@peaceroad/markdown-it-cjk-breaks-mod 0.1.4 → 0.1.6

package/README.md

@@ -4,9 +4,11 @@
 
 ### Punctuation spacing options
 
-Fine-tune the trigger list with `spaceAfterPunctuationTargets`. Provide either a single string or an array and every exact match becomes eligible for automatic spacing; this option replaces the defaults. Defaults remain `['！', '？', '⁉', '！？', '？！', '!?', '?!', '.', ':']`. To disable punctuation spacing while still setting `spaceAfterPunctuation`, pass `spaceAfterPunctuationTargets: []` (or `null`/`false`). Use `spaceAfterPunctuationTargetsAdd` to append triggers and `spaceAfterPunctuationTargetsRemove` to drop items from the resolved list.
+`spaceAfterPunctuation` inserts spacing only when this plugin suppresses a line break after punctuation. The second-pass matcher also covers inline markup starts (inline code, links/autolinks, images, inline HTML) as long as the raw source contains a visible newline boundary; if a `softbreak` remains, no spacing is injected.
 
-Use `spaceAfterPunctuation` to inject a space every time this plugin suppresses a line break after punctuation. Accepts `'half'` for ASCII space, `'full'` for an ideographic space, or any custom string via a literal value.
+`spaceAfterPunctuationTargets` lets you replace the default trigger list with a custom string or array. Defaults are `['！', '？', '⁉', '！？', '？！', '!?', '?!', '.', ':']`. To disable punctuation spacing while still setting `spaceAfterPunctuation`, pass `spaceAfterPunctuationTargets: []` (or `null`/`false`). Use `spaceAfterPunctuationTargetsAdd` to append triggers and `spaceAfterPunctuationTargetsRemove` to drop items from the resolved list.
+
+`spaceAfterPunctuation` accepts `'half'` for ASCII space, `'full'` for an ideographic space, or any custom string via a literal value. Raw matching is strict, so escapes or entities (e.g. `&`) right before the newline can prevent a match and skip spacing (safe-fail behavior).
 
 ```js
 import MarkdownIt from 'markdown-it';
@@ -28,7 +30,13 @@ const mdHalf = MarkdownIt({ html: true }).use(cjkBreaks, {
 mdHalf.render('こんにちは！\nWorld');
 // <p>こんにちは！ World</p>
 
-// Custom punctuation triggers
+// Inline code and links are supported when a raw newline is present
+mdHalf.render('漢！\n`code`');
+// <p>漢！ <code>code</code></p>
+mdHalf.render('漢！\n[link](url)');
+// <p>漢！ <a href="url">link</a></p>
+
+// Custom punctuation triggers (replaces defaults)
 const mdCustom = MarkdownIt({ html: true }).use(cjkBreaks, {
   spaceAfterPunctuation: 'half',
   spaceAfterPunctuationTargets: ['??']

package/index.js

@@ -12,7 +12,6 @@ const HANGUL_RE = /[\u1100-\u11FF\u302E\u302F\u3131-\u318E\u3200-\u321E\u3260-\u
 /* eslint-enable max-len */
 const WHITESPACE_RE = /\s/;
 const WHITESPACE_LEAD_RE = /^\s/;
-const WHITESPACE_TRAIL_RE = /\s$/;
 
 
 function is_surrogate(c1, c2) {
@@ -28,15 +27,18 @@ function is_hangul(c) {
 function create_punctuation_config(targets) {
   var sequences = new Set();
   var maxLength = 0;
+  var endCharMap = Object.create(null);
 
   for (var i = 0; i < targets.length; i++) {
     var value = targets[i];
     if (typeof value !== 'string' || value.length === 0) continue;
     sequences.add(value);
     if (value.length > maxLength) maxLength = value.length;
+    var endChar = get_last_char(value);
+    if (endChar) endCharMap[endChar] = true;
   }
 
-  return { sequences: sequences, maxLength: maxLength };
+  return { sequences: sequences, maxLength: maxLength, endCharMap: endCharMap };
 }
 
 
@@ -116,6 +118,11 @@ function matches_punctuation_sequence(trailing, punctuationConfig) {
   if (!trailing || !punctuationConfig || punctuationConfig.maxLength === 0) return false;
 
   var sequences = punctuationConfig.sequences;
+  var endCharMap = punctuationConfig.endCharMap;
+  if (endCharMap) {
+    var endChar = get_last_char(trailing);
