[C#] Strip Html Comments

2008-11-11T01:14:04+00:00
Procedural style code to strip HTML comments. No attempt was made to make this code more OO/patterned base so it could be readable/maintainable. ;)
public static class HtmlHelper {
    public static string StripHtmlComments(string html) {
        if (html == null) {
            throw new ArgumentNullException("html");
        }

        if (html.IndexOf("<!", StringComparison.Ordinal) < 0) {
            return html;
        }

        var cleanedHtml = new char[html.Length];
        bool inHtmlComment = false;
        bool inHtmlTag = false;
        int cleanCount = 0;

        for (int i = 0; i < html.Length; i++) {
            char current = html[i];

            if (!inHtmlComment && !inHtmlTag) {
                if (current == '<') {
                    if (i + 1 < html.Length) {
                        char nextChar = html[i + 1];
                        if (nextChar == '!') {
                                inHtmlComment = true;
                                continue;
                            }
                            else {
                                if (IsEnglishLetter(nextChar)) {
                                    inHtmlTag = true;
                                }
                            }
                        }
                    }
                }
                else if(inHtmlComment) {
                    if (current == '>') {
                        if (inHtmlComment) {
                            inHtmlComment = false;
                            continue;
                        }
                    }
                    continue;
                }
                else if (inHtmlTag) {
                    if (current == '>') {
                        inHtmlTag = false;
                    }
                }

                cleanedHtml[cleanCount++] = current;
            }

            return new String(cleanedHtml, 0, cleanCount);
        }

        private static bool IsEnglishLetter(char nextChar) {
            return ('a' <= nextChar && nextChar <= 'z') || ('A' <= nextChar && nextChar <= 'Z');
        }

}

//Unit Tests
[TestMethod]
public void NullStringReturnsThrowsArgumentNullException() {
    try {
        HtmlHelper.StripHtmlComments(null);
        Assert.Fail();
    }
    catch (ArgumentNullException) { 
    }
}

[TestMethod]
public void EmptyStringReturnsEmpty() {
    Assert.AreEqual(string.Empty, HtmlHelper.StripHtmlComments(string.Empty));
}

[TestMethod]
public void StringWithoutCommentReturnsSameString() {
    string s = "This has <strong>No Comments</strong>!";
    Assert.AreEqual(s, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void StringWithOnlyCommentReturnsEmptyString() {
    string s = "<!-- this go bye bye>";
    Assert.AreEqual(string.Empty, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithNonDashDashComment_ReturnsEmptyString() {
    string s = "<! this go bye bye>";
    Assert.AreEqual(string.Empty, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void StringWithTwoConsecutiveCommentsReturnsEmptyString() {
    string s = "<!-- this go bye bye><!-- another comment>";
    Assert.AreEqual(string.Empty, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void CommentWithStringBeforeReturnsString() {
    string s = "Hello<!-- this go bye bye -->";
    Assert.AreEqual("Hello", HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void CommentWithStringAfterReturnsString() {
    string s = "<!-- this go bye bye -->World";
    Assert.AreEqual("World", HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithAngleBracketsButNotHtml_NotSripped() {
    string s = "<$)*(@&$(@*>";
    Assert.AreEqual(s, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithCommentInterleavedWithText_RendersText() {
    string s = "Hello <!-- this go bye bye --> World <!--> This is fun";
    Assert.AreEqual("Hello  World  This is fun", HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithHtmlTags_DoesNotStripHtml() {
    string s = "<strong>Hello</strong><!this go bye bye>";
    Assert.AreEqual("<strong>Hello</strong>", HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithCommentInAttribute_DoesNotStripAttributeValue() {
    string s = "<img alt=\"<!-- This should remain -->\" />";
    Assert.AreEqual(s, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithCommentInSingleQuotedAttribute_DoesNotStripAttributeValue() {
    string s = "<img alt=\'<!-- This should remain -->\' />";
    Assert.AreEqual(s, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithCommentInNonQuotedAttribute_DoesNotStripAttributeValue() {
    string s = "<p title=<!--Thisshouldremain-->Test</p>";
    Assert.AreEqual(s, HtmlHelper.StripHtmlComments(s));
}

[TestMethod]
public void Html_WithCommentBetweenNonTagButLooksLikeTag_DoesStripComment() {
    string s = @"<ç123 title=""<!bc def>"">";
    Assert.AreEqual(@"<ç123 title="""">", HtmlHelper.StripHtmlComments(s));
}
GateKiller friends

[C#] Strip Html Comments
