fix(296): parse with lost repeated closing tag (#297)

* fix: extract node list ot separate file, fix nested parsing

* chore:  node list tests

* fix: nested tags parsing

* chore: changeset

* chore: remove unused files

* chore: disable publish on every commit

.changeset/purple-horses-shave.md

@@ -0,0 +1,27 @@
+---
+"@bbob/parser": major
+"@bbob/cli": major
+"@bbob/core": major
+"@bbob/html": major
+"@bbob/plugin-helper": major
+"@bbob/preset": major
+"@bbob/preset-html5": major
+"@bbob/preset-react": major
+"@bbob/preset-vue": major
+"@bbob/react": major
+"@bbob/types": major
+"@bbob/vue2": major
+"@bbob/vue3": major
+---
+
+Now `@bbob/parser` correctly parses nested tags like
+
+```html
+Hello World[u]Wrong underline[u] This is another text [u]and this, too[/u]
+```
+to
+```html
+Hello World[u]Wrong underline[u] This is another text <u>and this, too</u>
+```
+
+Fixes #296, #295

.github/workflows/pr.yml

@@ -1,5 +1,12 @@
 name: Pull Request
 on:
+  workflow_dispatch:
+    inputs:
+      test:
+        description: 'Build & Publish Pre-alpha Release'
+        required: true
+        type: boolean
+        default: true
 #  workflow_run:
 #    workflows:
 #      - Tests
@@ -23,7 +30,7 @@ permissions:
 
 jobs:
   build:
-    name: 'Build & pre-alpha release'
+    name: 'Build & Publish Pre-alpha Release'
 
     runs-on: ubuntu-latest
 

packages/bbob-parser/src/NodeList.ts

@@ -0,0 +1,36 @@
+class NodeList<Value> {
+  private readonly n: Value[];
+
+  constructor() {
+    this.n = [];
+  }
+
+  last() {
+    const len = this.n.length
+
+    if (len > 0) {
+      return this.n[len - 1];
+    }
+
+    return undefined;
+  }
+
+  has() {
+    return this.n.length > 0;
+  }
+
+  flush() {
+    return this.n.length ? this.n.pop() : undefined;
+  }
+
+  push(value: Value) {
+    this.n.push(value);
+  }
+
+  ref() {
+    return this.n;
+  }
+}
+
+export { NodeList };
+export default NodeList;

packages/bbob-parser/src/lexer.ts

@@ -91,6 +91,11 @@ export function createLexer(buffer: string, options: LexerOptions = {}): LexerTo
 
   /**
    * Emits newly created token to subscriber
+   *
+   * @param {number} type - 1 - word, 2 - tag, 3 - attr-name, 4 - attr-value, 5 - space, 6 - new-line
+   * @param {string} value - token value
+   * @param {number} startPos - start position
+   * @param {number} endPos - end position
    */
   function emitToken(type: number, value: string, startPos?: number, endPos?: number) {
     const token = createTokenOfType(type, value, row, prevCol, startPos, endPos);

packages/bbob-parser/src/parse.ts

@@ -6,37 +6,7 @@ import { createLexer } from "./lexer.js";
 
 import { Token, TYPE_ATTR_NAME, TYPE_ATTR_VALUE, TYPE_NEW_LINE, TYPE_SPACE, TYPE_TAG, TYPE_WORD } from "./Token.js";
 
-class NodeList<Value> {
-  private n: Value[];
-
-  constructor() {
-    this.n = [];
-  }
-
-  last() {
-    if (
-      Array.isArray(this.n) &&
-      this.n.length > 0 &&
-      typeof this.n[this.n.length - 1] !== "undefined"
-    ) {
-      return this.n[this.n.length - 1];
-    }
-
-    return null;
-  }
-
-  flush() {
-    return this.n.length ? this.n.pop() : false;
-  }
-
-  push(value: Value) {
-    this.n.push(value);
-  }
-
-  toArray() {
-    return this.n;
-  }
-}
+import { NodeList } from "./NodeList.js";
 
 const createList = <Type>() => new NodeList<Type>();
 
@@ -67,13 +37,13 @@ function parse(input: string, opts: ParseOptions = {}) {
    * @private
    * @type {NodeList}
    */
-  const tagNodes = createList<TagNode>();
+  let activeTagNode: TagNode | null = null;
   /**
    * Temp buffer of tag attributes
    * @private
    * @type {NodeList}
    */
-  const tagNodesAttrName = createList<string>();
+  let activeTagNodesAttrName: string | null = null;
 
