sourcecode 4.10.1__tar.gz → 4.10.3__tar.gz

{sourcecode-4.10.1 → sourcecode-4.10.3}/CHANGELOG.md

@@ -2,6 +2,50 @@
 
 ## [Unreleased]
 
+## [4.10.3] — 2026-08-05
+
+### Fixed
+
+- **`GATE-001` no longer reads Java control keywords as method calls.** The regex
+  body scan behind the gated-endpoint audit treated `if(...)`, `for(...)`,
+  `while(...)`, `switch(...)`, `catch(...)` and `synchronized(...)` as same-class
+  invocations, so a handler could be reported as `Controller#if`. Those symbols
+  reached the `risk` critical band and `audit-report` top risks. Control keywords
+  are now excluded from method-name matches; the real private-method gate finding
+  stays covered.
+- **`SEC-007` reports JavaMelody descriptor credentials in their field shape.**
+  `authorized-users` is now read directly from the named descriptor parameter even
+  when the value carries a digest (`{SHA-256}…`), spans multiple lines, or is
+  wrapped in `CDATA`, and the generic descriptor-value fallback is deduplicated so
+  one credential yields exactly one finding. The evidence line stays redacted.
+
+### Tests
+
+- Regression coverage binds both fixes through the full `spring-audit` CLI path,
+  including a `options/DPA/WEB-INF/web.xml` descriptor and a controller whose
+  handler body contains the control constructs that produced the false positive.
+
+## [4.10.2] — 2026-08-05
+
+### Fixed
+
+- **Durable run records are read-only by default.** Long repo-wide/deep commands
+  now write run records under `~/.sourcecode/runs/<repo-hash>/` instead of
+  `<repo>/.ask/runs`, with `ASK_RUNS_IN_REPO=1` as the explicit repo-local opt-in.
+- **`DEAD-001` comment scanning avoids ordinary files.** The commented-security
+  control pass now prefilters by security-control vocabulary before extracting
+  comment spans, preserving `DEAD-001` positives while reducing the cost added to
+  large `spring-audit` runs.
+- **Field regression guards for `spring-audit`.** End-to-end tests now lock that
+  reserved-keyword handler symbols such as `#if` cannot reach security findings,
+  and that JavaMelody `authorized-users` credentials in `web.xml` emit `SEC-007`.
+
+### Documentation
+
+- README, user guide, roadmap and defect ledger now record the 4.10.0 field
+  confirmation queue, the supervised heartbeat contract and the external run-record
+  location.
+
 ## [4.10.1] — 2026-08-05
 
 ### Fixed

{sourcecode-4.10.1 → sourcecode-4.10.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 4.10.1
+Version: 4.10.3
 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-4.10.1-blue)
+![Version](https://img.shields.io/badge/version-4.10.3-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`**
@@ -73,7 +73,14 @@ These numbers are cache-layer benchmarks, not a promise for every command. Per-s
 
 With the right command class, ASK Engine becomes **constant infrastructure** inside agent loops — call the fast/context surfaces before edits and reserve deep compositions for the points where their evidence is worth the runtime.
 
-**Large-repo controls.** Set `ASK_PROGRESS=1` to get line heartbeats in non-TTY supervised runs. Set `ASK_MAX_ANALYSIS_SECONDS=<seconds>` to fail before starting a repo-wide/deep command whose configured budget is below its safe floor. Use `ask audit-report --from-risk risk.json` to package an existing risk run instead of recomputing it.
+**Large-repo controls.** Set `ASK_PROGRESS=1` to get line heartbeats in non-TTY
+supervised runs. Long repo-wide/deep runs also write durable records under
+`~/.sourcecode/runs/<repo-hash>/` by default, not inside the audited repository;
+set `ASK_RUNS_IN_REPO=1` only when repo-local `.ask/runs` records are intentional.
+Set `ASK_MAX_ANALYSIS_SECONDS=<seconds>` to fail before starting a repo-wide/deep
+command whose configured budget is below its safe floor. Use
+`ask audit-report --from-risk risk.json` to package an existing risk run instead
+of recomputing it.
 
 **What a warm actually covers.** `ask cache warm` runs the compact analysis: it rebuilds the
 shared structural layers (L1/L2 + the Repository Intelligence Snapshot + the shared Canonical
@@ -99,7 +106,7 @@ brew tap haroundominique/sourcecode && brew install sourcecode
 # pip / pipx
 pipx install sourcecode        # or: pip install sourcecode
 
