sourcecode 5.5.1__py3-none-any.whl → 5.5.4__py3-none-any.whl

sourcecode/__init__.py

@@ -4,4 +4,4 @@ ASK Engine is the product. ``ask`` is the canonical CLI command; ``sourcecode``
 the legacy compatibility alias and the Python/PyPI package name. See
 docs/PRODUCT_IDENTITY.md (normative)."""
 
-__version__ = "5.5.1"
+__version__ = "5.5.4"

sourcecode/_docs/DEFECT-LEDGER.md

@@ -16,6 +16,17 @@ The facts these rows are keyed to are published: `ask schema facts-v1` prints th
 
 ---
 
+## Audit synchronization
+
+| ID | Defect | Found | Severity | Status |
+|---|---|---|---|---|
+| AS-01 | `risk`/`enrich` could omit `sql_taint.bounded_interpolations`, making `sinks_total` impossible to audit against its bounded evidence. | 5.4.2 re-audit | Medium | **closed 5.4.2** — both projections preserve the bounded rows; the paired fixture asserts the sink total and the retained bounded interpolation. |
+| AS-02 | Route projection emitted declarations from abstract Spring controllers, creating phantom routes and inflating the unprotected census. | 5.4.2 re-audit | Low | **closed 5.4.2** — abstract declarers are excluded while concrete subtype projection remains admitted; the route-surface regression covers both sides. |
+| AS-03 | The distributed ledger could leave `C3-123` open after the 5.4.1 build already excluded static assets. | 5.4.2 re-audit | Informational | **closed 5.4.2** — the stale state was synchronized to the wheel it documents. |
+| AS-04 | An unconstrained `@RequestMapping` was described by consumers as GET although Spring admits every verb. | 5.5.1 audit | Medium | **closed 5.5.1** — primary `method` is `ANY`; 5.5.4 adds the concrete `methods` list so consumers do not need to infer its expansion. |
+
+---
+
 ## Class C1 — One fact, two authorities
 
 | ID | Defect | Found | Severity | Status |

sourcecode/_docs/USER_GUIDE.md

@@ -136,7 +136,7 @@ pipx install sourcecode   # isolated install, no venv needed
 
 # Verify
 ask version
-# ask 5.5.1
+# ask 5.5.4
 ```
 
 Requires Python 3.9+.
@@ -261,7 +261,13 @@ ask endpoints /path/to/repo
 ask endpoints . --format yaml
 ```
 
-Output: list of `{method, path, controller, handler}`. Covers `@GetMapping`, `@PostMapping`, `@PutMapping`, `@DeleteMapping`, `@PatchMapping`, `@RequestMapping`, `@GET`, `@POST`, `@PUT`, `@DELETE`, `@PATCH` + `@Path`.
+Output: list of `{method, methods, path, controller, handler}`. `method` remains
+the backwards-compatible primary verb; `methods` is the complete declared set.
+An unconstrained `@RequestMapping` reports `method: "ANY"` and
+`methods: ["GET", "POST", "PUT", "DELETE", "PATCH"]`, because it accepts all
+of those verbs rather than defaulting to GET. Covers `@GetMapping`, `@PostMapping`,
+`@PutMapping`, `@DeleteMapping`, `@PatchMapping`, `@RequestMapping`, `@GET`,
+`@POST`, `@PUT`, `@DELETE`, `@PATCH` + `@Path`.
 
 Each endpoint carries a `security` policy. Standard annotations (`@PreAuthorize`, `@Secured`, `@RolesAllowed`, …) are recognized automatically; a custom authorization annotation reports `policy: "none_detected"` until declared in a `sourcecode.config.json` at the repo root:
 

sourcecode/cli.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import contextlib
+import errno
 import hashlib
 import json
 import difflib
@@ -1196,6 +1197,15 @@ def _safe_write_file(path: "Path", content: str) -> None:
             fh.flush()
             os.fsync(fh.fileno())
         tmp.replace(path)
+        # Atomic replacement alone does not prove that the destination is the
+        # requested artifact. Verify the postcondition before reporting a run
+        # output path, especially for non-empty JSON reports.
+        if not path.is_file() or (content and path.stat().st_size == 0):
+            raise OSError(
+                errno.EIO,
+                "atomic output replacement did not produce the requested file",
+                str(path),
+            )
         try:
             dir_fd = os.open(str(path.parent), os.O_RDONLY)
             try:
@@ -3125,6 +3135,7 @@ def main(
         raise typer.Exit()
 
     _t0 = time.monotonic()
+    _root_timing: dict[str, float] = {}
     no_tree: bool = False  # set True by --agent; --no-tree flag removed
 
     # Validate new flag choices
@@ -3947,6 +3958,7 @@ def main(
         if _advisory:
             typer.echo(_advisory, err=True)
 
+    _manifest_t0 = time.perf_counter()
     _perf_detection = perf.start()
     detector = ProjectDetector(build_default_detectors())
     workspace_analysis = WorkspaceAnalyzer().analyze(target, manifests)
@@ -4007,6 +4019,7 @@ def main(
         root_dependencies, root_summary = dependency_analyzer.analyze(target)
         dependency_records.extend(root_dependencies)
         dependency_summaries.append(root_summary)
+    _root_timing["manifests_ms"] = round((time.perf_counter() - _manifest_t0) * 1000, 2)
     module_graphs = []
     if graph_analyzer is not None:
         root_graph_tree = (
@@ -4652,6 +4665,13 @@ def main(
                 )
 
     perf.stop("analysis", _perf_analysis)
+    # Root timings are deliberately a small diagnostic envelope. They make cold
+    # manifest cost visible even when the optional perf recorder is disabled.
+    _root_timing.setdefault("context_ms", round((time.monotonic() - _t0) * 1000, 2))
+    _root_timing.setdefault("parse_ms", 0.0)
+    _root_timing.setdefault("cir_ms", 0.0)
+    _root_timing.setdefault("ris_ms", 0.0)
+    _root_timing.setdefault("snapshot_ms", 0.0)
     # 4. Serialize
     _progress.update("serializing output")
     _perf_serialize = perf.start()
@@ -4697,8 +4717,16 @@ def main(
         raw_dict = standard_view(sm, include_tree=tree and not no_tree)
         if not no_redact:
             raw_dict = redact_dict(raw_dict)
+        data = raw_dict
         content = _serialize_dict(raw_dict, format)
 
