sourcecode 1.60.0__tar.gz → 1.63.0__tar.gz

{sourcecode-1.60.0 → sourcecode-1.63.0}/CHANGELOG.md

@@ -1,5 +1,110 @@
 # Changelog
 
+## [1.63.0] — 2026-06-29
+
+### Added — actionable Hibernate rewrite targets (`migrate-check`)
+
+The Hibernate stratification went from **diagnostic** (how many files/occurrences
+per layer) to **actionable** (what to rewrite, where, and to what) so a migration
+agent can consume the output directly without re-parsing the repo. The `hibernate`
+section gains a sub-`schema_version` (now **2.0**); all existing keys preserved.
+
+- **`hibernate.rewrite_targets[]`** — one machine-readable target per call site:
+  `id`, `layer`, `source_file`, `line_start`/`line_end`, `current_pattern`,
+  `current_snippet`, `target_api` (the Hibernate-6 destination), `migration_kind`
+  (`manual_rewrite` / `assisted` / `mechanical` / `review`), `auto_migratable`,
+  `blocking_reason`, `symbol` (enclosing `Class#method`), `module`, `dynamic`.
+  Mapping rules: legacy `org.hibernate.Criteria`/`Restrictions`/`Projections` →
+  `CriteriaBuilder` (manual); `@Type(type=)` → `@Type(value=)`/`@JdbcTypeCode`
+  (mechanical, auto); `@TypeDef`/`@GenericGenerator` → assisted; concatenated HQL →
+  assisted + `runtime-resolved string`; `UserType`/`Interceptor`/`EventListener`/SPI
+  → its H6 contract (manual). Standard JPA annotations emit **no** targets.
+- **Per-layer migration-kind sub-counts** in `risk_matrix` (`manual_count`,
+  `assisted_count`, `mechanical_count`, `review_count`); Criteria adds
+  `static_count` vs `dynamic_count`; SPI adds `userType_rewrite_count` vs
+  `userType_resolvable_count` — so dynamic/reflexive cost is not blended into static.
+- **Honest effort** — each layer carries `effort_range` `{low, high, confidence}`
+  (legacy `estimated_effort` string kept); `hibernate.total_effort_range_days`
+  aggregates them; `hibernate.effort_model` exposes the auditable formula and the
+  caveat that layers may share files (the total is an upper bound, not deduplicated).
+- **`hibernate_readiness` (0–100)** — 4th readiness dimension alongside
+  jakarta/boot3/jdk. It does **not** sink the headline `readiness_score` (Hibernate
+  is an orthogonal rewrite axis), but in a rewrite zone `headline_blocker` is set to
+  `"hibernate_rewrite"` so a reader of "62/100" is not misled.
+- **`hibernate.golden_sql_hotspots[]`** — classes/methods that concentrate dynamic
+  query generation, ranked by reflexive-query volume — where to pin golden-SQL tests.
+
+### Fixed
+
+- `get_migration_readiness` MCP test asserted the pre-1.62 report `schema_version`
+  `1.3`; the v1.62.0 schema bump to `1.4` made it stale. Corrected to `1.4`.
+
+## [1.62.0] — 2026-06-29
+
+### Added — Hibernate 5.x → 6.x migration stratification (`migrate-check`)
+
+`migrate-check` no longer treats Hibernate as a single dependency-version bump.
+For enterprise monoliths that use dynamic persistence (verified against the
+Broadleaf Commerce CE Open Admin Platform pattern), a Hibernate major upgrade is
+a **persistence-layer rewrite problem in the dynamic-ORM areas, not an upgrade**.
+Output now carries a `hibernate` section (and a Hibernate block in `--format text`)
+built by the new `hibernate_strat` module.
+
+- **Four independent migration domains** — each classified on its own risk axis,
+  never aggregated into one score:
+  1. **JPA Annotation Layer** — LOW, escalates to HIGH on deprecated/reworked
+     annotations (`@Type(type=)`, `@TypeDef`, `@GenericGenerator`).
+  2. **Criteria API Layer** — HIGH, escalates to **CRITICAL** when Criteria is
+     built via reflection or abstraction DAOs (`DynamicEntityDao`, `GenericDao`,
+     `BasicPersistenceModule`). Legacy `org.hibernate.Criteria` /
+     `Restrictions` / `Projections` flagged (removed in Hibernate 6).
+  3. **HQL / String Queries** — MEDIUM, escalates to HIGH on string concatenation
+     (SQL shape not statically inferable under H6 parser changes).
+  4. **Hibernate SPI / Internal API** — CRITICAL blocker (`UserType`,
+     `CompositeUserType`, `Interceptor`, `EventListener`, `org.hibernate.engine.spi`).
+- **Hibernate Migration Impact Matrix** — per-layer risk, reason, and estimated effort.
+- **Module-level Hibernate exposure map** — per Maven/Gradle module: max risk,
+  layers present, and `dynamic-criteria` / `custom-SPI` / `reflection` tags.
+- **Critical call-chain detection** — flags dynamic query-generation paths
+  (`DynamicEntityDaoImpl`, `BasicPersistenceModule`, reflection-based DAOs).
+- **Stop-condition logic** — any of {dynamic Criteria, custom SPI, reflection-built
+  queries, concatenated query strings} forces the verdict
+  **"HIBERNATE 6 MIGRATION IS HIGH RISK REWRITE ZONE (NOT UPGRADE ZONE)"**;
+  otherwise UPGRADE-WITH-CARE or UPGRADE ZONE.
+- **Risk separation** — observable code risk vs inferred runtime risk reported distinctly.
+
+## [1.61.0] — 2026-06-29
+
+### Fixed — `migrate-check` false positives on already-migrated Spring Boot 3 repos
+
+Three defects made `migrate-check` report a Boot 3 / Jakarta repo as an
+un-migrated Boot 2 project (verified on Broadleaf Commerce CE 7.0.x —
+Boot 3.5.14, Java 17, `jakarta.*` fully adopted). The verdict contradicted the
+code; this corrects all three at the root and adds regression guardrails.
+Report `schema_version` bumped **1.2 → 1.3**.
+
+- **`spring_boot_2_detected` no longer defaults to `true`.** The detector now
+  resolves Maven `${...}` properties before reading versions and recognises
+  Boot in every form: `spring-boot-starter-parent`, a managed
+  `spring-boot-dependencies` BOM, a `spring-boot*` dependency versioned by
+  property, and the Gradle plugin. The old heuristic used a `DOTALL` regex that
+  matched any stray `2.x` library version anywhere in the pom. The field is now
+  **tri-state** — `true` (Boot 2 confirmed) / `false` (Boot 3+ confirmed) /
+  `null` (undetermined). **Absence of evidence is never reported as `true`.**
+  A new `spring_boot_version_detected` exposes the resolved version. Massive
+  `jakarta.*` import adoption vetoes any Boot-2 verdict.
+- **`readiness_score` no longer collapses to 0 on JDK-modernization volume.**
+  Blockers (critical/high) still floor a genuinely blocked repo, but orthogonal
+  JDK debt (`java.util.Date`, reflection) is capped so it can no longer sink a
+  jakarta-ready repo. The report now separates dimensions:
+  `jakarta_readiness`, `boot3_readiness`, and `jdk_modernization` (each 0–100).
+- **`javax.*` JDK/JSR namespaces are no longer flagged with
+  `javax-to-jakarta-migration-risk`.** `javax.cache` (JSR-107), `javax.sql`,
+  `javax.xml` (JAXP), `javax.naming`, `javax.management`, `javax.crypto`, etc.
+  keep the `javax` prefix forever and are now allowlisted. Only the Jakarta EE 9
+  renamed namespaces (`javax.servlet`, `javax.persistence`, `javax.validation`,
+  `javax.xml.bind`, …) still carry the flag.
+
 ## [1.60.0] — 2026-06-29
 
