code2llm 0.5.151__tar.gz → 0.5.152__tar.gz

{code2llm-0.5.151/code2llm.egg-info → code2llm-0.5.152}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code2llm
-Version: 0.5.151
+Version: 0.5.152
 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: STTS Project
@@ -66,11 +66,11 @@ Dynamic: requires-python
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.151-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-74.3h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.152-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-75.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
 
-- 🤖 **LLM usage:** $7.5000 (203 commits)
-- 👤 **Human dev:** ~$7429 (74.3h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $7.5000 (204 commits)
+- 👤 **Human dev:** ~$7584 (75.8h @ $100/h, 30min dedup)
 
 Generated on 2026-05-06 using [openrouter/qwen/qwen3-coder-next](https://openrouter.ai/qwen/qwen3-coder-next)
 

{code2llm-0.5.151 → code2llm-0.5.152}/README.md

@@ -2,11 +2,11 @@
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.151-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-74.3h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.152-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-75.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
 
-- 🤖 **LLM usage:** $7.5000 (203 commits)
-- 👤 **Human dev:** ~$7429 (74.3h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $7.5000 (204 commits)
+- 👤 **Human dev:** ~$7584 (75.8h @ $100/h, 30min dedup)
 
 Generated on 2026-05-06 using [openrouter/qwen/qwen3-coder-next](https://openrouter.ai/qwen/qwen3-coder-next)
 

code2llm-0.5.152/VERSION

@@ -0,0 +1 @@
+0.5.152

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/__init__.py

@@ -8,7 +8,7 @@ Includes NLP Processing Pipeline for query normalization, intent matching,
 and entity resolution with multilingual support.
 """
 
-__version__ = "0.5.151"
+__version__ = "0.5.152"
 __author__ = "STTS Project"
 
 # Core analysis components (lightweight, always needed)

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/cli_exports/orchestrator.py

@@ -85,6 +85,15 @@ def _show_dry_run_plan(formats: List[str], output_dir: Path, is_chunked: bool, r
     print(f"\n✅ Dry-run complete. Use without --dry-run to export.\n")
 
 
+def _should_skip_export_cache(args, is_chunked: bool) -> bool:
+    """Return True when this run must not read or write export cache."""
+    return (
+        is_chunked
+        or getattr(args, 'no_cache', False)
+        or getattr(args, 'force', False)
+    )
+
+
 def _run_exports(args, result, output_dir: Path, source_path: Optional[Path] = None):
     """Export analysis results in requested formats.
 
@@ -102,8 +111,9 @@ def _run_exports(args, result, output_dir: Path, source_path: Optional[Path] = N
         _show_dry_run_plan(formats, output_dir, is_chunked, result)
         return
 
-    # Skip cache for chunked or when explicitly disabled
-    skip_cache = is_chunked or getattr(args, 'no_cache', False)
+    # Skip cache for chunked runs or explicit cache bypass flags. --force is
+    # documented as a --no-cache alias, so it must bypass export cache too.
+    skip_cache = _should_skip_export_cache(args, is_chunked)
 
     if not skip_cache and source_path:
         cache = PersistentCache(str(source_path))

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/config.py

@@ -119,6 +119,15 @@ class FilterConfig:
         "*poetry.lock", "*Pipfile.lock", "*Cargo.lock", "*composer.lock",
         "*.terraform.lock.hcl", "*.tfstate", "*.tfstate.backup",
         "*.min.js", "*.min.css", "*.map",
+        # code2llm / analysis artefacts. These are useful outputs, but feeding
+        # them back into the next run makes reports rank generated files above
+        # real source files.
+        "*analysis.toon*", "*map.toon.yaml", "*evolution.toon*",
+        "*validation.toon.yaml", "*duplication.toon.yaml",
+        "*project.toon*", "*calls.yaml", "*calls.toon.yaml", "*calls.mmd",
+        "*compact_flow.mmd", "*flow.mmd", "*flow_detailed.mmd",
+        "*flow_full.mmd", "*mermaid.export", "*defscan-*.md",
+        "*SUMD.md", "*SUMR.md",
     ])
     include_patterns: List[str] = field(default_factory=list)
     min_function_lines: int = 1

