feat(r-graphics-pkg): rudiementary side effect support

src/dataflow/environments/built-in.ts

@@ -28,12 +28,10 @@ import type { ForceArguments } from '../internal/process/functions/call/common';
 import { processApply } from '../internal/process/functions/call/built-in/built-in-apply';
 import { registerBuiltInDefinitions } from './built-in-config';
 import { DefaultBuiltinConfig } from './default-builtin-config';
-import {LinkTo} from "../../queries/catalog/call-context-query/call-context-query-format";
-import {extractCFG} from "../../util/cfg/cfg";
+import type { LinkTo } from '../../queries/catalog/call-context-query/call-context-query-format';
+
+
 
-import {
-	identifyLinkToLastCallRelation
-} from "../../queries/catalog/call-context-query/identify-link-to-last-call-relation";
 
 export const BuiltIn = 'built-in';
 
@@ -64,11 +62,10 @@ export interface BuiltInIdentifierConstant<T = unknown> extends IdentifierRefere
 	value:     T
 }
 
-/* TODO: improve support for link to */
 export interface DefaultBuiltInProcessorConfiguration extends ForceArguments {
-	readonly returnsNthArgument?: number | 'last',
-	readonly cfg?: ExitPointType,
-	readonly readAllArguments?: boolean,
+	readonly returnsNthArgument?:    number | 'last',
+	readonly cfg?:                   ExitPointType,
+	readonly readAllArguments?:      boolean,
 	readonly hasUnknownSideEffects?: boolean | LinkTo<RegExp | string>
 }
 
@@ -96,7 +93,7 @@ function defaultBuiltInProcessor<OtherInfo>(
 
 	if(config.hasUnknownSideEffects) {
 		if(typeof config.hasUnknownSideEffects !== 'boolean') {
-			res.graph.markIdForUnknownSideEffects(rootId, config.hasUnknownSideEffects)
+			res.graph.markIdForUnknownSideEffects(rootId, config.hasUnknownSideEffects);
 		} else {
 			res.graph.markIdForUnknownSideEffects(rootId);
 		}

src/dataflow/environments/default-builtin-config.ts

@@ -39,11 +39,11 @@ export const DefaultBuiltinConfig: BuiltInDefinitions = [
 	{ type: 'function', names: ['apply', 'tapply', 'Tapply'],                  processor: 'builtin:apply',               config: { indexOfFunction: 2, nameOfFunctionArgument: 'FUN' },                        assumePrimitive: false },
 	{ type: 'function', names: ['print', 'message', 'warning'],                processor: 'builtin:default',             config: { returnsNthArgument: 0, forceArgs: 'all', hasUnknownSideEffects: { type: 'link-to-last-call', callName: /^sink$/ } },                                  assumePrimitive: false },
 	// graphics base
-	{ type: 'function', names: ["plot", "map", "image", "boxplot", "barplot", "matplot", "hist", "stem", "density", "smoothScatter", "contour", "persp"],
-		processor: 'builtin:default',             config: { forceArgs: 'all', hasUnknownSideEffects: { type: 'link-to-last-call', callName: /^pdf|jpeg|png|windows|postscript|xfig|bitmap|pictex|cairo_pdf|svg|bmp|tiff|X11|quartz$/ } }, assumePrimitive: true },
+	{ type:            'function', names:           ['plot', 'map', 'image', 'boxplot', 'barplot', 'matplot', 'hist', 'stem', 'density', 'smoothScatter', 'contour', 'persp'],
+		processor:       'builtin:default',             config:          { forceArgs: 'all', hasUnknownSideEffects: { type: 'link-to-last-call', callName: /^pdf|jpeg|png|windows|postscript|xfig|bitmap|pictex|cairo_pdf|svg|bmp|tiff|X11|quartz$/ } }, assumePrimitive: true },
 	// graphics addons
-	{ type: 'function', names: ["points", "abline", "lines", "text", "legend", "title", "axis", "polygon", "polypath", "pie", "rect", "segments", "arrows", "symbols", "tiplabels"],
-		processor: 'builtin:default',             config: { forceArgs: 'all', hasUnknownSideEffects: { type: 'link-to-last-call', callName: /^dev\.new|dev\.copy|plot|map|image|boxplot|barplot|matplot|hist|stem|density|smoothScatter|contour|persp$/ } }, assumePrimitive: true },
+	{ type:            'function', names:           ['points', 'abline', 'lines', 'text', 'legend', 'title', 'axis', 'polygon', 'polypath', 'pie', 'rect', 'segments', 'arrows', 'symbols', 'tiplabels'],
+		processor:       'builtin:default',             config:          { forceArgs: 'all', hasUnknownSideEffects: { type: 'link-to-last-call', callName: /^dev\.new|dev\.copy|plot|map|image|boxplot|barplot|matplot|hist|stem|density|smoothScatter|contour|persp$/ } }, assumePrimitive: true },
 	{ type: 'function', names: ['('],                                          processor: 'builtin:default',             config: { returnsNthArgument: 0 },                                                    assumePrimitive: true  },
 	{ type: 'function', names: ['load', 'load_all', 'setwd', 'set.seed'],      processor: 'builtin:default',             config: { hasUnknownSideEffects: true, forceArgs: [true] },                           assumePrimitive: false },
 	{ type: 'function', names: ['eval', 'body', 'formals', 'environment'],     processor: 'builtin:default',             config: { hasUnknownSideEffects: true, forceArgs: [true] },                           assumePrimitive: false },

