feat: Use a proper tag for !!merge << keys (#580)

eemeli · web-flow · commit 7a434f0713dd · 2024-10-05T17:34:51.000+03:00
diff --git a/docs/06_custom_tags.md b/docs/06_custom_tags.md
@@ -60,15 +60,16 @@ If including more than one custom tag from this set, make sure that the `'float'
 
 These tags are a part of the YAML 1.1 [language-independent types](https://yaml.org/type/), but are not a part of any default YAML 1.2 schema.
 
-| Identifier    | YAML Type                                             | JS Type      | Description                                                                                                                                                                       |
-| ------------- | ----------------------------------------------------- | ------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `'binary'`    | [`!!binary`](https://yaml.org/type/binary.html)       | `Uint8Array` | Binary data, represented in YAML as base64 encoded characters.                                                                                                                    |
-| `'floatTime'` | [`!!float`](https://yaml.org/type/float.html)         | `Number`     | Sexagesimal floating-point number format, e.g. `190:20:30.15`. To stringify with this tag, the node `format` must be `'TIME'`.                                                    |
-| `'intTime'`   | [`!!int`](https://yaml.org/type/int.html)             | `Number`     | Sexagesimal integer number format, e.g. `190:20:30`. To stringify with this tag, the node `format` must be `'TIME'`.                                                              |
-| `'omap'`      | [`!!omap`](https://yaml.org/type/omap.html)           | `Map`        | Ordered sequence of key: value pairs without duplicates. Using `mapAsMap: true` together with this tag is not recommended, as it makes the parse → stringify loop non-idempotent. |
-| `'pairs'`     | [`!!pairs`](https://yaml.org/type/pairs.html)         | `Array`      | Ordered sequence of key: value pairs allowing duplicates. To create from JS, use `doc.createNode(array, { tag: '!!pairs' })`.                                                     |
-| `'set'`       | [`!!set`](https://yaml.org/type/set.html)             | `Set`        | Unordered set of non-equal values.                                                                                                                                                |
-| `'timestamp'` | [`!!timestamp`](https://yaml.org/type/timestamp.html) | `Date`       | A point in time, e.g. `2001-12-15T02:59:43`.                                                                                                                                      |
+| Identifier    | YAML Type                                             | JS Type        | Description                                                                                                                                                                       |
+| ------------- | ----------------------------------------------------- | -------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `'binary'`    | [`!!binary`](https://yaml.org/type/binary.html)       | `Uint8Array`   | Binary data, represented in YAML as base64 encoded characters.                                                                                                                    |
+| `'floatTime'` | [`!!float`](https://yaml.org/type/float.html)         | `Number`       | Sexagesimal floating-point number format, e.g. `190:20:30.15`. To stringify with this tag, the node `format` must be `'TIME'`.                                                    |
+| `'intTime'`   | [`!!int`](https://yaml.org/type/int.html)             | `Number`       | Sexagesimal integer number format, e.g. `190:20:30`. To stringify with this tag, the node `format` must be `'TIME'`.                                                              |
+| `'merge'`     | [`!!merge`](https://yaml.org/type/merge.html)         | `Symbol('<<')` | A `<<` merge key which allows one or more mappings to be merged with the current one.                                                                                             |
+| `'omap'`      | [`!!omap`](https://yaml.org/type/omap.html)           | `Map`          | Ordered sequence of key: value pairs without duplicates. Using `mapAsMap: true` together with this tag is not recommended, as it makes the parse → stringify loop non-idempotent. |
+| `'pairs'`     | [`!!pairs`](https://yaml.org/type/pairs.html)         | `Array`        | Ordered sequence of key: value pairs allowing duplicates. To create from JS, use `doc.createNode(array, { tag: '!!pairs' })`.                                                     |
+| `'set'`       | [`!!set`](https://yaml.org/type/set.html)             | `Set`          | Unordered set of non-equal values.                                                                                                                                                |
+| `'timestamp'` | [`!!timestamp`](https://yaml.org/type/timestamp.html) | `Date`         | A point in time, e.g. `2001-12-15T02:59:43`.                                                                                                                                      |
 