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/file_analyzer.py

@@ -11,6 +11,7 @@ from .models import ClassInfo, FlowEdge, FlowNode, FunctionInfo, ModuleInfo
 from code2llm.analysis.dfg import DFGExtractor
 from code2llm.analysis.call_graph import CallGraphExtractor
 from .file_filter import FastFileFilter
+from .source_classifier import classify_source_path
 from .lang import (
     analyze_typescript_js, analyze_go, analyze_rust, analyze_java,
     analyze_cpp, analyze_csharp, analyze_php, analyze_ruby, analyze_generic,
@@ -76,6 +77,9 @@ class FileAnalyzer:
             return {}
 
         result = self._route_to_language_analyzer(content, file_path, module_name, ext)
+        if result and result.get('module'):
+            result['module'].line_count = len(content.splitlines())
+            result['module'].source_kind = classify_source_path(file_path)
 
         # Tag result with its source file so downstream callers
         # (e.g. PersistentCache in ProjectAnalyzer._store_to_persistent_cache)

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/file_filter.py

@@ -5,6 +5,11 @@ import re
 from pathlib import Path
 from .config import FilterConfig
 from .gitignore import load_gitignore_patterns
+from .source_classifier import (
+    ARCHIVE_DIR_NAMES,
+    GENERATED_OUTPUT_DIR_NAMES,
+    is_generated_artifact,
+)
 
 
 _SKIP_DIR_NAMES = frozenset({
@@ -19,13 +24,14 @@ _SKIP_DIR_NAMES = frozenset({
     'lib', 'lib64', 'site-packages', 'include', 'bin', 'share',
     '.code2llm_cache',
     'tests', 'test',
+    'coverage', '.nyc_output',
     # Backup and auto-generated directories that often contain venvs
     '.algitex', '.backup', 'backups', '.bak', 'bak',
     # Additional venv patterns
     'virtualenv', '.virtualenv', 'envs', '.envs', 'venv-', '.venv-',
     # CI/CD and deployment artifacts
     '.terraform', '.serverless', '.netlify', '.vercel',
-})
+}) | ARCHIVE_DIR_NAMES | GENERATED_OUTPUT_DIR_NAMES
 
 
 class FastFileFilter:
@@ -99,6 +105,9 @@ class FastFileFilter:
         path_lower = file_path.lower()
         basename_lower = Path(file_path).name.lower()
 
+        if is_generated_artifact(file_path, self.project_path):
+            return False
+
         return (
             self._passes_gitignore(file_path) and
             self._passes_excludes(path_lower, basename_lower) and

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/gitignore.py

@@ -1,7 +1,7 @@
 """Gitignore support for code2llm file filtering."""
 
 from pathlib import Path
-from typing import List
+from typing import Iterable, List
 import re
 
 
@@ -118,21 +118,25 @@ class GitIgnoreParser:
         return ignored
 
 
-def load_gitignore_patterns(project_path: Path) -> GitIgnoreParser:
-    """Load gitignore patterns from project directory.
-    
-    Searches up the directory tree from project_path until it finds a .gitignore file
-    or reaches the filesystem root. This ensures that gitignore rules are properly applied
-    even when analyzing subdirectories of a larger project.
-    """
+def _nearest_ignore_files(project_path: Path, names: Iterable[str]) -> List[Path]:
+    """Find nearest ignore files while walking upward from project_path."""
     current_path = project_path.resolve()
-    
-    while current_path != current_path.parent:  # Stop at filesystem root
-        gitignore_path = current_path / '.gitignore'
-        if gitignore_path.exists():
-            return GitIgnoreParser(gitignore_path)
+    while current_path != current_path.parent:
+        found = [current_path / name for name in names if (current_path / name).exists()]
+        if found:
+            return found
         current_path = current_path.parent
-    
-    # Check filesystem root as last resort
-    gitignore_path = current_path / '.gitignore'
-    return GitIgnoreParser(gitignore_path)
+    return [current_path / name for name in names if (current_path / name).exists()]
+
+
+def load_gitignore_patterns(project_path: Path) -> GitIgnoreParser:
+    """Load .gitignore and .code2llmignore patterns from the nearest parent.
+
+    .code2llmignore is loaded after .gitignore so project-specific analysis
+    rules can add exclusions or negations without changing version-control
+    ignores.
+    """
+    parser = GitIgnoreParser()
+    for ignore_path in _nearest_ignore_files(project_path, ('.gitignore', '.code2llmignore')):
+        parser._load_gitignore(ignore_path)
+    return parser