+    if isinstance(data, dict):
+        data.setdefault("_meta", {})["timing_ms"] = {
+            **_root_timing,
+            "total_ms": round((time.monotonic() - _t0) * 1000, 2),
+        }
+        content = _serialize_dict(data, format)
+
     perf.stop("serialize", _perf_serialize)
     perf.flush_recorder()
 
@@ -9221,6 +9249,17 @@ def spring_audit_cmd(
     elif format == "github-comment":
         output = _render_spring_audit_github_comment(combined, min_severity)
     else:
+        _bounded_default = (
+            not compact and output_path is None and format == "json" and not table
+        )
+        data = (
+            combined.to_compact_dict()
+            if compact or _bounded_default
+            else combined.to_dict()
+        )
+        if _bounded_default:
+            data.pop("mode", None)
+            data["default_output"] = "bounded"
         output = _serialize_dict(data, format)
 
     # F-AV: our own readers ask for the count under either spelling — the break
@@ -9808,6 +9847,9 @@ def risk_cmd(
             data, rule_filter=rule_filter, band_filter=band_filter, top_n=top_n
         )
     else:
+        if output_path is None:
+            from sourcecode.risk import compact_risk_output
+            data = compact_risk_output(data)
         output = _serialize_dict(data, format)
     _emit_command_output(
         output,

sourcecode/dependency_analyzer.py

@@ -162,6 +162,15 @@ class DependencyAnalyzer:
         *,
         workspace: str | None = None,
     ) -> tuple[list[DependencyRecord], DependencySummary]:
+        from sourcecode import manifest_cache
+
+        cached = manifest_cache.load(root)
+        if cached is not None:
+            records, summary = cached
+            if workspace is not None:
+                records = [replace(record, workspace=workspace) for record in records]
+                summary = replace(summary, dependencies=records)
+            return records, summary
         records: list[DependencyRecord] = []
         limitations: list[str] = []
 
@@ -185,7 +194,12 @@ class DependencyAnalyzer:
             replace(r, role=_infer_role(r.name, r.ecosystem, r.scope))
             for r in deduped
         ]
-        return deduped, self._build_summary(deduped, limitations)
+        summary = self._build_summary(deduped, limitations)
+        manifest_cache.save(root, deduped, summary)
+        if workspace is not None:
+            deduped = [replace(record, workspace=workspace) for record in deduped]
+            summary = replace(summary, dependencies=deduped)
+        return deduped, summary
 
     def merge_summaries(self, summaries: Iterable[DependencySummary]) -> DependencySummary:
         result = DependencySummary(requested=True)

sourcecode/deployment_prefix.py

@@ -23,7 +23,7 @@ from __future__ import annotations
 import re
 from dataclasses import dataclass, field
 from pathlib import Path
-from typing import Any, Optional
+from typing import Any, Iterable, Optional
 
 # Published Spring Boot property keys (DATA — never a predicate on a repo name).
 _CONTEXT_PATH_KEYS: tuple[str, ...] = (
@@ -163,25 +163,31 @@ def _profile_of_document(flat: "dict[str, str]", file_profile: Optional[str]) ->
     return file_profile
 
 
-def _config_files(root: Path) -> "list[Path]":
+def _config_files(root: Path, files: Optional[Iterable[Path]] = None) -> "list[Path]":
     """Spring config files under conventional resource directories.
 
     Bounded and sorted for determinism; test resources are excluded because a
     prefix declared for tests is not the deployed one.
     """
+    candidates = files if files is not None else root.rglob("application*.*")
     found: list[Path] = []
-    for pattern in ("**/src/main/resources/**/application*.*", "**/config/application*.*"):
-        for path in root.glob(pattern):
-            name = path.name
-            if name in _CONFIG_BASENAMES or _PROFILE_FILE_RE.match(name):
-                text = str(path).replace("\\", "/")
-                if "/src/test/" in text or "/target/" in text:
-                    continue
-                found.append(path)
+    for path in candidates:
+        path = Path(path)
+        name = path.name
+        if name not in _CONFIG_BASENAMES and not _PROFILE_FILE_RE.match(name):
+            continue
+        text = str(path).replace("\\", "/")
+        in_resources = "/src/main/resources/" in text
+        in_config = "/config/" in text
+        if not (in_resources or in_config) or "/src/test/" in text or "/target/" in text:
+            continue
+        found.append(path)
     return sorted(set(found))[:60]
 
 
-def detect_deployment_prefix(root: Path) -> DeploymentPrefix:
+def detect_deployment_prefix(
+    root: Path, *, files: Optional[Iterable[Path]] = None
+) -> DeploymentPrefix:
     """Read the deployment prefix from the repo's Spring configuration.
 
     Best effort and never raises: an unreadable or malformed config yields an
@@ -190,7 +196,7 @@ def detect_deployment_prefix(root: Path) -> DeploymentPrefix:
     """
     result = DeploymentPrefix()
     try:
-        candidates = _config_files(Path(root))
+        candidates = _config_files(Path(root), files)
     except Exception:
         return result
 

