@peaceroad/markdown-it-figure-with-p-caption 0.16.0 → 0.17.0

package/README.md

@@ -15,10 +15,10 @@ Optionally, you can auto-number image and table caption paragraphs starting from
 - Pure image paragraphs (`![...](...)`) become `<figure class="f-img">` blocks as soon as a caption paragraph (previous or next) or an auto-detected caption exists.
 - Auto detection runs per image paragraph when `autoCaptionDetection` is `true` (default). The priority is:
     1. Caption paragraphs immediately before or after the image (standard syntax).
-    2. Image `alt` text that already matches p7d-markdown-it-p-captions label formats (`Figure. `, `Figure 1. `, `図　`,`図1　`, etc.).
+    2. Image `alt` text that `p7d-markdown-it-p-captions` recognizes as an image caption start (`Figure. `, `Figure 1. `, `図　`, `図1　`, etc.).
     3. Image `title` attribute that matches the same labels.
     4. Optional fallbacks (`autoAltCaption`, `autoTitleCaption`) that inject the label when the alt/title lacks one.
-        - `autoAltCaption`: `false` (default), `true`, or a string label. `true` inspects the first sentence of the caption text and picks `Figure` / `図` based on detected language; a string uses that label verbatim.
+        - `autoAltCaption`: `false` (default), `true`, or a string label. `true` uses locale-aware generated-label defaults from `p7d-markdown-it-p-captions`, so the label text and punctuation stay aligned with the upstream caption language data. A string uses that label verbatim. Empty alt text does not generate a fallback caption.
         - `autoTitleCaption`: same behavior but sourced from the image `title`. It stays off by default so other plugins can keep using the `title` attribute for metadata.
 - Set `autoCaptionDetection: false` to disable the auto-caption workflow entirely.
 - Multi-image paragraphs are still wrapped as one figure when `multipleImages: true` (default). Layout-specific classes help with styling:
@@ -50,9 +50,9 @@ Optionally, you can auto-number image and table caption paragraphs starting from
 
 ### Embedded content by iframe
 
-- Inline HTML `<iframe>` elements become `<figure class="f-video">` when they point to known video hosts (YouTube `youtube.com` / `youtube-nocookie.com`, Vimeo `player.vimeo.com`).
+- Inline HTML `<iframe>` elements become `<figure class="f-video">` when they point to known video hosts (YouTube `youtube-nocookie.com`, Vimeo `player.vimeo.com`).
 - `<div>` wrappers are treated as iframe-type embeds only when the same HTML block contains an `<iframe ...>` tag (for example common video wrapper markup).
-- Blockquote-based social embeds (Twitter/X `twitter-tweet`, Mastodon `mastodon-embed`, Bluesky `bluesky-embed`, Instagram `instagram-media`, Tumblr `text-post-media`) are treated like iframe-type embeds when their `class` matches those providers. By default they become `<figure class="f-img">` so the caption label behaves like an image label (Labels can also use quote labels). You can override that figure class with `figureClassThatWrapsIframeTypeBlockquote` or the global `allIframeTypeFigureClassName`.
+- Blockquote-based social embeds (Twitter/X `twitter-tweet`, Mastodon `mastodon-embed`, Bluesky `bluesky-embed`, Instagram `instagram-media`, Tumblr `text-post-media`) are treated like iframe-type embeds when their class list contains one of those provider classes. Extra classes on the same blockquote do not block detection. By default they become `<figure class="f-img">` so the caption label behaves like an image label (Labels can also use quote labels). You can override that figure class with `figureClassThatWrapsIframeTypeBlockquote` or the global `allIframeTypeFigureClassName`.
 - `p7d-markdown-it-p-captions` ships with a `Slide.` label. When you use it (for example with Speaker Deck or SlideShare iframes), the `<figure>` wrapper automatically switches to `f-slide` (or whatever you set via `figureClassThatWrapsSlides`) so slides can get their own layout. If `allIframeTypeFigureClassName` is also configured, that class takes precedence even for slides, so you get a uniform embed wrapper without touching the slide option.
 - All other iframes fall back to `<figure class="f-iframe">` unless you override the class via `allIframeTypeFigureClassName`.
 
@@ -60,7 +60,7 @@ Optionally, you can auto-number image and table caption paragraphs starting from
 
 - The label inside the figcaption (the `span` element used for the label) is generated by `p7d-markdown-it-p-captions`, not by this plugin. By default the class name is formed by combining `classPrefix` with the mark name, producing names such as `f-img-label`, `f-video-label`, `f-blockquote-label`, and `f-slide-label`. 
 - With `markdown-it-attrs`, attributes attached to image-only paragraphs (for example `![...](...) {.foo #bar}`) are forwarded to the generated `<figure>`.
-- `styleProcess` controls parsing of trailing `{...}` from inline text in this plugin's own image scanner, but attributes already attached to paragraph tokens by `markdown-it-attrs` are still forwarded.
+- `styleProcess` controls parsing of a trailing `{...}` block from the last text token of an image-only paragraph in this plugin's own scanner. It is a narrow fallback parser, not full `markdown-it-attrs` parity, and attributes already attached to paragraph tokens by `markdown-it-attrs` are still forwarded.
 - Attributes attached to caption paragraphs stay on the converted `<figcaption>` token after paragraph-to-figcaption conversion.
 
 ## Behavior Customization
@@ -71,6 +71,7 @@ Optionally, you can auto-number image and table caption paragraphs starting from
 - `figureClassThatWrapsIframeTypeBlockquote`: override the class used when blockquote-based embeds (Twitter, Mastodon, Bluesky) are wrapped.
 - `figureClassThatWrapsSlides`: override the class assigned when a caption paragraph uses the `Slide.` label.
 - `classPrefix` (default `f`) controls the CSS namespace for every figure (`f-img`, `f-table`, etc.) so you can align with existing styles.
+- Wrapper/class-prefix options are trimmed during setup; whitespace-only values fall back to the default class for that option.
 
 ### Wrapping without captions
 
@@ -84,13 +85,15 @@ Every option below is forwarded verbatim to `p7d-markdown-it-p-captions`, which
 - `strongFilename` / `dquoteFilename`: pull out filenames from captions using `**filename**` or `"filename"` syntax and wrap them in `<strong class="f-*-filename">`.
 - `jointSpaceUseHalfWidth`: replace full-width space between Japanese labels and caption body with half-width space.
 - `bLabel` / `strongLabel`: emphasize the label span itself.
-- `removeUnnumberedLabel`: drop the leading “Figure. Etext entirely when no label number is present. Use `removeUnnumberedLabelExceptMarks` to keep specific labels (e.g., `['blockquote']` keeps `Quote. `).
+- `removeUnnumberedLabel`: drop the leading label entirely when no label number is present. Use `removeUnnumberedLabelExceptMarks` to keep specific labels (e.g., `['blockquote']` keeps `Quote. `).
 - `removeMarkNameInCaptionClass`: replace `.f-img-label` / `.f-table-label` with the generic `.f-label`.
 - `wrapCaptionBody`: wrap the non-label caption text in a span element.
 - `hasNumClass`: add a class attribute to label span element if it has a label number.
 - `labelClassFollowsFigure`: mirror the resolved `<figure>` class onto the `figcaption` spans (`f-embed-label`, `f-embed-label-joint`, `f-embed-body`, etc.) when you want captions styled alongside the wrapper.
 - `figureToLabelClassMap`: extend `labelClassFollowsFigure` by mapping specific figure classes (e.g., `f-embed`) to custom caption label classes such as `caption-embed caption-social` for fine-grained control. When this map is provided and `labelClassFollowsFigure` is not set explicitly, figure-following mode is enabled automatically.
 - `labelPrefixMarker`: allow a leading marker before labels (string or array, e.g., `*Figure. ...`). Arrays are limited to two markers; extras are ignored.
+- Automatic image-label fallback text and punctuation (`Figure. `, `図　`, etc.) are generated from `p7d-markdown-it-p-captions` locale metadata, not from a local hardcoded map in this plugin.
+- `preferredLanguages`: optional tie-break order for generated fallback labels. When omitted, this plugin derives the order once per render from `env.preferredLanguages`, `env.lang` / `env.locale`, then a cheap document-script heuristic that skips a leading hyphen-fenced frontmatter block (`---` or longer, spaces allowed before newline), and finally the raw `languages` order.
 
