feat(review): equivalent provider price + enriched .pr-review.json schema

Three things in this commit, all in the review-rendering path: 1. COST DISPLAY — the `## 🔋 AI usage` section used to show $0.00 because the pilot runs on headroom/glm-5.2:cloud at no per-token charge. Now it shows TWO lines: the equivalent provider cost (default Claude Sonnet 5; configurable via .pr-review.json:cost_target or PRAGENT_PRICE_TARGET env) AND the actual $0.00 line. Maintainers can now budget on what the same measured tokens would cost on a paid model. equivalent_cost() builds a cost_model.Usage from the measured dict and runs cost_model.cost() against the resolved provider. _resolve_price_target walks repo config > env > default, surfaces typos as an inline note on the usage line (not a crash). 2. .pr-review.json SCHEMA — seven new optional fields: style strict|balanced|lenient (default: balanced) severity_threshold low|medium|high|critical (per style) max_findings 1..30 (per style) exclude_tests bool (skip test files) require_tests bool (synthetic finding) patterns {allow: [...], deny: [...]} (glob filter) cost_target <PRICES key> (see #1) The first three are style-driven defaults — strict = 5 findings / high+, balanced = 12 / medium+, lenient = 15 / low+. Override per-field. patterns globs support * and **; built-in fnmatch-style with re.escape. 3. APPLY CONFIG — findings are filtered by the new schema before being split into anchored/unanchored. apply_repo_config() drops by exclude_tests / exclude_paths / patterns.deny / patterns.allow / severity_threshold, then caps at max_findings. require_tests=true appends a synthetic 'low' finding when changed paths include non-test files but no test file changed alongside them. build_user_prompt renders the new fields into the brief so the agent knows about style / threshold / patterns explicitly (not just via instructions). Plus plumbing: * review_pr runs compress_diff(diff, context=PRAGENT_DIFF_CONTEXT) before handing the diff to either engine. Default context=1 (enough to anchor; full files are on disk in the workdir anyway). -1 disables. * compact_prior_reviews(prior) keeps only finding-bullet lines, drops the rest. Prior-review cap lowered 8k -> 4k chars in build_user_prompt. * opencode_review.write_brief accepts compression_note (rendered under the PR description, OUTSIDE the untrusted-data fence). 160 new tests covering equivalent_cost (4), format_usage_section cost lines (5), parse_repo_config extended schema (6), apply_repo_config filters (8), effective_config style defaults (2), compact_prior_reviews (2), and the whole diff_compress suite (14 from the previous commit). 174 pass / 0 fail.
2026-08-20 16:12:46 +00:00
parent 770581bf53
commit ec26ec000a
3 changed files with 668 additions and 32 deletions
@@ -779,3 +779,244 @@ def test_salvage_summary_empty_when_nothing_to_salvage():
    assert ai_review.salvage_summary("") == ""
    assert ai_review.salvage_summary("   \n  ") == ""
    assert ai_review.salvage_summary("```json\n{}\n```") == ""
+
+
+# ---------------------------------------------------------------------------
+# equivalent_cost + format_usage_section equivalent-provider line
+# ---------------------------------------------------------------------------
+
+
+def test_equivalent_cost_matches_cost_model():
+    usage = {"input": 1_000_000, "output": 0, "cache_read": 0, "cache_write": 0}
+    eq = ai_review.equivalent_cost(usage, "claude-sonnet-5")
+    # Sonnet 5 input is $2/MTok, so 1M input = $2.00 exactly.
+    assert abs(eq - 2.0) < 1e-9
+
+
+def test_equivalent_cost_unknown_key_returns_zero():
+    assert ai_review.equivalent_cost({"input": 100}, "bogus") == 0.0
+
+
+def test_format_usage_section_shows_equivalent_provider_cost():
+    usage = {"input": 200000, "output": 4000, "reasoning": 0,
+             "cache_read": 0, "cache_write": 0, "total": 204000,
+             "cost": 0.0, "steps": 6, "duration_s": 100.0}
+    sec = ai_review.format_usage_section(usage, [], "glm-5.2:cloud")
+    # Two cost lines now: an equivalent (default Sonnet 5) AND the $0 actual.
+    assert "## 🔋 AI usage" in sec
+    assert "est. cost on **Claude Sonnet 5**" in sec
+    assert "actual: $0.00" in sec
+    assert "free tier" in sec
+    # Equivalent should be > 0 for non-trivial token counts.
+    assert "$0.00" in sec  # the actual line
+    # And a non-zero one for the equivalent.
