sourcecode 1.31.3__tar.gz → 1.31.5__tar.gz

{sourcecode-1.31.3 → sourcecode-1.31.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 1.31.3
+Version: 1.31.5
 Summary: Deterministic codebase context for AI coding agents
 License:                                  Apache License
                                    Version 2.0, January 2004
@@ -225,7 +225,7 @@ Description-Content-Type: text/markdown
 
 **Deterministic, behavior-aware codebase context for AI agents and PR review.**
 
-![Version](https://img.shields.io/badge/version-1.31.3-blue)
+![Version](https://img.shields.io/badge/version-1.31.5-blue)
 ![Python](https://img.shields.io/badge/python-3.10%2B-green)
 
 ---
@@ -261,7 +261,7 @@ pipx install sourcecode
 
 ```bash
 sourcecode version
-# sourcecode 1.31.3
+# sourcecode 1.31.5
 ```
 
 ---

{sourcecode-1.31.3 → sourcecode-1.31.5}/README.md

@@ -2,7 +2,7 @@
 
 **Deterministic, behavior-aware codebase context for AI agents and PR review.**
 
-![Version](https://img.shields.io/badge/version-1.31.3-blue)
+![Version](https://img.shields.io/badge/version-1.31.5-blue)
 ![Python](https://img.shields.io/badge/python-3.10%2B-green)
 
 ---
@@ -38,7 +38,7 @@ pipx install sourcecode
 
 ```bash
 sourcecode version
-# sourcecode 1.31.3
+# sourcecode 1.31.5
 ```
 
 ---

{sourcecode-1.31.3 → sourcecode-1.31.5}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "sourcecode"
-version = "1.31.3"
+version = "1.31.5"
 description = "Deterministic codebase context for AI coding agents"
 readme = "README.md"
 requires-python = ">=3.9"

{sourcecode-1.31.3 → sourcecode-1.31.5}/src/sourcecode/__init__.py

@@ -1,3 +1,3 @@
 """sourcecode — Deterministic codebase context maps for AI coding agents."""
 
-__version__ = "1.31.3"
+__version__ = "1.31.5"

{sourcecode-1.31.3 → sourcecode-1.31.5}/src/sourcecode/cli.py

@@ -401,7 +401,7 @@ def main(
         help=(
             "High-signal summary (typically 1000–3000 tokens depending on repo size): "
             "stacks, entry points, dependency summary, confidence, and gaps. "
-            "Includes security_surface, mybatis, and transactional_boundaries for Java projects. "
+            "Includes security_surface (when @M3FiltroSeguridad detected), mybatis (when MyBatis framework detected), and transactional_boundaries for Java projects. "
             "Use --agent for maximum signal or --slim (when available) for minimal token footprint."
         ),
     ),
@@ -1849,6 +1849,11 @@ def prepare_context_cmd(
         "--include-config",
         help="(generate-tests) Include tooling config files (*.conf.js, .eslintrc*, etc.) in test_gaps. Excluded by default.",
     ),
+    all_gaps: bool = typer.Option(
+        False,
+        "--all",
+        help="(generate-tests) Return the full test_gaps list without truncating to top 20.",
+    ),
 ) -> None:
     """Task-specific context for AI coding agents.
 
@@ -1930,7 +1935,7 @@ def prepare_context_cmd(
             _sys.stderr.flush()
     _t0 = _time.perf_counter()
     try:
-        output = builder.build(task, since=since, symptom=symptom, fast=fast, include_config=include_config)
+        output = builder.build(task, since=since, symptom=symptom, fast=fast, include_config=include_config, all_gaps=all_gaps)
     finally:
         _progress.finish()
     _t_total = (_time.perf_counter() - _t0) * 1000
@@ -1957,8 +1962,8 @@ def prepare_context_cmd(
         },
         "explain": {
             "project_summary": True, "architecture_summary": True,
-            "relevant_files": True, "key_dependencies": True,
-            "gaps": True, "confidence": True,
+            "relevant_files": False, "key_dependencies": True,
+            "gaps": False, "confidence": True,
             "suspected_areas": False, "improvement_opportunities": False,
             "test_gaps": False, "code_notes_summary": False,
             "changed_files": False, "affected_entry_points": False,
@@ -2413,14 +2418,20 @@ def repo_ir_cmd(
 