 ### Automatic numbering
 
@@ -437,13 +440,13 @@ A paragraph.
 
 ### Styles
 
-This example uses `classPrefix: 'custom'` and leaves `styleProcess: true` so a trailing `{.notice}` block moves onto the `<figure>` wrapper.
+This example uses `classPrefix: 'custom'` and leaves `styleProcess: true` so a trailing `{.notice}` block moves onto the `<figure>` wrapper. This fallback only handles the final trailing attrs block on an image-only paragraph; for broader attrs syntax support, keep using `markdown-it-attrs`.
 
 ```
 [Markdown]
-Figure. Highlighted cat. {.notice}
+Figure. Highlighted cat.
 
-![Highlighted cat](cat.jpg)
+![Highlighted cat](cat.jpg) {.notice}
 [HTML]
 <figure class="custom-img notice">
 <figcaption><span class="custom-img-label">Figure<span class="custom-img-label-joint">.</span></span> Highlighted cat.</figcaption>
@@ -453,7 +456,7 @@ Figure. Highlighted cat. {.notice}
 
 ### Automatic detection fallbacks
 
-`autoCaptionDetection` combined with `autoAltCaption` / `autoTitleCaption` can still generate caption text even when the original alt/title lacks labels. The corresponding attributes are cleared after conversion so the figcaption becomes the canonical source.
+`autoCaptionDetection` combined with `autoAltCaption` / `autoTitleCaption` can still generate caption text even when the original alt/title lacks labels, as long as the alt/title body is non-empty. The corresponding attributes are cleared after conversion so the figcaption becomes the canonical source. When these fallbacks are `true`, the generated label text and punctuation come from `p7d-markdown-it-p-captions` locale metadata rather than a local hardcoded map.
 
 ```
 [Markdown]

package/embeds/detect.js

@@ -0,0 +1,178 @@
+import {
+  BLOCKQUOTE_EMBED_CLASS_NAMES,
+  HTML_EMBED_CANDIDATES,
+  VIDEO_IFRAME_HOSTS,
+} from './providers.js'
+
+const htmlRegCache = new Map()
+const openingClassAttrReg = /^<[^>]*?\bclass=(?:"([^"]*)"|'([^']*)')/i
+const openingSrcAttrReg = /^<[^>]*?\bsrc=(?:"([^"]*)"|'([^']*)')/i
+const endBlockquoteScriptReg = /<\/blockquote> *<script[^>]*?><\/script>$/
+const iframeTagReg = /<iframe(?=[\s>])/i
+
+const getHtmlReg = (tag) => {
+  const cached = htmlRegCache.get(tag)
+  if (cached) return cached
+  const regexStr = `^<${tag} ?[^>]*?>[\\s\\S]*?<\\/${tag}>(\\n| *?)(<script [^>]*?>(?:<\\/script>)?)? *(\\n|$)`
+  const reg = new RegExp(regexStr)
+  htmlRegCache.set(tag, reg)
+  return reg
+}
+
+const getHtmlDetectionHints = (content) => {
+  const hasBlueskyHint = content.indexOf('bluesky-embed') !== -1
+  const hasVideoHint = content.indexOf('<video') !== -1
+  const hasAudioHint = content.indexOf('<audio') !== -1
+  const hasIframeHint = content.indexOf('<iframe') !== -1
+  const hasBlockquoteHint = content.indexOf('<blockquote') !== -1
+  const hasDivHint = content.indexOf('<div') !== -1
+  return {
+    hasBlueskyHint,
+    hasVideoHint,
+    hasAudioHint,
+    hasIframeHint,
+    hasBlockquoteHint,
+    hasDivHint,
+    hasIframeTag: hasIframeHint || (hasDivHint && iframeTagReg.test(content)),
+  }
+}
+
+const hasAnyHtmlDetectionHint = (hints) => {
+  return !!(
+    hints.hasBlueskyHint ||
+    hints.hasVideoHint ||
+    hints.hasAudioHint ||
+    hints.hasIframeHint ||
+    hints.hasBlockquoteHint ||
+    hints.hasDivHint
+  )
+}
+
+const appendHtmlBlockNewlineIfNeeded = (token, hasTag) => {
+  if ((hasTag[2] && hasTag[3] !== '\n') || (hasTag[1] !== '\n' && hasTag[2] === undefined)) {
+    token.content += '\n'
+  }
+}
+
+const consumeBlockquoteEmbedScript = (tokens, token, startIndex) => {
+  let addedContent = ''
+  let i = startIndex + 1
+  while (i < tokens.length) {
+    const nextToken = tokens[i]
+    if (nextToken.type === 'inline' && endBlockquoteScriptReg.test(nextToken.content)) {
+      addedContent += nextToken.content + '\n'
+      if (tokens[i + 1] && tokens[i + 1].type === 'paragraph_close') {
+        tokens.splice(i + 1, 1)
+      }
+      nextToken.content = ''
+      if (nextToken.children) {
+        for (let j = 0; j < nextToken.children.length; j++) {
+          nextToken.children[j].content = ''
+        }
+      }
+      break
+    }
+    if (nextToken.type === 'paragraph_open') {
+      addedContent += '\n'
+      tokens.splice(i, 1)
+      continue
+    }
+    i++
+  }
+  token.content += addedContent
+}
+
+const getOpeningAttrValue = (content, reg) => {
+  if (typeof content !== 'string' || content.charCodeAt(0) !== 0x3c) return ''
+  const match = content.match(reg)
+  if (!match) return ''
+  return match[1] || match[2] || ''
+}
+
+const hasKnownBlockquoteEmbedClass = (content) => {
+  const classAttr = getOpeningAttrValue(content, openingClassAttrReg)
+  if (!classAttr) return false
+  let start = 0
+  while (start < classAttr.length) {
+    while (start < classAttr.length && classAttr.charCodeAt(start) <= 0x20) start++
+    if (start >= classAttr.length) break
+    let end = start + 1
+    while (end < classAttr.length && classAttr.charCodeAt(end) > 0x20) end++
+    if (BLOCKQUOTE_EMBED_CLASS_NAMES.has(classAttr.slice(start, end))) return true
+    start = end + 1
+  }
+  return false
+}
+
+const isKnownVideoIframe = (content) => {
+  const src = getOpeningAttrValue(content, openingSrcAttrReg)
+  if (!src || src.slice(0, 8).toLowerCase() !== 'https://') return false
+  const slashIndex = src.indexOf('/', 8)
+  const host = (slashIndex === -1 ? src.slice(8) : src.slice(8, slashIndex)).toLowerCase()
+  return VIDEO_IFRAME_HOSTS.has(host)
+}
+
+const detectHtmlTagCandidate = (tokens, token, startIndex, detector, hints, result) => {
+  if (detector.requiresIframeTag && !hints.hasIframeTag) return ''
+  const hasTagHint = !!(detector.hintKey && hints[detector.hintKey])
+  const allowBlueskyFallback = detector.candidate === 'blockquote' && hints.hasBlueskyHint
+  if (!hasTagHint && !allowBlueskyFallback) return ''
+  const hasTag = hasTagHint ? token.content.match(getHtmlReg(detector.lookupTag)) : null
+  const isBlueskyFallback = detector.candidate === 'blockquote' && !hasTag && hints.hasBlueskyHint
+  if (!hasTag && !isBlueskyFallback) return ''
+  if (hasTag) {
+    appendHtmlBlockNewlineIfNeeded(token, hasTag)
+    if (detector.treatAsVideoIframe) {
+      result.isVideoIframe = true
+    }
+    return detector.matchedTag || detector.candidate
+  }
+  consumeBlockquoteEmbedScript(tokens, token, startIndex)
+  return 'blockquote'
+}
+
+const resolveHtmlWrapWithoutCaption = (matchedTag, result, htmlWrapWithoutCaption) => {
+  if (!htmlWrapWithoutCaption) return false
+  if (matchedTag === 'blockquote') {
+    return !!(result.isIframeTypeBlockquote && htmlWrapWithoutCaption.iframeTypeBlockquote)
+  }
+  return !!htmlWrapWithoutCaption[matchedTag]
+}
+
+export const detectHtmlFigureCandidate = (tokens, token, startIndex, htmlWrapWithoutCaption) => {
+  if (!token || token.type !== 'html_block') return null
+  const hints = getHtmlDetectionHints(token.content)
+  if (!hasAnyHtmlDetectionHint(hints)) return null
+
+  const result = {
+    isVideoIframe: false,
+    isIframeTypeBlockquote: false,
+  }
+
+  let matchedTag = ''
+  for (let i = 0; i < HTML_EMBED_CANDIDATES.length; i++) {
+    matchedTag = detectHtmlTagCandidate(tokens, token, startIndex, HTML_EMBED_CANDIDATES[i], hints, result)
+    if (matchedTag) break
+  }
+  if (!matchedTag) return null
+
+  if (matchedTag === 'blockquote') {
+    if (!hasKnownBlockquoteEmbedClass(token.content)) return null
+    result.isIframeTypeBlockquote = true
+  }
+
+  if (matchedTag === 'iframe' && isKnownVideoIframe(token.content)) {
+    result.isVideoIframe = true
+  }
+
+  return {
+    type: 'html',
+    tagName: matchedTag,
+    en: startIndex,
+    replaceInsteadOfWrap: false,
+    wrapWithoutCaption: resolveHtmlWrapWithoutCaption(matchedTag, result, htmlWrapWithoutCaption),
+    canWrap: true,
+    isVideoIframe: result.isVideoIframe,
+    isIframeTypeBlockquote: result.isIframeTypeBlockquote,
+  }
+}