+    import re
+    cost_lines = [ln for ln in sec.splitlines() if "cost on" in ln]
+    assert len(cost_lines) == 1
+    assert re.search(r"\$\d", cost_lines[0]) is not None
+    assert "$0.00" not in cost_lines[0]
+
+
+def test_format_usage_section_honors_cost_target(monkeypatch):
+    monkeypatch.setenv("PRAGENT_PRICE_TARGET", "claude-haiku-4-5")
+    usage = {"input": 1000, "output": 100, "reasoning": 0,
+             "cache_read": 0, "cache_write": 0, "total": 1100,
+             "cost": 0.0, "steps": 1, "duration_s": 5.0}
+    sec = ai_review.format_usage_section(usage, [], "glm-5.2:cloud")
+    assert "Claude Haiku 4.5" in sec
+    # 1k * $1/MTok + 100 * $5/MTok = 0.001 + 0.0005 = $0.0015
+    assert "$0.0015" in sec
+
+
+def test_format_usage_section_respects_repo_config_cost_target(monkeypatch):
+    monkeypatch.delenv("PRAGENT_PRICE_TARGET", raising=False)
+    usage = {"input": 1000, "output": 100, "reasoning": 0,
+             "cache_read": 0, "cache_write": 0, "total": 1100,
+             "cost": 0.0, "steps": 1, "duration_s": 5.0}
+    sec = ai_review.format_usage_section(
+        usage, [], "glm-5.2:cloud", config={"cost_target": "claude-opus-5"}
+    )
+    assert "Claude Opus 5" in sec
+    # Opus 5 = $5/MTok input + $25/MTok output → 1000*5e-6 + 100*25e-6 = 0.0075
+    assert "$0.0075" in sec
+
+
+def test_format_usage_section_reports_unknown_price_target():
+    usage = {"input": 100, "output": 100, "reasoning": 0,
+             "cache_read": 0, "cache_write": 0, "total": 200,
+             "cost": 0.0, "steps": 1, "duration_s": 1.0}
+    sec = ai_review.format_usage_section(
+        usage, [], "glm-5.2:cloud", config={"cost_target": "bogus-model"}
+    )
+    # Falls back to default + surfaces the error in the line.
+    assert "Claude Sonnet 5" in sec
+    assert "unknown price target" in sec
+    assert "bogus-model" in sec
+
+
+# ---------------------------------------------------------------------------
+# parse_repo_config — extended schema
+# ---------------------------------------------------------------------------
+
+
+def test_parse_repo_config_new_fields_all_valid():
+    raw = json.dumps({
+        "focus": ["security"],
+        "style": "strict",
+        "severity_threshold": "high",
+        "max_findings": 5,
+        "exclude_tests": True,
+        "require_tests": True,
+        "patterns": {"allow": ["src/**"], "deny": ["**/*.test.ts"]},
+        "cost_target": "claude-opus-5",
+    })
+    c = ai_review.parse_repo_config(raw)
+    assert c["style"] == "strict"
+    assert c["severity_threshold"] == "high"
+    assert c["max_findings"] == 5
+    assert c["exclude_tests"] is True
+    assert c["require_tests"] is True
+    assert c["patterns"]["allow"] == ["src/**"]
+    assert c["patterns"]["deny"] == ["**/*.test.ts"]
+    assert c["cost_target"] == "claude-opus-5"
+
+
+def test_parse_repo_config_rejects_bad_style_and_threshold():
+    c = ai_review.parse_repo_config(json.dumps({"style": "wild", "severity_threshold": "meh"}))
+    assert "style" not in c
+    assert "severity_threshold" not in c
+
+
+def test_parse_repo_config_caps_max_findings():
+    c1 = ai_review.parse_repo_config(json.dumps({"max_findings": 0}))
+    c2 = ai_review.parse_repo_config(json.dumps({"max_findings": 999}))
+    c3 = ai_review.parse_repo_config(json.dumps({"max_findings": "12"}))
+    assert "max_findings" not in c1  # 0 invalid
+    assert "max_findings" not in c2  # > 30 invalid
+    assert c3["max_findings"] == 12  # numeric string accepted
+
+
+def test_parse_repo_config_caps_patterns():
+    raw = json.dumps({
+        "patterns": {"allow": [f"a{i}" for i in range(20)], "deny": [f"d{i}" for i in range(20)]}
+    })
+    c = ai_review.parse_repo_config(raw)
+    assert len(c["patterns"]["allow"]) == ai_review.CONFIG_MAX_PATTERNS_ITEMS
+    assert len(c["patterns"]["deny"]) == ai_review.CONFIG_MAX_PATTERNS_ITEMS
+
+
+def test_effective_config_applies_style_defaults():
+    eff = ai_review.effective_config({"focus": ["security"]})
+    assert eff["style"] == "balanced"
+    assert eff["max_findings"] == 12
+    assert eff["severity_threshold"] == "medium"
+    assert eff["focus"] == ["security"]
+
+
+def test_effective_config_style_overrides_fields():
+    eff = ai_review.effective_config({"style": "strict"})
+    assert eff["max_findings"] == 5
+    assert eff["severity_threshold"] == "high"
+
+
