code2llm 0.5.171__tar.gz → 0.5.173__tar.gz

{code2llm-0.5.171/code2llm.egg-info → code2llm-0.5.173}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code2llm
-Version: 0.5.171
+Version: 0.5.173
 Summary: High-performance Python code flow analysis with optimized TOON format - CFG, DFG, call graphs, and intelligent code queries
 Home-page: https://github.com/wronai/stts
 Author: Tom Sapletta
@@ -32,6 +32,7 @@ Requires-Dist: numpy>=1.20
 Requires-Dist: jinja2>=3.0
 Requires-Dist: radon>=5.1
 Requires-Dist: astroid>=3.0
+Requires-Dist: psutil>=5.9
 Requires-Dist: code2logic
 Requires-Dist: vulture>=2.10
 Requires-Dist: tiktoken>=0.5
@@ -66,13 +67,13 @@ Dynamic: requires-python
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.171-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
-![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.96-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-91.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.173-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
+![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.73-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-93.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
 
-- 🤖 **LLM usage:** $17.9588 (245 commits)
-- 👤 **Human dev:** ~$9180 (91.8h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $17.7289 (247 commits)
+- 👤 **Human dev:** ~$9380 (93.8h @ $100/h, 30min dedup)
 
-Generated on 2026-07-05 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
+Generated on 2026-07-06 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
 
 ---
 

{code2llm-0.5.171 → code2llm-0.5.173}/README.md

@@ -2,13 +2,13 @@
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.171-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
-![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.96-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-91.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.173-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
+![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.73-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-93.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
 
-- 🤖 **LLM usage:** $17.9588 (245 commits)
-- 👤 **Human dev:** ~$9180 (91.8h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $17.7289 (247 commits)
+- 👤 **Human dev:** ~$9380 (93.8h @ $100/h, 30min dedup)
 
-Generated on 2026-07-05 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
+Generated on 2026-07-06 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
 
 ---
 

code2llm-0.5.173/VERSION

@@ -0,0 +1 @@
+0.5.173

{code2llm-0.5.171 → code2llm-0.5.173}/code2llm/__init__.py

@@ -8,7 +8,7 @@ Includes NLP Processing Pipeline for query normalization, intent matching,
 and entity resolution with multilingual support.
 """
 
-__version__ = "0.5.171"
+__version__ = "0.5.173"
 __author__ = "STTS Project"
 
 # Core analysis components (lightweight, always needed)

{code2llm-0.5.171 → code2llm-0.5.173}/code2llm/analysis/smells.py

@@ -131,18 +131,34 @@ class SmellDetector:
         return smells
 
     def _detect_data_clumps(self) -> List[CodeSmell]:
-        """Detect 3+ variables frequently passed together."""
+        """Detect 3+ variables frequently passed together.
+
+        Grouped by (file, argument-set), not argument-set alone: a common
+        constructor shape like `(self, project_root, config)` is a
+        conventional pattern reused independently by unrelated classes
+        across unrelated files (any two classes that each take a project
+        root and a config object) -- that's coincidence, not evidence the
+        two classes should share a bundled parameter object. Grouping by
+        bare argument-set alone falsely flagged such coincidences as data
+        clumps (reproduced: `wup.anomaly_detector.AnomalyDetector.__init__`
+        and `wup.testql_monitor.TestQLMonitor.__init__` -- two unrelated
+        classes in unrelated files -- flagged as clumped on `project_root,
+        self, config`). A genuine data clump -- args that should become one
+        object -- is a same-file (usually same-class) phenomenon; scoping
+        by file preserves that signal while dropping cross-file
+        coincidental-signature noise.
+        """
         smells = []
         # Simplified: find functions with same 3+ arguments
-        arg_sets = {}  # frozenset(args) -> List[func_names]
+        arg_sets = {}  # (file, frozenset(args)) -> List[func_names]
         for func_name, func_info in self.result.functions.items():
             if len(func_info.args) >= 3:
-                args = frozenset(func_info.args)
-                if args not in arg_sets:
-                    arg_sets[args] = []
-                arg_sets[args].append(func_name)
+                key = (func_info.file, frozenset(func_info.args))
+                if key not in arg_sets:
+                    arg_sets[key] = []
+                arg_sets[key].append(func_name)
 
