Adding tests for broken/incomplete syntax

jstest/BrokenSyntaxTests.js

@@ -0,0 +1,120 @@
+class BrokenSyntaxTests extends BaseTest {
+	#parser;
+
+	setUp() {
+		this.#parser = Markdown.completeParser;
+	}
+
+	/**
+	 * @param {string} markdown
+	 */
+	#iterateCharacters(markdown) {
+		for (var i = 1; i < markdown.length; i++) {
+			const portion = markdown.substring(0, i);
+			this.#parser.toHTML(portion);
+		}
+	}
+
+	test_brokenSyntax() {
+		// Just try a bunch of broken syntax and make sure no exceptions are
+		// thrown or infinite loops triggered. Don't care about what's rendered
+		// as most of these are incomplete or invalid syntax.
+		this.#iterateCharacters("# ");
+		this.#iterateCharacters("## ");
+		this.#iterateCharacters("Underlined\n-");
+		this.#iterateCharacters("* ");
+		this.#iterateCharacters(" * A");
+		this.#iterateCharacters("- A");
+		this.#iterateCharacters(" - A");
+		this.#iterateCharacters("+ A");
+		this.#iterateCharacters(" + A");
+		this.#iterateCharacters("1. A");
+		this.#iterateCharacters(" 1. A");
+		this.#iterateCharacters("1. A\n 1. B\n 2. C\n2. D");
+		this.#iterateCharacters("> A");
+		this.#iterateCharacters("> \nA");
+		this.#iterateCharacters("> \n>");
+		this.#iterateCharacters(">>");
+		this.#iterateCharacters("> > A");
+		this.#iterateCharacters("> \n> A");
+		this.#iterateCharacters("> > \n> A");
+		this.#iterateCharacters("```\ncode\n```");
+		this.#iterateCharacters("```java\ncode\n```");
+		this.#iterateCharacters("    code\n    code\n\n");
+		this.#iterateCharacters("------");
+		this.#iterateCharacters("  -  -    - --*");
+		this.#iterateCharacters("Header|Header\n--|--\nCell|Cell");
+		this.#iterateCharacters("Header\n--|--\nCell");
+		this.#iterateCharacters("|Header|\n|-|\n|Cell|");
+		this.#iterateCharacters("|Header|\n|--|--|--|\n|Cell|");
+		this.#iterateCharacters("|Header|Header|\n|--|\n|Cell|Cell|");
+		this.#iterateCharacters("|Header|\n|:-|\n|Cell|");
+		this.#iterateCharacters("|Header|\n|:-:|\n|Cell|");
+		this.#iterateCharacters("|Header|\n|-:|\n|Cell|");
+		this.#iterateCharacters("term\n: definition\nterm\n: definition");
+		this.#iterateCharacters("term\n:definition\nterm\n:definition");
+		this.#iterateCharacters("HTML\n\n*[HTML]: Hypertext");
+		this.#iterateCharacters("HTML\n*[HTML]: Hypertext");
+		this.#iterateCharacters("*[HTML]: Hypertext\nHTML");
+		this.#iterateCharacters("---{.foo}\n");
+		this.#iterateCharacters("---{#foo}\n");
+		this.#iterateCharacters("---{.foo #foo lang=en}\n");
+		this.#iterateCharacters("lorem *ipsum* dolor *sit* amet");
+		this.#iterateCharacters("*lorem* ipsum *dolor* sit *amet*");
+		this.#iterateCharacters("*lorem *ipsum *dolor*** sit");
+		this.#iterateCharacters("***lorem* ipsum* dolor* sit");
+		this.#iterateCharacters("lorem _ipsum_ dolor _sit_ amet");
+		this.#iterateCharacters("_lorem_ ipsum _dolor_ sit _amet_");
+		this.#iterateCharacters("_lorem _ipsum _dolor___ sit");
+		this.#iterateCharacters("___lorem_ ipsum_ dolor_ sit");
