weavatrix 0.2.12 → 0.2.14
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +90 -21
- package/SECURITY.md +2 -2
- package/docs/releases/v0.2.14.md +93 -0
- package/package.json +2 -2
- package/skill/SKILL.md +108 -43
- package/src/analysis/allowed-test-runner.js +20 -9
- package/src/analysis/architecture/contract-graph.js +119 -0
- package/src/analysis/architecture/contract-schema.js +110 -0
- package/src/analysis/architecture/contract-storage.js +35 -0
- package/src/analysis/architecture/contract-verification.js +168 -0
- package/src/analysis/architecture-contract.js +16 -343
- package/src/analysis/change-classification/diff-parser.js +103 -0
- package/src/analysis/change-classification/options.js +40 -0
- package/src/analysis/change-classification/symbol-classifier.js +177 -0
- package/src/analysis/change-classification.js +120 -519
- package/src/analysis/dead-code-review/policy.js +23 -0
- package/src/analysis/dead-code-review.js +9 -19
- package/src/analysis/dep-check.js +10 -57
- package/src/analysis/dep-rules.js +10 -303
- package/src/analysis/dependency/scoped-dependencies.js +40 -0
- package/src/analysis/endpoints/common.js +62 -0
- package/src/analysis/endpoints/extract.js +107 -0
- package/src/analysis/endpoints/inventory.js +84 -0
- package/src/analysis/endpoints/mounts.js +129 -0
- package/src/analysis/endpoints-java.js +80 -3
- package/src/analysis/endpoints.js +3 -448
- package/src/analysis/git-history/analytics.js +177 -0
- package/src/analysis/git-history/collector.js +109 -0
- package/src/analysis/git-history/options.js +68 -0
- package/src/analysis/git-history.js +128 -577
- package/src/analysis/health-capabilities.js +82 -0
- package/src/analysis/http-contracts/analysis.js +169 -0
- package/src/analysis/http-contracts/client-call-detection.js +81 -0
- package/src/analysis/http-contracts/client-call-parser.js +255 -0
- package/src/analysis/http-contracts/graph-context.js +143 -0
- package/src/analysis/http-contracts/matching.js +61 -0
- package/src/analysis/http-contracts/shared.js +86 -0
- package/src/analysis/http-contracts.js +6 -822
- package/src/analysis/internal-audit/dependency-health.js +111 -0
- package/src/analysis/internal-audit/python-manifests.js +65 -0
- package/src/analysis/internal-audit/repo-files.js +55 -0
- package/src/analysis/internal-audit/supply-chain.js +132 -0
- package/src/analysis/internal-audit.collect.js +5 -106
- package/src/analysis/internal-audit.run.js +113 -200
- package/src/analysis/jvm-dependency-evidence.js +69 -0
- package/src/analysis/source-correctness/retry-patterns.js +93 -0
- package/src/analysis/source-correctness.js +225 -0
- package/src/analysis/structure/dependency-graph.js +156 -0
- package/src/analysis/structure/findings.js +142 -0
- package/src/graph/builder/go-receiver-resolution.js +112 -0
- package/src/graph/builder/js/queries.js +48 -0
- package/src/graph/builder/lang-go.js +89 -4
- package/src/graph/builder/lang-js.js +4 -44
- package/src/graph/builder/pass2-resolution.js +68 -0
- package/src/graph/freshness-probe.js +2 -1
- package/src/graph/incremental-refresh.js +3 -2
- package/src/graph/internal-builder.build.js +22 -183
- package/src/graph/internal-builder.pass2.js +161 -0
- package/src/graph/internal-builder.resolvers.js +2 -105
- package/src/graph/resolvers/rust.js +117 -0
- package/src/mcp/actions/advisories.mjs +18 -0
- package/src/mcp/actions/graph-lifecycle.mjs +148 -0
- package/src/mcp/actions/graph-sync.mjs +195 -0
- package/src/mcp/actions/hosted-architecture.mjs +57 -0
- package/src/mcp/architecture-bootstrap.mjs +168 -0
- package/src/mcp/architecture-starter.mjs +234 -0
- package/src/mcp/catalog.mjs +22 -6
- package/src/mcp/evidence/bun-lock-graph.mjs +209 -0
- package/src/mcp/evidence/package-graph-common.mjs +115 -0
- package/src/mcp/evidence/package-lock-graph.mjs +92 -0
- package/src/mcp/evidence-snapshot.package-graph.mjs +5 -474
- package/src/mcp/graph/context-core.mjs +111 -0
- package/src/mcp/graph/context-seeds.mjs +208 -0
- package/src/mcp/graph/context-state.mjs +86 -0
- package/src/mcp/graph/tools-core.mjs +143 -0
- package/src/mcp/graph/tools-query.mjs +223 -0
- package/src/mcp/graph-context.mjs +4 -496
- package/src/mcp/health/audit-format.mjs +172 -0
- package/src/mcp/health/audit.mjs +171 -0
- package/src/mcp/health/dead-code.mjs +110 -0
- package/src/mcp/health/duplicates.mjs +83 -0
- package/src/mcp/health/endpoints.mjs +42 -0
- package/src/mcp/health/structure.mjs +219 -0
- package/src/mcp/runtime-version.mjs +32 -0
- package/src/mcp/server/auto-refresh.mjs +66 -0
- package/src/mcp/server/runtime-config.mjs +43 -0
- package/src/mcp/server/shutdown.mjs +40 -0
- package/src/mcp/sync/evidence-architecture.mjs +54 -0
- package/src/mcp/sync/evidence-common.mjs +88 -0
- package/src/mcp/sync/evidence-duplicates.mjs +100 -0
- package/src/mcp/sync/evidence-health.mjs +56 -0
- package/src/mcp/sync/evidence-packages.mjs +95 -0
- package/src/mcp/sync/payload-common.mjs +97 -0
- package/src/mcp/sync/payload-v2.mjs +53 -0
- package/src/mcp/sync/payload-v3.mjs +79 -0
- package/src/mcp/sync-evidence.mjs +7 -402
- package/src/mcp/sync-payload.mjs +10 -244
- package/src/mcp/tools-actions.mjs +18 -414
- package/src/mcp/tools-architecture.mjs +18 -70
- package/src/mcp/tools-context.mjs +56 -15
- package/src/mcp/tools-endpoints.mjs +4 -1
- package/src/mcp/tools-graph.mjs +17 -330
- package/src/mcp/tools-health.mjs +24 -705
- package/src/mcp/tools-verified-change.mjs +12 -4
- package/src/mcp-server.mjs +49 -146
- package/src/precision/lsp-client/constants.js +12 -0
- package/src/precision/lsp-client/environment.js +20 -0
- package/src/precision/lsp-client/errors.js +15 -0
- package/src/precision/lsp-client/lifecycle.js +127 -0
- package/src/precision/lsp-client/message-parser.js +81 -0
- package/src/precision/lsp-client/protocol.js +212 -0
- package/src/precision/lsp-client/registry.js +58 -0
- package/src/precision/lsp-client/repo-uri.js +60 -0
- package/src/precision/lsp-client/stdio-client.js +151 -0
- package/src/precision/lsp-client.js +10 -682
- package/src/precision/lsp-overlay/build.js +238 -0
- package/src/precision/lsp-overlay/contract.js +61 -0
- package/src/precision/lsp-overlay/reference-results.js +108 -0
- package/src/precision/lsp-overlay/semantic-inputs.js +62 -0
- package/src/precision/lsp-overlay/source-session.js +134 -0
- package/src/precision/lsp-overlay/store.js +150 -0
- package/src/precision/lsp-overlay/target-index.js +147 -0
- package/src/precision/lsp-overlay/target-query.js +55 -0
- package/src/precision/lsp-overlay.js +11 -868
- package/src/precision/typescript-lsp-provider.js +12 -682
- package/src/precision/typescript-provider/client.js +69 -0
- package/src/precision/typescript-provider/discovery.js +124 -0
- package/src/precision/typescript-provider/project-host.js +249 -0
- package/src/precision/typescript-provider/project-safety.js +161 -0
- package/src/precision/typescript-provider/reference-usage.js +53 -0
- package/src/version.js +7 -0
- package/docs/releases/v0.2.12.md +0 -45
|
@@ -0,0 +1,219 @@
|
|
|
1
|
+
import {degreeOf, rawGraph} from '../graph-context.mjs'
|
|
2
|
+
import {summarizeCommunities, aggregateGraph} from '../../analysis/graph-analysis.js'
|
|
3
|
+
import {computeStaticTestReachability} from '../../analysis/static-test-reachability.js'
|
|
4
|
+
import {computeHotPathReview} from '../../analysis/hot-path-review.js'
|
|
5
|
+
import {createPathClassifier, hasPathClass} from '../../path-classification.js'
|
|
6
|
+
import {toolResult} from '../tool-result.mjs'
|
|
7
|
+
|
|
8
|
+
export function tListCommunities(g, args, ctx) {
|
|
9
|
+
const max = Math.max(1, Math.min(100, Number(args.top_n) || 20))
|
|
10
|
+
const list = summarizeCommunities(ctx.graphPath, max)
|
|
11
|
+
if (!list.length) return 'No communities found in the graph.'