     if output_path:
         output_path.write_text(output, encoding="utf-8")
-        n_nodes = len((ir.get("graph") or {}).get("nodes") or [])
-        n_edges = len((ir.get("graph") or {}).get("edges") or [])
         size_kb = len(output.encode("utf-8")) // 1024
-        typer.echo(
-            f"IR written to {output_path} "
-            f"({size_kb}KB, {n_nodes} nodes, {n_edges} edges)",
-            err=True,
-        )
+        if summary_only:
+            typer.echo(
+                f"IR written to {output_path} ({size_kb}KB, graph omitted by --summary-only)",
+                err=True,
+            )
+        else:
+            n_nodes = len((ir.get("graph") or {}).get("nodes") or [])
+            n_edges = len((ir.get("graph") or {}).get("edges") or [])
+            typer.echo(
+                f"IR written to {output_path} "
+                f"({size_kb}KB, {n_nodes} nodes, {n_edges} edges)",
+                err=True,
+            )
     else:
         try:
             _sys.stdout.buffer.write(output.encode("utf-8"))
@@ -2457,7 +2468,11 @@ def _extract_java_endpoints(root: "Path") -> "dict[str, Any]":
         r'@(Get|Post|Put|Delete|Patch|Request)Mapping\s*'
         r'(?:\(\s*(?:value\s*=\s*)?(?:"([^"]*)"|\{[^}]*\}|[^)]*)\s*\))?',
     )
-    _CLASS_RE = _re.compile(r'(?:class|interface)\s+(\w+)')
+    _CLASS_RE = _re.compile(
+        r'^[ \t]*(?:(?:public|protected|private|abstract|final|@\w+)\s+)*'
+        r'(?:class|interface)\s+(\w+)',
+        _re.MULTILINE,
+    )
     _METHOD_RE = _re.compile(
         r'(?:public|protected|private)\s+\S+\s+(\w+)\s*\(',
     )
@@ -2467,6 +2482,9 @@ def _extract_java_endpoints(root: "Path") -> "dict[str, Any]":
     _CLASS_PATH_RE = _re.compile(
         r'@RequestMapping\s*\(\s*(?:value\s*=\s*)?["\']([^"\']+)["\']',
     )
+    # Handles array syntax: @RequestMapping({"/v1/foo", "/v1/bar"})
+    _CLASS_ARRAY_PATH_RE = _re.compile(r'@RequestMapping\s*\(\s*\{([^}]*)\}')
+    _QUOTED_STR_RE = _re.compile(r'"([^"]+)"')
     _REQUEST_METHOD_VERB_RE = _re.compile(r'method\s*=\s*RequestMethod\.([A-Z]+)')
     _VALUE_PATH_RE = _re.compile(r'value\s*=\s*"([^"]+)"')
 
@@ -2505,7 +2523,6 @@ def _extract_java_endpoints(root: "Path") -> "dict[str, Any]":
         class_name = cls_m.group(1) if cls_m else java_file.stem
 
         # Extract class-level base path and locate class body start
-        class_base = ""
         lines = content.splitlines()
 
         # First pass: find class/interface declaration line index
@@ -2518,26 +2535,91 @@ def _extract_java_endpoints(root: "Path") -> "dict[str, Any]":
                 class_body_start = i + 1
                 break
 