 ### Changed

{sourcecode-1.60.0 → sourcecode-1.63.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 1.60.0
+Version: 1.63.0
 Summary: Persistent structural context and ultra-fast repeated analysis for AI coding agents
 License-File: LICENSE
 Keywords: agents,ai,codebase,context,developer-tools,llm
@@ -657,6 +657,58 @@ Detects migration blockers across Java source files, Spring XML config files, an
 
 Each finding includes `severity`, `title`, `source_file`, `first_line`, `explanation`, `fix_hint`, `migration_target`, and `openrewrite_recipe` (when an automated recipe exists).
 
+#### Hibernate 5.x → 6.x stratification (the `hibernate` output section)
+
+A Hibernate major upgrade is **not** a single dependency bump for systems that use
+dynamic persistence. `migrate-check` stratifies Hibernate exposure into four
+independent migration domains — never one aggregated score — and emits **actionable,
+machine-readable rewrite targets** so a migration agent can consume the output
+directly instead of re-parsing the repo. Sub-`schema_version`: `2.0`.
+
+**Four layers** (each on its own risk axis):
+
+| Layer | Baseline | Escalates to |
+|-------|----------|--------------|
+| `jpa_annotations` | LOW (namespace handled by jakarta) | HIGH on deprecated `@Type(type=)` / `@TypeDef` / `@GenericGenerator` |
+| `criteria_api` | HIGH (JPA Criteria semantics changed; legacy `org.hibernate.Criteria` removed) | **CRITICAL** when built via reflection / abstraction DAOs (`DynamicEntityDao`, `GenericDao`, `BasicPersistenceModule`) |
+| `hql_string_queries` | MEDIUM (revalidate against H6 parser) | HIGH on string concatenation (SQL shape not statically inferable) |
+| `hibernate_spi_internal` | CRITICAL blocker | `UserType`, `CompositeUserType`, `Interceptor`, `EventListener`, `org.hibernate.engine.spi` |
+
+**Output keys (under `hibernate`):**
+
+- `classification` — `upgrade_zone` / `upgrade_with_care` / `rewrite_zone`. Any of
+  {dynamic Criteria, custom SPI, reflection-built queries, concatenated query
+  strings} forces **`rewrite_zone`** ("HIGH RISK REWRITE ZONE, NOT UPGRADE ZONE").
+- `risk_matrix[]` — per layer: `risk`, `reason`, `effort_range {low, high, confidence}`,
+  `file_count`, `occurrence_count`, migration-kind sub-counts (`manual_count`,
+  `assisted_count`, `mechanical_count`, `review_count`); Criteria adds
+  `static_count` vs `dynamic_count`; SPI adds `userType_rewrite_count` vs
+  `userType_resolvable_count`.
+- `rewrite_targets[]` — one actionable target per call site: `id`, `layer`,
+  `source_file`, `line_start`/`line_end`, `current_pattern`, `current_snippet`,
+  `target_api` (the Hibernate-6 destination), `migration_kind`
+  (`manual_rewrite` / `assisted` / `mechanical` / `review`), `auto_migratable`,
+  `blocking_reason`, `symbol` (enclosing `Class#method`), `module`, `dynamic`.
+- `module_exposure_map` — per Maven/Gradle module: `max_risk`, layers present, and
+  `dynamic-criteria` / `custom-SPI` / `reflection` tags.
+- `critical_call_chains[]` — dynamic query-generation paths (reflection-based DAOs).
+- `golden_sql_hotspots[]` — classes/methods ranked by dynamic-query volume — where
+  to pin golden-SQL behaviour tests before migrating.
+- `total_effort_range_days` + `effort_model` — aggregate range plus the auditable
+  formula (and the caveat that layers may share files, so the total is an upper bound).
+- `stop_conditions_triggered[]`, `risk_separation` (observable vs inferred runtime risk).
+
+The report also exposes **`hibernate_readiness`** (0–100) as a fourth readiness
+dimension alongside `jakarta_readiness` / `boot3_readiness` / `jdk_modernization`.
+Hibernate is an orthogonal rewrite axis, so it does not sink the headline
+`readiness_score`; instead, in a rewrite zone the top-level `headline_blocker` is set
+to `"hibernate_rewrite"` so a reader of the headline score is not misled.
+
+```bash
+# inspect only the Hibernate rewrite targets
