Wordpress source normaliser plugin extension

packages/gatsby-source-wordpress/README.md

@@ -88,6 +88,12 @@ plugins: [
         sourceUrl: "https://source-url.com",
         replacementUrl: "https://replacement-url.com",
       },
+      // Add plugins to modify WordPress entitys
+      plugins: [
+        {
+          resolve: `wordpress-normalizer-plugin`
+        },
+      ]
     },
   },
 ];
@@ -467,6 +473,26 @@ To learn more about image processing check
 * source code of [image processing example
   site](https://github.com/gatsbyjs/gatsby/tree/master/examples/image-processing).
 
+## WordPress Source Normalizer Plugins
+
+Write custom plugins to modify WordPress entities during the build process.
+
+The set method accepts two arguments, the normalizer and the priority in which this normalizer is set. Do not pass the priority if you wish your normalizer to be called last.
+
+**index.js Example**
+
+```javascript
+exports.normalize = function (normalizer) {
+
+  function modify(entities, args) {
+    // Modify entities
+    return entities
+  }
+
+  return normalizer.set(modify, 140)
+}
+```
+
 ## Site's `gatsby-node.js` example
 
 ```javascript

packages/gatsby-source-wordpress/src/__tests__/normalize.js

@@ -25,7 +25,7 @@ describe(`Process WordPress data`, () => {
   it(`creates Gatsby IDs for each entity`, () => {
     const createNodeId = jest.fn()
     createNodeId.mockReturnValue(`uuid-from-gatsby`)
-    entities = normalize.createGatsbyIds(createNodeId, entities)
+    entities = normalize.createGatsbyIds(entities, { createNodeId })
     expect(entities).toMatchSnapshot()
   })
   it(`Creates map of types`, () => {

packages/gatsby-source-wordpress/src/gatsby-node.js

@@ -1,5 +1,7 @@
 const fetch = require(`./fetch`)
 const normalize = require(`./normalize`)
+const Normalizer = require(`./normalizer`)
+const _ = require(`lodash`)
 
 const typePrefix = `wordpress__`
 const refactoredEntityTypes = {
@@ -28,6 +30,7 @@ exports.sourceNodes = async (
     verboseOutput,
     perPage = 100,
     searchAndReplaceContentUrls = {},
+    plugins = [],
   }
 ) => {
   const { createNode } = boundActionCreators
@@ -50,57 +53,44 @@ exports.sourceNodes = async (
     refactoredEntityTypes,
   })
 
-  // Normalize data & create nodes
-
-  // Remove ACF key if it's not an object
-  entities = normalize.normalizeACF(entities)
-
-  // Creates entities from object collections of entities
-  entities = normalize.normalizeEntities(entities)
-
-  // Standardizes ids & cleans keys
-  entities = normalize.standardizeKeys(entities)
-
-  // Converts to use only GMT dates
-  entities = normalize.standardizeDates(entities)
-
-  // Lifts all "rendered" fields to top-level.
-  entities = normalize.liftRenderedField(entities)
-
-  // Exclude entities of unknown shape
-  entities = normalize.excludeUnknownEntities(entities)
-
-  // Creates Gatsby IDs for each entity
-  entities = normalize.createGatsbyIds(createNodeId, entities)
-
-  // Creates links between authors and user entities
-  entities = normalize.mapAuthorsToUsers(entities)
-
-  // Creates links between posts and tags/categories.
-  entities = normalize.mapPostsToTagsCategories(entities)
-
-  // Creates links between tags/categories and taxonomies.
-  entities = normalize.mapTagsCategoriesToTaxonomies(entities)
-
-  // Creates links from entities to media nodes
-  entities = normalize.mapEntitiesToMedia(entities)
-
-  // Downloads media files and removes "sizes" data as useless in Gatsby context.
-  entities = await normalize.downloadMediaFiles({
-    entities,
+  var normalizer = new Normalizer(entities, {
+    createNodeId,
     store,
     cache,
     createNode,
     _auth,
-  })
-
-  // Search and replace Content Urls
-  entities = normalize.searchReplaceContentUrls({
-    entities,
     searchAndReplaceContentUrls,
   })
 
-  // creates nodes for each entry
+  normalizer
+    .set(normalize.normalizeACF, 10)
+    .set(normalize.normalizeEntities, 20)
+    .set(normalize.standardizeKeys, 30)
+    .set(normalize.standardizeDates, 40)
+    .set(normalize.liftRenderedField, 50)
+    .set(normalize.excludeUnknownEntities, 60)
+    .set(normalize.createGatsbyIds, 70)
+    .set(normalize.mapAuthorsToUsers, 80)
+    .set(normalize.mapPostsToTagsCategories, 90)
+    .set(normalize.mapTagsCategoriesToTaxonomies, 100)
+    .set(normalize.mapEntitiesToMedia, 110)
+    .set(normalize.downloadMediaFiles, 120)