sourcecode/manifest_cache.py

@@ -0,0 +1,62 @@
+"""Content-addressed cache for manifest dependency analysis."""
+from __future__ import annotations
+
+import hashlib
+import json
+import os
+from dataclasses import asdict
+from pathlib import Path
+
+from sourcecode.schema import DependencyRecord, DependencySummary
+
+
+def _base() -> Path:
+    configured = os.environ.get("SOURCECODE_CACHE_DIR", "")
+    return Path(configured) if configured else Path.home() / ".sourcecode" / "cache"
+
+
+def _key(root: Path) -> str:
+    digest = hashlib.sha256()
+    analyzer = Path(__file__).with_name("dependency_analyzer.py")
+    digest.update(analyzer.read_bytes())
+    names = {
+        "pom.xml", "build.gradle", "build.gradle.kts", "package.json",
+        "package-lock.json", "pyproject.toml", "Cargo.toml", "go.mod",
+        "composer.json", "Gemfile",
+    }
+    paths: set[Path] = set()
+    for name in names:
+        paths.update(root.rglob(name))
+    for path in sorted(paths):
+        if not path.is_file():
+            continue
+        try:
+            digest.update(path.relative_to(root).as_posix().encode())
+            digest.update(path.read_bytes())
+        except OSError:
+            continue
+    return digest.hexdigest()
+
+
+def load(root: Path) -> tuple[list[DependencyRecord], DependencySummary] | None:
+    try:
+        path = _base() / "manifest-cache-v1" / f"{_key(root)}.json"
+        data = json.loads(path.read_text(encoding="utf-8"))
+        records = [DependencyRecord(**row) for row in data["records"]]
+        summary = DependencySummary(**data["summary"])
+        summary.dependencies = records
+        return records, summary
+    except (OSError, ValueError, KeyError, TypeError):
+        return None
+
+
+def save(root: Path, records: list[DependencyRecord], summary: DependencySummary) -> None:
+    try:
+        path = _base() / "manifest-cache-v1" / f"{_key(root)}.json"
+        path.parent.mkdir(parents=True, exist_ok=True)
+        payload = {"records": [asdict(row) for row in records], "summary": asdict(summary)}
+        tmp = path.with_suffix(".tmp")
+        tmp.write_text(json.dumps(payload, sort_keys=True), encoding="utf-8")
+        tmp.replace(path)
+    except OSError:
+        return

sourcecode/migrate_check.py

@@ -2375,6 +2375,26 @@ class MigrationReport:
             compact, "findings", top_n,
             key=lambda f: SEVERITY_ORDER.get(f.get("severity"), 3),
         )
+        # Keep invariant rule prose once; each finding carries only location and
+        # rule identity. Full prose remains available in the non-compact report.
+        _rule_catalog: dict[str, dict] = {}
+        _lean_findings: list[dict] = []
+        for _finding in compact.get("findings", []):
+            _rule_id = _finding.get("rule_id", "unknown")
+            _rule_catalog.setdefault(_rule_id, {
+                "title": _finding.get("title", ""),
+                "explanation": _finding.get("explanation", ""),
+                "fix_hint": _finding.get("fix_hint", ""),
+            })
+            _lean_findings.append({
+                key: _finding[key]
+                for key in (
+                    "id", "rule_id", "severity", "source_file", "first_line",
+                    "migration_target", "imports_found", "code_context",
+                ) if key in _finding
+            })
+        compact["rule_catalog"] = _rule_catalog
+        compact["findings"] = _lean_findings
         hib = full.get("hibernate")
         if isinstance(hib, dict):
             h = dict(hib)

sourcecode/repository_ir.py

@@ -3772,17 +3772,40 @@ def _route_path_expression(args_str: str, constants: "Optional[dict[str, str]]")
     return args_str.strip()
 
 
-def _parse_route_http_method(ann_name: str, args_str: str) -> str:
-    """Derive HTTP method from annotation name or explicit method= arg."""
+_ANY_HTTP_METHODS = ("GET", "POST", "PUT", "DELETE", "PATCH")
+
+
+def _parse_route_http_methods(ann_name: str, args_str: str) -> tuple[str, ...]:
+    """Derive every declared HTTP method, preserving annotation order."""
     explicit = _HTTP_METHOD_MAP.get(ann_name)
     if explicit:
-        return explicit
-    m = re.search(r'method\s*=\s*(?:RequestMethod\.)?(\w+)', args_str or "")
+        return (explicit,)
+    # Spring accepts both `method=RequestMethod.GET` and the array spelling
+    # `method={RequestMethod.GET, RequestMethod.POST}`. The old expression
+    # expected the token immediately after `=`, so the braces form fell through
+    # to ANY and overstated the reachable verb surface.
+    m = re.search(
+        r'\bmethod\s*=\s*\{?\s*((?:RequestMethod\.)?\w+(?:\s*,\s*(?:RequestMethod\.)?\w+)*)',
+        args_str or "",
+    )
     if m:
-        return m.group(1).upper()
+        methods = re.findall(r'(?:RequestMethod\.)?(\w+)', m.group(1))
+        if methods:
+            return tuple(dict.fromkeys(method.upper() for method in methods))
     if ann_name == "@RequestMapping":
-        return "ANY"
-    return ""
+        return ("ANY",)
+    return ()
+
+
+def _parse_route_http_method(ann_name: str, args_str: str) -> str:
+    """Backward-compatible primary HTTP method for a route."""
+    methods = _parse_route_http_methods(ann_name, args_str)
+    return methods[0] if methods else ""
+
+
+def _published_http_methods(method: str) -> list[str]:
+    """Concrete verbs a consumer must consider for a route's method label."""
+    return list(_ANY_HTTP_METHODS if method == "ANY" else (method,))
 
 
 def _parse_route_extras(args_str: str) -> dict:
@@ -5014,7 +5037,8 @@ def _build_route_surface(
             _suffix_expr = _route_path_expression(args, constants)
             effective_prefixes = class_info.get(cls_simple, {}).get("prefixes", [""])
 