+sourcecode migrate-check . --format json | jq '.hibernate.rewrite_targets[]'
+```
+
 ### `rename-class` — Java class rename
 
 ```bash

{sourcecode-1.60.0 → sourcecode-1.63.0}/README.md

@@ -619,6 +619,58 @@ Detects migration blockers across Java source files, Spring XML config files, an
 
 Each finding includes `severity`, `title`, `source_file`, `first_line`, `explanation`, `fix_hint`, `migration_target`, and `openrewrite_recipe` (when an automated recipe exists).
 
+#### Hibernate 5.x → 6.x stratification (the `hibernate` output section)
+
+A Hibernate major upgrade is **not** a single dependency bump for systems that use
+dynamic persistence. `migrate-check` stratifies Hibernate exposure into four
+independent migration domains — never one aggregated score — and emits **actionable,
+machine-readable rewrite targets** so a migration agent can consume the output
+directly instead of re-parsing the repo. Sub-`schema_version`: `2.0`.
+
+**Four layers** (each on its own risk axis):
+
+| Layer | Baseline | Escalates to |
+|-------|----------|--------------|
+| `jpa_annotations` | LOW (namespace handled by jakarta) | HIGH on deprecated `@Type(type=)` / `@TypeDef` / `@GenericGenerator` |
+| `criteria_api` | HIGH (JPA Criteria semantics changed; legacy `org.hibernate.Criteria` removed) | **CRITICAL** when built via reflection / abstraction DAOs (`DynamicEntityDao`, `GenericDao`, `BasicPersistenceModule`) |
+| `hql_string_queries` | MEDIUM (revalidate against H6 parser) | HIGH on string concatenation (SQL shape not statically inferable) |
+| `hibernate_spi_internal` | CRITICAL blocker | `UserType`, `CompositeUserType`, `Interceptor`, `EventListener`, `org.hibernate.engine.spi` |
+
+**Output keys (under `hibernate`):**
+
+- `classification` — `upgrade_zone` / `upgrade_with_care` / `rewrite_zone`. Any of
+  {dynamic Criteria, custom SPI, reflection-built queries, concatenated query
+  strings} forces **`rewrite_zone`** ("HIGH RISK REWRITE ZONE, NOT UPGRADE ZONE").
+- `risk_matrix[]` — per layer: `risk`, `reason`, `effort_range {low, high, confidence}`,
+  `file_count`, `occurrence_count`, migration-kind sub-counts (`manual_count`,
+  `assisted_count`, `mechanical_count`, `review_count`); Criteria adds
+  `static_count` vs `dynamic_count`; SPI adds `userType_rewrite_count` vs
+  `userType_resolvable_count`.
+- `rewrite_targets[]` — one actionable target per call site: `id`, `layer`,
+  `source_file`, `line_start`/`line_end`, `current_pattern`, `current_snippet`,
+  `target_api` (the Hibernate-6 destination), `migration_kind`
+  (`manual_rewrite` / `assisted` / `mechanical` / `review`), `auto_migratable`,
+  `blocking_reason`, `symbol` (enclosing `Class#method`), `module`, `dynamic`.
+- `module_exposure_map` — per Maven/Gradle module: `max_risk`, layers present, and
+  `dynamic-criteria` / `custom-SPI` / `reflection` tags.
+- `critical_call_chains[]` — dynamic query-generation paths (reflection-based DAOs).
+- `golden_sql_hotspots[]` — classes/methods ranked by dynamic-query volume — where
+  to pin golden-SQL behaviour tests before migrating.
+- `total_effort_range_days` + `effort_model` — aggregate range plus the auditable
+  formula (and the caveat that layers may share files, so the total is an upper bound).
+- `stop_conditions_triggered[]`, `risk_separation` (observable vs inferred runtime risk).
+
+The report also exposes **`hibernate_readiness`** (0–100) as a fourth readiness
+dimension alongside `jakarta_readiness` / `boot3_readiness` / `jdk_modernization`.
+Hibernate is an orthogonal rewrite axis, so it does not sink the headline
+`readiness_score`; instead, in a rewrite zone the top-level `headline_blocker` is set
+to `"hibernate_rewrite"` so a reader of the headline score is not misled.
+
+```bash
+# inspect only the Hibernate rewrite targets
+sourcecode migrate-check . --format json | jq '.hibernate.rewrite_targets[]'
+```
+
 ### `rename-class` — Java class rename
 