package/embeds/providers.js

@@ -0,0 +1,27 @@
+export const HTML_EMBED_CANDIDATES = Object.freeze([
+  { candidate: 'video', lookupTag: 'video', hintKey: 'hasVideoHint' },
+  { candidate: 'audio', lookupTag: 'audio', hintKey: 'hasAudioHint' },
+  { candidate: 'iframe', lookupTag: 'iframe', hintKey: 'hasIframeHint' },
+  { candidate: 'blockquote', lookupTag: 'blockquote', hintKey: 'hasBlockquoteHint' },
+  {
+    candidate: 'div',
+    lookupTag: 'div',
+    hintKey: 'hasDivHint',
+    requiresIframeTag: true,
+    matchedTag: 'iframe',
+    treatAsVideoIframe: true,
+  },
+])
+
+export const BLOCKQUOTE_EMBED_CLASS_NAMES = new Set([
+  'twitter-tweet',
+  'instagram-media',
+  'text-post-media',
+  'bluesky-embed',
+  'mastodon-embed',
+])
+
+export const VIDEO_IFRAME_HOSTS = new Set([
+  'www.youtube-nocookie.com',
+  'player.vimeo.com',
+])

package/index.js

@@ -1,62 +1,176 @@
 import {
+  analyzeCaptionStart,
+  buildLabelClassLookup,
+  buildLabelPrefixMarkerRegFromMarkers,
+  getGeneratedLabelDefaults,
+  normalizeLabelPrefixMarkers,
   setCaptionParagraph,
   getMarkRegStateForLanguages,
+  stripLabelPrefixMarker,
 } from 'p7d-markdown-it-p-captions'
+import { detectHtmlFigureCandidate } from './embeds/detect.js'
 