|
|
12
|
+
return [
|
|
13
|
+
`Communities, largest first (list position = community_id for get_community):`,
|
|
14
|
+
...list.map((c, i) => `${String(i).padStart(3)}. ${c.name} — ${c.size} nodes (raw id ${c.id}; e.g. ${[...new Set(c.files)].join(', ')})`),
|
|
15
|
+
].join('\n')
|
|
16
|
+
}
|
|
17
|
+
|
|
18
|
+
// Folder-level architecture map: modules (top-two path segments) with file/symbol counts and the
|
|
19
|
+
// strongest module→module dependencies. Pure graph aggregation — no filesystem reads.
|
|
20
|
+
export function tModuleMap(g, args, ctx) {
|
|
21
|
+
const graph = rawGraph(ctx)
|
|
22
|
+
const testsOnly = graph.graphBuildMode === 'tests-only'
|
|
23
|
+
const includeNonProduct = args.include_non_product === true || testsOnly
|
|
24
|
+
const classifier = createPathClassifier(ctx?.repoRoot || null)
|
|
25
|
+
const nonProductFiles = new Set()
|
|
26
|
+
const classifiedFiles = new Set()
|
|
27
|
+
if (!includeNonProduct) {
|
|
28
|
+
for (const node of graph.nodes || []) {
|
|
29
|
+
if (!node?.source_file) continue
|
|
30
|
+
const sourceFile = String(node.source_file)
|
|
31
|
+
if (classifiedFiles.has(sourceFile)) continue
|
|
32
|
+
classifiedFiles.add(sourceFile)
|
|
33
|
+
const explanation = classifier.explain(sourceFile)
|
|
34
|
+
if (explanation.excluded || hasPathClass(explanation, 'test', 'e2e', 'generated', 'mock', 'story', 'docs', 'benchmark', 'temp')) {
|
|
35
|
+
nonProductFiles.add(sourceFile)
|
|
36
|
+
}
|
|
37
|
+
}
|
|
38
|
+
}
|
|
39
|
+
const visibleGraph = includeNonProduct || nonProductFiles.size === 0 ? graph : (() => {
|
|
40
|
+
const keep = new Set((graph.nodes || [])
|
|
41
|
+
.filter((node) => !node?.source_file || !nonProductFiles.has(String(node.source_file)))
|
|
42
|
+
.map((node) => String(node.id)))
|
|
43
|
+
const endpoint = (value) => String(value && typeof value === 'object' ? value.id : value)
|
|
44
|
+
return {
|
|
45
|
+
...graph,
|
|
46
|
+
nodes: (graph.nodes || []).filter((node) => keep.has(String(node.id))),
|
|
47
|
+
links: (graph.links || []).filter((link) => keep.has(endpoint(link.source)) && keep.has(endpoint(link.target))),
|
|
48
|
+
}
|
|
49
|
+
})()
|
|
50
|
+
const agg = aggregateGraph(visibleGraph, null)
|
|
51
|
+
const topN = Math.max(1, Math.min(60, Number(args.top_n) || 25))
|
|
52
|
+
const mods = agg.modules.slice(0, topN)
|
|
53
|
+
const edges = agg.moduleEdges.slice(0, Math.min(50, topN * 2))
|
|
54
|
+
const compileEdges = new Map()
|
|
55
|
+
const collectCompileEdges = (list, kind) => {
|
|
56
|
+
for (const edge of list || []) {
|
|
57
|
+
const key = `${edge.from}\0${edge.to}`
|
|
58
|
+
const current = compileEdges.get(key) || {from: edge.from, to: edge.to, count: 0, typeOnly: 0, compileOnly: 0}
|
|
59
|
+
current.count += edge.count
|
|
60
|
+
current[kind] += edge.count
|
|
61
|
+
compileEdges.set(key, current)
|
|
62
|
+
}
|
|
63
|
+
}
|
|
64
|
+
collectCompileEdges(agg.typeOnlyModuleEdges, 'typeOnly')
|
|
65
|
+
collectCompileEdges(agg.compileOnlyModuleEdges, 'compileOnly')
|
|
66
|
+
const compiled = [...compileEdges.values()].sort((a, b) => b.count - a.count).slice(0, Math.min(50, topN * 2))
|
|
67
|
+
return [
|
|
68
|
+
testsOnly
|
|
69
|
+
? 'Scope: tests-only graph; classified test/e2e/fixture surfaces are retained automatically.'
|
|
70
|
+
: includeNonProduct
|
|
71
|
+
? 'Scope: all indexed files, including classified non-product surfaces.'
|
|
72
|
+
: `Scope: production-only (default); excluded ${nonProductFiles.size} classified test/fixture/benchmark/generated/docs file(s). Pass include_non_product:true for the complete graph.`,
|
|
73
|
+
`Module map: ${agg.totals.files} files in ${agg.modules.length} folder-modules, ${agg.totals.moduleEdges} runtime module dependencies and ${agg.totals.compileTimeModuleEdges || 0} compile-time dependencies (${agg.totals.typeOnlyModuleEdges || 0} type-only, ${agg.totals.compileOnlyModuleEdges || 0} compile-only). Top ${mods.length}:`,
|
|
74
|
+
...mods.map((m) => ` ${m.name} — ${m.fileCount} files, ${m.symbolCount} symbols`),
|
|
75
|
+
``,
|
|
76
|
+
`Strongest runtime module dependencies:`,
|
|
77
|
+
...edges.map((e) => ` ${e.from} → ${e.to} (${e.count})`),
|
|
78
|
+
compiled.length ? `` : null,
|
|
79
|
+
compiled.length ? `Compile-time module dependencies (not runtime coupling):` : null,
|
|
80
|
+
...compiled.map((e) => ` ${e.from} → ${e.to} (${e.count}; ${e.typeOnly} type-only, ${e.compileOnly} compile-only)`),
|
|
81
|
+
].filter((line) => line != null).join('\n')
|
|
82
|
+
}
|
|
83
|
+
|
|
84
|
+
// Parser-backed local cost review. This keeps syntax, graph coupling and measured/static test
|
|
85
|
+
// evidence as separate fields so callers cannot mistake a ranking heuristic for profiler output.