-        # Second pass: extract class-level @RequestMapping path (only before class body)
+        # Second pass: extract class-level @RequestMapping path (only before class body).
+        # Supports both single-string and array syntax (Bug #1B).
         search_end = class_body_start if class_body_start else len(lines)
+        class_bases: list[str] = [""]  # default: no prefix
         for i in range(search_end):
+            sl = lines[i].strip()
+            if sl.startswith("//") or sl.startswith("*"):
+                continue
             if "@RequestMapping" in lines[i]:
-                block = "\n".join(lines[max(0, i - 1): i + 5])
+                # Cap block to search_end so method annotations inside the class body
+                # cannot be matched as the class-level prefix (Bug #1B).
+                block = "\n".join(lines[max(0, i - 1): min(i + 5, search_end + 1)])
                 if "class " in block or "interface " in block:
                     path_m = _CLASS_PATH_RE.search(block)
                     if path_m:
-                        class_base = path_m.group(1).rstrip("/")
-                    break
+                        captured = path_m.group(1).rstrip("/")
+                        # Handle string concat: @RequestMapping("lit" + ClassName.CONST)
+                        _CONCAT_CONST_RE = _re.compile(
+                            r'@RequestMapping\s*\(\s*(?:value\s*=\s*)?'
+                            r'["\']([^"\']*)["\'\s]*\+\s*(\w+)\.(\w+)'
+                        )
+                        cc_m = _CONCAT_CONST_RE.search(block)
+                        if cc_m:
+                            prefix_lit = cc_m.group(1)
+                            c_class = cc_m.group(2)
+                            c_field = cc_m.group(3)
+                            _SVAL_RE = _re.compile(
+                                r'static\s+final\s+String\s+'
+                                + _re.escape(c_field)
+                                + r'\s*=\s*"([^"]+)"'
+                            )
+                            resolved = None
+                            m_cur = _SVAL_RE.search(content)
+                            if m_cur:
+                                resolved = m_cur.group(1)
+                            else:
+                                for _jf in java_files:
+                                    if _jf.stem == c_class:
+                                        try:
+                                            _jf_txt = _jf.read_text(
+                                                encoding="utf-8", errors="replace"
+                                            )
+                                            m_jf = _SVAL_RE.search(_jf_txt)
+                                            if m_jf:
+                                                resolved = m_jf.group(1)
+                                                break
+                                        except OSError:
+                                            pass
+                            if resolved is not None:
+                                class_bases = [(prefix_lit + resolved).rstrip("/")]
+                            else:
+                                class_bases = [captured] if captured else [""]
+                        else:
+                            class_bases = [captured] if captured else [""]
+                    else:
+                        arr_m = _CLASS_ARRAY_PATH_RE.search(block)
+                        if arr_m:
+                            paths = _QUOTED_STR_RE.findall(arr_m.group(1))
+                            if paths:
+                                class_bases = [p.rstrip("/") for p in paths]
+                break
 
-        # Extract method-level endpoints starting from inside class body
-        # (skipping class-level annotations that appear before the class declaration)
+        # Extract method-level endpoints starting from inside class body.
+        # Bug #1A fix: track block comments and skip // lines before annotation checks.
         pending_annotations: list[tuple[str, str]] = []  # (http_verb, path_suffix)
         pending_filtro: Optional[str] = None
+        in_block_comment = False
 
         for i in range(class_body_start, len(lines)):
             line = lines[i]
             stripped = line.strip()
 
+            # Block comment state tracking (Bug #1A)
+            if in_block_comment:
+                if "*/" in stripped:
+                    in_block_comment = False
+                continue
+            if stripped.startswith("/*"):
+                if "*/" not in stripped:
+                    in_block_comment = True
+                continue
+            # Skip line comments (Bug #1A)
+            if stripped.startswith("//"):
+                continue
+
             # Check for @M3FiltroSeguridad
             fm = _FILTRO_RE.search(stripped)
             if fm:
@@ -2572,30 +2654,30 @@ def _extract_java_endpoints(root: "Path") -> "dict[str, Any]":
                 handler = mm.group(1) if mm else ""
                 if handler and not handler.startswith("class"):
                     for http_verb, path_suffix in pending_annotations:
-                        full_path = (class_base + "/" + path_suffix).replace("//", "/").rstrip("/") or "/"
-                        if not full_path.startswith("/"):
-                            full_path = "/" + full_path
-                        key = (class_name, handler, http_verb)
-                        if key not in seen:
-                            seen.add(key)
-                            entry: dict[str, Any] = {
-                                "method": http_verb,
-                                "path": full_path,
-                                "controller": class_name,
-                                "handler": handler,
-                            }
-                            if pending_filtro:
-                                entry["required_permission"] = pending_filtro
-                            endpoints.append(entry)
+                        for _cb in class_bases:  # Bug #1B: one endpoint per class prefix
+                            full_path = (_cb + "/" + path_suffix).replace("//", "/").rstrip("/") or "/"
+                            if not full_path.startswith("/"):
+                                full_path = "/" + full_path
+                            key = (class_name, handler, http_verb, _cb)
+                            if key not in seen:
+                                seen.add(key)
+                                entry: dict[str, Any] = {
+                                    "method": http_verb,
+                                    "path": full_path,
+                                    "controller": class_name,
+                                    "handler": handler,
+                                }
+                                if pending_filtro:
+                                    entry["required_permission"] = pending_filtro
+                                endpoints.append(entry)
                     pending_annotations = []
                     pending_filtro = None
                     continue
 