-        method = _parse_route_http_method(ann_name, args) or "GET"
+        route_methods = _parse_route_http_methods(ann_name, args) or ("GET",)
+        method = route_methods[0]
 
         if cls_simple in class_info:
             class_info[cls_simple]["own_endpoints"].append(
@@ -5070,6 +5094,7 @@ def _build_route_surface(
                     "effective_class": cls_fqn,
                     "path": full_path,
                     "method": method,
+                    "methods": _published_http_methods(method) if method == "ANY" else list(route_methods),
                     "return_type": (sym.return_type.strip() if sym.return_type else "void"),
                     "stable_id": sym.stable_id,
                     "inheritance_depth": 0,
@@ -5175,6 +5200,7 @@ def _build_route_surface(
                                     "effective_class": data["fqn"],
                                     "path": full_path,
                                     "method": verb,
+                                    "methods": _published_http_methods(verb),
                                     "return_type": _parent_rt_by_sym.get(declaring_sym, "void"),
                                     "stable_id": stable_id,
                                     "inheritance_depth": depth,
@@ -6411,10 +6437,31 @@ def extract_java_endpoints(
 
     java_files = sorted(
         p for p in root.rglob("*.java")
-        if not is_test_source_file(_rel(p))
+        if (include_tests or not is_test_source_file(_rel(p)))
         and "target/" not in str(p).replace("\\", "/")
         and not any(f in str(p).replace("\\", "/") for f in _CLIENT_PATH_FRAGMENTS)
     )
+    # `endpoints` needs XML both for Spring Security and WebScript detection.
+    # One inventory avoids seven name-specific walks plus a later desc.xml walk.
+    _xml_files = sorted(root.rglob("*.xml"))
+    # Deployment-prefix resolution formerly ran two independent `root.glob()`
+    # calls. One candidate list keeps that work bounded and reusable below.
+    _config_files = sorted(root.rglob("application*.*"))
+    _excluded_test_source_routes = 0
+    if not include_tests:
+        _route_marker = _re.compile(
+            r"@(GetMapping|PostMapping|PutMapping|DeleteMapping|PatchMapping|RequestMapping|GET|POST|PUT|DELETE|PATCH|Path)\b"
+        )
+        for _candidate in root.rglob("*.java"):
+            _candidate_rel = _rel(_candidate)
+            if not is_test_source_file(_candidate_rel):
+                continue
+            try:
+                _excluded_test_source_routes += len(
+                    _route_marker.findall(_candidate.read_text(encoding="utf-8", errors="replace"))
+                )
+            except OSError:
+                continue
 
     all_symbols: list[SymbolRecord] = []
     extends_map: dict[str, str] = {}
@@ -6620,6 +6667,7 @@ def extract_java_endpoints(
 
         entry: dict = {
             "method": route["method"],
+            "methods": route.get("methods", _published_http_methods(route["method"])),
             "path": route["path"],
             "controller": controller,
             # Two controllers can share a simple name across packages: keycloak
@@ -6776,23 +6824,23 @@ def extract_java_endpoints(
         re.IGNORECASE,
     )
     _xml_security_detected = False
-    _XML_GLOBS = (
-        "*security*.xml", "*Security*.xml",
-        "*applicationContext*.xml", "*-context.xml", "*Context.xml",
-        "*spring*.xml", "*Spring*.xml",
-    )
-    for _glob in _XML_GLOBS:
-        for _xf in root.rglob(_glob):
-            if "target/" in str(_xf).replace("\\", "/"):
-                continue
-            try:
-                _xt = _xf.read_text(encoding="utf-8", errors="replace")
-            except OSError:
-                continue
-            if _XML_SECURITY_RE.search(_xt):
-                _xml_security_detected = True
-                break
-        if _xml_security_detected:
+    for _xf in _xml_files:
+        _xml_name = _xf.name.lower()
+        if not (
+            "security" in _xml_name
+            or "spring" in _xml_name
+            or _xml_name.endswith("-context.xml")
+            or _xml_name.endswith("context.xml")
+        ):
+            continue
+        if "target/" in str(_xf).replace("\\", "/"):
+            continue
+        try:
+            _xt = _xf.read_text(encoding="utf-8", errors="replace")
+        except OSError:
+            continue
+        if _XML_SECURITY_RE.search(_xt):
+            _xml_security_detected = True
             break
 
     if _xml_security_detected:
@@ -6892,7 +6940,7 @@ def extract_java_endpoints(
     try:
         from sourcecode.deployment_prefix import detect_deployment_prefix
 
-        _prefix = detect_deployment_prefix(root)
+        _prefix = detect_deployment_prefix(root, files=_config_files)
         if _prefix.sources:
             result["deployment_prefix"] = _prefix.to_dict()
         if _prefix.prefix:
@@ -6912,6 +6960,12 @@ def extract_java_endpoints(
         )
     elif _test_util_endpoints:
         result["test_util_included"] = len(_test_util_endpoints)
+    if _excluded_test_source_routes:
+        result["test_source_excluded"] = _excluded_test_source_routes
+        result.setdefault("notes", []).append(
+            f"{_excluded_test_source_routes} route annotation(s) from test source trees "
+            "were excluded; use --include-tests to include them."
+        )
     # Surface incomplete-endpoint warnings (interface-defined controllers) only when
     # present, to keep output backward-compatible for the common case.
     if endpoint_warnings:
@@ -6949,7 +7003,9 @@ def extract_java_endpoints(
     # (*.desc.xml: <url>/<authentication>/<transaction>) in addition to the Java
     # signals gathered above.
     _webscript_descriptors = 0
-    for _df in root.rglob("*.desc.xml"):
+    for _df in _xml_files:
+        if not _df.name.endswith(".desc.xml"):
+            continue
         if "target/" in str(_df).replace("\\", "/"):
             continue
         try:
@@ -6977,7 +7033,7 @@ def extract_java_endpoints(
         from sourcecode.servlet_surface import build_servlet_surface
 
         _nonspring["servlets"] = int(
-            build_servlet_surface(root).get("declarations", 0)
+            build_servlet_surface(root, files=[*java_files, *_xml_files]).get("declarations", 0)
         )
     except Exception:
         pass
@@ -7419,16 +7475,22 @@ def compute_blast_radius(
 