 ## Writing Custom Tags
 
diff --git a/src/compose/compose-doc.ts b/src/compose/compose-doc.ts
@@ -28,6 +28,7 @@ export function composeDoc<
   const opts = Object.assign({ _directives: directives }, options)
   const doc = new Document(undefined, opts) as Document.Parsed<Contents, Strict>
   const ctx: ComposeContext = {
+    atKey: false,
     atRoot: true,
     directives: doc.directives,
     options: doc.options,
diff --git a/src/compose/compose-node.ts b/src/compose/compose-node.ts
@@ -11,6 +11,7 @@ import { resolveEnd } from './resolve-end.js'
 import { emptyScalarPosition } from './util-empty-scalar-position.js'
 
 export interface ComposeContext {
+  atKey: boolean
   atRoot: boolean
   directives: Directives
   options: Readonly<Required<Omit<ParseOptions, 'lineCounter'>>>
diff --git a/src/compose/compose-scalar.ts b/src/compose/compose-scalar.ts
@@ -87,14 +87,16 @@ function findScalarTagByName(
 }
 
 function findScalarTagByTest(
-  { directives, schema }: ComposeContext,
+  { atKey, directives, schema }: ComposeContext,
   value: string,
   token: FlowScalar,
   onError: ComposeErrorHandler
 ) {
   const tag =
     (schema.tags.find(
-      tag => tag.default && tag.test?.test(value)
+      tag =>
+        (tag.default === true || (atKey && tag.default === 'key')) &&
+        tag.test?.test(value)
     ) as ScalarTag) || schema[SCALAR]
 
   if (schema.compat) {
diff --git a/src/compose/resolve-block-map.ts b/src/compose/resolve-block-map.ts
@@ -69,11 +69,13 @@ export function resolveBlockMap(
     }
 
     // key value
+    ctx.atKey = true
     const keyStart = keyProps.end
     const keyNode = key
       ? composeNode(ctx, key, keyProps, onError)
       : composeEmptyNode(ctx, keyStart, start, null, keyProps, onError)
     if (ctx.schema.compat) flowIndentCheck(bm.indent, key, onError)
+    ctx.atKey = false
 
     if (mapIncludes(ctx, map.items, keyNode))
       onError(keyStart, 'DUPLICATE_KEY', 'Map keys must be unique')
diff --git a/src/compose/resolve-block-seq.ts b/src/compose/resolve-block-seq.ts
@@ -17,6 +17,7 @@ export function resolveBlockSeq(
   const seq = new NodeClass(ctx.schema) as YAMLSeq
 
   if (ctx.atRoot) ctx.atRoot = false
+  if (ctx.atKey) ctx.atKey = false
   let offset = bs.offset
   let commentEnd: number | null = null
   for (const { start, value } of bs.items) {
diff --git a/src/compose/resolve-flow-collection.ts b/src/compose/resolve-flow-collection.ts
@@ -32,6 +32,7 @@ export function resolveFlowCollection(
   coll.flow = true
   const atRoot = ctx.atRoot
   if (atRoot) ctx.atRoot = false
+  if (ctx.atKey) ctx.atKey = false
 
   let offset = fc.offset + fc.start.source.length
   for (let i = 0; i < fc.items.length; ++i) {
@@ -118,11 +119,13 @@ export function resolveFlowCollection(
       // item is a key+value pair
 
       // key value
+      ctx.atKey = true
       const keyStart = props.end
       const keyNode = key
         ? composeNode(ctx, key, props, onError)
         : composeEmptyNode(ctx, keyStart, start, null, props, onError)
       if (isBlock(key)) onError(keyNode.range, 'BLOCK_IN_FLOW', blockMsg)
+      ctx.atKey = false
 
       // value properties
       const valueProps = resolveProps(sep ?? [], {
diff --git a/src/compose/util-map-includes.ts b/src/compose/util-map-includes.ts
@@ -14,10 +14,6 @@ export function mapIncludes(
     typeof uniqueKeys === 'function'
       ? uniqueKeys
       : (a: ParsedNode, b: ParsedNode) =>
-          a === b ||
-          (isScalar(a) &&
-            isScalar(b) &&
-            a.value === b.value &&
-            !(a.value === '<<' && ctx.schema.merge))
+          a === b || (isScalar(a) && isScalar(b) && a.value === b.value)
   return items.some(pair => isEqual(pair.key, search))
 }
diff --git a/src/doc/Document.ts b/src/doc/Document.ts
@@ -383,13 +383,13 @@ export class Document<
       case '1.1':
         if (this.directives) this.directives.yaml.version = '1.1'
         else this.directives = new Directives({ version: '1.1' })
-        opt = { merge: true, resolveKnownTags: false, schema: 'yaml-1.1' }
+        opt = { resolveKnownTags: false, schema: 'yaml-1.1' }
         break
       case '1.2':
       case 'next':
         if (this.directives) this.directives.yaml.version = version
         else this.directives = new Directives({ version })
-        opt = { merge: false, resolveKnownTags: true, schema: 'core' }
+        opt = { resolveKnownTags: true, schema: 'core' }
         break
       case null:
         if (this.directives) delete this.directives
diff --git a/src/nodes/Node.ts b/src/nodes/Node.ts
@@ -7,7 +7,7 @@ import type { Alias } from './Alias.js'
 import { isDocument, NODE_TYPE } from './identity.js'
 import type { Scalar } from './Scalar.js'
 import { toJS, ToJSContext } from './toJS.js'
-import type { YAMLMap } from './YAMLMap.js'
+import type { MapLike, YAMLMap } from './YAMLMap.js'
 import type { YAMLSeq } from './YAMLSeq.js'
 
 export type Node<T = unknown> =
@@ -70,6 +70,16 @@ export abstract class NodeBase {
   /** A fully qualified tag, if required */
   declare tag?: string
 