-            # Non-annotation, non-method line — reset if it's a closing brace or blank
-            if stripped in ("}", "{", "") or stripped.startswith("//") or stripped.startswith("*"):
-                if stripped == "}":
-                    pending_annotations = []
-                    pending_filtro = None
+            # Non-annotation, non-method line — reset on closing brace
+            if stripped == "}":
+                pending_annotations = []
+                pending_filtro = None
 
     endpoints.sort(key=lambda e: (e.get("controller", ""), e.get("path", "")))
     undocumented = sum(1 for e in endpoints if "required_permission" not in e)

{sourcecode-1.31.3 → sourcecode-1.31.5}/src/sourcecode/detectors/heuristic.py

@@ -59,8 +59,13 @@ class HeuristicDetector(AbstractDetector):
         paths = flatten_file_tree(context.file_tree)
         counts: Counter[str] = Counter()
         for path in paths:
-            if path.startswith("."):
+            if path.startswith(".") or _is_auxiliary_path(path):
                 continue
+            # Skip JS/TS files bundled as Java static resources (not Node.js source)
+            if path.endswith((".js", ".ts", ".tsx", ".jsx", ".mjs")):
+                _np = path.replace("\\", "/")
+                if "src/main/resources/" in _np or "src/main/webapp/" in _np:
+                    continue
             for extension, stack in _EXTENSION_MAP.items():
                 if path.endswith(extension):
                     counts[stack] += 1

{sourcecode-1.31.3 → sourcecode-1.31.5}/src/sourcecode/prepare_context.py

@@ -721,7 +721,7 @@ class TaskContextBuilder:
     def __init__(self, root: Path) -> None:
         self.root = root
 
-    def build(self, task_name: str, *, since: Optional[str] = None, symptom: Optional[str] = None, fast: bool = False, include_config: bool = False) -> TaskOutput:
+    def build(self, task_name: str, *, since: Optional[str] = None, symptom: Optional[str] = None, fast: bool = False, include_config: bool = False, all_gaps: bool = False) -> TaskOutput:
         if task_name not in TASKS:
             raise ValueError(
                 f"Unknown task '{task_name}'. Available: {', '.join(TASKS)}"
@@ -1763,14 +1763,18 @@ class TaskContextBuilder:
                         "_rank": _pub_count + _ann_count * 2,
                     })
 