     # BFS for indirect callers
     indirect_callers: list[str] = []
+    _indirect_parent: dict[str, str] = {}
+    _hub_frontier_nodes: set[str] = set()
     visited: set[str] = set(matched_fqns) | set(direct_callers)
 
     while queue:
         node, depth = queue.pop(0)
         if depth >= _effective_depth:
             continue
-        for caller in _walk_callers(node):
+        _node_callers = _walk_callers(node)
+        if len(_node_callers) > _HUB_CALLER_THRESHOLD:
+            _hub_frontier_nodes.add(node)
+        for caller in _node_callers:
             if caller not in visited:
                 visited.add(caller)
                 all_affected[caller] = depth + 1
+                _indirect_parent[caller] = node
                 indirect_callers.append(caller)
                 queue.append((caller, depth + 1))
 
@@ -7658,6 +7720,29 @@ def compute_blast_radius(
     n_direct_classes = len(_direct_caller_classes)
     n_direct   = n_direct_symbols
     n_indirect = len(indirect_callers)
+
+    # The complete reverse walk is useful as a measured upper reach, but a
+    # service-locator hub can make it collapse to "the whole repository". Publish
+    # a second, conservative axis that stops at hub frontiers instead of forcing
+    # consumers to guess from a flat list of symbols.
+    _without_hubs: list[str] = []
+    _without_hubs_seen = set(matched_fqns) | set(direct_callers)
+    _without_hubs_queue: list[tuple[str, int]] = [
+        (c, 1) for c in direct_callers
+    ]
+    while _without_hubs_queue:
+        _node, _depth = _without_hubs_queue.pop(0)
+        if _depth >= _effective_depth:
+            continue
+        _next = _walk_callers(_node)
+        if len(_next) > _HUB_CALLER_THRESHOLD:
+            continue
+        for _caller in _next:
+            if _caller not in _without_hubs_seen:
+                _without_hubs_seen.add(_caller)
+                _without_hubs.append(_caller)
+                _without_hubs_queue.append((_caller, _depth + 1))
+
     n_ep       = len(endpoints_affected)
     n_txn      = len(txn_nodes)
     n_mappers  = len(mappers_affected)
@@ -7824,6 +7909,24 @@ def compute_blast_radius(
     _indirect_summary = (
         _indirect_ordered if _indirect_cap is None else _indirect_ordered[:_indirect_cap]
     )
+    _indirect_by_depth: dict[str, int] = {}
+    for _node in indirect_callers:
+        _depth_key = str(all_affected.get(_node, 0))
+        _indirect_by_depth[_depth_key] = _indirect_by_depth.get(_depth_key, 0) + 1
+    def _via_hub(_node: str) -> bool:
+        _seen_path: set[str] = set()
+        _cursor = _node
+        while _cursor in _indirect_parent and _cursor not in _seen_path:
+            _seen_path.add(_cursor)
+            _cursor = _indirect_parent[_cursor]
+            if _cursor in _hub_frontier_nodes:
+                return True
+        return False
+
+    _indirect_details = [
+        {"fqn": _node, "depth": all_affected.get(_node), "via_hub": _via_hub(_node)}
+        for _node in _indirect_summary
+    ]
 
     out: dict = {
         "target": target,
@@ -7870,6 +7973,10 @@ def compute_blast_radius(
             else _direct_caller_symbols[:_direct_cap]
         ),
         "indirect_callers": _indirect_summary,
+        "indirect_caller_details": _indirect_details,
+        "indirect_callers_unit": "reference symbols (methods or classes)",
+        "indirect_callers_by_depth": _indirect_by_depth,
+        "indirect_callers_without_hubs": _without_hubs,
         "container_wired": [w.to_dict() for w in _container_wired],
         "endpoints_affected": endpoints_affected,
         "mappers_affected": mappers_affected,
@@ -7889,6 +7996,10 @@ def compute_blast_radius(
             "indirect_caller_count": n_indirect,
             "indirect_callers_computed": not _bfs_truncated or _indirect_sampled,
             "indirect_callers_sampled": _indirect_sampled,
+            "indirect_caller_count_unit": "reference symbols (methods or classes)",
+            "indirect_callers_without_hubs_count": len(_without_hubs),
+            "indirect_callers_without_hubs_unit": "reference symbols (hub frontiers excluded)",
+            "hub_frontier_count": len(_hub_frontier_nodes),
             "endpoints_affected_count": n_ep,
             # C1-37: four figures called "endpoints" at one commit (3 364 /
             # 3 382 / 3 382 / 3 574). Each one now says which population it

sourcecode/risk.py

@@ -47,6 +47,31 @@ if TYPE_CHECKING:
 
 RISK_SCHEMA = "risk-v1"
 
+
+def compact_risk_output(data: dict) -> dict:
+    """Project stdout to stable rule prose plus small per-defect evidence."""
+    compact = dict(data)
+    rows = data.get("risks") or []
+    rules: dict[str, dict] = {}
+    lean: list[dict] = []
+    for row in rows:
+        rule_ids = row.get("rule_ids") or [row.get("rule_id", "unknown")]
+        for rule_id in rule_ids:
+            rules.setdefault(rule_id, {
+                "title": row.get("title", ""),
+                "statement": row.get("statement", ""),
+            })
+        lean.append({
+            key: row[key] for key in (
+                "defect_id", "symbol", "source_file", "first_line", "band",
+                "score", "rule_ids", "factors",
+            ) if key in row
+        })
+    compact["rules"] = rules
+    compact["risks"] = lean
+    compact["default_output"] = "bounded"
+    return compact
+
 #: The composition, in one place. The payload publishes it, `--help` renders it and
 #: the battery holds both to it: a hand-copied model line in the help screen went on
 #: advertising four factors for as long as it took someone to read it, which is the

sourcecode/servlet_surface.py

@@ -20,10 +20,11 @@ So the answer is a second, named population beside the first, with its own total
 
 from __future__ import annotations
 
+import os
 import re
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Optional
+from typing import Iterable, Optional
 from xml.etree import ElementTree as ET
 