+		this.#iterateCharacters("lorem **ipsum** dolor **sit** amet");
+		this.#iterateCharacters("**lorem** ipsum **dolor** sit **amet**");
+		this.#iterateCharacters("**lorem **ipsum **dolor****** sit");
+		this.#iterateCharacters("******lorem** ipsum** dolor** sit");
+		this.#iterateCharacters("lorem __ipsum__ dolor __sit__ amet");
+		this.#iterateCharacters("__lorem__ ipsum __dolor__ sit __amet__");
+		this.#iterateCharacters("lorem __ipsum__ dolor __sit__ amet");
+		this.#iterateCharacters("__lorem __ipsum __dolor______ sit");
+		this.#iterateCharacters("______lorem__ ipsum__ dolor__ sit");
+		this.#iterateCharacters("*_**`~~^==!__~*^!``**-+!**`==!~_^**``_");
+		this.#iterateCharacters("[link[(index.html(");
+		this.#iterateCharacters("[link[](index.html()");
+		this.#iterateCharacters("[link]](index.html))");
+		this.#iterateCharacters("[[link]((index.html)");
+		this.#iterateCharacters("]link])index.html)");
+		this.#iterateCharacters("[ link ] ( index.html )");
+		this.#iterateCharacters("(user@example.com)");
+		this.#iterateCharacters("(https://user@example.com)");
+		this.#iterateCharacters("(index.html \"title\")");
+		this.#iterateCharacters("(index.html \"title)");
+		this.#iterateCharacters("(index.html title\")");
+		this.#iterateCharacters("(index.html title)");
+		this.#iterateCharacters("![alt][image.jpg]");
+		this.#iterateCharacters("![alt[](image.jpg()");
+		this.#iterateCharacters("! [alt](image.jpg)");
+		this.#iterateCharacters("!(image.jpg)");
+		this.#iterateCharacters("[][]]][][[][][][[][[]][][][][][[[]]]][]][][");
+		this.#iterateCharacters("())())()()()(())(()((())))(()((()");
+		this.#iterateCharacters("<https://example.com>");
+		this.#iterateCharacters("<<https://example.com>");
+		this.#iterateCharacters("<https://example.com>>");
+		this.#iterateCharacters("[link][ref]\n\n[ref]: page.html");
+		this.#iterateCharacters("[link][ref]\n[ref]: page.html");
+		this.#iterateCharacters("[link][ref]\n\n[ref]:page.html");
+		this.#iterateCharacters("[ref]: page.html\n\n[link][ref]");
+		this.#iterateCharacters("[link][ref]\n\n[ref]: page.html \"title\"");
+		this.#iterateCharacters("[link][ref]\n\n[ref]: page.html title");
+		this.#iterateCharacters("Lorem[^1] ipsum[^abc] dolor[^-1]\n\n[^abc]: def\n[^1]: def\n[^-1]: def\n[^abc]: def");
+		this.#iterateCharacters("Lorem[^1]\n\n[^2]: def");
+		this.#iterateCharacters("[^1]] [[^2] [^3]] [[^4]] ]^5[");
+		this.#iterateCharacters("[^1]: Def\n\nLorem[^1]");
+		this.#iterateCharacters("[^1]: Def\n\nLorem[^2]");
+		this.#iterateCharacters("Lorem <foo> ipsum");
+		this.#iterateCharacters("Lorem <span> ipsum");
+		this.#iterateCharacters("Lorem <span class=\"foo\" id=\"foo\" lang=\"en\"> ipsum");
+		this.#iterateCharacters("Lorem </span> ipsum");
+		this.#iterateCharacters("Lorem </span foo=\"bar\">");
+		this.#iterateCharacters("Lorem <span foo='bar' baz=ipsum dolor> sit");
+		this.#iterateCharacters("\\*\\\\*\\\\\\*\\\\\\\\*\\*");
+	}
+}