   /**
    * Cache for nested tags checks
@@ -122,30 +92,31 @@ function parse(input: string, opts: ParseOptions = {}) {
    * Flushes temp tag nodes and its attributes buffers
    * @private
    */
-  function tagNodesFlush() {
-    if (tagNodes.flush()) {
-      tagNodesAttrName.flush();
+  function activeTagNodeFlush() {
+    if (activeTagNode) {
+      activeTagNode = null;
+      activeTagNodesAttrName = null;
     }
   }
 
   /**
    * @private
    */
-  function getNodes() {
+  function getNodesContent() {
     const lastNestedNode = nestedNodes.last();
 
     if (lastNestedNode && isTagNode(lastNestedNode)) {
       return lastNestedNode.content;
     }
 
-    return nodes.toArray();
+    return nodes.ref();
   }
 
   /**
    * @private
    */
   function nodesAppendAsString(
-    nodes?: TagNodeTree,
+    nodes: TagNodeTree,
     node?: TagNode,
     isNested = true
   ) {
@@ -167,7 +138,9 @@ function parse(input: string, opts: ParseOptions = {}) {
   /**
    * @private
    */
-  function nodesAppend(nodes?: TagNodeTree, node?: NodeContent) {
+  function nodesAppend(node: NodeContent) {
+    const nodes = getNodesContent() as TagNodeTree;
+
     if (Array.isArray(nodes) && typeof node !== "undefined") {
       if (isTagNode(node)) {
         if (isTagAllowed(node.tag)) {
@@ -186,18 +159,17 @@ function parse(input: string, opts: ParseOptions = {}) {
    * @param {Token} token
    */
   function tagHandleStart(token: Token) {
-    tagNodesFlush();
+    activeTagNodeFlush();
 
     const tagNode = TagNode.create(token.getValue(), {}, [], { from: token.getStart(), to: token.getEnd() });
     const isNested = isTokenNested(token);
 
-    tagNodes.push(tagNode);
+    activeTagNode = tagNode;
 
     if (isNested) {
       nestedNodes.push(tagNode);
     } else {
-      const nodes = getNodes();
-      nodesAppend(nodes, tagNode);
+      nodesAppend(tagNode);
     }
   }
 
@@ -209,20 +181,16 @@ function parse(input: string, opts: ParseOptions = {}) {
     const tagName = token.getValue().slice(1);
     const lastNestedNode = nestedNodes.flush();
 
-    tagNodesFlush();
+    activeTagNodeFlush();
 
     if (lastNestedNode) {
-      const nodes = getNodes()
-
       if (isTagNode(lastNestedNode)) {
         lastNestedNode.setEnd({ from: token.getStart(), to: token.getEnd() });
       }
 
-      nodesAppend(nodes, lastNestedNode);
+      nodesAppend(lastNestedNode);
     } else if (!isTagNested(tagName)) { // when we have only close tag [/some] without any open tag
-      const nodes = getNodes();
-
-      nodesAppend(nodes, token.toString({ openTag, closeTag }));
+      nodesAppend(token.toString({ openTag, closeTag }));
     } else if (typeof options.onError === "function") {
       const tag = token.getValue();
       const line = token.getLine();
@@ -241,31 +209,24 @@ function parse(input: string, opts: ParseOptions = {}) {
    * @param {Token} token
    */
   function nodeHandle(token: Token) {
-    /**
-     * @type {TagNode}
-     */
-    const activeTagNode = tagNodes.last();
+
     const tokenValue = token.getValue();
     const isNested = isTagNested(token.toString());
-    const nodes = getNodes();
 
-    if (activeTagNode !== null) {
+    if (activeTagNode) {
       switch (token.type) {
         case TYPE_ATTR_NAME:
-          tagNodesAttrName.push(tokenValue);
-          const attrName = tagNodesAttrName.last();
+          activeTagNodesAttrName = tokenValue;
 