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/lang/generic.py

@@ -4,6 +4,7 @@ import re
 from typing import Dict
 
 from code2llm.core.models import ClassInfo, FunctionInfo, ModuleInfo
+from code2llm.core.source_classifier import is_structural_only_file
 
 
 def analyze_generic(content: str, file_path: str, module_name: str,
@@ -19,7 +20,9 @@ def analyze_generic(content: str, file_path: str, module_name: str,
 
     # Count lines as basic metric
     lines = content.split('\n')
-    non_empty = len([l for l in lines if l.strip()])
+    if is_structural_only_file(file_path):
+        stats['files_processed'] += 1
+        return result
 
     # Try to detect function-like patterns
     func_patterns = [

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/models.py

@@ -100,6 +100,8 @@ class ModuleInfo(BaseModel):
     name: str
     file: str
     is_package: bool = False
+    source_kind: str = "source"
+    line_count: int = 0
     imports: List[str] = field(default_factory=list)
     functions: List[str] = field(default_factory=list)
     classes: List[str] = field(default_factory=list)

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/repo_files.py

@@ -5,6 +5,11 @@ from typing import List, Tuple, Optional
 
 from functools import lru_cache
 from .gitignore import load_gitignore_patterns, GitIgnoreParser
+from .source_classifier import (
+    ARCHIVE_DIR_NAMES,
+    GENERATED_OUTPUT_DIR_NAMES,
+    is_generated_artifact,
+)
 
 # Directories to skip during analysis
 SKIP_DIRS = {
@@ -14,7 +19,7 @@ SKIP_DIRS = {
     'build', 'dist', 'egg-info', '.eggs',
     'htmlcov', '.coverage', '.cache',
     'lib', 'lib64', 'site-packages', 'include', 'bin', 'share',  # venv internals
-}
+} | ARCHIVE_DIR_NAMES | GENERATED_OUTPUT_DIR_NAMES
 
 # Patterns that indicate a file should be skipped
 SKIP_PATTERNS = [
@@ -42,6 +47,8 @@ def should_skip_file(file_str: str, project_path: Optional[Path] = None,
     lower_path = file_str.lower()
     if any(pattern in lower_path for pattern in SKIP_PATTERNS):
         return True
+    if is_generated_artifact(file_str, project_path):
+        return True
     