-                _java_candidates.sort(key=lambda x: -x["_rank"])
+                _java_candidates.sort(
+                    key=lambda x: -(x["public_method_count"] * (1.5 if x["has_spring_annotations"] else 1.0))
+                )
+                _top = _java_candidates if all_gaps else _java_candidates[:20]
                 test_gaps = [
                     {
                         "path": c["path"],
                         "public_method_count": c["public_method_count"],
                         "has_spring_annotations": c["has_spring_annotations"],
+                        "rank_score": round(c["public_method_count"] * (1.5 if c["has_spring_annotations"] else 1.0), 1),
                     }
-                    for c in _java_candidates
+                    for c in _top
                 ]
             else:
                 # Non-Java algorithm (unchanged)
@@ -2206,8 +2210,67 @@ class TaskContextBuilder:
                 "refactor": max(15, min(30, _repo_size // 120)),
             }
             _budget = _task_budget.get(task_name, 15)
-            _selected = _ctx.select_subgraph(_ns, contracts=[], budget=_budget, min_score=0.15)
+            _selected = list(_ctx.select_subgraph(_ns, contracts=[], budget=_budget, min_score=0.15))
             _rf_map = {path: rf for _, path, rf in scored}
+
+            # Bug #3: onboard must cover ≥3 arch layers (controllers/services/domain/repositories).
+            if task_name == "onboard":
+                def _arch_layer(p: str) -> str:
+                    n = Path(p).name.lower()
+                    if "controller" in n:
+                        return "controllers"
+                    if "repository" in n or "mapper" in n or "dao" in n:
+                        return "repositories"
+                    if "service" in n:
+                        return "services"
+                    pn = p.replace("\\", "/")
+                    if "entity" in n or "/entity/" in pn or "/domain/" in pn or "/model/" in pn:
+                        return "domain"
+                    return "other"
+
+                _REQUIRED = {"controllers", "services", "repositories", "domain"}
+                _covered = {_arch_layer(p) for p in _selected} & _REQUIRED
+                _missing = _REQUIRED - _covered
+                if len(_covered) < 3 and _missing:
+                    _sel_set = set(_selected)
+                    # First pass: inject from already-scored files
+                    for _, _p, _ in sorted(scored, key=lambda x: -x[0]):
+                        if len(_covered) >= 3:
+                            break
+                        if _p in _sel_set or _p not in _rf_map:
+                            continue
+                        _layer = _arch_layer(_p)
+                        if _layer in _missing:
+                            _selected.append(_p)
+                            _sel_set.add(_p)
+                            _covered.add(_layer)
+                            _missing.discard(_layer)
+                    # Second pass: fallback scan of all_paths when scored files
+                    # don't cover enough layers (e.g. all Java files scored ≤ 0
+                    # due to auxiliary/example package detection).
+                    if len(_covered) < 3 and _missing:
+                        _NON_TEST = ("/test/", "/tests/", "/spec/")
+                        for _p in all_paths:
+                            if len(_covered) >= 3:
+                                break
+                            if _p in _sel_set:
+                                continue
+                            if any(s in _p.replace("\\", "/") for s in _NON_TEST):
+                                continue
+                            _layer = _arch_layer(_p)
+                            if _layer not in _missing:
+                                continue
+                            _rf_map[_p] = RelevantFile(
+                                path=_p,
+                                role="source",
+                                score=0.1,
+                                reason="layer coverage (onboard)",
+                            )
+                            _selected.append(_p)
+                            _sel_set.add(_p)
+                            _covered.add(_layer)
+                            _missing.discard(_layer)
+
             return [_rf_map[p] for p in _selected if p in _rf_map]
         except Exception:
             return [f for _, _, f in scored[:15]]

{sourcecode-1.31.3 → sourcecode-1.31.5}/src/sourcecode/serializer.py

@@ -409,7 +409,10 @@ def _spring_profiles_context(sm: "SourceMap") -> "Optional[dict[str, Any]]":
         else:
             matches = [
                 p for p in sm.file_paths
-                if pfx in Path(p).stem.lower() and p.endswith(".java")
+                if (Path(p).stem.lower() == pfx
+                    or Path(p).stem.lower().startswith(pfx + "-")
+                    or Path(p).stem.lower().endswith("-" + pfx))
+                and p.endswith(".java")
             ]
         if matches:
             per_profile[profile] = [Path(p).name for p in matches[:5]]
@@ -534,6 +537,8 @@ def _bootstrap_structured(eps: list) -> "Optional[dict[str, Any]]":
 
     for ep in eps:
         path = getattr(ep, "path", "")
+        if "/test/" in path or "/tests/" in path:
+            continue
         kind = getattr(ep, "kind", "")
         stem = _Path(path).stem
 
@@ -587,8 +592,9 @@ def _bootstrap_structured(eps: list) -> "Optional[dict[str, Any]]":
                 module_names.append(module)
 
         _ctrl_note = (
-            f"{controller_methods} @RequestMapping methods across "
+            f"{controller_methods} detected entry-point methods across "
             f"{controller_classes} controller classes"
+            f" (use 'sourcecode endpoints' for full surface)"
         )
         if len(module_names) > 30:
             # Group by first path segment under ddd/ (inferred domain area)