src/dataflow/extractor.ts

@@ -1,28 +1,29 @@
-import type {DataflowInformation} from './info';
-import type {DataflowProcessors} from './processor';
-import {processDataflowFor} from './processor';
-import {processUninterestingLeaf} from './internal/process/process-uninteresting-leaf';
-import {processSymbol} from './internal/process/process-symbol';
-import {processFunctionCall} from './internal/process/functions/call/default-call-handling';
-import {processFunctionParameter} from './internal/process/functions/process-parameter';
-import {processFunctionArgument} from './internal/process/functions/process-argument';
-import {processAsNamedCall} from './internal/process/process-named-call';
-import {processValue} from './internal/process/process-value';
-import {processNamedCall} from './internal/process/functions/call/named-call-handling';
-import {wrapArgumentsUnnamed} from './internal/process/functions/call/argument/make-argument';
-import {rangeFrom} from '../util/range';
-import type {NormalizedAst, ParentInformation} from '../r-bridge/lang-4.x/ast/model/processing/decorate';
-import {RType} from '../r-bridge/lang-4.x/ast/model/type';
-import type {RParseRequest, RParseRequests} from '../r-bridge/retriever';
-import {requestFingerprint} from '../r-bridge/retriever';
-import {initializeCleanEnvironments} from './environments/environment';
-import {standaloneSourceFile} from './internal/process/functions/call/built-in/built-in-source';
-import {DataflowGraph} from "./graph/graph";
-import {ControlFlowGraph, extractCFG} from "../util/cfg/cfg";
-import {EdgeType} from "./graph/edge";
+import type { DataflowInformation } from './info';
+import type { DataflowProcessors } from './processor';
+import { processDataflowFor } from './processor';
+import { processUninterestingLeaf } from './internal/process/process-uninteresting-leaf';
+import { processSymbol } from './internal/process/process-symbol';
+import { processFunctionCall } from './internal/process/functions/call/default-call-handling';
+import { processFunctionParameter } from './internal/process/functions/process-parameter';
+import { processFunctionArgument } from './internal/process/functions/process-argument';
+import { processAsNamedCall } from './internal/process/process-named-call';
+import { processValue } from './internal/process/process-value';
+import { processNamedCall } from './internal/process/functions/call/named-call-handling';
+import { wrapArgumentsUnnamed } from './internal/process/functions/call/argument/make-argument';
+import { rangeFrom } from '../util/range';
+import type { NormalizedAst, ParentInformation } from '../r-bridge/lang-4.x/ast/model/processing/decorate';
+import { RType } from '../r-bridge/lang-4.x/ast/model/type';
+import type { RParseRequest, RParseRequests } from '../r-bridge/retriever';
+import { requestFingerprint } from '../r-bridge/retriever';
+import { initializeCleanEnvironments } from './environments/environment';
+import { standaloneSourceFile } from './internal/process/functions/call/built-in/built-in-source';
+import type { DataflowGraph } from './graph/graph';
+import type { ControlFlowGraph } from '../util/cfg/cfg';
+import { extractCFG } from '../util/cfg/cfg';
+import { EdgeType } from './graph/edge';
 import {
 	identifyLinkToLastCallRelation
-} from "../queries/catalog/call-context-query/identify-link-to-last-call-relation";
+} from '../queries/catalog/call-context-query/identify-link-to-last-call-relation';
 