-ask version                    # ask 4.10.1 — and, on a build that has aged,
+ask version                    # ask 4.10.3 — and, on a build that has aged,
                                # how many releases have probably shipped since
 ```
 

{sourcecode-4.10.1 → sourcecode-4.10.3}/README.md

@@ -4,7 +4,7 @@
 
 **Context · Impact · Migration · Architecture · Review — everything from one structural model.**
 
-![Version](https://img.shields.io/badge/version-4.10.1-blue)
+![Version](https://img.shields.io/badge/version-4.10.3-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`**
@@ -35,7 +35,14 @@ These numbers are cache-layer benchmarks, not a promise for every command. Per-s
 
 With the right command class, ASK Engine becomes **constant infrastructure** inside agent loops — call the fast/context surfaces before edits and reserve deep compositions for the points where their evidence is worth the runtime.
 
-**Large-repo controls.** Set `ASK_PROGRESS=1` to get line heartbeats in non-TTY supervised runs. Set `ASK_MAX_ANALYSIS_SECONDS=<seconds>` to fail before starting a repo-wide/deep command whose configured budget is below its safe floor. Use `ask audit-report --from-risk risk.json` to package an existing risk run instead of recomputing it.
+**Large-repo controls.** Set `ASK_PROGRESS=1` to get line heartbeats in non-TTY
+supervised runs. Long repo-wide/deep runs also write durable records under
+`~/.sourcecode/runs/<repo-hash>/` by default, not inside the audited repository;
+set `ASK_RUNS_IN_REPO=1` only when repo-local `.ask/runs` records are intentional.
+Set `ASK_MAX_ANALYSIS_SECONDS=<seconds>` to fail before starting a repo-wide/deep
+command whose configured budget is below its safe floor. Use
+`ask audit-report --from-risk risk.json` to package an existing risk run instead
+of recomputing it.
 
 **What a warm actually covers.** `ask cache warm` runs the compact analysis: it rebuilds the
 shared structural layers (L1/L2 + the Repository Intelligence Snapshot + the shared Canonical
@@ -61,7 +68,7 @@ brew tap haroundominique/sourcecode && brew install sourcecode
 # pip / pipx
 pipx install sourcecode        # or: pip install sourcecode
 
-ask version                    # ask 4.10.1 — and, on a build that has aged,
+ask version                    # ask 4.10.3 — and, on a build that has aged,
                                # how many releases have probably shipped since
 ```
 

{sourcecode-4.10.1 → sourcecode-4.10.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "sourcecode"
-version = "4.10.1"
+version = "4.10.3"
 description = "Persistent structural context and ultra-fast repeated analysis for AI coding agents"
 readme = "README.md"
 requires-python = ">=3.9"

{sourcecode-4.10.1 → sourcecode-4.10.3}/src/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__ = "4.10.1"
+__version__ = "4.10.3"

{sourcecode-4.10.1 → sourcecode-4.10.3}/src/sourcecode/runs.py

@@ -2,6 +2,7 @@
 from __future__ import annotations
 
 import json
+import hashlib
 import os
 import time
 import uuid
@@ -18,7 +19,17 @@ def _now() -> float:
 
 
 def _runs_dir(root: Path) -> Path:
-    return Path(root).resolve() / ".ask" / "runs"
+    override = os.environ.get("ASK_RUNS_DIR", "").strip()
+    if override:
+        return Path(override).expanduser().resolve()
+    if os.environ.get("ASK_RUNS_IN_REPO", "").lower() in ("1", "true", "yes", "on"):
+        return Path(root).resolve() / ".ask" / "runs"
+    try:
+        key_source = str(Path(root).resolve())
+    except Exception:
+        key_source = str(Path(root).absolute())
+    repo_key = hashlib.sha256(key_source.encode("utf-8", errors="replace")).hexdigest()[:16]
+    return Path.home() / ".sourcecode" / "runs" / repo_key
 
 
 @dataclass

{sourcecode-4.10.1 → sourcecode-4.10.3}/src/sourcecode/security_config_scan.py

@@ -111,8 +111,8 @@ _IS_REFERENCE = re.compile(r"[$@]\{[^}]+\}|^\s*$|^\s*ENC\(|^\s*\*+\s*$")
 #: XML elements whose text is a configured VALUE. Servlet/JEE spec vocabulary.
 _DESCRIPTOR_VALUE = re.compile(
     r"<(?P<tag>param-value|env-entry-value|value|property|password)\b[^>]*>"
-    r"(?P<value>[^<]*)</(?P=tag)>",
-    re.IGNORECASE,
+    r"(?P<value>.*?)</(?P=tag)>",
+    re.IGNORECASE | re.DOTALL,
 )
 
 _DESCRIPTOR_INIT_PARAM = re.compile(
@@ -124,8 +124,8 @@ _DESCRIPTOR_PARAM_NAME = re.compile(
     re.IGNORECASE,
 )
 _DESCRIPTOR_PARAM_VALUE = re.compile(
-    r"<param-value\b[^>]*>(?P<value>[^<]*)</param-value>",
-    re.IGNORECASE,
+    r"<param-value\b[^>]*>(?P<value>.*?)</param-value>",
+    re.IGNORECASE | re.DOTALL,
 )
 _JAVAMELODY_CREDENTIAL_PARAMS: frozenset[str] = frozenset({
     "authorized-users",
@@ -212,6 +212,15 @@ _COMMENTED_CONFIG_CONTROLS: tuple[tuple[re.Pattern, str, str], ...] = (
     ),
 )
 
