Initial commit

.gitignore

@@ -0,0 +1 @@
+node_modules

index.js

@@ -0,0 +1,65 @@
+"use strict";
+
+const consumable = require("@joepie91/consumable");
+
+const mapFilter = require("@promistream/map-filter");
+const parseXML = require("@promistream/parse-xml");
+const pipe = require("@promistream/pipe");
+
+// NOTE: This is not a 100% strict parser. It assumes that the sitemap XML is well-formed, and tags appearing in strange places are ignored.
+
+module.exports = function createSitemapParsingStream() {
+	// The consumable is always filled with an object, so that mutating operations never fail. However, if a tag appears outside of an expected item context, its value will be stored on a placeholder object that gets thrown away as soon as an item context is encountered, effectively throwing away the tag's data itself.
+	let currentItem = consumable({});
+	let currentTag;
+
+	return pipe([
+		parseXML({ events: [ "opentag", "closetag", "text" ] }),
+		mapFilter((event) => {
+			if (event.type === "opentag") {
+				let tag = event.value;
+
+				if (tag.name === "sitemap") {
+					currentItem.replace({
+						type: "sitemap",
+						url: undefined,
+						lastModified: undefined
+					});
+				} else if (tag.name === "url") {
+					currentItem.replace({
+						type: "url",
+						url: undefined,
+						priority: undefined
+					});
+				}
+
+				currentTag = tag.name;
+
+				return mapFilter.NoValue;
+			} else if (event.type === "closetag") {
+				let tag = event.value;
+
+				// We need to ensure that we unset the currentTag once we're done with it, otherwise text elements *between* tags might erroneously end up in our items. Technically this is not correct as we've just moved to the parent tag, but since we're only interested in text which *directly* exists within a tag uninterrupted, we can cut some corners here.
+				currentTag = undefined;
+
+				if (tag.name === "sitemap" || tag.name === "url") {
+					return currentItem.replace({});
+				} else {
+					return mapFilter.NoValue;
+				}
+			} else {
+				let text = event.value;
+
+				if (currentTag === "loc") {
+					currentItem.peek().url = text;
+				} else if (currentTag === "lastmod") {
+					currentItem.peek().lastModified = text;
+				} else if (currentTag === "priority") {
+					currentItem.peek().currentTag = text;
+				}
+
+				return mapFilter.NoValue;
+			}
+		})
+	]);
+};

notes.txt

@@ -0,0 +1 @@
+TODO: Support for image sitemaps (https://developers.google.com/search/docs/advanced/sitemaps/image-sitemaps)

package.json

@@ -0,0 +1,21 @@
+{
+  "name": "@promistream/parse-sitemap",
+  "version": "0.1.0",
+  "main": "index.js",
+  "keywords": [
+    "promistream"
+  ],
+  "repository": "http://git.cryto.net/promistream/parse-sitemap.git",
+  "author": "Sven Slootweg <admin@cryto.net>",
+  "license": "WTFPL OR CC0-1.0",
+  "devDependencies": {
+    "@joepie91/eslint-config": "^1.1.0",
+    "eslint": "^6.8.0"
+  },
+  "dependencies": {
+    "@joepie91/consumable": "^1.0.1",
+    "@promistream/map-filter": "^0.1.0",
+    "@promistream/parse-xml": "^0.1.0",
+    "@promistream/pipe": "^0.1.4"
+  }
+}