-const htmlRegCache = new Map()
-const blueskyEmbedReg = /^<blockquote class="bluesky-embed"[^]*?>[\s\S]*?$/
-const videoIframeReg = /^<[^>]*? src="https:\/\/(?:www.youtube-nocookie.com|player.vimeo.com)\//i
-const classNameReg = /^<[^>]*? class="(twitter-tweet|instagram-media|text-post-media|bluesky-embed|mastodon-embed)"/
 const imageAttrsReg = /^ *\{(.*?)\} *$/
 const classAttrReg = /^\./
 const idAttrReg = /^#/
 const attrParseReg = /^(.*?)="?(.*)"?$/
 const sampLangReg = /^ *(?:samp|shell|console)(?:(?= )|$)/
-const endBlockquoteScriptReg = /<\/blockquote> *<script[^>]*?><\/script>$/
-const iframeTagReg = /<iframe(?=[\s>])/i
 const asciiLabelReg = /^[A-Za-z]/
 const CHECK_TYPE_TOKEN_MAP = {
   table_open: 'table',
   pre_open: 'pre',
   blockquote_open: 'blockquote',
 }
-const HTML_TAG_CANDIDATES = ['video', 'audio', 'iframe', 'blockquote', 'div']
-const fallbackLabelDefaults = {
-  img: { en: 'Figure', ja: '図' },
-  table: { en: 'Table', ja: '表' },
-}
-
 const escapeRegExp = (value) => value.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')
-const buildClassPrefix = (value) => (value ? value + '-' : '')
-const normalizeLanguages = (value) => {
-  if (!Array.isArray(value)) return ['en', 'ja']
-  const normalized = []
+const normalizeLanguageCode = (value) => {
+  if (value === null || value === undefined) return ''
+  const normalized = String(value).trim().toLowerCase()
+  if (!normalized) return ''
+  const separatorIndex = normalized.search(/[-_]/)
+  return separatorIndex === -1 ? normalized : normalized.slice(0, separatorIndex)
+}
+const normalizePreferredLanguages = (value, availableLanguages) => {
+  if (!Array.isArray(availableLanguages) || availableLanguages.length === 0) return []
+  const source = typeof value === 'string' ? [value] : (Array.isArray(value) ? value : [])
+  if (source.length === 0) return []
+  const allowed = new Set(availableLanguages)
+  const languages = []
   const seen = new Set()
-  for (let i = 0; i < value.length; i++) {
-    const lang = value[i]
-    if (typeof lang !== 'string') continue
-    const trimmed = lang.trim()
-    if (!trimmed || seen.has(trimmed)) continue
-    seen.add(trimmed)
-    normalized.push(trimmed)
+  for (let i = 0; i < source.length; i++) {
+    const lang = normalizeLanguageCode(source[i])
+    if (!lang || seen.has(lang) || !allowed.has(lang)) continue
+    seen.add(lang)
+    languages.push(lang)
   }
-  if (normalized.length === 0) return ['en', 'ja']
-  return normalized
+  return languages
 }
-const normalizeLabelPrefixMarkers = (value) => {
-  if (typeof value === 'string') {
-    return value ? [value] : []
+const prioritizeLanguage = (languages, preferredLanguage) => {
+  if (!preferredLanguage || languages.length === 0) return languages.slice()
+  const prioritized = []
+  prioritized.push(preferredLanguage)
+  for (let i = 0; i < languages.length; i++) {
+    if (languages[i] === preferredLanguage) continue
+    prioritized.push(languages[i])
+  }
+  return prioritized
+}
+const isAsciiAlphaCode = (code) => {
+  return (code >= 0x41 && code <= 0x5a) || (code >= 0x61 && code <= 0x7a)
+}
+const isJapaneseCharCode = (code) => {
+  return (
+    (code >= 0x3040 && code <= 0x30ff) ||
+    (code >= 0x31f0 && code <= 0x31ff) ||
+    (code >= 0x4e00 && code <= 0x9fff) ||
+    (code >= 0xff66 && code <= 0xff9f)
+  )
+}
+const isHyphenFenceLine = (src, lineStart) => {
+  if (typeof src !== 'string' || lineStart < 0 || lineStart >= src.length) return 0
+  let index = lineStart
+  let hyphenCount = 0
+  while (index < src.length && src.charCodeAt(index) === 0x2d) {
+    hyphenCount++
+    index++
+  }
+  if (hyphenCount < 3) return 0
+  while (index < src.length && src.charCodeAt(index) === 0x20) {
+    index++
+  }
+  if (index >= src.length || src.charCodeAt(index) !== 0x0a) return 0
+  return hyphenCount
+}
+const skipLeadingFrontmatter = (src) => {
+  if (typeof src !== 'string' || isHyphenFenceLine(src, 0) === 0) return src
+  let lineStart = src.indexOf('\n')
+  if (lineStart === -1) return src
+  lineStart++
+  while (lineStart < src.length) {
+    if (isHyphenFenceLine(src, lineStart) > 0) {
+      const nextLineStart = src.indexOf('\n', lineStart)
+      if (nextLineStart === -1) return ''
+      return src.slice(nextLineStart + 1)
+    }
+    const nextLineStart = src.indexOf('\n', lineStart)
+    if (nextLineStart === -1) break
+    lineStart = nextLineStart + 1
   }
-  if (Array.isArray(value)) {
-    const normalized = value.map(entry => String(entry)).filter(Boolean)
-    return normalized.length > 2 ? normalized.slice(0, 2) : normalized
+  return src
+}
+const detectDocumentPrimaryLanguage = (src, availableLanguages) => {
+  if (!src || availableLanguages.indexOf('ja') === -1) return ''
+  const body = skipLeadingFrontmatter(src)
+  const limit = Math.min(body.length, 8192)
+  let japaneseCount = 0
+  let asciiAlphaCount = 0
+  for (let i = 0; i < limit; i++) {
+    const code = body.charCodeAt(i)
+    if (isJapaneseCharCode(code)) {
+      japaneseCount++
+      continue
+    }
+    if (isAsciiAlphaCode(code)) {
+      asciiAlphaCount++
+    }
   }
-  return []
+  if (japaneseCount === 0) return ''
+  if (asciiAlphaCount === 0) return 'ja'
+  return japaneseCount * 2 >= asciiAlphaCount ? 'ja' : ''
+}
+const sourceMayNeedPreferredLanguages = (state) => {
+  const src = state && typeof state.src === 'string' ? state.src : ''
+  return src.indexOf('![') !== -1
+}
+const resolvePreferredLanguagesForState = (state, opt) => {
+  const availableLanguages = (
+    opt &&
+    opt.markRegState &&
+    Array.isArray(opt.markRegState.languages)
+  ) ? opt.markRegState.languages : []
+  if (availableLanguages.length === 0) return []
+
+  const explicitPreferred = opt && Array.isArray(opt.preferredLanguages)
+    ? opt.preferredLanguages
+    : []
+  if (explicitPreferred.length > 0) return explicitPreferred
+
+  const optionLanguages = opt && Array.isArray(opt.normalizedOptionLanguages)
+    ? opt.normalizedOptionLanguages
+    : []
+  const baseLanguages = optionLanguages.length > 0 ? optionLanguages : availableLanguages
+  const env = state && state.env ? state.env : null
+  const envPreferred = normalizePreferredLanguages(env && env.preferredLanguages, availableLanguages)
+  if (envPreferred.length > 0) return envPreferred
+
+  const envLanguage = normalizeLanguageCode(env && (env.preferredLanguage || env.lang || env.language || env.locale))
+  if (envLanguage && baseLanguages.indexOf(envLanguage) !== -1) {
+    return prioritizeLanguage(baseLanguages, envLanguage)
+  }
+
+  const detectedLanguage = detectDocumentPrimaryLanguage(state && state.src ? state.src : '', baseLanguages)
+  if (detectedLanguage) {
+    return prioritizeLanguage(baseLanguages, detectedLanguage)
+  }
+  return baseLanguages
+}
+const needsPreferredLanguagesResolution = (opt) => {
+  if (!opt || !opt.markRegState || !Array.isArray(opt.markRegState.languages)) return false
+  if (opt.markRegState.languages.length <= 1) return false
+  if (Array.isArray(opt.preferredLanguages) && opt.preferredLanguages.length > 0) return false
+  return opt.autoAltCaption === true || opt.autoTitleCaption === true
+}
+const normalizeOptionalClassName = (value) => {
+  if (value === null || value === undefined) return ''
+  const normalized = String(value).trim()
+  return normalized || ''
+}
+const buildClassPrefix = (value) => {
+  const normalized = normalizeOptionalClassName(value)
+  return normalized ? normalized + '-' : ''
 }
-const buildLabelPrefixMarkerRegFromList = (markers) => {
-  if (!markers || markers.length === 0) return null
-  const pattern = markers.map(escapeRegExp).join('|')
-  return new RegExp('^(?:' + pattern + ')(?:[ \\t　]+)?')
+const normalizeClassOptionWithFallback = (value, fallbackValue) => {
+  const normalized = normalizeOptionalClassName(value)
+  return normalized || fallbackValue
 }
 const resolveLabelPrefixMarkerPair = (markers) => {
   if (!markers || markers.length === 0) return { prev: [], next: [] }
@@ -65,26 +179,6 @@ const resolveLabelPrefixMarkerPair = (markers) => {
   }
   return { prev: [markers[0]], next: [markers[1]] }
 }
-const stripLeadingPrefix = (text, prefix) => {
-  if (typeof text !== 'string' || !text || !prefix) return text
-  if (text.startsWith(prefix)) return text.slice(prefix.length)
-  return text
-}
-const stripLabelPrefixMarkerFromInline = (inlineToken, markerText) => {
-  if (!inlineToken || !markerText) return
-  if (typeof inlineToken.content === 'string') {
-    inlineToken.content = stripLeadingPrefix(inlineToken.content, markerText)
-  }
-  if (inlineToken.children && inlineToken.children.length) {
-    for (let i = 0; i < inlineToken.children.length; i++) {
-      const child = inlineToken.children[i]
-      if (child && child.type === 'text' && typeof child.content === 'string') {
-        child.content = stripLeadingPrefix(child.content, markerText)
-        break
-      }
-    }
-  }
-}
 const getLabelPrefixMarkerMatch = (inlineToken, markerReg) => {
   if (!markerReg || !inlineToken || inlineToken.type !== 'inline') return null
   const content = typeof inlineToken.content === 'string' ? inlineToken.content : ''
@@ -127,20 +221,6 @@ const normalizeAutoLabelNumberSets = (value) => {
   return normalized
 }
 
-const buildLabelClassLookup = (opt) => {
-  const classPrefix = opt.classPrefix ? opt.classPrefix + '-' : ''
-  const defaultClasses = [classPrefix + 'label']
-  const withType = (type) => {
-    if (opt.removeMarkNameInCaptionClass) return defaultClasses
-    return [classPrefix + type + '-label', ...defaultClasses]
-  }
-  return {
-    img: withType('img'),
-    table: withType('table'),
-    default: defaultClasses,
-  }
-}
-
 const shouldApplyLabelNumbering = (captionType, opt) => {
   const setting = opt.autoLabelNumberSets
   if (!setting) return false
@@ -219,63 +299,23 @@ const getImageAltText = (token) => {
 
 const getImageTitleText = (token) => getTokenAttr(token, 'title')
 
