From 621affaa063550f086ea827811607bcf864f89e3 Mon Sep 17 00:00:00 2001
From: Blaise
Date: Sat, 29 Oct 2022 14:25:04 -0500
Subject: Only highlight text nodes

---
 proselight.js | 130 +++++++++++++++++++++++++++++++---------------------------
 1 file changed, 69 insertions(+), 61 deletions(-)

diff --git a/proselight.js b/proselight.js
index c79b8ce..a24ff32 100644
--- a/proselight.js
+++ b/proselight.js
@@ -26,6 +26,7 @@
 
         i,
         microlighted,
+        cn, //child nodes
         el; // current microlighted element to run through
 
 
@@ -35,70 +36,77 @@
         microlighted = _document.querySelectorAll('p');
 
         for (i = 0; el = microlighted[i++];) {
-            var text  = el.textContent,
-                pos   = 0,         // current position
-                next1 = text[0],   // next character
-                chr   = 1,         // current character
-                prev1,             // previous character
-                prev2,             // the one before the previous
-                token =            // current token content
-                el.innerHTML = '', // (and cleaning the node)
-                
-                // current token type:
-                tokenType = 0,
-
-                // flag determining if token is multi-character
-                multichar,
-                node;
-
-            // running through characters and highlighting
-            while (prev2 = prev1, prev1 = chr) {
-                chr = next1;
-                next1=text[++pos];
-                multichar = token.length > 1;
-
-                // checking if current token should be finalized
-                if (!chr  || // end of content
-                    [ // finalize conditions for other token types
-                        // 0: unformatted
-                        /[":;,\\.?!\])\/{}[(|]/[test](chr),
-                        // 1: parentesis or braces
-                        1,                // consist of a single character
-                        // 2: terminators
-                        1,                // consist of a single character
-                        // 3: separators
-                        1,                // consist of a single character
-                        // 4: quotes
-                        prev1 == '"' && multichar,
-                    ][tokenType]
-                ) {
-                    // appending the token to the result
-                    if (token) {
-                        // map token type into class
-                        el[appendChild](
-                            node = _document.createElement('span')
-                        ).setAttribute('class', 'ph'+tokenType);
-
-                        node[appendChild](_document.createTextNode(token));
+            cn = Array.from(el.childNodes);
+            for (j = 0; j < cn.length; j++) {
+                if (cn[j].nodeType != Node.TEXT_NODE) {
+                    el.appendChild(cn[j]);
+                    continue;
+                }
+                el.removeChild(cn[j]);
+                var text  = cn[j].textContent,
+                    pos   = 0,         // current position
+                    next1 = text[0],   // next character
+                    chr   = 1,         // current character
+                    prev1,             // previous character
+                    prev2,             // the one before the previous
+                    token = '',        // current token content
+                    
+                    // current token type:
+                    tokenType = 0,
+
+                    // flag determining if token is multi-character
+                    multichar,
+                    node;
+
+                // running through characters and highlighting
+                while (prev2 = prev1, prev1 = chr) {
+                    chr = next1;
+                    next1=text[++pos];
+                    multichar = token.length > 1;
+
+                    // checking if current token should be finalized
+                    if (!chr  || // end of content
+                        [ // finalize conditions for other token types
+                            // 0: unformatted
+                            /[":;,\\.?!\])\/{}[(|]/[test](chr),
+                            // 1: parentesis or braces
+                            1,                // consist of a single character
+                            // 2: terminators
+                            1,                // consist of a single character
+                            // 3: separators
+                            1,                // consist of a single character
+                            // 4: quotes
+                            prev1 == '"' && multichar,
+                        ][tokenType]
+                    ) {
+                        // appending the token to the result
+                        if (token) {
+                            // map token type into class
+                            el[appendChild](
+                                node = _document.createElement('span')
+                            ).setAttribute('class', 'ph'+tokenType);
+
+                            node[appendChild](_document.createTextNode(token));
+                        }
+
+                        // initializing a new token
+                        token = '';
+
+                        // determining the new token type (going up the
+                        // list until matching a token type start
+                        // condition)
+                        tokenType = 5;
+                        while (![
+                            1,                         //  0: unformatted
+                            /[\])\/{}[(|]/[test](chr), //  1: parenthesis or braces
+                            /[\\.?!]/[test](chr),      //  2: terminators
+                            /[:;,]/[test](chr),        //  3: separators
+                            chr == '"',                //  4: quotes
+                        ][--tokenType]);
                     }
 
-                    // initializing a new token
-                    token = '';
-
-                    // determining the new token type (going up the
-                    // list until matching a token type start
-                    // condition)
-                    tokenType = 5;
-                    while (![
-                        1,                         //  0: unformatted
-                        /[\])\/{}[(|]/[test](chr), //  1: parenthesis or braces
-                        /[\\.?!]/[test](chr),      //  2: terminators
-                        /[:;,]/[test](chr),        //  3: separators
-                        chr == '"',                //  4: quotes
-                    ][--tokenType]);
+                    token += chr;
                 }
-
-                token += chr;
             }
         }
     }
-- 
cgit v1.2.3