fix(single): Surround properly handles long chains of * or _

1 files changed, 16 insertions, 7 deletions

diff --git a/single.awk b/single.awk
index bf80054..2ac29cf 100644
--- a/single.awk
+++ b/single.awk
@@ -299,20 +299,29 @@ blankline {
 }
 
 # Emphasis
-function surround(brace, opening, closing) {
-	while ($0 ~ brace ".*" brace) {
-		sub(brace, opening)
-		sub(brace, closing)
+function surround(brace, opening, closing, blen, single_surround) {
+	if (blen == "") { blen = length(brace) }
+
+	while (match($0, brace "[^" brace "]+" brace)) {
+		elem = opening substr($0, RSTART + blen, RLENGTH - 2 * blen) closing
+		$0 = substr($0, 1, RSTART - 1) elem substr($0, RSTART + RLENGTH)
+
+		if (single_surround != "") { return; }
 	}
 }
 {
+	## Make sure <em> is on the inside for long chains
+	## E.g. *****a***** will be the same as ****_a_****
+	surround("___"      , "<strong><em>", "</em></strong>", 3, "oneshot")
+	surround("\\*\\*\\*", "<strong><em>", "</em></strong>", 3, "oneshot")
+
 	surround("__"    , "<strong>", "</strong>")
 	surround("_"     , "<em>"    , "</em>")
-	surround("\\*\\*", "<strong>", "</strong>")
-	surround("\\*"   , "<em>"    , "</em>")
+	surround("\\*\\*", "<strong>", "</strong>", 2)
+	surround("\\*"   , "<em>"    , "</em>", 1)
 	surround("~~"    , "<s>"     , "</s>")
 	surround("~"     , "<sub>"   , "</sub>")
-	surround("\\^"   , "<sup>"   , "</sup>")
+	surround("\\^"   , "<sup>"   , "</sup>", 1)
 	surround("``"    , "<code>"  , "</code>")
 	surround("`"     , "<code>"  , "</code>")
 }