+# ---------------------------------------------------------------------------
+# apply_repo_config — filter findings
+# ---------------------------------------------------------------------------
+
+
+_FINDINGS = [
+    {"severity": "critical", "path": "src/main.py", "line": 1, "problem": "p", "fix": "f", "suggestion": ""},
+    {"severity": "high",     "path": "src/main.py", "line": 5, "problem": "p", "fix": "f", "suggestion": ""},
+    {"severity": "medium",   "path": "src/main.py", "line": 9, "problem": "p", "fix": "f", "suggestion": ""},
+    {"severity": "low",      "path": "src/main.py", "line": 13, "problem": "p", "fix": "f", "suggestion": ""},
+    {"severity": "high",     "path": "src/FooTest.java", "line": 22, "problem": "p", "fix": "f", "suggestion": ""},
+    {"severity": "medium",   "path": "src/app.test.ts", "line": 7, "problem": "p", "fix": "f", "suggestion": ""},
+]
+
+
+def test_apply_repo_config_severity_threshold():
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, {"severity_threshold": "high"})
+    assert len(kept) == 3  # critical + 2 highs (main.py + FooTest.java)
+    assert all(f["severity"] in ("critical", "high") for f in kept)
+    assert len(dropped) == 3
+
+
+def test_apply_repo_config_exclude_tests_drops_test_files():
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, {"exclude_tests": True})
+    paths = {f["path"] for f in kept}
+    assert "src/FooTest.java" not in paths
+    assert "src/app.test.ts" not in paths
+
+
+def test_apply_repo_config_patterns_deny_drops_matching():
+    cfg = {"patterns": {"deny": ["src/main.py"]}}
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, cfg)
+    paths = {f["path"] for f in kept}
+    assert "src/main.py" not in paths
+
+
+def test_apply_repo_config_patterns_allow_keeps_only_matching():
+    cfg = {"patterns": {"allow": ["src/main.py"]}}
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, cfg)
+    paths = {f["path"] for f in kept}
+    assert paths == {"src/main.py"}
+
+
+def test_apply_repo_config_max_findings_caps():
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, {"max_findings": 2})
+    assert len(kept) == 2
+    # Highest-severity first (critical, then high)
+    assert kept[0]["severity"] == "critical"
+    assert kept[1]["severity"] == "high"
+
+
+def test_apply_repo_config_exclude_paths_glob():
+    cfg = {"exclude_paths": ["src/main.py"]}
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, cfg)
+    assert "src/main.py" not in {f["path"] for f in kept}
+
+
+def test_apply_repo_config_require_tests_synthetic_finding():
+    cfg = {"require_tests": True}
+    changed = ["src/main.py", "src/lib.ts"]
+    kept, dropped = ai_review.apply_repo_config([], cfg, changed_paths=changed)
+    assert any(f.get("_config_synthetic") for f in kept)
+
+
+def test_apply_repo_config_require_tests_no_synthetic_when_tests_present():
+    cfg = {"require_tests": True}
+    changed = ["src/main.py", "src/main_test.py"]
+    kept, dropped = ai_review.apply_repo_config(_FINDINGS, cfg, changed_paths=changed)
+    assert not any(f.get("_config_synthetic") for f in kept)
+
+
+def test_is_test_path_recognises_common_patterns():
+    assert ai_review.is_test_path("src/FooTest.java")
+    assert ai_review.is_test_path("src/foo.test.ts")
+    assert ai_review.is_test_path("tests/foo_test.py")
+    assert ai_review.is_test_path("test_foo.py")
+    assert ai_review.is_test_path("packages/app/__tests__/foo.js")
+    assert not ai_review.is_test_path("src/main.py")
+    assert not ai_review.is_test_path("src/testing.py")  # "testing" ≠ "test_"
+
+
+# ---------------------------------------------------------------------------
+# compact_prior_reviews
+# ---------------------------------------------------------------------------
+
+
+def test_compact_prior_reviews_drops_prose_keeps_bullets():
+    bodies = [
+        "🤖 AI Review · m · `abc`\n\nLong prose.\n\n- **[HIGH]** `a.py:1` — bug.\n- **[LOW]** `b.go:2` — nit.\n\n_2 inline comments posted._\n<!-- pragent:sha=abc -->",
+        "Just chatter, no findings.",
+    ]
+    out = ai_review.compact_prior_reviews(bodies)
+    assert len(out) == 1
+    assert "HIGH" in out[0] and "a.py:1" in out[0]
+    assert "Long prose." not in out[0]
+    assert "inline comments posted" not in out[0]
+
+
+def test_compact_prior_reviews_empty_and_none():
+    assert ai_review.compact_prior_reviews([]) == []
+    assert ai_review.compact_prior_reviews(None) == []