 export const processors: DataflowProcessors<ParentInformation> = {
 	[RType.Number]:             processValue,
@@ -57,21 +58,21 @@ export const processors: DataflowProcessors<ParentInformation> = {
 
 
 function resolveLinkToSideEffects(ast: NormalizedAst, graph: DataflowGraph) {
-	let cfg: ControlFlowGraph | undefined = undefined
+	let cfg: ControlFlowGraph | undefined = undefined;
 	for(const s of graph.unknownSideEffects) {
 		if(typeof s !== 'object') {
-			continue
+			continue;
 		}
 		if(!cfg) {
-			cfg = extractCFG(ast).graph
+			cfg = extractCFG(ast).graph;
 		}
 		/* this has to change whenever we add a new link to relations because we currently offer no abstraction for the type */
-		const potentials = identifyLinkToLastCallRelation(s.id, cfg, graph, s.linkTo.callName)
+		const potentials = identifyLinkToLastCallRelation(s.id, cfg, graph, s.linkTo.callName);
 		for(const pot of potentials) {
 			graph.addEdge(s.id, pot, EdgeType.Reads);
 		}
 		if(potentials.length > 0) {
-			graph.unknownSideEffects.delete(s)
+			graph.unknownSideEffects.delete(s);
 		}
 	}
 }
@@ -103,6 +104,6 @@ export function produceDataFlowGraph<OtherInfo>(
 		}
 	}
 
-	resolveLinkToSideEffects(ast, df.graph)
+	resolveLinkToSideEffects(ast, df.graph);
 	return df;
 }

src/dataflow/graph/dataflowgraph-builder.ts

@@ -12,6 +12,8 @@ import { EmptyArgument } from '../../r-bridge/lang-4.x/ast/model/nodes/r-functio
 import { BuiltIn } from '../environments/built-in';
 import { EdgeType } from './edge';
 import type { ControlDependency } from '../info';
+import type { LinkTo } from '../../queries/catalog/call-context-query/call-context-query-format';
+import { DefaultBuiltinConfig } from '../environments/default-builtin-config';
 
 export function emptyGraph(idMap?: AstIdMap) {
 	return new DataflowGraphBuilder(idMap);
@@ -277,3 +279,11 @@ export class DataflowGraphBuilder extends DataflowGraph {
 		return this;
 	}
 }
+
+export function getBuiltInSideEffect(name: string): LinkTo<RegExp> | undefined {
+	const got = DefaultBuiltinConfig.find(e => e.names.includes(name));
+	if(got?.type !== 'function') {
+		return undefined;
+	}
+	return (got?.config as { hasUnknownSideEffects: LinkTo<RegExp> | undefined }).hasUnknownSideEffects;
+}

src/dataflow/graph/diff.ts

@@ -130,7 +130,10 @@ function diffOutgoingEdges(ctx: DataflowDiffContext): void {
 
 function diffRootVertices(ctx: DataflowDiffContext): void {
 	setDifference(ctx.left.rootIds(), ctx.right.rootIds(), { ...ctx, position: `${ctx.position}Root vertices differ in graphs. ` });
-	setDifference(ctx.left.unknownSideEffects, ctx.right.unknownSideEffects, { ...ctx, position: `${ctx.position}Unknown side effects differ in graphs. ` });
+	setDifference(
+		new Set([...ctx.left.unknownSideEffects].map(n => typeof n === 'object' ? n.id : n)),
+		new Set([...ctx.right.unknownSideEffects].map(n => typeof n === 'object' ? n.id : n)),
+		{ ...ctx, position: `${ctx.position}Unknown side effects differ in graphs. ` });
 }
 