-          if (attrName) {
-            activeTagNode.attr(attrName, "");
+          if (tokenValue) {
+            activeTagNode.attr(tokenValue, "");
           }
           break;
 
         case TYPE_ATTR_VALUE:
-          const attrValName = tagNodesAttrName.last();
-
-          if (attrValName) {
-            activeTagNode.attr(attrValName, tokenValue);
-            tagNodesAttrName.flush();
+          if (activeTagNodesAttrName) {
+            activeTagNode.attr(activeTagNodesAttrName, tokenValue);
+            activeTagNodesAttrName = null;
           } else {
             activeTagNode.attr(tokenValue, tokenValue);
           }
@@ -277,20 +238,20 @@ function parse(input: string, opts: ParseOptions = {}) {
           if (isNested) {
             activeTagNode.append(tokenValue);
           } else {
-            nodesAppend(nodes, tokenValue);
+            nodesAppend(tokenValue);
           }
           break;
 
         case TYPE_TAG:
           // if tag is not allowed, just pass it as is
-          nodesAppend(nodes, token.toString({ openTag, closeTag }));
+          nodesAppend(token.toString({ openTag, closeTag }));
           break;
       }
     } else if (token.isText()) {
-      nodesAppend(nodes, tokenValue);
+      nodesAppend(tokenValue);
     } else if (token.isTag()) {
       // if tag is not allowed, just pass it as is
-      nodesAppend(nodes, token.toString({ openTag, closeTag }));
+      nodesAppend(token.toString({ openTag, closeTag }));
     }
   }
 
@@ -330,15 +291,20 @@ function parse(input: string, opts: ParseOptions = {}) {
   // eslint-disable-next-line no-unused-vars
   const tokens = tokenizer.tokenize();
 
-  // handles situations where we open tag, but forgot close them
+  // handles situations where we opened tag, but forget to close them
   // for ex [q]test[/q][u]some[/u][q]some [u]some[/u] // forgot to close [/q]
   // so we need to flush nested content to nodes array
-  const lastNestedNode = nestedNodes.flush();
-  if (isTagNode(lastNestedNode) && isTagNested(lastNestedNode.tag)) {
-    nodesAppendAsString(getNodes(), lastNestedNode, false);
-  }
+  do {
+    const node = nestedNodes.flush();
 
-  return nodes.toArray();
+    if (isTagNode(node) && isTagNested(node.tag)) {
+      nodesAppendAsString(getNodesContent(), node, false);
+    } else if (typeof node !== 'undefined') {
+      nodesAppend(node);
+    }
+  } while (nestedNodes.has());
+
+  return nodes.ref();
 }
 
 export { parse };

packages/bbob-parser/test/NodeList.test.ts

@@ -0,0 +1,35 @@
+import { NodeList } from '../src/NodeList';
+
+describe('NodeList', () => {
+  test('push', () => {
+    const list = new NodeList();
+
+    list.push('a');
+    list.push('b');
+    list.push('c');
+
+    expect(list.ref()).toEqual(['a', 'b', 'c']);
+  });
+
+  test('last', () => {
+    const list = new NodeList();
+
+    list.push('a');
+    list.push('b');
+    list.push('c');
+
+    expect(list.last()).toBe('c');
+    expect(list.ref()).toEqual(['a', 'b', 'c']);
+  });
+
+  test('flush', () => {
+    const list = new NodeList();
+
+    list.push('a');
+    list.push('b');
+    list.push('c');
+
+    expect(list.flush()).toBe('c');
+    expect(list.ref()).toEqual(['a', 'b']);
+  });
+});

packages/bbob-parser/test/parse.test.ts

@@ -962,6 +962,36 @@ sdfasdfasdf
     );
   });
 
+  test('parse with lost repeated closing tag', () => {
+    const input = parse('Hello World[u]Wrong underline[u] This is another text [u]and this, too[/u]');
+
+    expect(input).toBeMatchAST([
+      'Hello',
+      ' ',
+      'World',
+      '[u]',
+      'Wrong',
+      ' ',
+      'underline',
+      '[u]',
+      ' ',
+      'This',
+      ' ',
+      'is',
+      ' ',
+      'another',
+      ' ',
+      'text',
+      ' ',
+      {
+        tag: 'u',
+        attrs: {},
+        content: ['and', ' ', 'this,', ' ', 'too'],
+      }
+    ]);
+  });
+
+
   test('parse with url in tag content', () => {
     const input = parse('[img]https://tw.greywool.com/i/e3Ph5.png[/img]');