+    if (!endChar || !endCharMap[endChar]) return false;
+  }
   var maxLength = Math.min(trailing.length, punctuationConfig.maxLength);
 
   for (var len = maxLength; len > 0; len--) {
@@ -126,6 +133,16 @@ function matches_punctuation_sequence(trailing, punctuationConfig) {
 }
 
 
+function get_last_char(text) {
+  if (!text) return '';
+  var len = text.length;
+  if (len === 1) return text;
+  var c1 = text.charCodeAt(len - 2);
+  var c2 = text.charCodeAt(len - 1);
+  return is_surrogate(c1, c2) ? text.slice(-2) : text.slice(-1);
+}
+
+
 function is_printable_ascii(ch) {
   if (!ch) return false;
   var code = ch.charCodeAt(0);
@@ -176,7 +193,7 @@ function build_next_text_info(tokens, trackSkippedEmpty) {
 }
 
 
-function process_inlines(tokens, state, ctx, inlineToken) {
+function process_inlines(tokens, ctx, inlineToken) {
   var i, last, next, c1, c2, remove_break;
   var either = ctx.either;
   var normalizeSoftBreaks = ctx.normalizeSoftBreaks;
@@ -185,6 +202,7 @@ function process_inlines(tokens, state, ctx, inlineToken) {
   var maxPunctuationLength = ctx.maxPunctuationLength;
   var considerInlineBoundaries = ctx.considerInlineBoundaries;
   var needsPunctuation = punctuationSpace && punctuationConfig && maxPunctuationLength > 0;
+  var punctuationEndCharMap = punctuationConfig ? punctuationConfig.endCharMap : null;
 
   if (!tokens || tokens.length === 0) return;
   if (normalizeSoftBreaks) normalize_text_tokens(tokens);
@@ -283,14 +301,16 @@ function process_inlines(tokens, state, ctx, inlineToken) {
 
       if (remove_break) {
         var insertPunctuationSpace = false;
-        if (needsPunctuation && last && next && next !== '\u200b') {
-          var trailing = hasLastText ? lastTextContent.slice(-maxPunctuationLength) : '';
-          if (matches_punctuation_sequence(trailing, punctuationConfig)) {
-            if (!nextWidthComputed) {
-              nextWidthClass = get_cached_width_class(next);
+        if (needsPunctuation && hasLastText && last && next && next !== '\u200b') {
+          if (!punctuationEndCharMap || punctuationEndCharMap[last]) {
+            var trailing = lastTextContent.slice(-maxPunctuationLength);
+            if (matches_punctuation_sequence(trailing, punctuationConfig)) {
+              if (!nextWidthComputed) {
+                nextWidthClass = get_cached_width_class(next);
+              }
+              var nextIsFullwidthOrWide = nextWidthClass === 'F' || nextWidthClass === 'W';
+              if (nextIsFullwidthOrWide || is_printable_ascii(next)) insertPunctuationSpace = true;
             }
-            var nextIsFullwidthOrWide = nextWidthClass === 'F' || nextWidthClass === 'W';
-            if (nextIsFullwidthOrWide || is_printable_ascii(next)) insertPunctuationSpace = true;
           }
         }
         token.type    = 'text';
@@ -365,7 +385,6 @@ function split_text_token(token) {
     parts.push(clone_text_token(TokenConstructor, token, content.slice(start)));
   }
 
-  if (parts.length === 0) parts.push(token);
   return parts;
 }
 
@@ -407,6 +426,7 @@ function apply_missing_punctuation_spacing(tokens, inlineToken, punctuationSpace
   if (!tokens || tokens.length === 0) return;
   var maxPunctuationLength = punctuationConfig.maxLength;
   if (maxPunctuationLength <= 0) return;
+  var endCharMap = punctuationConfig.endCharMap;
 
   var rawSearchState = { pos: 0 };
 
@@ -414,13 +434,17 @@ function apply_missing_punctuation_spacing(tokens, inlineToken, punctuationSpace
     var current = tokens[idx];
     if (!current || current.type !== 'text' || !current.content) continue;
 
+    if (endCharMap) {
+      var endChar = get_last_char(current.content);
+      if (!endChar || !endCharMap[endChar]) continue;
+    }
     var trailing = current.content.slice(-maxPunctuationLength);
     if (!matches_punctuation_sequence(trailing, punctuationConfig)) continue;