 

src/dataflow/graph/graph.ts

@@ -24,7 +24,7 @@ import { cloneEnvironmentInformation } from '../environments/clone';
 import { jsonReplacer } from '../../util/json';
 import { BuiltIn } from '../environments/built-in';
 import { dataflowLogger } from '../logger';
-import {LinkTo} from "../../queries/catalog/call-context-query/call-context-query-format";
+import type { LinkTo } from '../../queries/catalog/call-context-query/call-context-query-format';
 
 export type DataflowFunctionFlowInformation = Omit<DataflowInformation, 'graph' | 'exitPoints'>  & { graph: Set<NodeId> }
 
@@ -410,7 +410,7 @@ export class DataflowGraph<
 	/** Marks the given node as having unknown side effects */
 	public markIdForUnknownSideEffects(id: NodeId, target?: LinkTo<RegExp | string>): this {
 		if(target) {
-			this._unknownSideEffects.add({ id: normalizeIdToNumberIfPossible(id), linkTo: typeof target.callName === 'string' ? { ...target, callName: new RegExp(target.callName) } : target as LinkTo<RegExp> })
+			this._unknownSideEffects.add({ id: normalizeIdToNumberIfPossible(id), linkTo: typeof target.callName === 'string' ? { ...target, callName: new RegExp(target.callName) } : target as LinkTo<RegExp> });
 			return this;
 		}
 		this._unknownSideEffects.add(normalizeIdToNumberIfPossible(id));

src/documentation/data/server/doc-data-server-messages.ts

@@ -26,8 +26,8 @@ import {
 	responseQueryMessage
 } from '../../../cli/repl/server/messages/message-query';
 import { exampleQueryCode } from '../query/example-query-code';
-import { CallTargets } from '../../../queries/catalog/call-context-query/call-context-query-format';
 import { requestLineageMessage, responseLineageMessage } from '../../../cli/repl/server/messages/message-lineage';
+import { CallTargets } from '../../../queries/catalog/call-context-query/identify-link-to-last-call-relation';
 
 export function documentAllServerMessages() {
 
@@ -229,7 +229,7 @@ While the context is derived from the \`filename\`, we currently offer no way to
 	});
 
 	const deprecatedByQuery = `(<a href="${FlowrWikiBaseRef}/Query%20API">deprecated</a>)`;
-	
+
 	documentServerMessage({
 		title:                  'Slice',
 		type:                   'request',

src/documentation/print-query-wiki.ts

@@ -11,7 +11,6 @@ import {
 	showQuery,
 	tocForQueryType
 } from './doc-util/doc-query';
-import { CallTargets } from '../queries/catalog/call-context-query/call-context-query-format';
 import { describeSchema } from '../util/schema';
 import { markdownFormatter } from '../util/ansi';
 import { executeCallContextQueries } from '../queries/catalog/call-context-query/call-context-query-executor';
@@ -30,6 +29,7 @@ import { executeDependenciesQuery } from '../queries/catalog/dependencies-query/
 import { getReplCommand } from './doc-util/doc-cli-option';
 import { NewIssueUrl } from './doc-util/doc-issue';
 import { executeLocationMapQuery } from '../queries/catalog/location-map-query/location-map-query-executor';
+import { CallTargets } from '../queries/catalog/call-context-query/identify-link-to-last-call-relation';
 