-const detectCaptionLanguage = (text) => {
-  const target = (text || '').trim()
-  if (!target) return 'en'
-  for (let i = 0; i < target.length; i++) {
-    const char = target[i]
-    const code = target.charCodeAt(i)
-    if (isJapaneseCharCode(code)) return 'ja'
-    if (isSentenceBoundaryChar(char) || char === '\n') break
-  }
-  return 'en'
-}
-
-const isJapaneseCharCode = (code) => {
-  return (
-    (code >= 0x3040 && code <= 0x30ff) || // Hiragana + Katakana
-    (code >= 0x31f0 && code <= 0x31ff) || // Katakana extensions
-    (code >= 0x4e00 && code <= 0x9fff) || // CJK Unified Ideographs
-    (code >= 0xff66 && code <= 0xff9f)    // Half-width Katakana
-  )
-}
-
-const isSentenceBoundaryChar = (char) => {
-  return char === '.' || char === '!' || char === '?' || char === '。' || char === '！' || char === '？'
-}
-
-const getAutoFallbackLabel = (text, captionType) => {
-  const type = captionType === 'table' ? 'table' : 'img'
-  const lang = detectCaptionLanguage(text)
-  const defaults = fallbackLabelDefaults[type] || fallbackLabelDefaults.img
-  if (lang === 'ja') return defaults.ja || defaults.en || ''
-  return defaults.en || defaults.ja || ''
-}
-
-const getPersistedFallbackLabel = (text, captionType, fallbackState) => {
-  const type = captionType === 'table' ? 'table' : 'img'
-  if (!fallbackState) return getAutoFallbackLabel(text, type)
-  if (fallbackState[type]) return fallbackState[type]
-  const resolved = getAutoFallbackLabel(text, type)
-  fallbackState[type] = resolved
-  return resolved
-}
-
-const buildCaptionWithFallback = (text, fallbackOption, captionType = 'img', fallbackState) => {
+const buildCaptionWithFallback = (text, fallbackOption, mark, markRegState, preferredLanguages) => {
   const trimmedText = (text || '').trim()
   if (!fallbackOption) return ''
+  if (!trimmedText) return ''
   let label = ''
+  let generatedDefaults = null
   if (typeof fallbackOption === 'string') {
     label = fallbackOption.trim()
   } else if (fallbackOption === true) {
-    label = getPersistedFallbackLabel(trimmedText, captionType, fallbackState)
+    generatedDefaults = getGeneratedLabelDefaults(mark, trimmedText, markRegState, preferredLanguages)
+    label = generatedDefaults && generatedDefaults.label ? generatedDefaults.label : ''
   }
-  if (!label) return trimmedText
-  const isAsciiLabel = asciiLabelReg.test(label)
-  if (!trimmedText) {
-    return isAsciiLabel ? label + '.' : label
+  if (!label) return fallbackOption === true ? '' : trimmedText
+  if (generatedDefaults) {
+    return label + (generatedDefaults.joint || '') + (generatedDefaults.space || '') + trimmedText
   }
-  return label + (isAsciiLabel ? '. ' : '　') + trimmedText
+  return label + (asciiLabelReg.test(label) ? '. ' : '　') + trimmedText
 }
 
 const createAutoCaptionParagraph = (captionText, TokenConstructor) => {
@@ -401,73 +441,71 @@ const ensureAutoFigureNumbering = (tokens, range, caption, figureNumberState, op
   updateInlineTokenContent(inlineToken, originalText, newLabelText)
 }
 
-const matchAutoCaptionText = (text, reg) => {
-  if (!text || !reg) return ''
+const matchAutoCaptionText = (text, opt, preferredMark = 'img') => {
+  if (!text || !opt || !opt.markRegState) return ''
   const trimmed = text.trim()
-  if (trimmed && reg.test(trimmed)) return trimmed
+  if (!trimmed) return ''
+  const analysis = analyzeCaptionStart(trimmed, {
+    markRegState: opt.markRegState,
+    preferredMark,
+  })
+  if (analysis) return trimmed
   return ''
 }
 
-const getAutoCaptionFromImage = (imageToken, opt, fallbackLabelState) => {
-  const imgCaptionMarkReg = opt && opt.imgCaptionMarkReg ? opt.imgCaptionMarkReg : null
+const getAutoCaptionFromImage = (imageToken, opt) => {
   if (!opt.autoCaptionDetection) return ''
-  if (!imgCaptionMarkReg && !opt.autoAltCaption && !opt.autoTitleCaption) return ''
+  if (!opt.autoAltCaption && !opt.autoTitleCaption && !(opt.markRegState && opt.markRegState.markReg && opt.markRegState.markReg.img)) return ''
 
   const altText = getImageAltText(imageToken)
-  let caption = matchAutoCaptionText(altText, imgCaptionMarkReg)
+  let caption = matchAutoCaptionText(altText, opt)
   if (caption) {
     clearImageAltAttr(imageToken)
     return caption
   }
   if (!caption && opt.autoAltCaption) {
     const altForFallback = altText || ''
-    caption = buildCaptionWithFallback(altForFallback, opt.autoAltCaption, 'img', fallbackLabelState)
-    if (imageToken) {
+    const fallbackCaption = buildCaptionWithFallback(altForFallback, opt.autoAltCaption, 'img', opt.markRegState, opt.preferredLanguages)
+    if (fallbackCaption && imageToken) {
       clearImageAltAttr(imageToken)
     }
+    caption = fallbackCaption
   }
   if (caption) return caption
 
   const titleText = getImageTitleText(imageToken)
-  caption = matchAutoCaptionText(titleText, imgCaptionMarkReg)
+  caption = matchAutoCaptionText(titleText, opt)
   if (caption) {
     clearImageTitleAttr(imageToken)
     return caption
   }
   if (!caption && opt.autoTitleCaption) {
     const titleForFallback = titleText || ''
-    caption = buildCaptionWithFallback(titleForFallback, opt.autoTitleCaption, 'img', fallbackLabelState)
-    if (imageToken) {
+    const fallbackCaption = buildCaptionWithFallback(titleForFallback, opt.autoTitleCaption, 'img', opt.markRegState, opt.preferredLanguages)
+    if (fallbackCaption && imageToken) {
       clearImageTitleAttr(imageToken)
     }
+    caption = fallbackCaption
   }
   return caption
 }
 
-const getHtmlReg = (tag) => {
-  if (htmlRegCache.has(tag)) return htmlRegCache.get(tag)
-  const regexStr = `^<${tag} ?[^>]*?>[\\s\\S]*?<\\/${tag}>(\\n| *?)(<script [^>]*?>(?:<\\/script>)?)? *(\\n|$)`
-  const reg = new RegExp(regexStr)
-  htmlRegCache.set(tag, reg)
-  return reg
-}
-
-const checkPrevCaption = (tokens, n, caption, fNum, sp, opt, captionState) => {
+const checkPrevCaption = (tokens, n, caption, sp, opt, captionState) => {
   if(n < 3) return caption
   const captionStartToken = tokens[n-3]
   const captionInlineToken = tokens[n-2]
   const captionEndToken = tokens[n-1]
   if (captionStartToken === undefined || captionEndToken === undefined) return
   if (captionStartToken.type !== 'paragraph_open' || captionEndToken.type !== 'paragraph_close') return
-  setCaptionParagraph(n-3, captionState, caption, fNum, sp, opt)
+  setCaptionParagraph(n-3, captionState, caption, null, sp, opt)
   const captionName = sp && sp.captionDecision ? sp.captionDecision.mark : ''
   if(!captionName) {
     if (opt.labelPrefixMarkerWithoutLabelPrevReg) {
-      const markerMatch = getLabelPrefixMarkerMatch(captionInlineToken, opt.labelPrefixMarkerWithoutLabelPrevReg)
-      if (markerMatch) {
-        stripLabelPrefixMarkerFromInline(captionInlineToken, markerMatch)
-        caption.isPrev = true
-      }
+        const markerMatch = getLabelPrefixMarkerMatch(captionInlineToken, opt.labelPrefixMarkerWithoutLabelPrevReg)
+        if (markerMatch) {
+          stripLabelPrefixMarker(captionInlineToken, markerMatch)
+          caption.isPrev = true
+        }
     }
     return
   }
@@ -476,22 +514,22 @@ const checkPrevCaption = (tokens, n, caption, fNum, sp, opt, captionState) => {
   return
 }
 