+    .set(normalize.searchReplaceContentUrls, 130)
+
+  for (let i = 0; i < plugins.length; i++) {
+    var requiredPlugin = require(plugins[i].resolve)
+
+    if (_.isFunction(requiredPlugin.normalize)) {
+      var pluginNormalizer = requiredPlugin.normalize(normalizer)
+      if (pluginNormalizer instanceof Normalizer) {
+        normalizer = pluginNormalizer
+      }
+    }
+
+  }
+
+  entities = await normalizer.normalize()
+
   normalize.createNodesFromEntities({ entities, createNode })
 
   return

packages/gatsby-source-wordpress/src/normalize.js

@@ -140,9 +140,9 @@ exports.liftRenderedField = entities =>
 exports.excludeUnknownEntities = entities =>
   entities.filter(e => e.wordpress_id) // Excluding entities without ID
 
-exports.createGatsbyIds = (createNodeId, entities) =>
+exports.createGatsbyIds = (entities, args) =>
   entities.map(e => {
-    e.id = createNodeId(`${e.__type}-${e.wordpress_id.toString()}`)
+    e.id = args.createNodeId(`${e.__type}-${e.wordpress_id.toString()}`)
     return e
   })
 
@@ -223,21 +223,18 @@ exports.mapTagsCategoriesToTaxonomies = entities =>
     return e
   })
 
-exports.searchReplaceContentUrls = function({
-  entities,
-  searchAndReplaceContentUrls,
-}) {
+exports.searchReplaceContentUrls = function(entities, args) {
   if (
-    !_.isPlainObject(searchAndReplaceContentUrls) ||
-    !_.has(searchAndReplaceContentUrls, `sourceUrl`) ||
-    !_.has(searchAndReplaceContentUrls, `replacementUrl`) ||
-    typeof searchAndReplaceContentUrls.sourceUrl !== `string` ||
-    typeof searchAndReplaceContentUrls.replacementUrl !== `string`
+    !_.isPlainObject(args.searchAndReplaceContentUrls) ||
+    !_.has(args.searchAndReplaceContentUrls, `sourceUrl`) ||
+    !_.has(args.searchAndReplaceContentUrls, `replacementUrl`) ||
+    typeof args.searchAndReplaceContentUrls.sourceUrl !== `string` ||
+    typeof args.searchAndReplaceContentUrls.replacementUrl !== `string`
   ) {
     return entities
   }
 
-  const { sourceUrl, replacementUrl } = searchAndReplaceContentUrls
+  const { sourceUrl, replacementUrl } = args.searchAndReplaceContentUrls
 
   const _blacklist = [`_links`, `__type`]
 
@@ -379,24 +376,18 @@ exports.mapEntitiesToMedia = entities => {
 }
 
 // Downloads media files and removes "sizes" data as useless in Gatsby context.
-exports.downloadMediaFiles = async ({
-  entities,
-  store,
-  cache,
-  createNode,
-  _auth,
-}) =>
+exports.downloadMediaFiles = async (entities, args) =>
   Promise.all(
     entities.map(async e => {
       let fileNode
       if (e.__type === `wordpress__wp_media`) {
         try {
           fileNode = await createRemoteFileNode({
             url: e.source_url,
-            store,
-            cache,
-            createNode,
-            auth: _auth,
+            store: args.store,
+            cache: args.cache,
+            createNode: args.createNode,
+            auth: args._auth,
           })
         } catch (e) {
           // Ignore

packages/gatsby-source-wordpress/src/normalizer.js

@@ -0,0 +1,65 @@
+const _ = require(`lodash`)
+
+class Normalizer {
+
+    /**
+     * Normalizer Constructor
+     * 
+     * @param  {Array} entitys
+     * @param  {Object} args
+     * @return {Object} this
+     */
+    constructor(entitys, args) {
+        this.entitys = entitys
+        this.args = args
+        this.normalizers = []
+        this.queue = []
+    }
+
+    /**
+     * Set Normalizers
+     * 
+     * @param {Function} normalizer
+     * @param {null|Number} priority
+     * @return {Object}
+     */
+    set(normalizer, priority = null) {
+        let property = priority === null ? `queue` : `normalizers`
+
+        this[property].push({
+            normalizer,
+            priority,
+        })
+
+        return this
+    }
+
+    /**
+     * Normalize the entities
+     * 
+     * @return {Array}
+     */
+    async normalize() {
+        var normalizers = this.getNormalizers()
+        for (let i = 0; i < normalizers.length; i++) {
+            this.entitys = await normalizers[i].normalizer(this.entitys, this.args)
+        }
+
+        return this.entitys
+    }
+
+    /**
+     * Concat the queue and prioritised normalizers
+     * 
+     * @return {Array}
+     */
+    getNormalizers() {
+        return _.concat(
+            this.normalizers.sort((a, b) => a.priority - b.priority),
+            this.queue
+        )
+    }
+
+}
+
+module.exports = Normalizer