|
|
86
|
+
export function tHotPathReview(g, args, ctx) {
|
|
87
|
+
const review = computeHotPathReview(rawGraph(ctx), {
|
|
88
|
+
repoRoot: ctx?.repoRoot || null,
|
|
89
|
+
path: args.path,
|
|
90
|
+
includeTests: args.include_tests === true,
|
|
91
|
+
includeClassified: args.include_classified === true,
|
|
92
|
+
topN: args.top_n,
|
|
93
|
+
cyclomaticThreshold: args.cyclomatic_threshold,
|
|
94
|
+
callThreshold: args.call_threshold,
|
|
95
|
+
loopDepthThreshold: args.loop_depth_threshold,
|
|
96
|
+
timeRankThreshold: args.time_rank_threshold,
|
|
97
|
+
minScore: args.min_score,
|
|
98
|
+
})
|
|
99
|
+
if (!review.ok) return toolResult(`Hot-path review refused: ${review.error}.`, review)
|
|
100
|
+
const pct = (value) => typeof value === 'number' ? `${Math.round(value * 100)}%` : String(value || 'NOT_AVAILABLE')
|
|
101
|
+
const coverageLine = review.coverage.actualCoverage === 'AVAILABLE'
|
|
102
|
+
? `Measured coverage: ${review.coverage.measuredFiles} file(s) from ${review.coverage.sources.join(', ') || 'coverage report'}.`
|
|
103
|
+
: review.coverage.staticReachability
|
|
104
|
+
? `actualCoverage: NOT_AVAILABLE; static test reachability ${review.coverage.staticReachability.reachableFiles}/${review.coverage.staticReachability.productFiles} product file(s).`
|
|
105
|
+
: 'actualCoverage: NOT_AVAILABLE; no measured or static test evidence was available.'
|
|
106
|
+
const text = [
|
|
107
|
+
`Local hot-path review: ${review.candidateSymbols} candidate(s) from ${review.analyzedSymbols} analyzed symbol(s); showing ${review.hotspots.length}.`,
|
|
108
|
+
`Thresholds: time rank >=${review.thresholds.timeRank}, cyclomatic >=${review.thresholds.cyclomatic}, calls >=${review.thresholds.calls}, loop depth >=${review.thresholds.loopDepth}, score >=${review.thresholds.minScore}.`,
|
|
109
|
+
`Selection: ${review.selectionPolicy.mode}${review.selectionPolicy.strongLocalFallback ? '; strong local sort/recursion/deep-loop evidence can pass below the blended score gate' : '; strict explicit score gate'}.`,
|
|
110
|
+
coverageLine,
|
|
111
|
+
'Local syntax cost and graph coupling are separate. Scores are review priority, not measured runtime.',
|
|
112
|
+
'',
|
|
113
|
+
...(review.hotspots.length ? review.hotspots.flatMap((item, index) => {
|
|
114
|
+
const tests = item.testEvidence.actualCoverage === 'NOT_AVAILABLE'
|
|
115
|
+
? item.testEvidence.staticReachable ? `static-test d${item.testEvidence.distance}` : 'no-static-test-path'
|
|
116
|
+
: `coverage ${pct(item.testEvidence.actualCoverage)}`
|
|
117
|
+
const pointer = `${item.file}${item.startLine ? `:${item.startLine}${item.endLine > item.startLine ? `-${item.endLine}` : ''}` : ''}`
|
|
118
|
+
const evidence = item.sourceEvidence.length
|
|
119
|
+
? `\n evidence: ${item.sourceEvidence.map((entry) => `${entry.kind}@L${entry.line || '?'}${entry.detail ? ` (${entry.detail})` : ''}`).join('; ')}`
|
|
120
|
+
: ''
|
|
121
|
+
return [
|
|
122
|
+
` ${String(index + 1).padStart(2)}. score ${String(item.score).padStart(5)} syntax ${String(item.localSyntax.score).padStart(5)} graph ${String(item.graphRisk.score).padStart(5)} ${item.confidence}`,
|
|
123
|
+
` ${item.label} (${pointer}; fan-in ${item.graphRisk.fanIn}, fan-out ${item.graphRisk.fanOut}; ${tests})`,
|
|
124
|
+
` ${item.reasons.slice(0, 5).join('; ')}${evidence}`,
|
|
125
|
+
]
|
|
126
|
+
}) : [' (none at the selected thresholds)']),
|
|
127
|
+
review.bounds.truncated ? `... +${review.bounds.totalCandidates - review.bounds.returned} more (raise top_n to display more, raise min_score or narrow path to tighten; lower min_score to broaden).` : null,
|
|
128
|
+
'',
|
|
129
|
+
'Caveat: no interprocedural Big-O, recursion-bound, CFG, dead-store or taint-flow claim is made.',
|
|
130
|
+
].filter((line) => line != null).join('\n')
|
|
131
|
+
return toolResult(text, review, {
|
|
132
|
+
completeness: {
|
|
133
|
+
symbols: 'COMPLETE_FOR_INDEXED_GRAPH',
|
|
134
|
+
output: review.bounds.truncated ? 'BOUNDED' : 'COMPLETE',
|
|
135
|
+
coverage: review.coverage.actualCoverage,
|
|
136
|
+
},
|
|
137
|
+
})
|
|
138
|
+
}
|
|
139
|
+
|
|
140
|
+
// Coverage × graph: map an EXISTING coverage report (istanbul/lcov/coverage.py/Go — read offline,
|
|
141
|
+
// tests are never executed here) onto files and symbols, then rank refactor risk as
|
|
142
|
+
// connectivity × uncovered share. Pairs with get_dependents: many dependents + low coverage ⇒ write
|
|
143
|
+
// tests before changing. Coverage pcts in this layer are fractions (0..1).
|
|
144
|
+
export function tCoverageMap(g, args, ctx) {
|
|
145
|
+
if (!ctx.repoRoot) return 'Coverage mapping needs the repo root (not provided to this server).'
|
|
146
|
+
const agg = aggregateGraph(rawGraph(ctx), ctx.repoRoot)
|
|
147
|
+
const pathFilter = args.path ? String(args.path).replace(/\\/g, '/').replace(/\/+$/, '') : null
|
|
148
|
+
const inScope = (p) => !pathFilter || p === pathFilter || String(p).startsWith(`${pathFilter}/`)
|
|
149
|
+
const allFiles = agg.modules.flatMap((m) => m.files.filter((f) => inScope(f.path)))
|
|
150
|
+
const measured = allFiles.filter((f) => f.coverage != null)
|
|
151
|
+
if (!measured.length) {
|
|
152
|
+
const fallback = computeStaticTestReachability(rawGraph(ctx), {repoRoot: ctx.repoRoot, path: pathFilter || ''})
|
|
153
|
+
const topN = Math.max(1, Math.min(50, Number(args.top_n) || 15))
|
|
154
|
+
const reachable = fallback.reachable.slice(0, topN)
|
|
155
|
+
const unreachable = fallback.unreachable.slice(0, topN)
|
|
156
|
+
return [
|
|
157
|
+
`Static test reachability${pathFilter ? ` for ${pathFilter}` : ''}: ${fallback.reachableFiles}/${fallback.productFiles} product file(s) have a runtime graph path from ${fallback.testFiles} indexed test file(s).`,
|
|
158
|
+
`actualCoverage: ${fallback.actualCoverage}. This is NOT coverage: imports/calls only show that a test can statically reach a file, never that a line, branch or symbol executed.`,
|
|
159
|
+
