feat: add tagNameProcessor and attributeNameProcessor

Author: lpatiny

README.md

@@ -59,7 +59,7 @@ object = {
 | cdataTagName                | If specified, parser parse CDATA as nested tag instead of adding it's value to parent tag.                                                                                                                                                                                                                                                                                                                         | `false`                                               |
 | arrayMode                   | When `false`, a tag with single occurrence is parsed as an object but as an array in case of multiple occurences. When `true`, a tag will be parsed as an array always excluding leaf nodes. When `strict`, all the tags will be parsed as array only. When instance of `RegEx`, only tags will be parsed as array that match the regex. When `function` a tag name is passed to the callback that can be checked. | `false`                                               |
 | tagValueProcessor           | Process tag value during transformation. Like HTML decoding, word capitalization, etc. Applicable in case of string only.                                                                                                                                                                                                                                                                                          | `(value) => decoder.decode(value).replace(/\r/g, '')` |
-| attributeValueProcessor     | Process attribute value during transformation. Like HTML decoding, word capitalization, etc. Applicable in case of string only.                                                                                                                                                                                                                                                                                    | `(value) => value`                                    |
+| attributeValueProcessor     | Process attribute value during transformation. Like HTML decoding, word capitalization, etc.                                                                                                                                                                                                                                                                                                                       | `(value) => value`                                    |
 | stopNodes                   | an array of tag names which are not required to be parsed. They are kept as Uint8Array.                                                                                                                                                                                                                                                                                                                            | `[]`                                                  |
 
 ## [API Documentation](https://cheminfo.github.io/arraybuffer-xml-parser/)

package.json

@@ -35,19 +35,19 @@
   },
   "devDependencies": {
     "@babel/plugin-transform-modules-commonjs": "^7.15.4",
-    "@types/jest": "^27.0.1",
+    "@types/jest": "^27.0.2",
     "cheminfo-build": "^1.1.11",
-    "eslint": "^7.32.0",
-    "eslint-config-cheminfo": "^5.4.0",
+    "eslint": "^8.0.1",
+    "eslint-config-cheminfo": "^7.1.1",
     "he": "^1.2.0",
-    "iobuffer": "^5.0.3",
-    "jest": "^27.1.0",
+    "iobuffer": "^5.0.4",
+    "jest": "^27.3.1",
     "pako": "^2.0.4",
-    "prettier": "^2.3.2",
-    "rollup": "^2.56.3",
+    "prettier": "^2.4.1",
+    "rollup": "^2.58.0",
     "uint8-base64": "^0.1.1"
   },
   "dependencies": {
-    "dynamic-typing": "^0.1.2"
+    "dynamic-typing": "^0.1.3"
   }
 }

src/__tests__/tagProcessors.test.js

@@ -0,0 +1,32 @@
+/* eslint-disable camelcase */
+/* eslint-disable no-tabs */
+import he from 'he';
+
+import { parse } from '../parse';
+
+describe('XMLParser', () => {
+  it('tag name processor', () => {
+    const xmlData = `<?xml version='1.0'?>
+        <any_name>
+            <person>
+                end
+            </person>
+        </any_name>`;
+
+    const result = parse(xmlData, {
+      tagNameProcessor: (name) => name.toUpperCase(),
+    });
+    expect(result).toStrictEqual({ ANY_NAME: { PERSON: 'end' } });
+  });
+
+  it('attribute name processor', () => {
+    const xmlData = `<?xml version='1.0'?>
+	    <ab param1="abc" param2="def">
+	    </ab>`;
+
+    const result = parse(xmlData, {
+      attributeNameProcessor: (name) => name.toUpperCase(),
+    });
+    expect(result).toStrictEqual({ ab: { $PARAM1: 'abc', $PARAM2: 'def' } });
+  });
+});

src/parse.js

@@ -4,22 +4,24 @@ import { traversableToJSON } from './traversableToJSON';
 
