docrev 0.11.1 → 0.11.3

package/lib/slides.ts

@@ -1,756 +1,756 @@
-/**
- * Slide processing for Beamer and PPTX output
- *
- * Handles:
- * - ::: step blocks for incremental reveals
- * - ::: buildup blocks for progressive bullet reveals with greying
- * - ::: notes blocks for speaker notes
- * - Slide boundaries (---)
- * - Slide styles: {.dark}, {.light}, {.accent}, {.inverse}
- * - Special slides: {.cover}, {.thanks}, {.section}, {.plain}
- *
- * Syntax examples:
- *   ## Title {.dark}           - Dark background slide
- *   ## Welcome {.cover}        - Cover slide (no numbering, centered)
- *   ## Thank You {.thanks}     - Thanks slide (no numbering)
- *   # Part 1 {.section}        - Section divider slide
- *   ## Image {.plain}          - No header/footer, full content
- *   ## Highlight {.accent .nonumber}  - Accent color, no slide number
- *
- * Buildup syntax:
- *   ::: buildup
- *   - First point
- *     - Sub A
- *     - Sub B
- *   - Second point
- *   :::
- *
- * Generates slides where current point is colored, previous are greyed out.
- * Subpoints appear sequentially within their parent.
- */
-
-interface Step {
-  index: number;
-  content: string;
-}
-
-interface SlideStyle {
-  background: string | null;
-  type: string | null;
-  nonumber: boolean;
-  center: boolean;
-  classes: string[];
-}
-
-interface Slide {
-  title: string;
-  titleLevel: number;
-  steps: Step[];
-  notes: string | null;
-  preamble: string;
-  style: SlideStyle;
-  _frontmatter?: string;
-}
-
-interface BulletItem {
-  text: string;
-  indent: number;
-  children: BulletItem[];
-}
-
-interface BuildupStep {
-  itemIndex: number;
-  subIndex: number | null;
-}
-
-interface BuildupResult {
-  content: string;
-  afterContent?: string;
-  buildupSteps: string[] | null;
-}
-
-interface StepWithBuildup extends Step {
-  beforeBuildup?: string;
-  afterBuildup?: string;
-  buildupSteps?: string[] | null;
-}
-
-/**
- * Grey color for "completed" buildup items
- */
-const GREY_COLOR = '#888888';
-
-/**
- * Background styles mapped to Beamer options
- */
-const BEAMER_BACKGROUNDS: Record<string, string> = {
-  dark: '\\setbeamercolor{background canvas}{bg=black}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
-  light: '\\setbeamercolor{background canvas}{bg=white}\\setbeamercolor{normal text}{fg=black}\\usebeamercolor[fg]{normal text}',
-  accent: '\\setbeamercolor{background canvas}{bg=structure.fg}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
-  inverse: '\\setbeamercolor{background canvas}{bg=structure.fg!90!black}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
-};
-
-/**
- * Parse a bullet list into a tree structure
- */
-function parseBulletList(content: string): BulletItem[] {
-  const lines = content.split('\n');
-  const items: BulletItem[] = [];
-  const stack: Array<BulletItem & { children: BulletItem[] }> = [{ children: items, indent: -1 } as any];
-
-  for (const line of lines) {
-    // Match bullet lines: "- text" or "  - text" etc.