fallback.bounds.truncated ? `Traversal was bounded/truncated (${fallback.bounds.traversedStates}/${fallback.bounds.maxStates} states, depth ≤${fallback.bounds.maxDepth}, ${fallback.testFiles}/${fallback.totalTestFiles} test files).` : `Traversal: ${fallback.bounds.traversedStates} bounded state(s), depth ≤${fallback.bounds.maxDepth}.`,
|
|
160
|
+
'',
|
|
161
|
+
'Nearest runtime paths from tests:',
|
|
162
|
+
...(reachable.length ? reachable.map((entry) => {
|
|
163
|
+
const nearest = entry.nearestTests[0]
|
|
164
|
+
return ` ${nearest.confidence.padStart(6)} d${nearest.distance} ${entry.file} ← ${nearest.test}\n path: ${nearest.path.join(' → ')}`
|
|
165
|
+
}) : [' (none)']),
|
|
166
|
+
'',
|
|
167
|
+
`No runtime path from an indexed test (${fallback.unreachableFiles}; not proof of no tests):`,
|
|
168
|
+
...(unreachable.length ? unreachable.map((file) => ` ${file}`) : [' (none)']),
|
|
169
|
+
fallback.unreachableFiles > unreachable.length ? ` … +${fallback.unreachableFiles - unreachable.length} more (raise top_n or narrow path)` : null,
|
|
170
|
+
'',
|
|
171
|
+
'No coverage report found — generate one for measured coverage:',
|
|
172
|
+
'Generate one with the repo\'s own test runner, then call coverage_map again:',
|
|
173
|
+
' JS/TS: npx vitest run --coverage (or jest --coverage)',
|
|
174
|
+
' Python: pytest --cov --cov-report=json',
|
|
175
|
+
' Go: go test ./... -coverprofile=coverage.out',
|
|
176
|
+
'Read locations: coverage/coverage-summary.json, coverage/coverage-final.json, (coverage/)lcov.info, coverage.json, coverage.out.',
|
|
177
|
+
].filter((line) => line != null).join('\n')
|
|
178
|
+
}
|
|
179
|
+
const pctStr = (v) => (v == null ? 'n/a' : `${Math.round(v * 100)}%`)
|
|
180
|
+
const sources = [...new Set(measured.map((f) => f.coverageSource).filter(Boolean))]
|
|
181
|
+
const avg = measured.reduce((s, f) => s + f.coverage, 0) / measured.length
|
|
182
|
+
const rollup = agg.modules
|
|
183
|
+
.map((m) => {
|
|
184
|
+
const withCov = m.files.filter((f) => f.coverage != null && inScope(f.path))
|
|
185
|
+
if (!withCov.length) return null
|
|
186
|
+
return {
|
|
187
|
+
name: m.name,
|
|
188
|
+
measured: withCov.length,
|
|
189
|
+
total: m.files.filter((f) => inScope(f.path)).length,
|
|
190
|
+
avg: withCov.reduce((s, f) => s + f.coverage, 0) / withCov.length,
|
|
191
|
+
}
|
|
192
|
+
})
|
|
193
|
+
.filter(Boolean)
|
|
194
|
+
.sort((a, b) => a.avg - b.avg)
|
|
195
|
+
const topN = Math.max(1, Math.min(50, Number(args.top_n) || 15))
|
|
196
|
+
// risk = graph degree × uncovered share; only symbols below 80% matter
|
|
197
|
+
const risky = agg.symbols
|
|
198
|
+
.filter((s) => s.coverage != null && s.coverage < 0.8 && inScope(s.file))
|
|
199
|
+
.map((s) => ({...s, degree: degreeOf(g, s.id)}))
|
|
200
|
+
.filter((s) => s.degree > 0)
|
|
201
|
+
.sort((a, b) => b.degree * (1 - b.coverage) - a.degree * (1 - a.coverage))
|
|
202
|
+
.slice(0, topN)
|
|
203
|
+
return [
|
|
204
|
+
`Coverage map (${measured.length}/${allFiles.length} files measured, avg ${pctStr(avg)}; report: ${sources.join(', ') || 'unknown'}${pathFilter ? `; filter ${pathFilter}` : ''}).`,
|
|
205
|
+
``,
|
|
206
|
+
`Modules by average coverage (worst first):`,
|
|
207
|
+
...rollup.slice(0, 20).map((m) => ` ${pctStr(m.avg).padStart(5)} ${m.name} (${m.measured}/${m.total} files measured)`),
|
|
208
|
+
``,
|
|
209
|
+
`Refactor-risk hotspots — connected symbols with low coverage (ranked by degree × uncovered):`,
|
|
210
|
+
...(risky.length
|
|
211
|
+
? risky.map((s) => ` ${pctStr(s.coverage).padStart(5)} deg ${String(s.degree).padStart(3)} ${s.label} (${s.file}${s.line ? `:${s.line}` : ''})`)
|
|
212
|
+
: [' (none — every connected symbol is ≥80% covered or unmeasured)']),
|
|
213
|
+
``,
|
|
214
|
+
`Tip: before refactoring a hotspot, run get_dependents on it — low coverage × many dependents means write tests first.`,
|
|
215
|
+
].join('\n')
|
|
216
|
+
}
|
|
217
|
+
|
|
218
|
+
// HTTP endpoint inventory: Express/Fastify/Nest/Flask/FastAPI/Go/Rust/Spring route definitions.
|
|
219
|
+
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
import {readFileSync} from 'node:fs'
|
|
2
|
+
import process from 'node:process'
|
|
3
|
+
|
|
4
|
+
export const STALE_RUNTIME_OVERRIDE_ENV = 'WEAVATRIX_ALLOW_STALE_RUNTIME'
|
|
5
|
+
|
|
6
|
+
export function runtimeVersionStatus({runningVersion, packageJsonPath, allowStale} = {}) {
|
|
7
|
+
let diskVersion = null, packageVersionError = null
|
|
8
|
+
try {
|
|
9
|
+
const parsed = JSON.parse(readFileSync(packageJsonPath, 'utf8'))
|
|
10
|
+
diskVersion = typeof parsed?.version === 'string' && parsed.version.trim() ? parsed.version.trim() : null
|
|
11
|
+
if (!diskVersion) packageVersionError = 'package.json has no valid version'
|
|
12
|
+
} catch (error) {
|
|
13
|
+
packageVersionError = `package.json is unreadable: ${error.message}`
|
|
14
|
+
}
|
|
15
|
+
const staleRuntime = diskVersion !== String(runningVersion || '')
|
|
16
|
+
const staleRuntimeAllowed = allowStale == null
|
|
17
|
+
? process.env[STALE_RUNTIME_OVERRIDE_ENV] === '1'
|
|
18
|
+
: allowStale === true
|
|
19
|
+
return {
|
|
20
|
+
version: String(runningVersion || ''),
|
|
21
|
+
diskVersion,
|
|
22
|
+
staleRuntime,
|
|
23
|
+
staleRuntimeAllowed,
|
|
24
|
+
...(packageVersionError ? {packageVersionError} : {}),
|
|
25
|
+
}
|
|
26
|
+
}
|
|
27
|
+
|
|
28
|
+
export function staleRuntimeMessage(status) {
|
|
29
|
+
const disk = status.diskVersion || 'unavailable'
|
|
30
|
+
return `STALE_RUNTIME: running Weavatrix ${status.version || 'unknown'} but package.json on disk is ${disk}. Restart/reconnect the MCP server before using tools. For deliberate source-development only, set ${STALE_RUNTIME_OVERRIDE_ENV}=1.`
|
|
31
|
+
}
|
|
32
|
+
|
|
@@ -0,0 +1,66 @@
|
|
|
1
|
+
import {dirname} from 'node:path'
|
|
2
|
+
import {buildGraphForRepo, defaultPrecisionMode} from '../../build-graph.js'
|
|
3
|
+
import {persistedFreshnessMatches, repositoryFreshnessProbe} from '../../graph/freshness-probe.js'
|
|
4
|
+