 POPULATION = "servlet"
@@ -84,7 +85,8 @@ def _child_text(elem: "ET.Element", name: str) -> str:
 
 def _from_web_xml(xml_path: Path, relative: str) -> "list[ServletMapping]":
     try:
-        root = ET.parse(str(xml_path)).getroot()
+        text = xml_path.read_text(encoding="utf-8", errors="replace")
+        root = ET.fromstring(text)
     except Exception:
         return []
     classes: "dict[str, str]" = {}
@@ -110,6 +112,27 @@ def _from_web_xml(xml_path: Path, relative: str) -> "list[ServletMapping]":
                 source_file=relative,
                 source="web.xml",
             ))
+    # Keep a tolerant lexical fallback for descriptors containing legacy XML
+    # constructs that ElementTree rejects or partially normalizes. Comments are
+    # removed first so commented deployment examples never become live routes.
+    try:
+        active = re.sub(r"<!--.*?-->", "", text, flags=re.DOTALL)
+        for block in re.findall(r"<servlet-mapping\b[^>]*>(.*?)</servlet-mapping>", active, re.DOTALL | re.IGNORECASE):
+            name_match = re.search(r"<servlet-name\b[^>]*>\s*([^<]+?)\s*</servlet-name>", block, re.DOTALL | re.IGNORECASE)
+            pattern_match = re.search(r"<url-pattern\b[^>]*>\s*([^<]+?)\s*</url-pattern>", block, re.DOTALL | re.IGNORECASE)
+            if not pattern_match:
+                continue
+            candidate = ServletMapping(
+                url_pattern=pattern_match.group(1).strip(),
+                servlet_name=name_match.group(1).strip() if name_match else "",
+                servlet_class=classes.get(name_match.group(1).strip(), "") if name_match else "",
+                source_file=relative,
+                source="web.xml",
+            )
+            if candidate not in out:
+                out.append(candidate)
+    except Exception:
+        pass
     return out
 
 
@@ -144,14 +167,36 @@ def _from_registration_beans(java_path: Path, relative: str) -> "list[ServletMap
     return out
 
 
-def build_servlet_surface(root: Path, *, limit: Optional[int] = None) -> dict:
+def build_servlet_surface(
+    root: Path,
+    *,
+    limit: Optional[int] = None,
+    files: Optional[Iterable[Path]] = None,
+) -> dict:
     """Every servlet-mapped URL pattern under `root`, as a population of its own."""
     from sourcecode import non_coverage
 
     root = Path(root)
     mappings: "list[ServletMapping]" = []
-    for path in sorted(root.rglob("*")):
-        if any(part in _SKIP_DIRS for part in path.parts) or not path.is_file():
+    if files is None:
+        candidates: list[Path] = []
+        for directory, dirs, names in os.walk(root):
+            dirs[:] = sorted(name for name in dirs if name not in _SKIP_DIRS)
+            candidates.extend(
+                Path(directory) / name
+                for name in sorted(names)
+                if name == "web.xml" or name.endswith(".java")
+            )
+    else:
+        candidates = sorted(
+            {
+                Path(path)
+                for path in files
+                if Path(path).name == "web.xml" or Path(path).suffix == ".java"
+            }
+        )
+    for path in candidates:
+        if not path.is_file():
             continue
         try:
             # POSIX, like every other path this CLI publishes. C2-18 fixed this
@@ -168,9 +213,16 @@ def build_servlet_surface(root: Path, *, limit: Optional[int] = None) -> dict:
         elif path.suffix == ".java":
             mappings.extend(_from_registration_beans(path, relative))
 
-    # One URL pattern declared twice is two declarations of one reachable path; the
-    # census counts patterns, the rows keep every declaration.
-    patterns = sorted({m.url_pattern for m in mappings})
+    # A URL pattern is the observable HTTP surface. Multiple descriptors can
+    # repeat it, but publishing every duplicate made `declarations` disagree with
+    # the pattern census and with endpoint-side reconciliation. Keep the first
+    # deterministic declaration as provenance and expose one row per reachable
+    # pattern.
+    _unique: dict[str, ServletMapping] = {}
+    for mapping in sorted(mappings, key=lambda m: (m.url_pattern, m.source_file, m.source)):
+        _unique.setdefault(mapping.url_pattern, mapping)
+    mappings = list(_unique.values())
+    patterns = sorted(_unique)
     by_source: "dict[str, int]" = {}
     for mapping in mappings:
         by_source[mapping.source] = by_source.get(mapping.source, 0) + 1

{sourcecode-5.5.1.dist-info → sourcecode-5.5.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 5.5.1
+Version: 5.5.4
 Summary: Persistent structural context and ultra-fast repeated analysis for AI coding agents
 License-File: LICENSE
 Keywords: agents,ai,codebase,context,developer-tools,llm
@@ -42,7 +42,7 @@ Description-Content-Type: text/markdown
 
 **Context · Impact · Migration · Architecture · Review — everything from one structural model.**
 
-![Version](https://img.shields.io/badge/version-5.5.1-blue)
+![Version](https://img.shields.io/badge/version-5.5.4-blue)
 ![Python](https://img.shields.io/badge/python-3.9%2B-green)
 
 > **ASK Engine** is the product. The CLI command is **`ask`**. The legacy **`sourcecode`**
@@ -126,7 +126,7 @@ brew tap haroundominique/sourcecode && brew install sourcecode
 # pip / pipx
 pipx install sourcecode        # or: pip install sourcecode
 