-const checkNextCaption = (tokens, en, caption, fNum, sp, opt, captionState) => {
+const checkNextCaption = (tokens, en, caption, sp, opt, captionState) => {
   if (en + 2 > tokens.length) return
   const captionStartToken = tokens[en+1]
   const captionInlineToken = tokens[en+2]
   const captionEndToken = tokens[en+3]
   if (captionStartToken === undefined || captionEndToken === undefined) return
   if (captionStartToken.type !== 'paragraph_open' || captionEndToken.type !== 'paragraph_close') return
-  setCaptionParagraph(en+1, captionState, caption, fNum, sp, opt)
+  setCaptionParagraph(en+1, captionState, caption, null, sp, opt)
   const captionName = sp && sp.captionDecision ? sp.captionDecision.mark : ''
   if(!captionName) {
     if (opt.labelPrefixMarkerWithoutLabelNextReg) {
-      const markerMatch = getLabelPrefixMarkerMatch(captionInlineToken, opt.labelPrefixMarkerWithoutLabelNextReg)
-      if (markerMatch) {
-        stripLabelPrefixMarkerFromInline(captionInlineToken, markerMatch)
-        caption.isNext = true
-      }
+        const markerMatch = getLabelPrefixMarkerMatch(captionInlineToken, opt.labelPrefixMarkerWithoutLabelNextReg)
+        if (markerMatch) {
+          stripLabelPrefixMarker(captionInlineToken, markerMatch)
+          caption.isNext = true
+        }
     }
     return
   }
@@ -613,18 +651,18 @@ const wrapWithFigure = (tokens, range, checkTokenTagName, caption, replaceInstea
     breakToken.content = '\n'
     return breakToken
   }
-  if (opt.styleProcess && caption.isNext && sp.attrs.length > 0) {
-    for (let i = 0; i < sp.attrs.length; i++) {
-      const attr = sp.attrs[i]
-      figureStartToken.attrJoin(attr[0], attr[1])
-    }
-  }
-  // For vsce
-  if (caption.name === 'img' && tokens[n].attrs) {
-    for (let i = 0; i < tokens[n].attrs.length; i++) {
-      const attr = tokens[n].attrs[i]
-      figureStartToken.attrJoin(attr[0], attr[1])
+  if (caption.name === 'img') {
+    const joinAttrs = (attrs) => {
+      if (!attrs || attrs.length === 0) return
+      for (let i = 0; i < attrs.length; i++) {
+        const attr = attrs[i]
+        figureStartToken.attrJoin(attr[0], attr[1])
+      }
     }
+    // `styleProcess` should keep working even when markdown-it-attrs is absent.
+    if (opt.styleProcess) joinAttrs(sp.attrs)
+    // Forward attrs already materialized by markdown-it-attrs on the image paragraph.
+    joinAttrs(tokens[n].attrs)
   }
   if (replaceInsteadOfWrap) {
     tokens.splice(en, 1, createBreakToken(), figureEndToken, createBreakToken())
@@ -640,10 +678,10 @@ const wrapWithFigure = (tokens, range, checkTokenTagName, caption, replaceInstea
   return
 }
 
-const checkCaption = (tokens, n, en, caption, fNum, sp, opt, captionState) => {
-  checkPrevCaption(tokens, n, caption, fNum, sp, opt, captionState)
+const checkCaption = (tokens, n, en, caption, sp, opt, captionState) => {
+  checkPrevCaption(tokens, n, caption, sp, opt, captionState)
   if (caption.isPrev) return
-  checkNextCaption(tokens, en, caption, fNum, sp, opt, captionState)
+  checkNextCaption(tokens, en, caption, sp, opt, captionState)
   return
 }
 
@@ -736,129 +774,38 @@ const detectFenceToken = (token, n, caption) => {
   }
 }
 
-const detectHtmlBlockToken = (tokens, token, n, caption, sp, opt) => {
-  if (!token || token.type !== 'html_block') return null
-  const content = token.content
-  const hasBlueskyHint = content.indexOf('bluesky-embed') !== -1
-  const hasVideoHint = content.indexOf('<video') !== -1
-  const hasAudioHint = content.indexOf('<audio') !== -1
-  const hasIframeHint = content.indexOf('<iframe') !== -1
-  const hasBlockquoteHint = content.indexOf('<blockquote') !== -1
-  const hasDivHint = content.indexOf('<div') !== -1
-  const hasIframeTag = hasIframeHint || (hasDivHint && iframeTagReg.test(content))
-  const hasBlueskyEmbed = hasBlueskyHint && blueskyEmbedReg.test(content)
-  if (!hasBlueskyHint
-      && !hasVideoHint
-      && !hasAudioHint
-      && !hasIframeHint
-      && !hasBlockquoteHint
-      && !hasDivHint) {
-    return null
-  }
-  let matchedTag = ''
-  for (let i = 0; i < HTML_TAG_CANDIDATES.length; i++) {
-    const candidate = HTML_TAG_CANDIDATES[i]
-    const treatDivAsIframe = candidate === 'div'
-    const lookupTag = treatDivAsIframe ? 'div' : candidate
-    let hasTagHint = false
-    if (candidate === 'video') {
-      hasTagHint = hasVideoHint
-    } else if (candidate === 'audio') {
-      hasTagHint = hasAudioHint
-    } else if (candidate === 'iframe') {
-      hasTagHint = hasIframeHint
-    } else if (candidate === 'blockquote') {
-      hasTagHint = hasBlockquoteHint
-    } else {
-      hasTagHint = hasDivHint
-    }
-    if (candidate === 'div' && !hasIframeTag) continue
-    if (!hasTagHint && !(candidate === 'blockquote' && hasBlueskyEmbed)) continue
-    const hasTag = hasTagHint ? content.match(getHtmlReg(lookupTag)) : null
-    const isBlueskyBlockquote = !hasTag && hasBlueskyEmbed && candidate === 'blockquote'
-    if (!(hasTag || isBlueskyBlockquote)) continue
-    if (hasTag) {
-      if ((hasTag[2] && hasTag[3] !== '\n') || (hasTag[1] !== '\n' && hasTag[2] === undefined)) {
-        token.content += '\n'
-      }
-      matchedTag = treatDivAsIframe ? 'iframe' : candidate
-      if (treatDivAsIframe) {
-        sp.isVideoIframe = true
-      }
-    } else {
-      let addedCont = ''
-      let j = n + 1
-      while (j < tokens.length) {
-        const nextToken = tokens[j]
-        if (nextToken.type === 'inline' && endBlockquoteScriptReg.test(nextToken.content)) {
-          addedCont += nextToken.content + '\n'
-          if (tokens[j + 1] && tokens[j + 1].type === 'paragraph_close') {
-            tokens.splice(j + 1, 1)
-          }
-          nextToken.content = ''
-          if (nextToken.children) {
-            for (let k = 0; k < nextToken.children.length; k++) {
-              nextToken.children[k].content = ''
-            }
-          }
-          break
-        }
-        if (nextToken.type === 'paragraph_open') {
-          addedCont += '\n'
-          tokens.splice(j, 1)
-          continue
-        }
-        j++
-      }
-      token.content += addedCont
-      matchedTag = 'blockquote'
-    }
-    break
-  }
-  if (!matchedTag) return null
-  if (matchedTag === 'blockquote') {
-    if (token.content.indexOf('class="') !== -1 && classNameReg.test(token.content)) {
-      sp.isIframeTypeBlockquote = true
-    } else {
-      return null
-    }
-  }
-  if (matchedTag === 'iframe' && videoIframeReg.test(token.content)) {
-    sp.isVideoIframe = true
-  }
-  caption.name = matchedTag
-  let wrapWithoutCaption = false
-  const htmlWrapWithoutCaption = opt.htmlWrapWithoutCaption
-  if (matchedTag === 'blockquote') {
-    wrapWithoutCaption = !!(sp.isIframeTypeBlockquote && htmlWrapWithoutCaption && htmlWrapWithoutCaption.iframeTypeBlockquote)
-  } else if (htmlWrapWithoutCaption) {
-    wrapWithoutCaption = !!htmlWrapWithoutCaption[matchedTag]
-  }
-  return {
-    type: 'html',
-    tagName: matchedTag,
-    en: n,
-    replaceInsteadOfWrap: false,
-    wrapWithoutCaption,
-    canWrap: true,
-  }
+const hasLeadingImageChild = (token) => {
+  return !!(token &&
+    token.type === 'inline' &&
+    token.children &&
+    token.children.length > 0 &&
+    token.children[0] &&
+    token.children[0].type === 'image')
 }
 