php/markdown.php

@@ -120,6 +120,15 @@ class MDUtils {
 		$tn = gettype($value);
 		return ($tn === 'object') ? get_class($value) : $tn;
 	}
+
+	/**
+	 * Returns the value if it's a string, otherwise returns an empty string.
+	 * For wrapping regex capture groups which may be `false` instead of a
+	 * string.
+	 */
+	public static function makeString(mixed $text, ?string $fallback=''): ?string {
+		return gettype($text) === 'string' && $text !== '' ? $text : $fallback;
+	}
 }
 
 /**
@@ -1481,7 +1490,7 @@ class MDTagModifier {
 			if (!$found) return [ $line, null ];
 		}
 		if (!mb_eregi(self::trailingClassRegex, $line, $groups)) return [ $line, null ];
-		$bareLine = $groups[1];
+		$bareLine = MDUtils::makeString($groups[1]);
 		$mod = self::fromContents($groups[2]);
 		return [ $bareLine, $mod ];
 	}
@@ -2029,7 +2038,7 @@ class MDFencedCodeBlockReader extends MDReader {
 		$openFenceLine = $state->lines[$p++];
 		[$openFenceLine, $modifier] = MDTagModifier::fromLine($openFenceLine, $state);
 		if (!mb_eregi('```\\s*([a-z0-9]*)\\s*$', $openFenceLine, $groups)) return null;
-		$language = $groups[1] !== false && mb_strlen($groups[1]) > 0 ? $groups[1] : null;
+		$language = MDUtils::makeString($groups[1], null);
 		$codeLines = [];
 		while ($state->hasLines(1, $p)) {
 			$line = $state->lines[$p++];
@@ -2255,7 +2264,8 @@ class MDDefinitionListReader extends MDReader {
 		if ($termCount == 0 || $definitionCount == 0) return null;
 		$blocks = array_map(function($line) use ($state) {
 			if (mb_eregi('^:\\s+(.*?)$', $line, $groups)) {
-				return new MDDefinitionListDefinitionNode($state->inlineMarkdownToNodes($groups[1]));
+				$content = MDUtils::makeString($groups[1]);
+				return new MDDefinitionListDefinitionNode($state->inlineMarkdownToNodes($content));
 			} else {
 				return new MDDefinitionListTermNode($state->inlineMarkdownToNodes($line));
 			}
@@ -2315,8 +2325,8 @@ class MDFootnoteReader extends MDReader {
 	public function readBlock(MDState $state): ?MDBlockNode {
 		$p = $state->p;
 		if (!mb_eregi('^\\s*\\[\\^\\s*([^\\]]+)\\s*\\]:\\s+(.*)\\s*$', $state->lines[$p++], $groups)) return null;
-		$symbol = $groups[1];
-		$def = $groups[2];
+		$symbol = MDUtils::makeString($groups[1]);
+		$def = MDUtils::makeString($groups[2]);
 		while ($state->hasLines(1, $p)) {
 			$line = $state->lines[$p++];
 			if (mb_eregi('^\\s+', $line)) {
@@ -2414,8 +2424,8 @@ class MDAbbreviationReader extends MDReader {
 		$p = $state->p;
 		$line = $state->lines[$p++];
 		if (!mb_eregi('^\\s*\\*\\[([^\\]]+?)\\]:\\s+(.*?)\\s*$', $line, $groups)) return null;
-		$abbrev = $groups[1];
-		$def = $groups[2];
+		$abbrev = MDUtils::makeString($groups[1]);
+		$def = MDUtils::makeString($groups[2]);
 		$this->defineAbbreviation($state, $abbrev, $def);
 		$state->p = $p;
 		return new MDBlockNode(); // empty