-ask version                    # ask 5.5.1 — and, on a build that has aged,
+ask version                    # ask 5.5.4 — and, on a build that has aged,
                                # how many releases have probably shipped since
 ```
 

{sourcecode-5.5.1.dist-info → sourcecode-5.5.4.dist-info}/RECORD

@@ -1,4 +1,4 @@
-sourcecode/__init__.py,sha256=naF38QL0K2KmWPa4bwGr7p-7cwqXbbiWTHF3C5Vp1Wo,308
+sourcecode/__init__.py,sha256=YOYdzwhKTESrTz_n8s2dcmsijpZC-UZTrzT3notGeRw,308
 sourcecode/adaptive_scanner.py,sha256=yJBKjNpkY6bpueYJ2YnRezen3sYZDecEt7WaaNWdqug,9466
 sourcecode/archetype.py,sha256=CZvRLpkHot_D8D3JFQVorr-EHDJyh0BbS7RnSTqBigM,40499
 sourcecode/architectural_baseline.py,sha256=4GiMVBLJVHvRKSWQGf2ZVurI1-R0qk5pmaevgvGSgVA,26325
@@ -19,7 +19,7 @@ sourcecode/chain_rules.py,sha256=Bi6UHfgd-GxWswmnHRcPz5jdbAuqka3Zkz_P-MTvqhw,127
 sourcecode/change_plan.py,sha256=aX1mp2XnlDN-8R2BcTiu8HlkDwQ_4fN1ZLckj0VMXHM,9945
 sourcecode/cir_graphs.py,sha256=9G0HHj1kw2325IDyzo2OpX73BNswEckecf4MZUXB4JM,12078
 sourcecode/classifier.py,sha256=JBzPwSSrDG-tUHAbcKB678HRbjLpD-ohzbzzO62mgpo,20114
-sourcecode/cli.py,sha256=AtpKG_Adheo-jF8xyyjzdV2KJRZgJxzkKlCXMutKbmU,647905
+sourcecode/cli.py,sha256=e5KKOjoGJw71izHDZAP04IHjZIVDhtfAS-DHeZIJpbg,649752
 sourcecode/client_calls.py,sha256=5_sBfFVXi6RgGiyWrlJKH4x4EF21do6gKlAOuFCKAoA,17709
 sourcecode/code_notes_analyzer.py,sha256=EJemNCNc9Dn-1RZYu-aNbK0ELzmsyC4s6FdHi3XyNEI,9392
 sourcecode/compare.py,sha256=2GdDy0qkDSgmkkpgvoCHitoLj3mt30EeFgJ8PGEXJng,14712
@@ -41,8 +41,8 @@ sourcecode/data_labels.py,sha256=c_hmgudTE68cVRdywfbfonr7nUWIL-JH2HOXB8Pplv0,845
 sourcecode/declarations.py,sha256=Jr7oJDg004sFw0TvJySNVggXFT2jB6hQgbeQK2qreVo,7000
 sourcecode/defect_identity.py,sha256=XS2VFDImg9NLaRImybKTeyk-CoDSsXmchldg0dwvM7c,7028
 sourcecode/degradation.py,sha256=BtuGUKFgkzI3Z6fwmUdxEUEaLFAcPQ_bHhU2w9ItDxw,18376
-sourcecode/dependency_analyzer.py,sha256=fQCaWQ7_BNcVgZv8gweJCwJ1CGlXsz8nw_tqN3aHJNE,63952
-sourcecode/deployment_prefix.py,sha256=g0_4bE_m9svjhnapRpOp8BJkj6Ul94Ij_iVezygMZ0I,11956
+sourcecode/dependency_analyzer.py,sha256=MRv6eh1TG5E9Y5Db-bNdZHDIJ3gWvvL3bBU1-9_QSZI,64593
+sourcecode/deployment_prefix.py,sha256=kPHPT6ZLg6CnMqCp9QwhbNNjDyS0sXAfKlPtSHUMggs,12180
 sourcecode/detach.py,sha256=PvkCcbIuTaDKKt5TGqtQLahmsmCkflui33jLnMEWkvg,8006
 sourcecode/doc_analyzer.py,sha256=05bjTUbDbmnbajD_cgRnACzS8T7xxBKVX4CjkJlhZg8,24411
 sourcecode/dynamic_argument_surface.py,sha256=u6zJoY-zChYVw1-SrgiyAq-rUwqMZ5ZqBwLfbPJ_8ng,2960
@@ -71,9 +71,10 @@ sourcecode/identity_fallback.py,sha256=uRC_j_80uaI4Ue4tf-OhKX4h472dmufYcjEFGSxY2
 sourcecode/integration_coordinates.py,sha256=7vGFxN4tCn_KTisYFJuixa8XkyHBoasEFB3fqM62OBI,7407
 sourcecode/jdk_exports.py,sha256=fCrlwNAXUT9gge_joq6kMnY3zJxYB2pxqy-0w3o3MJI,874
 sourcecode/license.py,sha256=vvgE1vBUWh8ozWI7USYuteh6noxntnjuzDQIPF-l4rA,27795
+sourcecode/manifest_cache.py,sha256=NkVtg0T5YfHdIiM63jStvcLcsznM6_bnTHib4t8_FW8,2158
 sourcecode/mcp_nudge.py,sha256=lKemOqK_wny2u7Ymcr2Idi5Kx8pXY02jCi-_nJYLGMg,2992
 sourcecode/metrics_analyzer.py,sha256=gLoRWKygF18jLgwsqmGXSWopw-f5iO1VbM7Jjdif6ag,22809
-sourcecode/migrate_check.py,sha256=8KKgsiiWu21XwCXgFV1bB8AVsB-pl6KNBTmMW6NYGB8,149028
+sourcecode/migrate_check.py,sha256=vdJ_2hOR9XOIQsz4czzamQMKKzHnNErg8vPdzzYqfMY,150019
 sourcecode/migration_blast.py,sha256=OPryWkM6-PbvcG3eJFdcWUnlpXBje41nu2BkrZnE2hg,9359
 sourcecode/non_coverage.py,sha256=sKp0J4_hKzKb_HMoXtdP91D0lMfEAPrOx44sqdQc18Y,23520
