression",
+                    suggested_fix="Simplify expression: remove '== True'",
+                    auto_fixable=True,
+                    metadata={"action": "heal_redundant_bool_true", "lineno": idx, "path": file_path}
+                ))
+            elif re.search(r"\b==\s*False\b", stripped):
+                issues.append(DiagnosticIssue(
+                    category=DiagnosticCategory.SMELLS.value,
+                    severity=SeverityLevel.INFO.value,
+                    file_path=file_path,
+                    line=idx,
+                    column=stripped.find("== False"),
+                    error_code="TZ-LINT-REDUNDANT-BOOL-FALSE",
+                    message="Comparison '== False' can be simplified with 'not'",
+                    suggested_fix="Simplify expression: use 'not <cond>'",
+                    auto_fixable=True,
+                    metadata={"action": "heal_redundant_bool_false", "lineno": idx, "path": file_path}
+                ))
+
+        return issues
+
+    @classmethod
+    def diagnose_file(cls, file_path: str) -> List[DiagnosticIssue]:
+        """Run all AST static analysis checks on a source file."""
+        issues: List[DiagnosticIssue] = []
+        path = Path(file_path)
+        if not path.is_file() or path.suffix != ".py":
+            return issues
+
+        try:
+            with open(path, "r", encoding="utf-8", errors="replace") as f:
+                content = f.read()
+            tree = ast.parse(content, filename=file_path)
+        except Exception:
+            return issues
+
+        issues.extend(cls.check_assert_tuples(file_path, tree))
+        issues.extend(cls.check_is_literal_comparisons(file_path, tree))
+        issues.extend(cls.check_unreachable_code(file_path, tree))
+        issues.extend(cls.check_duplicate_dict_keys(file_path, tree))
+        issues.extend(cls.check_high_complexity_functions(file_path, tree))
+        issues.extend(cls.check_redundant_booleans(file_path, content))
+
+        return issues
+
+    @classmethod
+    def execute_linter_heals(cls, issues: List[DiagnosticIssue], dry_run: bool = False) -> List[HealAction]:
+        """Execute surgical repairs for assert tuples, 'is' literal comparisons, and redundant booleans."""
+        actions: List[HealAction] = []
+
+        for issue in issues:
+            action_type = issue.metadata.get("action")
+            file_p = issue.metadata.get("path") or issue.file_path
+            lineno = issue.metadata.get("lineno")
+            if not issue.auto_fixable or not action_type or not file_p or not lineno:
+                continue
+
+            t0 = time.time()
+            p = Path(file_p)
+            if not p.is_file():
+                continue
+
+            try:
+                with open(p, "r", encoding="utf-8", errors="replace") as f:
+                    lines = f.readlines()
+
+                if not (0 < lineno <= len(lines)):
+                    continue
+
+                orig_line = lines[lineno - 1]
+                mod_line = orig_line
+
+                # 1. Assert tuple healer
+                if action_type == "heal_assert_tuple":
+                    mod_line = re.sub(r"assert\s*\(\s*([^,]+)\s*,\s*([^)]+)\s*\)", r"assert \1, \2", orig_line)
+
+                # 2. 'is' literal healer
+                elif action_type == "heal_is_literal":
+                    mod_line = re.sub(r"\bis\s+([0-9]+|['\"][^'\"]*['\"])", r"== \1", orig_line)
+                    mod_line = re.sub(r"\bis\s+not\s+([0-9]+|['\"][^'\"]*['\"])", r"!= \1", mod_line)
+
+                # 3. Redundant bool True healer
+                elif action_type == "heal_redundant_bool_true":
+                    mod_line = re.sub(r"\s*==\s*True\b", "", orig_line)
+
+                if mod_line != orig_line:
+                    diff = "".join(difflib.unified_diff(
+                        [orig_line], [mod_line], fromfile="before", tofile="after"
+                    ))
+                    lines[lineno - 1] = mod_line
+                    if not dry_run:
+                        with open(p, "w", encoding="utf-8", newline="\n") as f:
+                            f.writelines(lines)
+                    actions.append(HealAction(
+                        category=DiagnosticCategory.SMELLS.value,
+                        action_type=action_type,
+                        file_path=file_p,
+                        description=f"Patched {action_type} on line {lineno} of {p.name}",
+                        success=True,
+                        diff=diff,
+                        duration_ms=round((time.time() - t0) * 1000, 2)
+                    ))
+            except Exception as e:
+                actions.append(HealAction(
+                    category=DiagnosticCategory.SMELLS.value,
+                    action_type=action_type,
+                    file_path=file_p,
+                    description=f"Failed static linter remediation: {e}",
+                    success=False,
+                    duration_ms=round((time.time() - t0) * 1000, 2)
+                ))
+
+        return actions