 ```bash

{sourcecode-1.60.0 → sourcecode-1.63.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "sourcecode"
-version = "1.60.0"
+version = "1.63.0"
 description = "Persistent structural context and ultra-fast repeated analysis for AI coding agents"
 readme = "README.md"
 requires-python = ">=3.9"

{sourcecode-1.60.0 → sourcecode-1.63.0}/src/sourcecode/__init__.py

@@ -1,3 +1,3 @@
 """sourcecode — Deterministic codebase context maps for AI coding agents."""
 
-__version__ = "1.60.0"
+__version__ = "1.63.0"

{sourcecode-1.60.0 → sourcecode-1.63.0}/src/sourcecode/cli.py

@@ -4838,6 +4838,16 @@ def migrate_check_cmd(
       MIG-007  javax.inject import (MEDIUM)
       MIG-008  javax.ws.rs import (MEDIUM — JAX-RS changed)
 
+    \b
+    Hibernate 5→6 stratification (in the 'hibernate' output section):
+      4 independent layers (JPA annotations / Criteria / HQL / SPI), a per-layer
+      risk matrix with effort ranges, a module exposure map, critical call-chain
+      detection, golden-SQL hotspots, a hibernate_readiness score, and an UPGRADE
+      vs REWRITE verdict (dynamic Criteria, custom SPI, reflection-built queries,
+      or concatenated query strings force the REWRITE classification).
+      Emits actionable rewrite_targets[] (call-site line ranges → target_api +
+      migration_kind) so a migration agent can consume the output directly.
+
     \b
     Examples:
       sourcecode migrate-check .