+  /**
+   * Customize the way that a key-value pair is resolved.
+   * Used for YAML 1.1 !!merge << handling.
+   */
+  declare addToJSMap?: (
+    ctx: ToJSContext | undefined,
+    map: MapLike,
+    value: unknown
+  ) => void
+
   /** A plain JS representation of this node */
   abstract toJSON(): any
 
diff --git a/src/nodes/addPairToJSMap.ts b/src/nodes/addPairToJSMap.ts
@@ -1,25 +1,20 @@
 import { warn } from '../log.js'
+import { addMergeToJSMap, isMergeKey } from '../schema/yaml-1.1/merge.js'
 import { createStringifyContext } from '../stringify/stringify.js'
-import { isAlias, isMap, isNode, isScalar, isSeq } from './identity.js'
+import { isNode } from './identity.js'
 import type { Pair } from './Pair.js'
-import { Scalar } from './Scalar.js'
 import { toJS, ToJSContext } from './toJS.js'
 import type { MapLike } from './YAMLMap.js'
 
-const MERGE_KEY = '<<'
-
 export function addPairToJSMap(
   ctx: ToJSContext | undefined,
   map: MapLike,
   { key, value }: Pair
 ) {
-  if (ctx?.doc.schema.merge && isMergeKey(key)) {
-    value = isAlias(value) ? value.resolve(ctx.doc) : value
-    if (isSeq(value)) for (const it of value.items) mergeToJSMap(ctx, map, it)
-    else if (Array.isArray(value))
-      for (const it of value) mergeToJSMap(ctx, map, it)
-    else mergeToJSMap(ctx, map, value)
-  } else {
+  if (isNode(key) && key.addToJSMap) key.addToJSMap(ctx, map, value)
+  // TODO: Should drop this special case for bare << handling
+  else if (isMergeKey(ctx, key)) addMergeToJSMap(ctx, map, value)
+  else {
     const jsKey = toJS(key, '', ctx)
     if (map instanceof Map) {
       map.set(jsKey, toJS(value, jsKey, ctx))
@@ -41,45 +36,6 @@ export function addPairToJSMap(
   return map
 }
 
-const isMergeKey = (key: unknown) =>
-  key === MERGE_KEY ||
-  (isScalar(key) &&
-    key.value === MERGE_KEY &&
-    (!key.type || key.type === Scalar.PLAIN))
-
-// If the value associated with a merge key is a single mapping node, each of
-// its key/value pairs is inserted into the current mapping, unless the key
-// already exists in it. If the value associated with the merge key is a
-// sequence, then this sequence is expected to contain mapping nodes and each
-// of these nodes is merged in turn according to its order in the sequence.
-// Keys in mapping nodes earlier in the sequence override keys specified in
-// later mapping nodes. -- http://yaml.org/type/merge.html
-function mergeToJSMap(
-  ctx: ToJSContext | undefined,
-  map: MapLike,
-  value: unknown
-) {
-  const source = ctx && isAlias(value) ? value.resolve(ctx.doc) : value
-  if (!isMap(source))
-    throw new Error('Merge sources must be maps or map aliases')
-  const srcMap = source.toJSON(null, ctx, Map)
-  for (const [key, value] of srcMap) {
-    if (map instanceof Map) {
-      if (!map.has(key)) map.set(key, value)
-    } else if (map instanceof Set) {
-      map.add(key)
-    } else if (!Object.prototype.hasOwnProperty.call(map, key)) {
-      Object.defineProperty(map, key, {
-        value,
-        writable: true,
-        enumerable: true,
-        configurable: true
-      })
-    }
-  }
-  return map
-}
-
 function stringifyKey(
   key: unknown,
   jsKey: unknown,
diff --git a/src/schema/Schema.ts b/src/schema/Schema.ts
@@ -13,7 +13,6 @@ const sortMapEntriesByKey = (a: Pair<any>, b: Pair<any>) =>
 export class Schema {
   compat: Array<CollectionTag | ScalarTag> | null
   knownTags: Record<string, CollectionTag | ScalarTag>
-  merge: boolean
   name: string
   sortMapEntries: ((a: Pair, b: Pair) => number) | null
   tags: Array<CollectionTag | ScalarTag>
@@ -38,10 +37,9 @@ export class Schema {
       : compat
         ? getTags(null, compat)
         : null
-    this.merge = !!merge
     this.name = (typeof schema === 'string' && schema) || 'core'
     this.knownTags = resolveKnownTags ? coreKnownTags : {}
-    this.tags = getTags(customTags, this.name)
+    this.tags = getTags(customTags, this.name, merge)
     this.toStringOptions = toStringDefaults ?? null
 
