[Repo Assist] Improve CommonMark ATX heading compliance

RELEASE_NOTES.md

@@ -14,6 +14,7 @@
 * Strip `#if SYMBOL` / `#endif // SYMBOL` marker lines from `LiterateCode` source before syntax-highlighting so they do not appear in formatted output. [#693](https://github.com/fsprojects/FSharp.Formatting/issues/693)
 * Fix incorrect column ranges for inline spans (links, images, inline code) in the Markdown parser — spans and subsequent literals now report correct `StartColumn`/`EndColumn` values. [#744](https://github.com/fsprojects/FSharp.Formatting/issues/744)
 * Normalize `--projects` paths to absolute paths before passing to the project cracker, fixing failures when relative paths are supplied. [#793](https://github.com/fsprojects/FSharp.Formatting/issues/793)
+* Improve CommonMark compliance for ATX headings: reject `#` not followed by a space (e.g. `#NoSpace` is now a paragraph), reject more than 6 `#` characters as a heading, support 0–3 leading spaces before the opening `#` sequence, and fix empty content when the entire header body is a closing `###` sequence. [#191](https://github.com/fsprojects/FSharp.Formatting/issues/191)
 
 ### Changed
 * Update FCS to 43.10.100. [#935](https://github.com/fsprojects/FSharp.Formatting/pull/966)

src/FSharp.Formatting.Markdown/MarkdownParser.fs

@@ -572,21 +572,49 @@
     | ((StringPosition.TrimBoth header) as line1) :: ((StringPosition.TrimEnd(StringPosition.EqualsRepeated("-",
                                                                                                             MarkdownRange.zero))) as line2) :: rest ->
         Some(2, header, [ line1; line2 ], rest)
-    | (StringPosition.StartsWithRepeated "#" (n, StringPosition.TrimBoth(header, ln)) as line1) :: rest ->
-        let header =
-            // Drop "##" at the end, but only when it is preceded by some whitespace
-            // (For example "## Hello F#" should be "Hello F#")
-            if header.EndsWith '#' then
-                let noHash = header.TrimEnd [| '#' |]
-
-                if noHash.Length > 0 && Char.IsWhiteSpace(noHash.Chars(noHash.Length - 1)) then
-                    noHash
+    | ((line1text, ln1) as line1) :: rest ->
+        // ATX heading (CommonMark): optional 0–3 leading spaces, then 1–6 '#' characters,
+        // then a space or end of line (a tab or other char after '#' is not valid).
+        let mutable i = 0
+
+        while i < 3 && i < line1text.Length && line1text.[i] = ' ' do
+            i <- i + 1
+
+        let hstart = i
+
+        while i < line1text.Length && line1text.[i] = '#' do
+            i <- i + 1
+
+        let n = i - hstart
+
+        if n < 1 || n > 6 || (i < line1text.Length && line1text.[i] <> ' ') then
+            None
+        else
+            let contentStart = if i < line1text.Length then i + 1 else i
+            let content = (line1text.Substring(contentStart)).TrimEnd()
+            // Remove optional closing sequence of '#' preceded by space (or empty '#'-only content).
+            // For example "## Hello F#" keeps the '#' because it is not preceded by a space.
+            let header =
+                if content.EndsWith('#') then
+                    let noHash = content.TrimEnd([| '#' |])
+
+                    if noHash = "" || (noHash.Length > 0 && noHash.[noHash.Length - 1] = ' ') then
+                        noHash.Trim()
+                    else
+                        content.Trim()
                 else
-                    header
-            else
-                header
+                    content.Trim()
+
+            let rawContent = line1text.Substring(contentStart)
+            let leadingContentSpaces = rawContent.Length - rawContent.TrimStart(' ').Length
+            let headerStart = ln1.StartColumn + contentStart + leadingContentSpaces
+
+            let headerLn =
+                { ln1 with
+                    StartColumn = headerStart
+                    EndColumn = headerStart + header.Length }
 
-        Some(n, (header, ln), [ line1 ], rest)
+            Some(n, (header, headerLn), [ line1 ], rest)
     | _rest -> None
 
 let (|YamlFrontmatter|_|) lines =

tests/Benchmarks/testfiles/mstest-0.1/line-endings-cr.html

@@ -1,4 +1,4 @@
-<h2>Header</h2>
+<p>##Header</p>
 
 <hr />
 

tests/Benchmarks/testfiles/mstest-0.1/line-endings-crlf.html

@@ -1,4 +1,4 @@
-<h2>Header</h2>
+<p>##Header</p>
 
 <hr />
 

tests/Benchmarks/testfiles/mstest-0.1/line-endings-lf.html

@@ -1,4 +1,4 @@
-<h2>Header</h2>
+<p>##Header</p>
 
 <hr />
 

tests/Benchmarks/testfiles/php-markdown/Tight blocks.html

@@ -2,20 +2,19 @@
 * ciao</p>
 
 <p>Paragraph and 1 space:
- * ciao</p>
+* ciao</p>
 
 <p>Paragraph and 3 spaces:
-  * ciao</p>
+* ciao</p>
 
 <p>Paragraph and 4 spaces:
-   * ciao</p>
-
-<p>Paragraph before header:</p>
+* ciao</p>
 
-<h1>Header</h1>
+<p>Paragraph before header:
+#Header</p>
 
 <p>Paragraph before blockquote:</p>
 
 <blockquote>
-  <p>Some quote.</p>
+<p>Some quote.</p>
 </blockquote>

tests/FSharp.Markdown.Tests/CommonMarkSpecTest.fs

@@ -21,7 +21,17 @@ open FSharp.Formatting.Markdown
 let properNewLines (text: string) =
     text.Replace("\r\n", "\n").Replace("\n", System.Environment.NewLine)
 
-let enabledSections = [ "Fenced code blocks"; "Indented code blocks"; "Paragraphs"; "Precedence"; "Tabs" ]
+let enabledSections =
+    [ "Fenced code blocks"
+      "Indented code blocks"
+      "Paragraphs"
+      "Precedence"
+      "Tabs"
+      "Blank lines"
+      "Inlines"
+      "Soft line breaks"
+      "Textual content"
+      "ATX headings" ]
 
 let getTests () =
     sample

tests/FSharp.Markdown.Tests/Markdown.fs

@@ -582,15 +582,15 @@ let ``Transform bulleted lists correctly`` () =
 let ``Transform header 1 correctly`` () =
     let doc = "#Header 1\nHeader 1\r\n========"
 
-    let expected = "<h1>Header 1</h1>\r\n<h1>Header 1</h1>\r\n" |> properNewLines
+    let expected = "<p>#Header 1</p>\r\n<h1>Header 1</h1>\r\n" |> properNewLines
 
     Markdown.ToHtml doc |> shouldEqual expected
 
 [<Test>]
 let ``Transform header 2 correctly`` () =
     let doc = "##Header 2\nHeader 2\r\n--------"
 
-    let expected = "<h2>Header 2</h2>\r\n<h2>Header 2</h2>\r\n" |> properNewLines
+    let expected = "<p>##Header 2</p>\r\n<h2>Header 2</h2>\r\n" |> properNewLines
 
     Markdown.ToHtml doc |> shouldEqual expected