 sourcecode/openapi_surface.py,sha256=BTt0K-woZbkbWTN77IkqeBm_Okag9owR0848fmot8sk,16207
@@ -105,9 +106,9 @@ sourcecode/relevance_scorer.py,sha256=0AgEt4KrV73nioMqBgjhGjtY7L2C7L7cSyKtj3IKcr
 sourcecode/remedies.py,sha256=V2_cD4pLq8bKRkzF-3RGFoiu1xWYZS384WDVp1MrlZw,10225
 sourcecode/rename_refactor.py,sha256=h6dNFlB9aZ_3q6heeHBkgXQeXaT03nvPSsYH6P8qxFg,12965
 sourcecode/repo_classifier.py,sha256=FG1vaWKdWXsWdl-S8hjVMiTqcwgaRXkDyvK4rPcOGtQ,22681
-sourcecode/repository_ir.py,sha256=PXaggQQvuHD633B4BTBEXryxANSQTSKFHD869Ww6Bxk,381935
+sourcecode/repository_ir.py,sha256=f2dLU59dOopIowKey_lBiQNP5bFJ86h92XCKApH0d8A,387247
 sourcecode/ris.py,sha256=Xbx34dkJI_uXil9xsgw-fA_Apcm4nbnTjQgFm66dP1g,27489
-sourcecode/risk.py,sha256=432XEiL9xp8JBaxidwQInZBN4F98QJLAsqgZrh3CKow,81790
+sourcecode/risk.py,sha256=u2V7nG5yGW8VV4sFDSgFkZqdI6ltmilWhlQwlIH7d2o,82660
 sourcecode/rule_catalog.py,sha256=pTkgkQZ1u7atB3fkQJ9BPWo8lNiamvEuRN7MtRgle0c,5263
 sourcecode/rule_pass.py,sha256=DTNxD-nAecMJhIzI1LCd9FyRumInQSZjitacNh4Zoc8,14671
 sourcecode/runs.py,sha256=o4oXkB3WFJGScbK_C10OAWzsE18B9Zc55u_IE8_yl4M,6291
@@ -124,7 +125,7 @@ sourcecode/semantic_impact_engine.py,sha256=t09IirGC3JjQDy33JZd1_WKzQVKXkoNl3-XE
 sourcecode/semantic_integration_engine.py,sha256=7a0WqAInOv39f0Yr_94TYo_JP_8QpeI9KGaknpzAyQU,18899
 sourcecode/semantic_services.py,sha256=nbUuPv-F01USTt_9CHT8iy_ucCIw3fz4W3Aquea_pd4,10782
 sourcecode/serializer.py,sha256=cK4I3WzX0H-oBPbqzBsUBRux5GcYmH_yPDQwxvkf8yQ,141598
-sourcecode/servlet_surface.py,sha256=aOeRne07man8DwHsbog4FTOyF3sYV-oP_N1e4gt3Uvk,9115
+sourcecode/servlet_surface.py,sha256=2MLMMUQI1tbNq88iR7ahWm_17ZO7uI5mxXGLe9wthPQ,11387
 sourcecode/source_text.py,sha256=R52gmSljz7L6YVPbixv5W-LOZ5eKUVvSUP3S-QH_IWc,9047
 sourcecode/spring_event_topology.py,sha256=5_ON_21Le5zbG-1GRc5GLIi5HJfy_QjcXLVPC5WeUGQ,18055
 sourcecode/spring_findings.py,sha256=cW6NiQAFUM9mDYPitJr6MrQg--YBI41LxrSEMI9aeIQ,25272
@@ -212,10 +213,10 @@ sourcecode/telemetry/consent.py,sha256=pQdl-QeLl6Gcibn0eWHSKZrm-HYSsjpVqOnjrgFp8
 sourcecode/telemetry/events.py,sha256=4_yeO58U-Cwc1Qb27VB0_EjhmroY0k91n3_VGxeALB8,2776
 sourcecode/telemetry/filters.py,sha256=RzxauTz8HliO4BllQnXEXc7zTeqdCZi5MgqGEDuW7OQ,6570
 sourcecode/telemetry/transport.py,sha256=4gGHsq0WeY9VywEZXA3vUxykfiYnw9uuqfjAAec7F8o,1681
-sourcecode/_docs/DEFECT-LEDGER.md,sha256=sTFPEV2xx-L9n4mrG8UO4LxeETMJpGKHXyKQBZlqNog,497112
-sourcecode/_docs/USER_GUIDE.md,sha256=UUbSLDQlPTGbyn8n2eNC-bV1Am6iD0tGXoOQieHPDBg,79546
-sourcecode-5.5.1.dist-info/METADATA,sha256=i5TGIE9kswmF5uIu344XnclOs9X_Q9nIksMxuAMYUq4,47104
-sourcecode-5.5.1.dist-info/WHEEL,sha256=QccIxa26bgl1E6uMy58deGWi-0aeIkkangHcxk2kWfw,87
-sourcecode-5.5.1.dist-info/entry_points.txt,sha256=-JEAdChrK5We51kZcb7OaDcyil-dHBjBPL-NhuO-QY8,89
-sourcecode-5.5.1.dist-info/licenses/LICENSE,sha256=7DdHrU9Z_3e7dSvq4ISijZNjnuHo5NIHNiHDouMQ9JU,10491
-sourcecode-5.5.1.dist-info/RECORD,,
+sourcecode/_docs/DEFECT-LEDGER.md,sha256=ToyTn-J_Lqg64X2DeQQByyvoHCy_hYGK0FdeF8e4-cU,498363
+sourcecode/_docs/USER_GUIDE.md,sha256=vKyfdo74dvMxo761kpDcdCKPzM6QFK13UlidCitQAoQ,79837
+sourcecode-5.5.4.dist-info/METADATA,sha256=JDRjnkUfXvwtMmLyyheXfsu6nxX-_DEk1YQz8-FX7uI,47104
+sourcecode-5.5.4.dist-info/WHEEL,sha256=QccIxa26bgl1E6uMy58deGWi-0aeIkkangHcxk2kWfw,87
+sourcecode-5.5.4.dist-info/entry_points.txt,sha256=-JEAdChrK5We51kZcb7OaDcyil-dHBjBPL-NhuO-QY8,89
+sourcecode-5.5.4.dist-info/licenses/LICENSE,sha256=7DdHrU9Z_3e7dSvq4ISijZNjnuHo5NIHNiHDouMQ9JU,10491
+sourcecode-5.5.4.dist-info/RECORD,,