phptest/BrokenSyntaxTests.php

@@ -0,0 +1,132 @@
+<?php
+declare(strict_types=1);
+
+use PHPUnit\Framework\TestCase;
+
+require_once __DIR__ . '/../php/markdown.php';
+
+final class BrokenSyntaxTests extends TestCase {
+	private ?Markdown $parser = null;
+
+	protected function setUp(): void {
+		parent::setUp();
+		$this->parser = Markdown::completeParser();
+	}
+
+	private function iterateCharacters(string $markdown) {
+		for ($i = 1; $i < mb_strlen($markdown); $i++) {
+			$portion = mb_substr($markdown, 0, $i);
+			try {
+				$this->parser->toHTML($portion);
+			} catch (Error $e) {
+				print("Broken portion is:\n\"{$portion}\"\n");
+				throw $e;
+			}
+		}
+	}
+
+	public function test_brokenSyntax() {
+		// Just try a bunch of broken syntax and make sure no exceptions are
+		// thrown or infinite loops triggered. Don't care about what's rendered
+		// as most of these are incomplete or invalid syntax.
+		$this->iterateCharacters("# ");
+		$this->iterateCharacters("## ");
+		$this->iterateCharacters("Underlined\n-");
+		$this->iterateCharacters("* ");
+		$this->iterateCharacters(" * A");
+		$this->iterateCharacters("- A");
+		$this->iterateCharacters(" - A");
+		$this->iterateCharacters("+ A");
+		$this->iterateCharacters(" + A");
+		$this->iterateCharacters("1. A");
+		$this->iterateCharacters(" 1. A");
+		$this->iterateCharacters("1. A\n 1. B\n 2. C\n2. D");
+		$this->iterateCharacters("> A");
+		$this->iterateCharacters("> \nA");
+		$this->iterateCharacters("> \n>");
+		$this->iterateCharacters(">>");
+		$this->iterateCharacters("> > A");
+		$this->iterateCharacters("> \n> A");
+		$this->iterateCharacters("> > \n> A");
+		$this->iterateCharacters("```\ncode\n```");
+		$this->iterateCharacters("```java\ncode\n```");
+		$this->iterateCharacters("    code\n    code\n\n");
+		$this->iterateCharacters("------");
+		$this->iterateCharacters("  -  -    - --*");
+		$this->iterateCharacters("Header|Header\n--|--\nCell|Cell");
+		$this->iterateCharacters("Header\n--|--\nCell");
+		$this->iterateCharacters("|Header|\n|-|\n|Cell|");
+		$this->iterateCharacters("|Header|\n|--|--|--|\n|Cell|");
+		$this->iterateCharacters("|Header|Header|\n|--|\n|Cell|Cell|");
+		$this->iterateCharacters("|Header|\n|:-|\n|Cell|");
+		$this->iterateCharacters("|Header|\n|:-:|\n|Cell|");
+		$this->iterateCharacters("|Header|\n|-:|\n|Cell|");
+		$this->iterateCharacters("term\n: definition\nterm\n: definition");
+		$this->iterateCharacters("term\n:definition\nterm\n:definition");
+		$this->iterateCharacters("HTML\n\n*[HTML]: Hypertext");
+		$this->iterateCharacters("HTML\n*[HTML]: Hypertext");
+		$this->iterateCharacters("*[HTML]: Hypertext\nHTML");
+		$this->iterateCharacters("---{.foo}\n");
+		$this->iterateCharacters("---{#foo}\n");
+		$this->iterateCharacters("---{.foo #foo lang=en}\n");
+		$this->iterateCharacters("lorem *ipsum* dolor *sit* amet");
+		$this->iterateCharacters("*lorem* ipsum *dolor* sit *amet*");
+		$this->iterateCharacters("*lorem *ipsum *dolor*** sit");
+		$this->iterateCharacters("***lorem* ipsum* dolor* sit");
+		$this->iterateCharacters("lorem _ipsum_ dolor _sit_ amet");
+		$this->iterateCharacters("_lorem_ ipsum _dolor_ sit _amet_");
+		$this->iterateCharacters("_lorem _ipsum _dolor___ sit");
+		$this->iterateCharacters("___lorem_ ipsum_ dolor_ sit");
+		$this->iterateCharacters("lorem **ipsum** dolor **sit** amet");
+		$this->iterateCharacters("**lorem** ipsum **dolor** sit **amet**");
+		$this->iterateCharacters("**lorem **ipsum **dolor****** sit");
+		$this->iterateCharacters("******lorem** ipsum** dolor** sit");
+		$this->iterateCharacters("lorem __ipsum__ dolor __sit__ amet");
