Fix GFM tables to require a non-pipe in header row

Related-to: GH-20.

Author: wooorm

src/construct/gfm_table.rs

@@ -312,6 +312,8 @@ pub fn head_row_start(tokenizer: &mut Tokenizer) -> State {
         Some(b'|') => State::Retry(StateName::GfmTableHeadRowBreak),
         _ => {
             tokenizer.tokenize_state.seen = true;
+            // Count the first character, that isn’t a pipe, double.
+            tokenizer.tokenize_state.size_b += 1;
             State::Retry(StateName::GfmTableHeadRowBreak)
         }
     }
@@ -332,22 +334,34 @@ pub fn head_row_break(tokenizer: &mut Tokenizer) -> State {
         None => {
             tokenizer.tokenize_state.seen = false;
             tokenizer.tokenize_state.size = 0;
+            tokenizer.tokenize_state.size_b = 0;
             State::Nok
         }
         Some(b'\n') => {
-            // Feel free to interrupt:
-            tokenizer.interrupt = true;
-            tokenizer.exit(Name::GfmTableRow);
-            tokenizer.enter(Name::LineEnding);
-            tokenizer.consume();
-            tokenizer.exit(Name::LineEnding);
-            State::Next(StateName::GfmTableHeadDelimiterStart)
+            // If anything other than one pipe (ignoring whitespace) was used, it’s fine.
+            if tokenizer.tokenize_state.size_b > 1 {
+                tokenizer.tokenize_state.size_b = 0;
+                // Feel free to interrupt:
+                tokenizer.interrupt = true;
+                tokenizer.exit(Name::GfmTableRow);
+                tokenizer.enter(Name::LineEnding);
+                tokenizer.consume();
+                tokenizer.exit(Name::LineEnding);
+                State::Next(StateName::GfmTableHeadDelimiterStart)
+            } else {
+                tokenizer.tokenize_state.seen = false;
+                tokenizer.tokenize_state.size = 0;
+                tokenizer.tokenize_state.size_b = 0;
+                State::Nok
+            }
         }
         Some(b'\t' | b' ') => {
             tokenizer.attempt(State::Next(StateName::GfmTableHeadRowBreak), State::Nok);
             State::Retry(space_or_tab(tokenizer))
         }
         _ => {
+            tokenizer.tokenize_state.size_b += 1;
+
             // Whether a delimiter was seen.
             if tokenizer.tokenize_state.seen {
                 tokenizer.tokenize_state.seen = false;

tests/gfm_table.rs

@@ -129,6 +129,24 @@ fn gfm_table() -> Result<(), String> {
         "should support empty body cells"
     );
 
+    assert_eq!(
+        to_html_with_options(":\n|-|\n|a|\n\nb\n|-|\n|c|\n\n|\n|-|\n|d|\n\n|\n|-|\n|e|\n\n|:\n|-|\n|f|\n\n||\n|-|\n|g|\n\n| |\n|-|\n|h|\n", &Options::gfm())?,
+        "<table>\n<thead>\n<tr>\n<th>:</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>a</td>\n</tr>\n</tbody>\n</table>\n<table>\n<thead>\n<tr>\n<th>b</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>c</td>\n</tr>\n</tbody>\n</table>\n<p>|\n|-|\n|d|</p>\n<p>|\n|-|\n|e|</p>\n<table>\n<thead>\n<tr>\n<th>:</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>f</td>\n</tr>\n</tbody>\n</table>\n<table>\n<thead>\n<tr>\n<th></th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>g</td>\n</tr>\n</tbody>\n</table>\n<table>\n<thead>\n<tr>\n<th></th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>h</td>\n</tr>\n</tbody>\n</table>\n",
+        "should need any character other than a single pipe in the header row"
+    );
+
+    assert_eq!(
+        to_html_with_options("a\n|-\n\nb\n||\n\nc\n|-|\n\nd\n|:|\n\ne\n| |\n\nf\n| -|\n\ng\n|- |\n", &Options::gfm())?,
+        "<table>\n<thead>\n<tr>\n<th>a</th>\n</tr>\n</thead>\n</table>\n<p>b\n||</p>\n<table>\n<thead>\n<tr>\n<th>c</th>\n</tr>\n</thead>\n</table>\n<p>d\n|:|</p>\n<p>e\n| |</p>\n<table>\n<thead>\n<tr>\n<th>f</th>\n</tr>\n</thead>\n</table>\n<table>\n<thead>\n<tr>\n<th>g</th>\n</tr>\n</thead>\n</table>\n",
+        "should need a dash in the delimimter row"
+    );
+
+    assert_eq!(
+        to_html_with_options("|\n|", &Options::gfm())?,
+        "<p>|\n|</p>",
+        "should need something"
+    );
+
     assert_eq!(
         to_html_with_options("| a |\n| - |\n- b", &Options::gfm())?,
         "<table>\n<thead>\n<tr>\n<th>a</th>\n</tr>\n</thead>\n</table>\n<ul>\n<li>b</li>\n</ul>",