-const detectImageParagraph = (tokens, token, nextToken, n, caption, sp, opt) => {
-  if (!token || token.type !== 'paragraph_open') return null
-  if (!nextToken || nextToken.type !== 'inline' || !nextToken.children || nextToken.children.length === 0) return null
-  if (nextToken.children[0].type !== 'image') return null
-
+const detectImageParagraph = (nextToken, n, caption, sp, opt) => {
   const multipleImagesEnabled = !!opt.multipleImages
   const styleProcessEnabled = !!opt.styleProcess
   const allowSingleImageWithoutCaption = !!opt.oneImageWithoutCaption
+  const children = nextToken.children
+  const imageToken = children[0]
+  const childrenLength = children.length
   let imageNum = 1
   let isMultipleImagesHorizontal = true
   let isMultipleImagesVertical = true
   let isValid = true
   caption.name = 'img'
-  const children = nextToken.children
-  const childrenLength = children.length
+  if (childrenLength === 1) {
+    return {
+      type: 'image',
+      tagName: 'img',
+      en: n + 2,
+      replaceInsteadOfWrap: true,
+      wrapWithoutCaption: allowSingleImageWithoutCaption,
+      canWrap: true,
+      imageToken,
+    }
+  }
   if (!multipleImagesEnabled && childrenLength > 2) {
     return {
       type: 'image',
@@ -867,7 +814,7 @@ const detectImageParagraph = (tokens, token, nextToken, n, caption, sp, opt) =>
       replaceInsteadOfWrap: true,
       wrapWithoutCaption: false,
       canWrap: false,
-      imageToken: children[0]
+      imageToken,
     }
   }
   for (let childIndex = 1; childIndex < childrenLength; childIndex++) {
@@ -882,6 +829,7 @@ const detectImageParagraph = (tokens, token, nextToken, n, caption, sp, opt) =>
             for (let i = 0; i < parsedAttrs.length; i++) {
               sp.attrs.push(parsedAttrs[i])
             }
+            child.content = ''
           }
           break
         }
@@ -936,31 +884,29 @@ const detectImageParagraph = (tokens, token, nextToken, n, caption, sp, opt) =>
     replaceInsteadOfWrap: true,
     wrapWithoutCaption: isValid && allowSingleImageWithoutCaption,
     canWrap: isValid,
-    imageToken: children[0]
+    imageToken,
   }
 }
 
 const figureWithCaption = (state, opt) => {
-  let fNum = {
-    img: 0,
-    table: 0,
-  }
-
   const figureNumberState = {
     img: 0,
     table: 0,
   }
 
-  const fallbackLabelState = {
-    img: null,
-    table: null,
-  }
-
   const captionState = { tokens: state.tokens, Token: state.Token }
-  figureWithCaptionCore(state.tokens, opt, fNum, figureNumberState, fallbackLabelState, state.Token, captionState, null, 0)
+  const shouldResolvePreferredLanguages = !!(
+    opt.shouldResolvePreferredLanguages &&
+    sourceMayNeedPreferredLanguages(state)
+  )
+  const renderOpt = shouldResolvePreferredLanguages ? Object.create(opt) : opt
+  if (shouldResolvePreferredLanguages) {
+    renderOpt.preferredLanguages = resolvePreferredLanguagesForState(state, opt)
+  }
+  figureWithCaptionCore(state.tokens, renderOpt, figureNumberState, state.Token, captionState, null, 0)
 }
 
-const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLabelState, TokenConstructor, captionState, parentType = null, startIndex = 0) => {
+const figureWithCaptionCore = (tokens, opt, figureNumberState, TokenConstructor, captionState, parentType = null, startIndex = 0) => {
   const rRange = { start: startIndex, end: startIndex }
   const rCaption = {
     name: '', nameSuffix: '', isPrev: false, isNext: false
@@ -978,7 +924,7 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
     const containerType = getNestedContainerType(token)
 
     if (containerType && containerType !== 'blockquote') {
-      const closeIndex = figureWithCaptionCore(tokens, opt, fNum, figureNumberState, fallbackLabelState, TokenConstructor, captionState, containerType, n + 1)
+      const closeIndex = figureWithCaptionCore(tokens, opt, figureNumberState, TokenConstructor, captionState, containerType, n + 1)
       n = (typeof closeIndex === 'number' ? closeIndex : n) + 1
       continue
     }
@@ -992,16 +938,23 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
     const tokenType = token.type
     const blockType = CHECK_TYPE_TOKEN_MAP[tokenType]
     if (tokenType === 'paragraph_open') {
-      resetRangeState(rRange, n)
-      resetCaptionState(rCaption)
-      resetSpecialState(rSp)
       const nextToken = tokens[n + 1]
-      detection = detectImageParagraph(tokens, token, nextToken, n, rCaption, rSp, opt)
+      if (hasLeadingImageChild(nextToken)) {
+        resetRangeState(rRange, n)
+        resetCaptionState(rCaption)
+        resetSpecialState(rSp)
+        detection = detectImageParagraph(nextToken, n, rCaption, rSp, opt)
+      }
     } else if (tokenType === 'html_block') {
       resetRangeState(rRange, n)
       resetCaptionState(rCaption)
       resetSpecialState(rSp)
-      detection = detectHtmlBlockToken(tokens, token, n, rCaption, rSp, opt)
+      detection = detectHtmlFigureCandidate(tokens, token, n, opt.htmlWrapWithoutCaption)
+      if (detection) {
+        rCaption.name = detection.tagName
+        rSp.isVideoIframe = !!detection.isVideoIframe
+        rSp.isIframeTypeBlockquote = !!detection.isIframeTypeBlockquote
+      }
     } else if (tokenType === 'fence') {
       resetRangeState(rRange, n)
       resetCaptionState(rCaption)
@@ -1016,7 +969,7 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
 
     if (!detection) {
       if (containerType === 'blockquote') {
-        const closeIndex = figureWithCaptionCore(tokens, opt, fNum, figureNumberState, fallbackLabelState, TokenConstructor, captionState, containerType, n + 1)
+        const closeIndex = figureWithCaptionCore(tokens, opt, figureNumberState, TokenConstructor, captionState, containerType, n + 1)
         n = (typeof closeIndex === 'number' ? closeIndex : n) + 1
       } else {
         n++
@@ -1027,13 +980,13 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
     rRange.end = detection.en
 
     rSp.figureClassName = resolveFigureClassName(detection.tagName, rSp, opt)
-    checkCaption(tokens, rRange.start, rRange.end, rCaption, fNum, rSp, opt, captionState)
+    checkCaption(tokens, rRange.start, rRange.end, rCaption, rSp, opt, captionState)
     applyCaptionDrivenFigureClass(rCaption, rSp, opt)
 