-        for args, funcs in arg_sets.items():
+        for (file, args), funcs in arg_sets.items():
             if len(funcs) >= 2:
                 for func_name in funcs:
                     func_info = self.result.functions[func_name]
@@ -163,16 +179,33 @@ class SmellDetector:
         return smells
 
     def _detect_shotgun_surgery(self) -> List[CodeSmell]:
-        """Detect variables whose mutation requires changes across many functions."""
+        """Detect variables whose mutation requires changes across many functions.
+
+        Grouped by (file, variable), not variable name alone: a generic local
+        name like `args`, `data`, or `client` is conventionally reused as an
+        independent local in unrelated functions across unrelated files/
+        packages (every argparse `main()` has an `args`) -- that's naming
+        coincidence, not a variable whose logic is genuinely coupled across
+        call sites. Grouping by bare name alone falsely flagged every such
+        coincidence as "shotgun surgery" (reproduced: 40+ false-positive
+        tickets in a single monorepo, e.g. `args` in a 40-line, one-function
+        CLI file reported as "spans 8 functions" -- the other 7 were
+        unrelated `args` locals in unrelated files elsewhere in the tree).
+        Real shotgun surgery -- one variable/attribute whose change genuinely
+        ripples across many functions -- is a same-file (or same-class)
+        phenomenon; scoping by file preserves that signal while dropping the
+        cross-file name-collision noise.
+        """
         smells = []
-        var_mutators = {}  # variable -> set(functions)
+        var_mutators = {}  # (file, variable) -> set(functions)
 
         for mutation in self.result.mutations:
-            if mutation.variable not in var_mutators:
-                var_mutators[mutation.variable] = set()
-            var_mutators[mutation.variable].add(mutation.scope)
+            key = (mutation.file, mutation.variable)
+            if key not in var_mutators:
+                var_mutators[key] = set()
+            var_mutators[key].add(mutation.scope)
 
-        for var, funcs in var_mutators.items():
+        for (file, var), funcs in var_mutators.items():
             if len(funcs) >= 5:
                 # Find a representative function to report the smell
                 func_name = list(funcs)[0]
@@ -188,10 +221,11 @@ class SmellDetector:
                         line=func_info.line,
                         severity=0.8,
                         description=(
-                            f"Mutation of variable '{var}' spans {len(funcs)} functions."
+                            f"Mutation of variable '{var}' spans {len(funcs)} functions"
+                            f" in {file}."
                             " Changing this logic requires work in many places."
                         ),
-                        context={"variable": var, "affected_functions": list(funcs)},
+                        context={"variable": var, "file": file, "affected_functions": list(funcs)},
                     )
                 )
         return smells

{code2llm-0.5.171 → code2llm-0.5.173}/code2llm/nlp/__init__.py

@@ -4,7 +4,7 @@ Provides query normalization, intent matching, and entity resolution
 with multilingual support and fuzzy matching.
 """
 
-__version__ = "0.5.171"
+__version__ = "0.5.173"
 
 from .pipeline import NLPPipeline
 from .normalization import QueryNormalizer

{code2llm-0.5.171 → code2llm-0.5.173/code2llm.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code2llm
-Version: 0.5.171
+Version: 0.5.173
 Summary: High-performance Python code flow analysis with optimized TOON format - CFG, DFG, call graphs, and intelligent code queries
 Home-page: https://github.com/wronai/stts
 Author: Tom Sapletta
@@ -32,6 +32,7 @@ Requires-Dist: numpy>=1.20
 Requires-Dist: jinja2>=3.0
 Requires-Dist: radon>=5.1
 Requires-Dist: astroid>=3.0
+Requires-Dist: psutil>=5.9
 Requires-Dist: code2logic
 Requires-Dist: vulture>=2.10
 Requires-Dist: tiktoken>=0.5
@@ -66,13 +67,13 @@ Dynamic: requires-python
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.171-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
-![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.96-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-91.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.173-blue) ![Python](https://img.shields.io/badge/python-3.9+-blue) ![License](https://img.shields.io/badge/license-Apache--2.0-green)
+![AI Cost](https://img.shields.io/badge/AI%20Cost-$17.73-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-93.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fdeep%2Fdeep--v4--pro-lightgrey)
 
-- 🤖 **LLM usage:** $17.9588 (245 commits)
-- 👤 **Human dev:** ~$9180 (91.8h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $17.7289 (247 commits)
+- 👤 **Human dev:** ~$9380 (93.8h @ $100/h, 30min dedup)
 
-Generated on 2026-07-05 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
+Generated on 2026-07-06 using [openrouter/deep/deep-v4-pro](https://openrouter.ai/deep/deep-v4-pro)
 