import {PRECISION_OVERLAY_V, precisionSemanticInputsMatch, readPrecisionOverlay} from '../../precision/lsp-overlay.js'
|
|
5
|
+
import {loadGraph} from '../graph-context.mjs'
|
|
6
|
+
|
|
7
|
+
export function createAutoRefresh(getApi) {
|
|
8
|
+
const probeCache = new Map()
|
|
9
|
+
const configured = process.env.WEAVATRIX_AUTO_REFRESH_DEBOUNCE_MS == null
|
|
10
|
+
? 2_000 : Number(process.env.WEAVATRIX_AUTO_REFRESH_DEBOUNCE_MS)
|
|
11
|
+
const debounceMs = Math.max(0, Math.min(5_000, Number.isFinite(configured) ? configured : 2_000))
|
|
12
|
+
return async function autoRefresh(callCtx, currentGraph) {
|
|
13
|
+
if (!callCtx?.repoRoot || !callCtx?.graphPath) return {graph: null, refresh: null}
|
|
14
|
+
const activePrecision = currentGraph?.graphPrecisionMode || defaultPrecisionMode()
|
|
15
|
+
const probeKey = `${callCtx.graphPath}\0${currentGraph?.graphBuildMode || 'full'}\0${activePrecision}`
|
|
16
|
+
let semanticInputsChanged = false
|
|
17
|
+
if (activePrecision === 'lsp' && currentGraph) {
|
|
18
|
+
try {
|
|
19
|
+
const overlay = readPrecisionOverlay(callCtx.graphPath, currentGraph)
|
|
20
|
+
semanticInputsChanged = typeof overlay?.semanticInputFingerprint === 'string'
|
|
21
|
+
&& !precisionSemanticInputsMatch(overlay, callCtx.repoRoot, currentGraph)
|
|
22
|
+
} catch {
|
|
23
|
+
semanticInputsChanged = true
|
|
24
|
+
}
|
|
25
|
+
}
|
|
26
|
+
const cached = probeCache.get(probeKey)
|
|
27
|
+
if (!semanticInputsChanged && currentGraph && cached && Date.now() - cached.checkedAt < debounceMs) {
|
|
28
|
+
return {graph: currentGraph, refresh: {kind: 'none', revision: currentGraph.graphRevision || null, changedFiles: 0}}
|
|
29
|
+
}
|
|
30
|
+
const beforeProbe = repositoryFreshnessProbe(callCtx.repoRoot)
|
|
31
|
+
const precisionMissing = activePrecision === 'lsp' && (
|
|
32
|
+
Number(currentGraph?.precisionOverlayV) !== PRECISION_OVERLAY_V || semanticInputsChanged
|
|
33
|
+
)
|
|
34
|
+
if (!precisionMissing && beforeProbe && currentGraph && (
|
|
35
|
+
cached?.probe === beforeProbe
|
|
36
|
+
|| persistedFreshnessMatches(currentGraph, beforeProbe, currentGraph.graphBuildMode || 'full')
|
|
37
|
+
)) {
|
|
38
|
+
probeCache.set(probeKey, {probe: beforeProbe, checkedAt: Date.now()})
|
|
39
|
+
return {graph: currentGraph, refresh: {kind: 'none', revision: currentGraph.graphRevision || null, changedFiles: 0}}
|
|
40
|
+
}
|
|
41
|
+
const result = await buildGraphForRepo(callCtx.repoRoot, {
|
|
42
|
+
mode: currentGraph?.graphBuildMode || 'full',
|
|
43
|
+
precision: activePrecision,
|
|
44
|
+
scope: '',
|
|
45
|
+
outDir: dirname(callCtx.graphPath),
|
|
46
|
+
})
|
|
47
|
+
if (!result.ok) throw new Error(result.error || 'automatic graph refresh failed')
|
|
48
|
+
getApi().resetStalenessCache()
|
|
49
|
+
const fresh = loadGraph(callCtx.graphPath, {repoRoot: callCtx.repoRoot})
|
|
50
|
+
const afterProbe = repositoryFreshnessProbe(callCtx.repoRoot)
|
|
51
|
+
if (afterProbe && afterProbe === beforeProbe) probeCache.set(probeKey, {probe: afterProbe, checkedAt: Date.now()})
|
|
52
|
+
else probeCache.delete(probeKey)
|
|
53
|
+
const update = result.refresh || {kind: 'full', changedFiles: [], reason: 'automatic-refresh'}
|
|
54
|
+
return {
|
|
55
|
+
graph: fresh,
|
|
56
|
+
refresh: {
|
|
57
|
+
kind: update.kind,
|
|
58
|
+
revision: update.revision || fresh.graphRevision || null,
|
|
59
|
+
changedFiles: Array.isArray(update.changedFiles) ? update.changedFiles.length : 0,
|
|
60
|
+
notice: update.kind === 'none'
|
|
61
|
+
? undefined
|
|
62
|
+
: `Graph ${update.kind === 'incremental' ? 'incrementally refreshed' : 'rebuilt'} before this answer (${update.reason || 'repository changed'}).`,
|
|
63
|
+
},
|
|
64
|
+
}
|
|
65
|
+
}
|
|
66
|
+
}
|
|
@@ -0,0 +1,43 @@
|
|
|
1
|
+
import {createRequire} from 'node:module'
|
|
2
|
+
import {existsSync, realpathSync, statSync} from 'node:fs'
|
|
3
|
+
import {dirname, join} from 'node:path'
|
|
4
|
+
import {fileURLToPath} from 'node:url'
|
|
5
|
+
import {graphOutDirForRepo} from '../../graph/layout.js'
|
|
6
|
+
|
|
7
|
+
const SOURCE_DIR = join(dirname(fileURLToPath(import.meta.url)), '..', '..')
|
|
8
|
+
const MCP_DIR = join(SOURCE_DIR, 'mcp')
|
|
9
|
+
const CATALOG_URL = new URL('../catalog.mjs', import.meta.url)
|
|
10
|
+
|
|
11
|
+
export const PACKAGE_JSON_PATH = join(SOURCE_DIR, '..', 'package.json')
|
|
12
|
+
export const PACKAGE_VERSION = (() => {
|
|
13
|
+
try { return createRequire(import.meta.url)('../../../package.json').version } catch { return '0.0.0' }
|
|
14
|
+
})()
|
|
15
|
+
export const SERVER_INFO = {name: 'weavatrix', version: PACKAGE_VERSION}
|
|
16
|
+
|
|
17
|
+
export const loadServerCatalog = (version = 0) => import(version ? `${CATALOG_URL.href}?v=${version}` : CATALOG_URL.href)
|
|
18
|
+
|
|
19
|
+
export function hotCatalogVersion(hotFiles) {
|
|
20
|
+
let version = 0
|
|
21
|
+
for (const file of hotFiles) {
|
|
22
|
+
try {
|
|
23
|
+
const modified = statSync(join(MCP_DIR, file)).mtimeMs
|
|
24
|
+
if (modified > version) version = modified
|
|
25
|
+
} catch { /* a missing file does not bump the version */ }
|
|
26
|
+
}
|
|
27
|
+
return version
|
|
28
|
+
}
|
|
29
|
+
|
|
30
|
+
export function resolveServerTarget(argv, log = () => {}) {
|
|
31
|
+
let graphPath = argv[2], repoArg = argv[3], capabilities = argv[4]
|
|
32
|
+
try {
|
|
33
|
+
if (graphPath && statSync(graphPath).isDirectory()) {
|
|
34
|
+
repoArg = realpathSync.native(graphPath)
|
|
35
|
+
capabilities = argv[3]
|
|
36
|
+
graphPath = join(graphOutDirForRepo(repoArg), 'graph.json')
|
|
37
|
+
if (!existsSync(graphPath)) log(`no graph built yet for ${repoArg} — ask the agent to call rebuild_graph; it builds into the standard weavatrix-graphs layout`)
|
|
38
|
+
}
|
|
39
|
+
} catch { /* argv[2] is a graph path or unavailable */ }
|
|
40
|
+
let repoRoot = null
|
|
41
|
+
try { if (repoArg && statSync(repoArg).isDirectory()) repoRoot = realpathSync.native(repoArg) } catch { /* invalid repo root */ }
|
|
42
|
+
return {graphPath, repoRoot, capabilities}
|
|
43
|
+
}
|
|
@@ -0,0 +1,40 @@