     let hasCaption = rCaption.isPrev || rCaption.isNext
     let pendingAutoCaption = ''
     if (!hasCaption && detection.type === 'image' && opt.autoCaptionDetection) {
-      pendingAutoCaption = getAutoCaptionFromImage(detection.imageToken, opt, fallbackLabelState)
+      pendingAutoCaption = getAutoCaptionFromImage(detection.imageToken, opt)
       if (pendingAutoCaption) {
         hasCaption = true
       }
@@ -1042,7 +995,7 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
     if (detection.canWrap === false) {
       let nextIndex = rRange.end + 1
       if (containerType === 'blockquote') {
-        const closeIndex = figureWithCaptionCore(tokens, opt, fNum, figureNumberState, fallbackLabelState, TokenConstructor, captionState, containerType, rRange.start + 1)
+        const closeIndex = figureWithCaptionCore(tokens, opt, figureNumberState, TokenConstructor, captionState, containerType, rRange.start + 1)
         nextIndex = Math.max(nextIndex, (typeof closeIndex === 'number' ? closeIndex : rRange.end) + 1)
       }
       n = nextIndex
@@ -1069,7 +1022,7 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
         rRange.start += insertedLength
         rRange.end += insertedLength
         n += insertedLength
-        checkCaption(tokens, rRange.start, rRange.end, rCaption, fNum, rSp, opt, captionState)
+        checkCaption(tokens, rRange.start, rRange.end, rCaption, rSp, opt, captionState)
         applyCaptionDrivenFigureClass(rCaption, rSp, opt)
       }
       ensureAutoFigureNumbering(tokens, rRange, rCaption, figureNumberState, opt)
@@ -1097,7 +1050,7 @@ const figureWithCaptionCore = (tokens, opt, fNum, figureNumberState, fallbackLab
     }
 
     if (containerType === 'blockquote') {
-      const closeIndex = figureWithCaptionCore(tokens, opt, fNum, figureNumberState, fallbackLabelState, TokenConstructor, captionState, containerType, rRange.start + 1)
+      const closeIndex = figureWithCaptionCore(tokens, opt, figureNumberState, TokenConstructor, captionState, containerType, rRange.start + 1)
       const fallbackIndex = rCaption.name ? rRange.end : n
       nextIndex = Math.max(nextIndex, (typeof closeIndex === 'number' ? closeIndex : fallbackIndex) + 1)
     }
@@ -1111,12 +1064,13 @@ const mditFigureWithPCaption = (md, option) => {
   let opt = {
     // Caption languages delegated to p-captions.
     languages: ['en', 'ja'],
+    preferredLanguages: null, // optional tie-break order for generated fallback labels; defaults to inferred document order / languages
 
     // --- figure-wrapper behavior ---
     classPrefix: 'f',
     figureClassThatWrapsIframeTypeBlockquote: null,
     figureClassThatWrapsSlides: null,
-    styleProcess : true,
+    styleProcess: true,
     oneImageWithoutCaption: false,
     iframeWithoutCaption: false,
     videoWithoutCaption: false,
@@ -1163,11 +1117,13 @@ const mditFigureWithPCaption = (md, option) => {
   if (!hasExplicitLabelClassFollowsFigure && opt.figureToLabelClassMap) {
     opt.labelClassFollowsFigure = true
   }
-  opt.languages = normalizeLanguages(opt.languages)
+  opt.classPrefix = normalizeOptionalClassName(opt.classPrefix)
+  opt.allIframeTypeFigureClassName = normalizeOptionalClassName(opt.allIframeTypeFigureClassName)
   opt.markRegState = getMarkRegStateForLanguages(opt.languages)
-  opt.imgCaptionMarkReg = opt.markRegState && opt.markRegState.markReg
-    ? opt.markRegState.markReg.img
-    : null
+  opt.preferredLanguages = normalizePreferredLanguages(opt.preferredLanguages, opt.markRegState.languages)
+  if (opt.preferredLanguages.length === 0) opt.preferredLanguages = null
+  opt.normalizedOptionLanguages = normalizePreferredLanguages(opt.languages, opt.markRegState.languages)
+  opt.shouldResolvePreferredLanguages = needsPreferredLanguagesResolution(opt)
   opt.htmlWrapWithoutCaption = {
     iframe: !!opt.iframeWithoutCaption,
     video: !!opt.videoWithoutCaption,
@@ -1183,21 +1139,33 @@ const mditFigureWithPCaption = (md, option) => {
   const classPrefix = buildClassPrefix(opt.classPrefix)
   opt.figureClassPrefix = classPrefix
   opt.captionClassPrefix = classPrefix
+  const defaultIframeTypeBlockquoteClass = classPrefix + 'img'
+  const defaultSlideFigureClass = classPrefix + 'slide'
   if (!hasExplicitFigureClassThatWrapsIframeTypeBlockquote) {
-    opt.figureClassThatWrapsIframeTypeBlockquote = classPrefix + 'img'
+    opt.figureClassThatWrapsIframeTypeBlockquote = defaultIframeTypeBlockquoteClass
+  } else {
+    opt.figureClassThatWrapsIframeTypeBlockquote = normalizeClassOptionWithFallback(
+      opt.figureClassThatWrapsIframeTypeBlockquote,
+      defaultIframeTypeBlockquoteClass,
+    )
   }
   if (!hasExplicitFigureClassThatWrapsSlides) {
-    opt.figureClassThatWrapsSlides = classPrefix + 'slide'
+    opt.figureClassThatWrapsSlides = defaultSlideFigureClass
+  } else {
+    opt.figureClassThatWrapsSlides = normalizeClassOptionWithFallback(
+      opt.figureClassThatWrapsSlides,
+      defaultSlideFigureClass,
+    )
   }
   // Precompute label-class permutations so numbering lookup doesn't rebuild arrays per caption.
   opt.labelClassLookup = buildLabelClassLookup(opt)
   const markerList = normalizeLabelPrefixMarkers(opt.labelPrefixMarker)
-  opt.labelPrefixMarkerReg = buildLabelPrefixMarkerRegFromList(markerList)
+  opt.labelPrefixMarkerReg = buildLabelPrefixMarkerRegFromMarkers(markerList)
   opt.cleanCaptionRegCache = new Map()
   if (opt.allowLabelPrefixMarkerWithoutLabel === true) {
     const markerPair = resolveLabelPrefixMarkerPair(markerList)
-    opt.labelPrefixMarkerWithoutLabelPrevReg = buildLabelPrefixMarkerRegFromList(markerPair.prev)
-    opt.labelPrefixMarkerWithoutLabelNextReg = buildLabelPrefixMarkerRegFromList(markerPair.next)
+    opt.labelPrefixMarkerWithoutLabelPrevReg = buildLabelPrefixMarkerRegFromMarkers(markerPair.prev)
+    opt.labelPrefixMarkerWithoutLabelNextReg = buildLabelPrefixMarkerRegFromMarkers(markerPair.next)
   } else {
     opt.labelPrefixMarkerWithoutLabelPrevReg = null
     opt.labelPrefixMarkerWithoutLabelNextReg = null

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@peaceroad/markdown-it-figure-with-p-caption",
-  "version": "0.16.0",
+  "version": "0.17.0",
   "description": "A markdown-it plugin. For a paragraph with only one image, a table or code block or blockquote, and by writing a caption paragraph immediately before or after, they are converted into the figure element with the figcaption element.",
   "main": "index.js",
   "type": "module",
@@ -20,19 +20,20 @@
     "url": "https://github.com/peaceroad/p7d-markdown-it-figure-with-p-caption/issues"
   },
   "devDependencies": {
-    "@peaceroad/markdown-it-cjk-breaks-mod": "^0.1.6",
-    "@peaceroad/markdown-it-renderer-fence": "^0.4.1",
-    "@peaceroad/markdown-it-renderer-image": "^0.9.1",
-    "@peaceroad/markdown-it-strong-ja": "^0.7.2",
+    "@peaceroad/markdown-it-cjk-breaks-mod": "^0.1.10",
+    "@peaceroad/markdown-it-renderer-fence": "^0.6.1",
+    "@peaceroad/markdown-it-renderer-image": "^0.13.0",
+    "@peaceroad/markdown-it-strong-ja": "^0.9.0",
     "highlight.js": "^11.11.1",
     "markdown-it": "^14.1.0",
     "markdown-it-attrs": "^4.3.1"
   },
   "dependencies": {
-    "p7d-markdown-it-p-captions": "^0.21.0"
+    "p7d-markdown-it-p-captions": "0.22.0"
   },
   "files": [
     "index.js",
+    "embeds/",
     "README.md",
     "LICENSE"
   ]