+_COMMENTED_JAVA_HINTS = (
+    "@PreAuthorize", "@PostAuthorize", "@Secured", "@RolesAllowed", "@DenyAll",
+    "@PermitAll", "FiltroSeguridad", "Security", "Authz", "Authorize",
+    "Authorized", "setSecure", "csrfTokenRepository", "csrf(",
+)
+_COMMENTED_XML_HINTS = (
+    "<transport-guarantee", "CONFIDENTIAL", "<security-constraint", "<intercept-url",
+)
+
 
 @dataclass(frozen=True)
 class SecurityConfigObservation:
@@ -355,7 +364,14 @@ def _scan_descriptor(text: str, rel: str) -> "list[SecurityConfigObservation]":
     holds a reference and is not reported; one that carries a digest literal
     holds the credential itself, in a file under version control.
     """
+    def _xml_value(raw: str) -> str:
+        value = raw.strip()
+        if value.startswith("<![CDATA[") and value.endswith("]]>"):
+            value = value[len("<![CDATA["):-len("]]>")]
+        return value.strip()
+
     out: "list[SecurityConfigObservation]" = []
+    javamelody_value_offsets: set[int] = set()
     for param in _DESCRIPTOR_INIT_PARAM.finditer(text):
         body = param.group("body")
         name_m = _DESCRIPTOR_PARAM_NAME.search(body)
@@ -363,22 +379,29 @@ def _scan_descriptor(text: str, rel: str) -> "list[SecurityConfigObservation]":
         if name_m is None or value_m is None:
             continue
         name = name_m.group("value").strip().lower()
-        value = value_m.group("value").strip()
+        value = _xml_value(value_m.group("value"))
         if name not in _JAVAMELODY_CREDENTIAL_PARAMS:
             continue
         if not value or _IS_REFERENCE.search(value):
             continue
         if ":" not in value:
             continue
-        if _DIGEST_LITERAL.search(value):
-            continue
         user = value.split(":", 1)[0].strip() or "user"
+        value_offset = param.start("body") + value_m.start("value")
+        javamelody_value_offsets.add(value_offset)
+        digest = _DIGEST_LITERAL.search(value)
+        evidence = (
+            f"<param-value>{user}:…{{{digest.group('algo')}}}"
+            f"{digest.group('digest')[:8]}…</param-value>"
+            if digest is not None
+            else f"<param-value>{user}:…</param-value>"
+        )
         out.append(SecurityConfigObservation(
             rule="SEC-007",
             file=rel,
-            line=_line_of(text, param.start() + value_m.start()),
+            line=_line_of(text, value_offset),
             symbol=rel,
-            evidence=f"<param-value>{user}:…</param-value>",
+            evidence=evidence,
             detail=(
                 "JavaMelody authorized-users stores a monitoring-console credential "
                 "directly in the deployment descriptor. The value is redacted here; "
@@ -386,7 +409,9 @@ def _scan_descriptor(text: str, rel: str) -> "list[SecurityConfigObservation]":
             ),
         ))
     for match in _DESCRIPTOR_VALUE.finditer(text):
-        value = match.group("value").strip()
+        if match.start("value") in javamelody_value_offsets:
+            continue
+        value = _xml_value(match.group("value"))
         if not value or _IS_REFERENCE.search(value):
             continue
         digest = _DIGEST_LITERAL.search(value)
@@ -456,6 +481,9 @@ def _scan_commented_controls(
     from sourcecode.source_text import commented_spans
 
     rules = _COMMENTED_XML_CONTROLS if xml else _COMMENTED_JAVA_CONTROLS
+    hints = _COMMENTED_XML_HINTS if xml else _COMMENTED_JAVA_HINTS
+    if not any(hint in source for hint in hints):
+        return []
     out: "list[SecurityConfigObservation]" = []
     seen: set[tuple[str, int]] = set()
     for start, end in commented_spans(source, xml=xml):

{sourcecode-4.10.1 → sourcecode-4.10.3}/src/sourcecode/spring_security_audit.py

@@ -508,7 +508,8 @@ def _line_of(source: str, offset: int) -> int:
 _JAVA_METHOD_RE = re.compile(
     r"(?P<mods>(?:(?:public|protected|private|static|final|synchronized|abstract|native)\s+)*)"
     r"(?P<ret>[\w.$<>\[\], ?]+)\s+"
-    r"(?P<name>[a-zA-Z_][\w$]*)\s*\([^;{}]*\)\s*(?:throws\s+[^{]+)?\{",
+    r"(?P<name>(?!(?:if|for|while|switch|catch|synchronized|return|new|throw|try|else|do)\b)"
+    r"[a-zA-Z_][\w$]*)\s*\([^;{}]*\)\s*(?:throws\s+[^{]+)?\{",
     re.MULTILINE,
 )