     # Check gitignore if parser provided
     if gitignore_parser and project_path:

code2llm-0.5.152/code2llm/core/source_classifier.py

@@ -0,0 +1,185 @@
+"""Source path classification and default noise filtering.
+
+The analyzer intentionally tracks some non-code files so export caches are
+invalidated when project manifests or docs change. Those files should not be
+treated the same as production source, and generated code2llm artefacts should
+not be fed back into the next analysis run.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Optional
+
+
+CODE_EXTENSIONS = {
+    ".py", ".ts", ".tsx", ".js", ".jsx", ".mjs", ".cjs", ".go", ".rs",
+    ".java", ".cpp", ".cc", ".cxx", ".hpp", ".hh", ".h", ".c", ".cs",
+    ".rb", ".rbw", ".php", ".swift", ".kt", ".kts", ".scala", ".r",
+    ".m", ".sh", ".bash", ".zsh", ".dart", ".ex", ".exs", ".clj",
+    ".cljs", ".hs", ".lua", ".pl", ".pm", ".raku", ".rakumod",
+}
+
+DOC_EXTENSIONS = {".md", ".mdx", ".markdown", ".rst", ".adoc", ".asciidoc", ".txt"}
+
+STRUCTURAL_ONLY_EXTENSIONS = {
+    ".yaml", ".yml", ".toml", ".ini", ".cfg", ".conf", ".properties",
+    ".xml", ".json", ".json5", ".jsonc", ".env", ".proto", ".graphql",
+    ".gql", ".avsc", ".prisma", ".tf", ".tfvars", ".hcl", ".bicep",
+    ".nix",
+    *DOC_EXTENSIONS,
+}
+
+ARCHIVE_DIR_NAMES = frozenset({
+    "_archive", "archive", "archives", "_archives", "legacy", "old",
+})
+
+GENERATED_OUTPUT_DIR_NAMES = frozenset({
+    ".code2llm_cache",
+    "code2llm_output",
+    "output_toon",
+    "output_all",
+    "output_comparison",
+    "output_modes",
+    "analysis_all",
+    "compare",
+    "calls_output",
+    "refactor_output",
+    "playwright-report",
+    "test-results",
+    "html-report",
+})
+
+GENERATED_EXACT_FILENAMES = frozenset({
+    "analysis.toon",
+    "analysis.toon.yaml",
+    "analysis_toon.yaml",
+    ".code2llm_incremental.json",
+    "map.toon",
+    "map.toon.yaml",
+    "evolution.toon",
+    "evolution.toon.yaml",
+    "validation.toon.yaml",
+    "duplication.toon.yaml",
+    "calls.toon.yaml",
+    "calls.yaml",
+    "calls.mmd",
+    "compact_flow.mmd",
+    "flow.mmd",
+    "flow_detailed.mmd",
+    "flow_full.mmd",
+    "mermaid.export",
+    "project.toon",
+    "project.toon.yaml",
+    "project.toon.txt",
+    "defscan-classes-py.md",
+    "defscan-classes-ts.md",
+    "sumd.md",
+    "sumr.md",
+})
+
+GENERATED_PREFIXES = ("defscan-",)
+
+GENERATED_SNIFF_FILENAMES = frozenset({
+    "context.md",
+    "readme.md",
+    "index.html",
+    "prompt.txt",
+})
+
+GENERATED_MARKERS = (
+    "generated by: `code2llm`",
+    "generated by code2llm",
+    "code2llm - generated analysis files",
+    "<title>code2llm analysis results</title>",
+    "# system architecture analysis\n<!-- generated",
+    "use the attached/generated files as the authoritative context.",
+)
+
+
+def _relative_parts(path: Path, project_root: Optional[Path]) -> tuple[str, ...]:
+    """Return normalized relative path parts for classification."""
+    try:
+        if project_root:
+            rel = path.resolve().relative_to(project_root.resolve())
+        else:
+            rel = path
+    except (OSError, ValueError, RuntimeError):
+        rel = path
+    return tuple(part.lower() for part in rel.parts)
+
+
+def _looks_like_generated_content(path: Path) -> bool:
+    """Cheap content sniff for generic generated filenames such as README.md."""
+    try:
+        text = path.read_text(encoding="utf-8", errors="ignore")[:4096].lower()
+    except (OSError, UnicodeDecodeError):
+        return False
+    return any(marker in text for marker in GENERATED_MARKERS)
+
+
+def _has_code2llm_output_manifest(path: Path) -> bool:
+    """Return True if a directory looks like a code2llm output folder."""
+    return (
+        (path / "analysis.toon.yaml").exists()
+        or (path / "analysis.toon").exists()
+        or (path / "map.toon.yaml").exists()
+    )
+
+
+def is_generated_artifact(path: str | Path, project_root: Optional[Path] = None) -> bool:
+    """Return True when *path* is a generated analysis/report artefact."""
+    p = Path(path)
+    parts = _relative_parts(p, project_root)
+    if any(part in GENERATED_OUTPUT_DIR_NAMES for part in parts[:-1]):
+        return True
+
+    if project_root:
+        rel_parts = parts[:-1]
+        for idx, part in enumerate(rel_parts):
+            if part != "project":
+                continue
+            candidate = Path(project_root).joinpath(*parts[: idx + 1])
+            if _has_code2llm_output_manifest(candidate):
+                return True
+
+    name = p.name.lower()
+    if name in GENERATED_EXACT_FILENAMES:
+        return True
+    if any(name.startswith(prefix) for prefix in GENERATED_PREFIXES):
+        return True
+    if name in GENERATED_SNIFF_FILENAMES and _looks_like_generated_content(p):
+        return True
+
+    # The default code2llm output folder is commonly named "project". Avoid
+    # treating arbitrary project/ directories as generated unless the filename
+    # itself is a known code2llm artefact.
+    if "project" in parts[:-1] and name in GENERATED_EXACT_FILENAMES:
+        return True
+
+    return False
+
+
+def classify_source_path(path: str | Path, project_root: Optional[Path] = None) -> str:
+    """Classify a path for reporting and filtering."""
+    p = Path(path)
+    parts = _relative_parts(p, project_root)
+    if is_generated_artifact(p, project_root):
+        return "generated"
+    if any(part in ARCHIVE_DIR_NAMES for part in parts[:-1]):
+        return "archive"
+    if any(part in {"tests", "test", "testing", "__tests__"} for part in parts[:-1]):
+        return "test"
+    ext = p.suffix.lower()
+    if ext in DOC_EXTENSIONS:
+        return "docs"
+    if ext in STRUCTURAL_ONLY_EXTENSIONS:
+        return "config"
+    if ext in CODE_EXTENSIONS:
+        return "source"
+    return "other"
+
+
+def is_structural_only_file(path: str | Path) -> bool:
+    """Return True for tracked non-code text files that should emit no symbols."""
+    return Path(path).suffix.lower() in STRUCTURAL_ONLY_EXTENSIONS