|
|
|
1
|
+
import process from 'node:process'
|
|
2
|
+
import {
|
|
3
|
+
activeLspClientCount,
|
|
4
|
+
beginLspClientShutdown,
|
|
5
|
+
shutdownActiveLspClients,
|
|
6
|
+
} from '../../precision/lsp-client.js'
|
|
7
|
+
|
|
8
|
+
async function settleWithin(promise, timeoutMs) {
|
|
9
|
+
let timer
|
|
10
|
+
const settled = await Promise.race([
|
|
11
|
+
Promise.resolve(promise).then(() => true, () => true),
|
|
12
|
+
new Promise((resolve) => { timer = setTimeout(() => resolve(false), timeoutMs) }),
|
|
13
|
+
])
|
|
14
|
+
if (timer) clearTimeout(timer)
|
|
15
|
+
return settled
|
|
16
|
+
}
|
|
17
|
+
|
|
18
|
+
export function createShutdownController({log, targetMutation}) {
|
|
19
|
+
let shuttingDown = false, shutdownPromise = null
|
|
20
|
+
const request = (reason, exitCode = 0) => {
|
|
21
|
+
if (shutdownPromise) return shutdownPromise
|
|
22
|
+
shuttingDown = true
|
|
23
|
+
beginLspClientShutdown()
|
|
24
|
+
process.stdin.pause()
|
|
25
|
+
const activeAtStart = activeLspClientCount()
|
|
26
|
+
log(`shutdown requested (${reason}); draining graph work and ${activeAtStart} semantic provider(s)`)
|
|
27
|
+
shutdownPromise = (async () => {
|
|
28
|
+
const initiallyDrained = await settleWithin(targetMutation(), 2_500)
|
|
29
|
+
const semantic = await shutdownActiveLspClients({timeoutMs: 3_000})
|
|
30
|
+
const fullyDrained = initiallyDrained || await settleWithin(targetMutation(), 1_500)
|
|
31
|
+
log(`shutdown cleanup: graph=${fullyDrained ? 'drained' : 'bounded-timeout'}, semantic=${semantic.requested} requested/${semantic.remaining} remaining${semantic.timedOut ? ' (forced)' : ''}`)
|
|
32
|
+
})().catch((error) => {
|
|
33
|
+
log(`shutdown cleanup failed: ${error.stack || error.message}`)
|
|
34
|
+
}).finally(() => {
|
|
35
|
+
process.exit(exitCode)
|
|
36
|
+
})
|
|
37
|
+
return shutdownPromise
|
|
38
|
+
}
|
|
39
|
+
return {isShuttingDown: () => shuttingDown, request}
|
|
40
|
+
}
|
|
@@ -0,0 +1,54 @@
|
|
|
1
|
+
import {
|
|
2
|
+
CAPS, SEVERITIES, bool, compare, count, int, list, moduleId, path,
|
|
3
|
+
reasons, state, text, token, verdict,
|
|
4
|
+
} from './evidence-common.mjs'
|
|
5
|
+
|
|
6
|
+
function moduleFact(value) {
|
|
7
|
+
const id = moduleId(value?.id || value?.name)
|
|
8
|
+
return id ? {id, fileCount: int(value.fileCount), nodeCount: int(value.nodeCount), symbolCount: int(value.symbolCount)} : null
|
|
9
|
+
}
|
|
10
|
+
|
|
11
|
+
function dependency(value) {
|
|
12
|
+
const from = moduleId(value?.from), to = moduleId(value?.to)
|
|
13
|
+
return from && to && from !== to ? {from, to, count: int(value.count)} : null
|
|
14
|
+
}
|
|
15
|
+
|
|
16
|
+
function cycleFact(value) {
|
|
17
|
+
const id = text(value?.id, 64), kind = value?.kind
|
|
18
|
+
if (!id || !/^[a-f0-9]{16,64}$/i.test(id) || !['runtime', 'compile-time'].includes(kind)) return null
|
|
19
|
+
const members = [...new Set((value.members || []).map((item) => path(item)).filter(Boolean))].sort(compare).slice(0, 200)
|
|
20
|
+
const representativePath = (value.representativePath || []).map((item) => path(item)).filter(Boolean).slice(0, 201)
|
|
21
|
+
if (members.length < 2 || representativePath.length < 2) return null
|
|
22
|
+
return {id, kind, size: Math.max(int(value.size), members.length), members, membersTruncated: bool(value.membersTruncated) || int(value.size) > members.length, representativePath}
|
|
23
|
+
}
|
|
24
|
+
|
|
25
|
+
function boundaryFact(value) {
|
|
26
|
+
const ruleId = token(value?.ruleId, 96), from = path(value?.from), to = path(value?.to)
|
|
27
|
+
if (!ruleId || !from || !to || !['forbidden', 'allowedOnly'].includes(value?.kind) || !SEVERITIES.has(value?.severity)) return null
|
|
28
|
+
return {kind: value.kind, ruleId, severity: value.severity, from, to}
|
|
29
|
+
}
|
|
30
|
+
|
|
31
|
+
export function sanitizeArchitecture(value) {
|
|
32
|
+
const modules = list(value?.modules, CAPS.modules, moduleFact, (a, b) => compare(a.id, b.id))
|
|
33
|
+
const dependencies = {}
|
|
34
|
+
let truncated = modules.truncated
|
|
35
|
+
const dependencyMeta = {}
|
|
36
|
+
for (const kind of ['runtime', 'typeOnly', 'compileOnly']) {
|
|
37
|
+
const result = list(value?.dependencies?.[kind], CAPS.dependencies, dependency, (a, b) => compare(a.from, b.from) || compare(a.to, b.to))
|
|
38
|
+
dependencies[kind] = result.items
|
|
39
|
+
const completenessKey = `${kind}Dependencies`
|
|
40
|
+
dependencyMeta[completenessKey] = count(value?.completeness?.[completenessKey], result.total, result.items.length)
|
|
41
|
+
truncated ||= result.truncated
|
|
42
|
+
}
|
|
43
|
+
const cycles = list(value?.cycles, CAPS.findings, cycleFact, (a, b) => compare(a.kind, b.kind) || b.size - a.size || compare(a.id, b.id))
|
|
44
|
+
const boundaries = list(value?.boundaryViolations, CAPS.findings, boundaryFact, (a, b) => compare(a.ruleId, b.ruleId) || compare(a.from, b.from) || compare(a.to, b.to))
|
|
45
|
+
truncated ||= cycles.truncated || boundaries.truncated
|
|
46
|
+
const outState = state(value?.state)
|
|
47
|
+
return {
|
|
48
|
+
state: truncated && outState === 'COMPLETE' ? 'PARTIAL' : outState,
|
|
49
|
+
verdict: verdict(value?.verdict),
|
|
50
|
+
completeness: {modules: count(value?.completeness?.modules, modules.total, modules.items.length), ...dependencyMeta, cycles: count(value?.completeness?.cycles, cycles.total, cycles.items.length), boundaryViolations: count(value?.completeness?.boundaryViolations, boundaries.total, boundaries.items.length), reasons: reasons(value?.completeness?.reasons)},
|
|
51
|
+
modules: modules.items, dependencies, cycles: cycles.items, boundaryViolations: boundaries.items,
|
|
52
|
+
boundaryRulesState: state(value?.boundaryRulesState),
|
|
53
|
+
}
|
|
54
|
+
}
|
|
@@ -0,0 +1,88 @@
|
|
|
1
|
+
export const SEVERITIES = new Set(['critical', 'high', 'medium', 'low', 'info'])
|
|
2
|
+
export const CONFIDENCE = new Set(['high', 'medium', 'low'])
|
|
3
|
+
export const CATEGORIES = new Set(['unused', 'structure', 'vulnerability', 'malware'])
|
|
4
|
+