     Object.defineProperty(this, MAP, { value: map })
diff --git a/src/schema/tags.ts b/src/schema/tags.ts
@@ -9,6 +9,7 @@ import { int, intHex, intOct } from './core/int.js'
 import { schema as core } from './core/schema.js'
 import { schema as json } from './json/schema.js'
 import { binary } from './yaml-1.1/binary.js'
+import { merge } from './yaml-1.1/merge.js'
 import { omap } from './yaml-1.1/omap.js'
 import { pairs } from './yaml-1.1/pairs.js'
 import { schema as yaml11 } from './yaml-1.1/schema.js'
@@ -36,6 +37,7 @@ const tagsByName = {
   intOct,
   intTime,
   map,
+  merge,
   null: nullTag,
   omap,
   pairs,
@@ -50,6 +52,7 @@ export type Tags = Array<ScalarTag | CollectionTag | TagId>
 
 export const coreKnownTags = {
   'tag:yaml.org,2002:binary': binary,
+  'tag:yaml.org,2002:merge': merge,
   'tag:yaml.org,2002:omap': omap,
   'tag:yaml.org,2002:pairs': pairs,
   'tag:yaml.org,2002:set': set,
@@ -58,9 +61,17 @@ export const coreKnownTags = {
 
 export function getTags(
   customTags: SchemaOptions['customTags'] | undefined,
-  schemaName: string
+  schemaName: string,
+  addMergeTag?: boolean
 ) {
-  let tags: Tags | undefined = schemas.get(schemaName)
+  const schemaTags = schemas.get(schemaName)
+  if (schemaTags && !customTags) {
+    return addMergeTag && !schemaTags.includes(merge)
+      ? schemaTags.concat(merge)
+      : schemaTags.slice()
+  }
+
+  let tags: Tags | undefined = schemaTags
   if (!tags) {
     if (Array.isArray(customTags)) tags = []
     else {
@@ -79,14 +90,18 @@ export function getTags(
   } else if (typeof customTags === 'function') {
     tags = customTags(tags.slice())
   }
+  if (addMergeTag) tags = tags.concat(merge)
 
-  return tags.map(tag => {
-    if (typeof tag !== 'string') return tag
-    const tagObj = tagsByName[tag]
-    if (tagObj) return tagObj
-    const keys = Object.keys(tagsByName)
-      .map(key => JSON.stringify(key))
-      .join(', ')
-    throw new Error(`Unknown custom tag "${tag}"; use one of ${keys}`)
-  })
+  return tags.reduce<(CollectionTag | ScalarTag)[]>((tags, tag) => {
+    const tagObj = typeof tag === 'string' ? tagsByName[tag] : tag
+    if (!tagObj) {
+      const tagName = JSON.stringify(tag)
+      const keys = Object.keys(tagsByName)
+        .map(key => JSON.stringify(key))
+        .join(', ')
+      throw new Error(`Unknown custom tag ${tagName}; use one of ${keys}`)
+    }
+    if (!tags.includes(tagObj)) tags.push(tagObj)
+    return tags
+  }, [])
 }
diff --git a/src/schema/types.ts b/src/schema/types.ts
@@ -14,11 +14,13 @@ interface TagBase {
   createNode?: (schema: Schema, value: unknown, ctx: CreateNodeContext) => Node
 
   /**
-   * If `true`, together with `test` allows for values to be stringified without
-   * an explicit tag. For most cases, it's unlikely that you'll actually want to
-   * use this, even if you first think you do.
+   * If `true`, allows for values to be stringified without
+   * an explicit tag together with `test`.
+   * If `'key'`, this only applies if the value is used as a mapping key.
+   * For most cases, it's unlikely that you'll actually want to use this,
+   * even if you first think you do.
    */
-  default?: boolean
+  default?: boolean | 'key'
 
   /**
    * If a tag has multiple forms that should be parsed and/or stringified
diff --git a/src/schema/yaml-1.1/merge.ts b/src/schema/yaml-1.1/merge.ts
diff --git a/src/schema/yaml-1.1/schema.ts b/src/schema/yaml-1.1/schema.ts
diff --git a/src/stringify/stringify.ts b/src/stringify/stringify.ts
diff --git a/tests/doc/anchors.ts b/tests/doc/anchors.ts