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/core/streaming/scanner.py

@@ -5,6 +5,7 @@ from pathlib import Path
 from typing import Dict, List, Optional, Tuple
 
 from code2llm.core.models import ClassInfo, FlowNode, FunctionInfo, ModuleInfo
+from code2llm.core.source_classifier import classify_source_path, is_generated_artifact
 
 from .cache import StreamingFileCache
 from .prioritizer import FilePriority
@@ -45,7 +46,9 @@ class StreamingScanner:
         result = {
             'module': ModuleInfo(
                 name=priority.module_name,
-                file=priority.file_path
+                file=priority.file_path,
+                source_kind=classify_source_path(priority.file_path),
+                line_count=len(content.splitlines()),
             ),
             'functions': {},
             'classes': {},
@@ -187,6 +190,8 @@ class StreamingScanner:
             
             if any(x in file_str.lower() for x in ['__pycache__', '.venv', 'venv']):
                 continue
+            if is_generated_artifact(py_file, project_path):
+                continue
             
             # Calculate module name
             rel_path = py_file.relative_to(project_path)

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm/nlp/__init__.py

@@ -4,7 +4,7 @@ Provides query normalization, intent matching, and entity resolution
 with multilingual support and fuzzy matching.
 """
 
-__version__ = "0.5.151"
+__version__ = "0.5.152"
 
 from .pipeline import NLPPipeline
 from .normalization import QueryNormalizer

{code2llm-0.5.151 → code2llm-0.5.152/code2llm.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code2llm
-Version: 0.5.151
+Version: 0.5.152
 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: STTS Project
@@ -66,11 +66,11 @@ Dynamic: requires-python
 
 ## AI Cost Tracking
 
-![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.151-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-74.3h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
+![PyPI](https://img.shields.io/badge/pypi-costs-blue) ![Version](https://img.shields.io/badge/version-0.5.152-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-$7.50-orange) ![Human Time](https://img.shields.io/badge/Human%20Time-75.8h-blue) ![Model](https://img.shields.io/badge/Model-openrouter%2Fqwen%2Fqwen3--coder--next-lightgrey)
 
-- 🤖 **LLM usage:** $7.5000 (203 commits)
-- 👤 **Human dev:** ~$7429 (74.3h @ $100/h, 30min dedup)
+- 🤖 **LLM usage:** $7.5000 (204 commits)
+- 👤 **Human dev:** ~$7584 (75.8h @ $100/h, 30min dedup)
 
 Generated on 2026-05-06 using [openrouter/qwen/qwen3-coder-next](https://openrouter.ai/qwen/qwen3-coder-next)
 

{code2llm-0.5.151 → code2llm-0.5.152}/code2llm.egg-info/SOURCES.txt

@@ -54,6 +54,7 @@ code2llm/core/models.py
 code2llm/core/persistent_cache.py
 code2llm/core/refactoring.py
 code2llm/core/repo_files.py
+code2llm/core/source_classifier.py
 code2llm/core/streaming_analyzer.py
 code2llm/core/toon_size_manager.py
 code2llm/core/lang/__init__.py
@@ -178,6 +179,7 @@ tests/test_calls_toon_export.py
 tests/test_declarative_collection.py
 tests/test_deep_analysis.py
 tests/test_edge_cases.py
+tests/test_export_cache_flags.py
 tests/test_file_analyzer_tagging.py
 tests/test_flow_exporter.py
 tests/test_format_quality.py

{code2llm-0.5.151 → code2llm-0.5.152}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "code2llm"
-version = "0.5.151"
+version = "0.5.152"
 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"

{code2llm-0.5.151 → code2llm-0.5.152}/tests/test_declarative_collection.py

@@ -20,6 +20,7 @@ from code2llm.core.config import (
     ALL_EXTENSIONS, ALL_FILENAMES, DECLARATIVE_EXTENSIONS, Config,
     LANGUAGE_FILENAMES,
 )
+from code2llm.core.file_analyzer import FileAnalyzer
 from code2llm.core.persistent_cache import PersistentCache
 