 
 registerQueryDocumentation('call-context', {

src/queries/catalog/call-context-query/call-context-query-executor.ts

@@ -1,79 +1,21 @@
-import type {DataflowGraph} from '../../../dataflow/graph/graph';
+import type { DataflowGraph } from '../../../dataflow/graph/graph';
 import type {
 	CallContextQuery,
 	CallContextQueryKindResult,
 	CallContextQueryResult,
 	CallContextQuerySubKindResult,
 	SubCallContextQueryFormat
 } from './call-context-query-format';
-import {CallTargets} from './call-context-query-format';
-import type {NodeId} from '../../../r-bridge/lang-4.x/ast/model/processing/node-id';
-import {recoverContent} from '../../../r-bridge/lang-4.x/ast/model/processing/node-id';
-import {VertexType} from '../../../dataflow/graph/vertex';
-import {assertUnreachable} from '../../../util/assert';
-import {edgeIncludesType, EdgeType} from '../../../dataflow/graph/edge';
-import {resolveByName} from '../../../dataflow/environments/resolve-by-name';
-import {BuiltIn} from '../../../dataflow/environments/built-in';
-import {extractCFG} from '../../../util/cfg/cfg';
-import {TwoLayerCollector} from '../../two-layer-collector';
-import {compactRecord} from '../../../util/objects';
-import {ReferenceType} from '../../../dataflow/environments/identifier';
-
-import type {BasicQueryData} from '../../base-query-format';
-import {identifyLinkToLastCallRelation} from "./identify-link-to-last-call-relation";
-
-function satisfiesCallTargets(id: NodeId, graph: DataflowGraph, callTarget: CallTargets): NodeId[] | 'no'  {
-	const callVertex = graph.get(id);
-	if(callVertex === undefined || callVertex[0].tag !== VertexType.FunctionCall) {
-		return 'no';
-	}
-	const [info,outgoing] = callVertex;
-	const callTargets = [...outgoing]
-		.filter(([, e]) => edgeIncludesType(e.types, EdgeType.Calls))
-		.map(([t]) => t)
-	;
-
-	let builtIn = false;
-
-	if(info.environment === undefined) {
-		/* if we have a call with an unbound environment,
-		 * this only happens if we are sure of built-in relations and want to save references
-		 */
-		builtIn = true;
-	} else {
-		/*
-		 * for performance and scoping reasons, flowR will not identify the global linkage,
-		 * including any potential built-in mapping.
-		 */
-		const reResolved = resolveByName(info.name, info.environment, ReferenceType.Unknown);
-		if(reResolved?.some(t => t.definedAt === BuiltIn)) {
-			builtIn = true;
-		}
-	}
-
-	switch(callTarget) {
-		case CallTargets.Any:
-			return callTargets;
-		case CallTargets.OnlyGlobal:
-			if(callTargets.length === 0) {
-				return builtIn ? [BuiltIn] : [];
-			} else {
-				return 'no';
-			}
-		case CallTargets.MustIncludeGlobal:
-			return builtIn || callTargets.length === 0 ? [...callTargets, BuiltIn] : 'no';
-		case CallTargets.OnlyLocal:
-			return !builtIn && callTargets.length > 0 ? callTargets : 'no';
-		case CallTargets.MustIncludeLocal:
-			if(callTargets.length > 0) {
-				return builtIn ? [...callTargets, BuiltIn] : callTargets;
-			} else {
-				return 'no';
-			}
-		default:
-			assertUnreachable(callTarget);
-	}
-}
+import type { NodeId } from '../../../r-bridge/lang-4.x/ast/model/processing/node-id';
+import { recoverContent } from '../../../r-bridge/lang-4.x/ast/model/processing/node-id';
+import { VertexType } from '../../../dataflow/graph/vertex';
+import { edgeIncludesType, EdgeType } from '../../../dataflow/graph/edge';
+import { extractCFG } from '../../../util/cfg/cfg';
+import { TwoLayerCollector } from '../../two-layer-collector';
+import { compactRecord } from '../../../util/objects';
+
+import type { BasicQueryData } from '../../base-query-format';
+import { identifyLinkToLastCallRelation, satisfiesCallTargets } from './identify-link-to-last-call-relation';
 
 /* if the node is effected by nse, we have an ingoing nse edge */
 function isQuoted(node: NodeId, graph: DataflowGraph): boolean {

src/queries/catalog/call-context-query/call-context-query-format.ts

@@ -9,19 +9,7 @@ import Joi from 'joi';
 import { asciiCallContext } from '../../query-print';
 import type { PipelineOutput } from '../../../core/steps/pipeline/pipeline';
 import type { DEFAULT_DATAFLOW_PIPELINE } from '../../../core/steps/pipeline/default-pipelines';
-
-export enum CallTargets {
-	/** call targets a function that is not defined locally (e.g., the call targets a library function) */
-	OnlyGlobal = 'global',
-	/** call targets a function that is defined locally or globally, but must include a global function */
-	MustIncludeGlobal = 'must-include-global',
-	/** call targets a function that is defined locally  */
-	OnlyLocal = 'local',
-	/** call targets a function that is defined locally or globally, but must include a local function */
-	MustIncludeLocal = 'must-include-local',
-	/** call targets a function that is defined locally or globally */
-	Any = 'any'
-}
+import { CallTargets } from './identify-link-to-last-call-relation';
 
 export interface DefaultCallContextQueryFormat<CallName extends RegExp | string> extends BaseQueryFormat {
 	readonly type:            'call-context';