+		$this->iterateCharacters("__lorem__ ipsum __dolor__ sit __amet__");
+		$this->iterateCharacters("lorem __ipsum__ dolor __sit__ amet");
+		$this->iterateCharacters("__lorem __ipsum __dolor______ sit");
+		$this->iterateCharacters("______lorem__ ipsum__ dolor__ sit");
+		$this->iterateCharacters("*_**`~~^==!__~*^!``**-+!**`==!~_^**``_");
+		$this->iterateCharacters("[link[(index.html(");
+		$this->iterateCharacters("[link[](index.html()");
+		$this->iterateCharacters("[link]](index.html))");
+		$this->iterateCharacters("[[link]((index.html)");
+		$this->iterateCharacters("]link])index.html)");
+		$this->iterateCharacters("[ link ] ( index.html )");
+		$this->iterateCharacters("(user@example.com)");
+		$this->iterateCharacters("(https://user@example.com)");
+		$this->iterateCharacters("(index.html \"title\")");
+		$this->iterateCharacters("(index.html \"title)");
+		$this->iterateCharacters("(index.html title\")");
+		$this->iterateCharacters("(index.html title)");
+		$this->iterateCharacters("![alt][image.jpg]");
+		$this->iterateCharacters("![alt[](image.jpg()");
+		$this->iterateCharacters("! [alt](image.jpg)");
+		$this->iterateCharacters("!(image.jpg)");
+		$this->iterateCharacters("[][]]][][[][][][[][[]][][][][][[[]]]][]][][");
+		$this->iterateCharacters("())())()()()(())(()((())))(()((()");
+		$this->iterateCharacters("<https://example.com>");
+		$this->iterateCharacters("<<https://example.com>");
+		$this->iterateCharacters("<https://example.com>>");
+		$this->iterateCharacters("[link][ref]\n\n[ref]: page.html");
+		$this->iterateCharacters("[link][ref]\n[ref]: page.html");
+		$this->iterateCharacters("[link][ref]\n\n[ref]:page.html");
+		$this->iterateCharacters("[ref]: page.html\n\n[link][ref]");
+		$this->iterateCharacters("[link][ref]\n\n[ref]: page.html \"title\"");
+		$this->iterateCharacters("[link][ref]\n\n[ref]: page.html title");
+		$this->iterateCharacters("Lorem[^1] ipsum[^abc] dolor[^-1]\n\n[^abc]: def\n[^1]: def\n[^-1]: def\n[^abc]: def");
+		$this->iterateCharacters("Lorem[^1]\n\n[^2]: def");
+		$this->iterateCharacters("[^1]] [[^2] [^3]] [[^4]] ]^5[");
+		$this->iterateCharacters("[^1]: Def\n\nLorem[^1]");
+		$this->iterateCharacters("[^1]: Def\n\nLorem[^2]");
+		$this->iterateCharacters("Lorem <foo> ipsum");
+		$this->iterateCharacters("Lorem <span> ipsum");
+		$this->iterateCharacters("Lorem <span class=\"foo\" id=\"foo\" lang=\"en\"> ipsum");
+		$this->iterateCharacters("Lorem </span> ipsum");
+		$this->iterateCharacters("Lorem </span foo=\"bar\">");
+		$this->iterateCharacters("Lorem <span foo='bar' baz=ipsum dolor> sit");
+		$this->iterateCharacters("\\*\\\\*\\\\\\*\\\\\\\\*\\*");
+		$this->assertTrue(true); // to suppress risky test warning
+	}
+}
+?>

runphptests.sh

@@ -4,6 +4,7 @@ php lib/phpunit.phar --display-warnings --display-notices \
 	phptest/TokenTests.php \
 	phptest/InlineTests.php \
 	phptest/BlockTests.php \
+	phptest/BrokenSyntaxTests.php \
 	phptest/spreadsheet/CellAddressRangeTests.php \
 	phptest/spreadsheet/CellValueTests.php \
 	phptest/spreadsheet/ExpressionSetTests.php \

testjs.html

@@ -442,6 +442,7 @@
 		<script src="jstest/UtilsTests.js"></script>
 		<script src="jstest/InlineTests.js"></script>
 		<script src="jstest/BlockTests.js"></script>
+		<script src="jstest/BrokenSyntaxTests.js"></script>
 		<script src="jstest/spreadsheet/CellValueTests.js"></script>
 		<script src="jstest/spreadsheet/CellAddressRangeTests.js"></script>
 		<script src="jstest/spreadsheet/ExpressionSetTests.js"></script>
@@ -453,6 +454,7 @@
 					UtilsTests,
 					InlineTests,
 					BlockTests,
+					BrokenSyntaxTests,
 					CellValueTests,
 					CellAddressRangeTests,
 					ExpressionSetTests,