 ---
 

{code2llm-0.5.171 → code2llm-0.5.173}/code2llm.egg-info/SOURCES.txt

@@ -202,4 +202,6 @@ tests/test_project_toon_export.py
 tests/test_prompt_engine.py
 tests/test_prompt_txt.py
 tests/test_refactoring_engine.py
+tests/test_smells_data_clumps.py
+tests/test_smells_shotgun_surgery.py
 tests/test_toon_v2.py

{code2llm-0.5.171 → code2llm-0.5.173}/code2llm.egg-info/requires.txt

@@ -5,6 +5,7 @@ numpy>=1.20
 jinja2>=3.0
 radon>=5.1
 astroid>=3.0
+psutil>=5.9
 code2logic
 vulture>=2.10
 tiktoken>=0.5

{code2llm-0.5.171 → code2llm-0.5.173}/pyproject.toml

@@ -4,10 +4,10 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "code2llm"
-version = "0.5.171"
+version = "0.5.173"
 description = "High-performance Python code flow analysis with optimized TOON format - CFG, DFG, call graphs, and intelligent code queries"
 readme = "README.md"
-requires-python = ">=3.8"
+requires-python = ">=3.10"
 authors = [
     {name = "Tom Sapletta", email = "tom@sapletta.com"},
 ]
@@ -45,6 +45,7 @@ dependencies = [
     "jinja2>=3.0",
     "radon>=5.1",
     "astroid>=3.0",
+    "psutil>=5.9",
     "code2logic",
     "vulture>=2.10",
     "tiktoken>=0.5",

code2llm-0.5.173/tests/test_smells_data_clumps.py

@@ -0,0 +1,83 @@
+"""Tests for SmellDetector._detect_data_clumps.
+
+2026-07-06 incident (same root cause as the Shotgun Surgery fix): grouping
+functions by their argument-name set alone, ignoring file, meant a common
+constructor shape like `(self, project_root, config)` -- conventionally
+reused independently by unrelated classes across unrelated files -- falsely
+triggered "Data Clump". Reproduced live:
+`wup.anomaly_detector.AnomalyDetector.__init__` and
+`wup.testql_monitor.TestQLMonitor.__init__` are two unrelated classes in
+unrelated files, flagged as clumped on `project_root, self, config` purely
+because both happen to take a project root and a config object.
+"""
+
+from __future__ import annotations
+
+from code2llm.analysis.smells import SmellDetector
+from code2llm.core.models import AnalysisResult, FunctionInfo
+
+
+def _function(name: str, file: str, args: list[str], line: int = 1) -> FunctionInfo:
+    return FunctionInfo(
+        name=name, qualified_name=f"{file}::{name}", file=file, line=line, args=args
+    )
+
+
+def test_same_args_across_unrelated_files_is_not_a_data_clump():
+    """The core regression: two unrelated classes in unrelated files each
+    taking (self, project_root, config) must NOT be flagged."""
+    result = AnalysisResult()
+    result.functions["a::AnomalyDetector.__init__"] = _function(
+        "AnomalyDetector.__init__", "wup/anomaly_detector.py", ["self", "project_root", "config"]
+    )
+    result.functions["b::TestQLMonitor.__init__"] = _function(
+        "TestQLMonitor.__init__", "wup/testql_monitor.py", ["self", "project_root", "config"]
+    )
+
+    smells = SmellDetector(result)._detect_data_clumps()
+
+    assert smells == []
+
+
+def test_genuine_same_file_data_clump_still_detected():
+    """Two methods in the SAME file sharing a 3+ arg signature is the real
+    smell and must still be caught."""
+    result = AnalysisResult()
+    file = "core/session.py"
+    result.functions["core.session::start"] = _function(
+        "start", file, ["self", "user_id", "token", "scope"]
+    )
+    result.functions["core.session::refresh"] = _function(
+        "refresh", file, ["self", "user_id", "token", "scope"]
+    )
+
+    smells = SmellDetector(result)._detect_data_clumps()
+
+    assert len(smells) == 2
+    assert all(s.name == "Data Clump: self, user_id, token, scope"
+               or set(s.context["clump"]) == {"self", "user_id", "token", "scope"}
+               for s in smells)
+    assert all(s.file == file for s in smells)
+
+
+def test_single_function_with_matching_args_in_other_file_not_flagged():
+    """Only one function per file matches the arg-set -- no clump within
+    either file, and cross-file coincidence must not count."""
+    result = AnalysisResult()
+    result.functions["a::f"] = _function("f", "a.py", ["self", "x", "y", "z"])
+    result.functions["b::g"] = _function("g", "b.py", ["self", "x", "y", "z"])
+
+    smells = SmellDetector(result)._detect_data_clumps()
+
+    assert smells == []
+
+
+def test_fewer_than_three_args_never_counted():
+    result = AnalysisResult()
+    file = "core/session.py"
+    result.functions["core.session::start"] = _function("start", file, ["self", "user_id"])
+    result.functions["core.session::refresh"] = _function("refresh", file, ["self", "user_id"])
+
+    smells = SmellDetector(result)._detect_data_clumps()
+
+    assert smells == []