 /**
  * Parse an ArrayBuffer or Uint8Array representing an XML
- * @param {ArrayBuffer|Uint8Arra} xmlData
+ * @param {string|ArrayBuffer|Uint8Array} xmlData
  * @param {object} [options={}]
- * @param {string} [attributeNamePrefix='$']
- * @param {boolean} [attributesNodeName=false]
- * @param {string} [textNodeName='#text']
- * @param {boolean} [trimValues=true] should we remove ascii < 32
- * @param {boolean} [ignoreAttributes=false] skip attributes
- * @param {boolean} [ignoreNameSpace=false]
- * @param {boolean} [dynamicTypingAttributeValue=true] Parse attribute values that looks like number or boolean
- * @param {boolean} [allowBooleanAttributes=false]
- * @param {boolean} [dynamicTypingNodeValue=true] Parse tag values that looks like number or boolean
- * @param {boolean} [arrayMode=false]
- * @param {boolean} [cdataTagName=false]
- * @param {function} [tagValueProcessor=(v, node) => decoder.decode(v)] Tag values can be modified during parsing. By default we decode the tag value (a Uint8Array) using TextDecoder
- * @param {function} [attributeValueProcessor=(v) => v] Attribute values can be modified during parsing
- * @param {boolean} [stopNodes=[]] prevent further parsing
+ * @param {string} [options.attributeNamePrefix='$']
+ * @param {boolean} [options.attributesNodeName=false]
+ * @param {string} [options.textNodeName='#text']
+ * @param {boolean} [options.trimValues=true] should we remove ascii < 32
+ * @param {boolean} [options.ignoreAttributes=false] skip attributes
+ * @param {boolean} [options.ignoreNameSpace=false]
+ * @param {boolean} [options.dynamicTypingAttributeValue=true] Parse attribute values that looks like number or boolean
+ * @param {boolean} [options.allowBooleanAttributes=false]
+ * @param {boolean} [options.dynamicTypingNodeValue=true] Parse tag values that looks like number or boolean
+ * @param {boolean} [options.arrayMode=false]
+ * @param {boolean} [options.cdataTagName=false]
+ * @param {function} [options.tagValueProcessor=(v, node) => decoder.decode(v)] Tag values can be modified during parsing. By default we decode the tag value (a Uint8Array) using TextDecoder
+ * @param {function} [options.attributeValueProcessor=(v) => v] Attribute values can be modified during parsing
+ * @param {function} [options.tagNameProcessor=(v) => v] Callback allowing to rename tag names
+ * @param {function} [options.attributeNameProcessor=(v) => v] Callback allowing to rename attribute name
+ * @param {boolean} [options.stopNodes=[]] prevent further parsing
  *
  * @returns {object}
  */

src/traversableToJSON.js

@@ -10,7 +10,13 @@ import { isTagNameInArrayMode, merge, isEmptyObject } from './util';
  * @returns
  */
 export function traversableToJSON(node, options, parentTagName) {
-  const { dynamicTypingNodeValue, tagValueProcessor, arrayMode } = options;
+  const {
+    dynamicTypingNodeValue,
+    tagValueProcessor,
+    arrayMode,
+    tagNameProcessor,
+    attributeNameProcessor,
+  } = options;
   const result = {};
 
   if (tagValueProcessor) {
@@ -43,9 +49,12 @@ export function traversableToJSON(node, options, parentTagName) {
     if (options.attributeNamePrefix) {
       // need to rename the attributes
       const renamedAttributes = {};
-      for (let key in node.attributes) {
-        renamedAttributes[options.attributeNamePrefix + key] =
-          node.attributes[key];
+      for (let attributeName in node.attributes) {
+        const newAttributeName = attributeNameProcessor
+          ? attributeNameProcessor(attributeName)
+          : attributeName;
+        renamedAttributes[options.attributeNamePrefix + newAttributeName] =
+          node.attributes[attributeName];
       }
       attributes = renamedAttributes;
     }
@@ -57,14 +66,13 @@ export function traversableToJSON(node, options, parentTagName) {
     merge(result, attributes, arrayMode);
   }
 
-  const keys = Object.keys(node.children);
-  for (let index = 0; index < keys.length; index++) {
-    const tagName = keys[index];
+  for (const tagName in node.children) {
+    const newTagName = tagNameProcessor ? tagNameProcessor(tagName) : tagName;
     if (node.children[tagName] && node.children[tagName].length > 1) {
       result[tagName] = [];
       for (let tag in node.children[tagName]) {
         if (Object.prototype.hasOwnProperty.call(node.children[tagName], tag)) {
-          result[tagName].push(
+          result[newTagName].push(
             traversableToJSON(node.children[tagName][tag], options, tagName),
           );
         }
@@ -78,7 +86,7 @@ export function traversableToJSON(node, options, parentTagName) {
       const asArray =
         (arrayMode === true && typeof subResult === 'object') ||
         isTagNameInArrayMode(tagName, arrayMode, parentTagName);
-      result[tagName] = asArray ? [subResult] : subResult;
+      result[newTagName] = asArray ? [subResult] : subResult;
     }
   }