export const CHECK_KEYS = ['osv', 'malware']
|
|
5
|
+
export const PACKAGE_DEPENDENCY_KINDS = new Set(['runtime', 'dev', 'optional', 'peer', 'optional-peer'])
|
|
6
|
+
export const CAPS = Object.freeze({
|
|
7
|
+
modules: 500, dependencies: 2000, findings: 500, hotspots: 250, badges: 100,
|
|
8
|
+
packages: 5000, usage: 1000, files: 20, packageGraphNodes: 5000, packageGraphEdges: 20000,
|
|
9
|
+
duplicateGroups: 100, duplicateMembers: 12, divergenceCandidates: 100,
|
|
10
|
+
})
|
|
11
|
+
export const DUPLICATE_THRESHOLDS = Object.freeze({
|
|
12
|
+
clones: Object.freeze({mode: 'renamed', minSimilarityPercent: 80, minTokens: 50}),
|
|
13
|
+
divergence: Object.freeze({sameName: true, maxSimilarityPercent: 45, minTokens: 50, maxImplementationsPerName: 12}),
|
|
14
|
+
})
|
|
15
|
+
|
|
16
|
+
const STATES = new Set(['COMPLETE', 'PARTIAL', 'NOT_CHECKED', 'NOT_APPLICABLE', 'ERROR'])
|
|
17
|
+
const VERDICTS = new Set(['PASS', 'FAIL', 'UNKNOWN'])
|
|
18
|
+
const CONTROL = /[\u0000-\u001f\u007f]/
|
|
19
|
+
const ABSOLUTE_PATH_FRAGMENT = /(?:^|[\/\s"'`(=])[a-z]:[\\/]|(?:^|[\s"'`(=])(?:\\\\[^\\/\s]+(?:[\\/]|$)|file:(?:\/\/)?[\\/]|\/(?!\/)[^\s])/i
|
|
20
|
+
const TOKEN = /^[\p{L}\p{N}_.:@+\-#$<>()\[\],]+$/u
|
|
21
|
+
const PACKAGE = /^(?:@[a-z0-9._-]+\/)?[a-z0-9][a-z0-9._-]*$/i
|
|
22
|
+
|
|
23
|
+
export const int = (value) => Number.isFinite(value) && value >= 0 ? Math.trunc(value) : 0
|
|
24
|
+
export const bool = (value) => value === true
|
|
25
|
+
export const text = (value, max = 256) => typeof value === 'string' && value.length > 0 && value.length <= max && !CONTROL.test(value) ? value : undefined
|
|
26
|
+
export const token = (value, max = 256) => { const valueText = text(value, max); return valueText && TOKEN.test(valueText) ? valueText : undefined }
|
|
27
|
+
export const privacySafeText = (value, max = 256) => { const valueText = text(value, max); return valueText && !ABSOLUTE_PATH_FRAGMENT.test(valueText) ? valueText : undefined }
|
|
28
|
+
export const packageName = (value) => { const valueText = text(value, 256); return valueText && PACKAGE.test(valueText) ? valueText : undefined }
|
|
29
|
+
export const packageVersion = (value) => { const valueText = text(value, 128); return valueText && /^[A-Za-z0-9][A-Za-z0-9._+~-]*$/.test(valueText) ? valueText : undefined }
|
|
30
|
+
export const state = (value) => STATES.has(value) ? value : 'ERROR'
|
|
31
|
+
export const verdict = (value) => VERDICTS.has(value) ? value : 'UNKNOWN'
|
|
32
|
+
export const compare = (a, b) => String(a).localeCompare(String(b), 'en')
|
|
33
|
+
|
|
34
|
+
export function path(value, max = 4096) {
|
|
35
|
+
const raw = text(value, max)
|
|
36
|
+
if (!raw) return undefined
|
|
37
|
+
const normalized = raw.replace(/\\/g, '/')
|
|
38
|
+
if (/^(?:[a-z][a-z0-9+.-]*:|\/)/i.test(normalized)) return undefined
|
|
39
|
+
const parts = normalized.split('/')
|
|
40
|
+
return parts.length && parts.every((part) => part && part !== '.' && part !== '..') ? normalized : undefined
|
|
41
|
+
}
|
|
42
|
+
|
|
43
|
+
export function graphId(value) {
|
|
44
|
+
const id = text(value, 4096)
|
|
45
|
+
if (!id) return undefined
|
|
46
|
+
const hash = id.indexOf('#')
|
|
47
|
+
const file = hash < 0 ? id : id.slice(0, hash)
|
|
48
|
+
const safeFile = path(file)
|
|
49
|
+
if (!safeFile) return undefined
|
|
50
|
+
if (hash < 0) return safeFile
|
|
51
|
+
const suffix = id.slice(hash)
|
|
52
|
+
return suffix.length <= 512 && /^#[^\\/\s\u0000-\u001f\u007f]{1,511}$/u.test(suffix) ? `${safeFile}${suffix}` : undefined
|
|
53
|
+
}
|
|
54
|
+
|
|
55
|
+
export function moduleId(value) { return value === '(root)' ? value : path(value) }
|
|
56
|
+
export function set(out, key, value) { if (value !== undefined) out[key] = value }
|
|
57
|
+
|
|
58
|
+
export function stableStringify(value) {
|
|
59
|
+
if (Array.isArray(value)) return `[${value.map(stableStringify).join(',')}]`
|
|
60
|
+
if (value && typeof value === 'object') return `{${Object.keys(value).sort(compare).map((key) => `${JSON.stringify(key)}:${stableStringify(value[key])}`).join(',')}}`
|
|
61
|
+
return JSON.stringify(value)
|
|
62
|
+
}
|
|
63
|
+
|
|
64
|
+
export function list(values, cap, mapper, sorter) {
|
|
65
|
+
const all = (Array.isArray(values) ? values : []).map(mapper).filter(Boolean).sort(sorter)
|
|
66
|
+
return {items: all.slice(0, cap), total: all.length, truncated: all.length > cap}
|
|
67
|
+
}
|
|
68
|
+
|
|
69
|
+
export function count(value, fallbackTotal, returned) {
|
|
70
|
+
const total = Math.max(int(value?.total), fallbackTotal)
|
|
71
|
+
return {total, returned, truncated: bool(value?.truncated) || total > returned}
|
|
72
|
+
}
|
|
73
|
+
|
|
74
|
+
export function reasons(values) {
|
|
75
|
+
return [...new Set((Array.isArray(values) ? values : []).map((value) => token(value, 96)).filter(Boolean))].sort(compare).slice(0, 32)
|
|
76
|
+
}
|
|
77
|
+
|
|
78
|
+
export function numericRecord(value, keys) {
|
|
79
|
+
const out = {}
|
|
80
|
+
for (const key of keys) out[key] = int(value?.[key])
|
|
81
|
+
return out
|
|
82
|
+
}
|
|
83
|
+
|
|
84
|
+
export function checks(value) {
|
|
85
|
+
const out = {}
|
|
86
|
+
for (const key of CHECK_KEYS) out[key] = state(value?.[key])
|
|
87
|
+
return out
|
|
88
|
+
}
|
|
@@ -0,0 +1,100 @@
|
|
|
1
|
+
import {
|
|
2
|
+
CAPS, DUPLICATE_THRESHOLDS, bool, compare, count, graphId, int, list,
|
|
3
|
+
path, reasons, state, text, verdict,
|
|
4
|
+
} from './evidence-common.mjs'
|
|
5
|
+
|
|
6
|
+
function duplicateEvidenceId(value) {
|
|
7
|
+
const id = text(value, 64)
|
|
8
|
+
return id && /^[a-f0-9]{24,64}$/i.test(id) ? id : undefined
|
|
9
|
+
}
|
|
10
|
+
|
|
11
|
+
function duplicateMember(value) {
|
|
12
|
+
if (!value || typeof value !== 'object' || Array.isArray(value)) return null
|
|
13
|
+
const file = path(value.file)
|
|
14
|
+
const startLine = int(value.startLine), endLine = int(value.endLine), tokens = int(value.tokens)
|
|
15
|
+
if (!file || startLine < 1 || endLine < startLine || tokens < DUPLICATE_THRESHOLDS.clones.minTokens) return null