 
@@ -202,3 +203,70 @@ def test_lockfiles_excluded_by_default(iac_project, tmp_path):
     assert "Cargo.lock" not in names
     # But the non-lock package.json stays.
     assert "package.json" in names
+
+
+def test_generated_analysis_artifacts_are_excluded_by_default(tmp_path):
+    """code2llm outputs must not be fed back into the next analysis run."""
+    p = tmp_path / "proj"
+    p.mkdir()
+    (p / "app.py").write_text("def run():\n    return 1\n")
+    out = p / "project"
+    out.mkdir()
+    (out / "analysis.toon.yaml").write_text("# code2llm\nHEALTH[0]: ok\n")
+    (out / "map.toon.yaml").write_text("# generated map\n")
+    (out / "index.html").write_text("<title>code2llm Analysis Results</title>\n")
+    batch = out / "batch_1"
+    batch.mkdir()
+    (batch / "context.md").write_text("# System Architecture Analysis\n<!-- generated in 0.01s -->\n")
+    (p / ".code2llm_incremental.json").write_text("{}\n")
+    (p / "SUMD.md").write_text("# generated summary\n")
+    (p / "defscan-classes-py.md").write_text("class Noise:\n    pass\n")
+
+    cfg = Config()
+    cfg.performance.parallel_enabled = False
+    files = ProjectAnalyzer(cfg, p)._collect_files(p)
+    names = {str(Path(fp).relative_to(p)) for fp, _ in files}
+
+    assert "app.py" in names
+    assert "project/analysis.toon.yaml" not in names
+    assert "project/map.toon.yaml" not in names
+    assert "project/index.html" not in names
+    assert "project/batch_1/context.md" not in names
+    assert ".code2llm_incremental.json" not in names
+    assert "SUMD.md" not in names
+    assert "defscan-classes-py.md" not in names
+
+
+def test_code2llmignore_is_applied(tmp_path):
+    """Project-local .code2llmignore should refine analysis scope."""
+    p = tmp_path / "proj"
+    p.mkdir()
+    (p / ".code2llmignore").write_text("docs/\n")
+    (p / "app.py").write_text("def run():\n    return 1\n")
+    docs = p / "docs"
+    docs.mkdir()
+    (docs / "README.md").write_text("# Docs\n")
+
+    cfg = Config()
+    cfg.performance.parallel_enabled = False
+    files = ProjectAnalyzer(cfg, p)._collect_files(p)
+    names = {Path(fp).name for fp, _ in files}
+
+    assert "app.py" in names
+    assert "README.md" not in names
+
+
+def test_markdown_and_config_do_not_emit_fake_symbols(iac_project):
+    """Tracked docs/config invalidate cache but should not create classes/functions."""
+    cfg = Config()
+    analyzer = FileAnalyzer(cfg)
+
+    markdown = analyzer.analyze_file(str(iac_project / "README.md"), "README")
+    yaml_result = analyzer.analyze_file(str(iac_project / "k8s.yaml"), "k8s")
+
+    assert markdown["functions"] == {}
+    assert markdown["classes"] == {}
+    assert yaml_result["functions"] == {}
+    assert yaml_result["classes"] == {}
+    assert markdown["module"].source_kind == "docs"
+    assert yaml_result["module"].source_kind == "config"

code2llm-0.5.152/tests/test_export_cache_flags.py

@@ -0,0 +1,27 @@
+from types import SimpleNamespace
+
+from code2llm.cli_exports.orchestrator import _should_skip_export_cache
+
+
+def test_force_skips_export_cache():
+    args = SimpleNamespace(no_cache=False, force=True)
+
+    assert _should_skip_export_cache(args, is_chunked=False) is True
+
+
+def test_no_cache_skips_export_cache():
+    args = SimpleNamespace(no_cache=True, force=False)
+
+    assert _should_skip_export_cache(args, is_chunked=False) is True
+
+
+def test_chunked_skips_export_cache():
+    args = SimpleNamespace(no_cache=False, force=False)
+
+    assert _should_skip_export_cache(args, is_chunked=True) is True
+
+
+def test_standard_run_can_use_export_cache():
+    args = SimpleNamespace(no_cache=False, force=False)
+
+    assert _should_skip_export_cache(args, is_chunked=False) is False

code2llm-0.5.151/VERSION

@@ -1 +0,0 @@
-0.5.151