-    if (WHITESPACE_TRAIL_RE.test(current.content)) continue;
 
     var nextInfo = find_next_visible_token(tokens, idx + 1);
     if (!nextInfo) continue;
     if (nextInfo.token.type === 'text' && WHITESPACE_LEAD_RE.test(nextInfo.token.content || '')) continue;
+    if (has_active_break(tokens, idx, nextInfo.index)) continue;
 
     if (!raw_boundary_includes_newline(inlineToken.content, tokens, idx, nextInfo.index, nextInfo.fragment, rawSearchState)) {
       continue;
@@ -435,20 +459,35 @@ function apply_missing_punctuation_spacing(tokens, inlineToken, punctuationSpace
   }
 }
 
+function has_active_break(tokens, fromIdx, nextIdx) {
+  for (var idx = fromIdx + 1; idx < nextIdx; idx++) {
+    var token = tokens[idx];
+    if (!token) continue;
+    if (token.type === 'softbreak') return true;
+    if (token.type === 'text' && token.content === '\n') return true;
+  }
+  return false;
+}
+
 
 function raw_boundary_includes_newline(source, tokens, fromIdx, nextIdx, afterFragment, state) {
   if (!source || !afterFragment) return false;
+  var fragments = Array.isArray(afterFragment) ? afterFragment : [afterFragment];
   var beforeFragment = tokens[fromIdx].content || '';
   var betweenFragment = '';
   for (var k = fromIdx + 1; k < nextIdx; k++) {
     if (tokens[k].markup) betweenFragment += tokens[k].markup;
   }
-  var candidate = beforeFragment + betweenFragment + '\n' + afterFragment;
-  if (!candidate) return false;
-  var startPos = source.indexOf(candidate, state.pos);
-  if (startPos === -1) return false;
-  state.pos = startPos + candidate.length - afterFragment.length;
-  return true;
+  for (var i = 0; i < fragments.length; i++) {
+    var fragment = fragments[i];
+    if (!fragment) continue;
+    var candidate = beforeFragment + betweenFragment + '\n' + fragment;
+    var startPos = source.indexOf(candidate, state.pos);
+    if (startPos === -1) continue;
+    state.pos = startPos + candidate.length - fragment.length;
+    return true;
+  }
+  return false;
 }
 
 
@@ -467,10 +506,18 @@ function find_next_visible_token(tokens, startIdx) {
 function derive_after_fragment(token) {
   if (!token) return '';
   if (token.type === 'text' || token.type === 'html_inline' || token.type === 'code_inline') {
-    return token.content || '';
+    if (token.type !== 'code_inline') return token.content || '';
+    var fragments = [];
+    var markup = token.markup || '';
+    var content = token.content || '';
+    if (markup && content) fragments.push(markup + content);
+    if (markup) fragments.push(markup);
+    if (content) fragments.push(content);
+    return fragments;
   }
   if (token.type === 'image') return '![';
-  if (token.type === 'link_open') return '[';
+  if (token.type === 'link_open') return token.markup || '[';
+  if (token.nesting === 1 && token.markup) return token.markup;
   if (token.type === 'inline') return token.content || '';
   return '';
 }
@@ -594,7 +641,7 @@ export default function cjk_breaks_plugin(md, opts) {
   function cjk_breaks(state) {
     for (var blkIdx = state.tokens.length - 1; blkIdx >= 0; blkIdx--) {
       if (state.tokens[blkIdx].type !== 'inline') continue;
-      process_inlines(state.tokens[blkIdx].children, state, ctx, state.tokens[blkIdx]);
+      process_inlines(state.tokens[blkIdx].children, ctx, state.tokens[blkIdx]);
     }
   }
   if (!md || !md.core || !md.core.ruler) return;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@peaceroad/markdown-it-cjk-breaks-mod",
-  "version": "0.1.4",
+  "version": "0.1.6",
   "type": "module",
   "description": "Suppress linebreaks between east asian (Especially Japanese) characters",
   "repository": {
@@ -22,7 +22,7 @@
     "eastasianwidth": "^0.3.0"
   },
   "devDependencies": {
-    "@peaceroad/markdown-it-strong-ja": "^0.5.5",
+    "@peaceroad/markdown-it-strong-ja": "^0.7.2",
     "markdown-it": "^14.1.0"
   }
 }