|
|
16
|
+
const out = {file, startLine, endLine, tokens}
|
|
17
|
+
const nodeId = graphId(value.graphNodeId)
|
|
18
|
+
if (nodeId?.startsWith(`${file}#`)) out.graphNodeId = nodeId
|
|
19
|
+
return out
|
|
20
|
+
}
|
|
21
|
+
|
|
22
|
+
function compareDuplicateMember(a, b) {
|
|
23
|
+
return compare(a.file, b.file) || a.startLine - b.startLine || a.endLine - b.endLine || compare(a.graphNodeId || '', b.graphNodeId || '')
|
|
24
|
+
}
|
|
25
|
+
|
|
26
|
+
function duplicateMembers(values, cap = CAPS.duplicateMembers) {
|
|
27
|
+
const raw = Array.isArray(values) ? values : []
|
|
28
|
+
const unique = new Map()
|
|
29
|
+
for (const value of raw) {
|
|
30
|
+
const item = duplicateMember(value)
|
|
31
|
+
if (!item) continue
|
|
32
|
+
unique.set(`${item.file}\0${item.startLine}\0${item.endLine}\0${item.graphNodeId || ''}`, item)
|
|
33
|
+
}
|
|
34
|
+
const all = [...unique.values()].sort(compareDuplicateMember)
|
|
35
|
+
return {items: all.slice(0, cap), total: all.length, invalid: raw.length - all.length, truncated: all.length > cap}
|
|
36
|
+
}
|
|
37
|
+
|
|
38
|
+
function cloneGroup(value) {
|
|
39
|
+
const id = duplicateEvidenceId(value?.id)
|
|
40
|
+
const members = duplicateMembers(value?.members)
|
|
41
|
+
const rawStrongestSimilarity = Number(value?.strongestSimilarity)
|
|
42
|
+
const rawWeakestLinkedSimilarity = Number(value?.weakestLinkedSimilarity)
|
|
43
|
+
const strongestSimilarity = Math.trunc(rawStrongestSimilarity)
|
|
44
|
+
const weakestLinkedSimilarity = Math.trunc(rawWeakestLinkedSimilarity)
|
|
45
|
+
if (!id || members.items.length < 2 || strongestSimilarity < DUPLICATE_THRESHOLDS.clones.minSimilarityPercent ||
|
|
46
|
+
!Number.isFinite(rawStrongestSimilarity) || rawStrongestSimilarity > 100 ||
|
|
47
|
+
!Number.isFinite(rawWeakestLinkedSimilarity) || weakestLinkedSimilarity < DUPLICATE_THRESHOLDS.clones.minSimilarityPercent ||
|
|
48
|
+
rawWeakestLinkedSimilarity > 100 || weakestLinkedSimilarity > strongestSimilarity) return null
|
|
49
|
+
const memberCount = Math.max(int(value?.memberCount), members.total)
|
|
50
|
+
const returnedTokens = members.items.reduce((sum, member) => sum + member.tokens, 0)
|
|
51
|
+
return {
|
|
52
|
+
id, memberCount, totalTokens: Math.max(int(value?.totalTokens), returnedTokens),
|
|
53
|
+
strongestSimilarity, weakestLinkedSimilarity,
|
|
54
|
+
membersTruncated: bool(value?.membersTruncated) || members.truncated || members.invalid > 0 || memberCount > members.items.length,
|
|
55
|
+
members: members.items,
|
|
56
|
+
}
|
|
57
|
+
}
|
|
58
|
+
|
|
59
|
+
function divergenceCandidate(value) {
|
|
60
|
+
const id = duplicateEvidenceId(value?.id)
|
|
61
|
+
const symbol = text(value?.symbol, 256)
|
|
62
|
+
const rawSimilarity = Number(value?.similarity)
|
|
63
|
+
const similarity = Math.trunc(rawSimilarity)
|
|
64
|
+
const members = duplicateMembers(value?.members, 2)
|
|
65
|
+
if (!id || !symbol || !/^[A-Za-z_$][A-Za-z0-9_$]*$/.test(symbol) || members.total !== 2 || members.invalid > 0 ||
|
|
66
|
+
!Number.isFinite(rawSimilarity) || rawSimilarity < 0 || rawSimilarity > DUPLICATE_THRESHOLDS.divergence.maxSimilarityPercent ||
|
|
67
|
+
members.items[0].file === members.items[1].file) return null
|
|
68
|
+
return {id, symbol, similarity, totalTokens: Math.max(int(value?.totalTokens), members.items[0].tokens + members.items[1].tokens), members: members.items}
|
|
69
|
+
}
|
|
70
|
+
|
|
71
|
+
export function sanitizeDuplicates(value) {
|
|
72
|
+
const rawGroups = Array.isArray(value?.cloneGroups) ? value.cloneGroups : []
|
|
73
|
+
const rawDivergence = Array.isArray(value?.divergenceCandidates) ? value.divergenceCandidates : []
|
|
74
|
+
const cloneGroups = list(rawGroups, CAPS.duplicateGroups, cloneGroup,
|
|
75
|
+
(a, b) => b.totalTokens - a.totalTokens || b.memberCount - a.memberCount || compare(a.id, b.id))
|
|
76
|
+
const divergence = list(rawDivergence, CAPS.divergenceCandidates, divergenceCandidate,
|
|
77
|
+
(a, b) => b.totalTokens - a.totalTokens || a.similarity - b.similarity || compare(a.symbol, b.symbol) || compare(a.id, b.id))
|
|
78
|
+
const invalid = rawGroups.length - cloneGroups.total + rawDivergence.length - divergence.total
|
|
79
|
+
const membersTruncated = cloneGroups.items.some((group) => group.membersTruncated)
|
|
80
|
+
const groupCompleteness = count(value?.completeness?.cloneGroups, rawGroups.length, cloneGroups.items.length)
|
|
81
|
+
const divergenceCompleteness = count(value?.completeness?.divergenceCandidates, rawDivergence.length, divergence.items.length)
|
|
82
|
+
const truncated = cloneGroups.truncated || divergence.truncated || groupCompleteness.truncated || divergenceCompleteness.truncated || membersTruncated || invalid > 0
|
|
83
|
+
const outReasons = reasons([
|
|
84
|
+
...(Array.isArray(value?.completeness?.reasons) ? value.completeness.reasons : []),
|
|
85
|
+
...(invalid > 0 ? ['INVALID_DUPLICATE_EVIDENCE_DROPPED'] : []),
|
|
86
|
+
...(membersTruncated ? ['CLONE_MEMBERS_TRUNCATED'] : []),
|
|
87
|
+
...(cloneGroups.truncated ? ['CLONE_GROUPS_TRUNCATED'] : []),
|
|
88
|
+
...(divergence.truncated ? ['DIVERGENCE_CANDIDATES_TRUNCATED'] : []),
|
|
89
|
+
])
|
|
90
|
+
const rawFragments = value?.completeness?.fragments
|
|
91
|
+
const eligible = int(rawFragments?.eligible), filtered = int(rawFragments?.filtered)
|
|
92
|
+
const total = Math.max(int(rawFragments?.total), eligible + filtered)
|
|
93
|
+
const outState = state(value?.state)
|
|
94
|
+
return {
|
|
95
|
+
state: truncated && outState === 'COMPLETE' ? 'PARTIAL' : outState,
|
|
96
|
+
verdict: verdict(value?.verdict), thresholds: DUPLICATE_THRESHOLDS,
|
|
97
|
+
completeness: {fragments: {total, eligible: Math.min(eligible, total), filtered: Math.max(filtered, total - eligible)}, cloneGroups: groupCompleteness, divergenceCandidates: divergenceCompleteness, reasons: outReasons},
|
|
98
|
+
cloneGroups: cloneGroups.items, divergenceCandidates: divergence.items,
|
|
99
|
+
}
|
|
100
|
+
}
|