code2llm-0.5.173/tests/test_smells_shotgun_surgery.py

@@ -0,0 +1,97 @@
+"""Tests for SmellDetector._detect_shotgun_surgery.
+
+2026-07-06 incident: grouping mutations by variable *name* alone (ignoring
+file) meant a generic local name like `args`/`data`/`client` -- conventionally
+reused as an independent local in unrelated functions across unrelated
+files -- falsely triggered "Shotgun Surgery" whenever 5+ unrelated functions
+anywhere in the analyzed tree happened to share that name. Reproduced live:
+40+ false-positive tickets in one monorepo, including `args` in a 40-line,
+one-function CLI file reported as "spans 8 functions" (the other 7 were
+unrelated `args` locals in unrelated files elsewhere in the tree).
+"""
+
+from __future__ import annotations
+
+from code2llm.analysis.smells import SmellDetector
+from code2llm.core.models import AnalysisResult, FunctionInfo, Mutation
+
+
+def _function(name: str, file: str, line: int = 1) -> FunctionInfo:
+    return FunctionInfo(name=name, qualified_name=f"{file}::{name}", file=file, line=line)
+
+
+def _mutation(variable: str, file: str, scope: str, line: int = 1) -> Mutation:
+    return Mutation(
+        variable=variable, file=file, line=line, type="assign", scope=scope, context=""
+    )
+
+
+def test_same_name_across_unrelated_files_is_not_shotgun_surgery():
+    """The core regression: `args` reused independently in 6 unrelated
+    files/functions must NOT be flagged -- no single file has more than one
+    mutating function."""
+    result = AnalysisResult()
+    for i in range(6):
+        file = f"pkg{i}/cli.py"
+        scope = f"pkg{i}.cli::main"
+        result.functions[scope] = _function("main", file)
+        result.mutations.append(_mutation("args", file, scope))
+
+    smells = SmellDetector(result)._detect_shotgun_surgery()
+
+    assert smells == []
+
+
+def test_genuine_same_file_shotgun_surgery_still_detected():
+    """5+ functions *within the same file* mutating the same variable is the
+    real smell and must still be caught."""
+    result = AnalysisResult()
+    file = "core/session.py"
+    for i in range(5):
+        scope = f"core.session::step_{i}"
+        result.functions[scope] = _function(f"step_{i}", file, line=10 * i)
+        result.mutations.append(_mutation("state", file, scope))
+
+    smells = SmellDetector(result)._detect_shotgun_surgery()
+
+    assert len(smells) == 1
+    assert smells[0].name == "Shotgun Surgery: state"
+    assert set(smells[0].context["affected_functions"]) == {
+        f"core.session::step_{i}" for i in range(5)
+    }
+
+
+def test_four_functions_in_one_file_below_threshold():
+    result = AnalysisResult()
+    file = "core/session.py"
+    for i in range(4):
+        scope = f"core.session::step_{i}"
+        result.functions[scope] = _function(f"step_{i}", file)
+        result.mutations.append(_mutation("state", file, scope))
+
+    smells = SmellDetector(result)._detect_shotgun_surgery()
+
+    assert smells == []
+
+
+def test_same_variable_name_in_two_files_counts_separately_per_file():
+    """5 mutators of 'data' in file A (real smell) plus unrelated mutators of
+    'data' in file B must not combine into a bigger, less accurate count,
+    and file B (below threshold) must not itself be flagged."""
+    result = AnalysisResult()
+    file_a = "a.py"
+    for i in range(5):
+        scope = f"a::fn_{i}"
+        result.functions[scope] = _function(f"fn_{i}", file_a)
+        result.mutations.append(_mutation("data", file_a, scope))
+
+    file_b = "b.py"
+    for i in range(3):
+        scope = f"b::fn_{i}"
+        result.functions[scope] = _function(f"fn_{i}", file_b)
+        result.mutations.append(_mutation("data", file_b, scope))
+
+    smells = SmellDetector(result)._detect_shotgun_surgery()
+
+    assert len(smells) == 1
+    assert smells[0].context["file"] == file_a

code2llm-0.5.171/VERSION

@@ -1 +0,0 @@
-0.5.171