-    const match = line.match(/^(\s*)[-*]\s+(.+)$/);
-    if (!match) continue;
-
-    const indent = (match[1] || '').length;
-    const text = (match[2] || '').trim();
-
-    const item: BulletItem = { text, indent, children: [] };
-
-    // Find parent based on indentation
-    while (stack.length > 1) {
-      const top = stack[stack.length - 1];
-      if (!top || top.indent < indent) break;
-      stack.pop();
-    }
-
-    const parent = stack[stack.length - 1];
-    if (parent) {
-      parent.children.push(item);
-    }
-    stack.push(item as any);
-  }
-
-  return items;
-}
-
-/**
- * Flatten bullet tree into sequential reveal steps
- * Each step is: { itemIndex, subIndex, isSubItem }
- */
-function flattenBuildupSteps(items: BulletItem[]): BuildupStep[] {
-  const steps: BuildupStep[] = [];
-
-  for (let i = 0; i < items.length; i++) {
-    const item = items[i];
-    if (!item) continue;
-
-    if (item.children.length === 0) {
-      // No children - single step for this item
-      steps.push({ itemIndex: i, subIndex: null });
-    } else {
-      // Has children - first show parent, then each child
-      steps.push({ itemIndex: i, subIndex: -1 }); // Parent only
-
-      for (let j = 0; j < item.children.length; j++) {
-        steps.push({ itemIndex: i, subIndex: j });
-      }
-    }
-  }
-
-  return steps;
-}
-
-/**
- * Render a bullet item with optional color
- */
-function renderBulletItem(
-  item: BulletItem,
-  color: string | null,
-  indentLevel: number = 0,
-  showChildrenUpTo: number | null = null
-): string {
-  const indent = '  '.repeat(indentLevel);
-  const lines: string[] = [];
-
-  // Render the main item
-  const text = color ? `[${item.text}]{color=${color}}` : item.text;
-  lines.push(`${indent}- ${text}`);
-
-  // Render children if any should be shown
-  if (showChildrenUpTo !== null && showChildrenUpTo >= 0 && item.children) {
-    for (let i = 0; i <= showChildrenUpTo && i < item.children.length; i++) {
-      const child = item.children[i];
-      if (!child) continue;
-      const childColor = color; // Children inherit parent's color state
-      const childText = childColor ? `[${child.text}]{color=${childColor}}` : child.text;
-      lines.push(`${indent}  - ${childText}`);
-    }
-  }
-
-  return lines.join('\n');
-}
-
-/**
- * Generate buildup slides from a bullet list
- */
-function expandBuildup(content: string, format: string): string[] {
-  const items = parseBulletList(content);
-  if (items.length === 0) return [content];
-
-  const steps = flattenBuildupSteps(items);
-  const slideContents: string[] = [];
-
-  for (const step of steps) {
-    const lines: string[] = [];
-
-    for (let i = 0; i < items.length; i++) {
-      const item = items[i];
-      if (!item) continue;
-
-      if (i < step.itemIndex) {
-        // Previous top-level item - greyed out with all children
-        const allChildrenIndex = item.children.length > 0 ? item.children.length - 1 : null;
-        lines.push(renderBulletItem(item, GREY_COLOR, 0, allChildrenIndex));
-      } else if (i === step.itemIndex) {
-        // Current top-level item - colored
-        if (step.subIndex === null) {
-          // No children case - just show item
-          lines.push(renderBulletItem(item, null, 0, null));
-        } else if (step.subIndex === -1) {
-          // Has children but showing parent only first
-          lines.push(renderBulletItem(item, null, 0, null));
-        } else {
-          // Showing parent + children up to subIndex
-          lines.push(renderBulletItem(item, null, 0, step.subIndex));
-        }
-      }
-      // Future items (i > step.itemIndex) - not shown yet
-    }
-
-    slideContents.push(lines.join('\n'));
-  }
-
-  return slideContents;
-}
-
-/**
- * Process ::: buildup blocks in content and expand to steps
- */
-function processBuildupBlocks(content: string, format: string): BuildupResult {
-  const buildupMatch = content.match(/^:::\s*buildup\s*\n([\s\S]*?)\n:::\s*$/m);
-
-  if (!buildupMatch || buildupMatch.index === undefined) {
-    return { content, buildupSteps: null };
-  }
-
-  const buildupContent = buildupMatch[1] || '';
-  const buildupSteps = expandBuildup(buildupContent, format);
-
-  // Remove the buildup block from content (will be replaced by steps)
-  const beforeBuildup = content.slice(0, buildupMatch.index).trim();
-  const afterBuildup = content.slice(buildupMatch.index + buildupMatch[0].length).trim();
-
-  return {
-    content: beforeBuildup,
-    afterContent: afterBuildup,
-    buildupSteps,
-  };
-}
-
-/**
- * Parse slide style attributes from heading
- */
-function parseSlideStyle(heading: string): { title: string; style: SlideStyle } {
-  const style: SlideStyle = {
-    background: null,
-    type: null,
-    nonumber: false,
-    center: false,
-    classes: [],
-  };
-
-  // Match {.class1 .class2} at end of heading
-  const attrMatch = heading.match(/\s*\{([^}]+)\}\s*$/);
-  if (!attrMatch || attrMatch.index === undefined) {
-    return { title: heading.trim(), style };
-  }
-
-  const title = heading.slice(0, attrMatch.index).trim();
-  const attrs = attrMatch[1] || '';
-
-  // Parse each .class
-  const classMatches = attrs.matchAll(/\.(\w+)/g);
-  for (const match of classMatches) {
-    if (!match[1]) continue;
-    const cls = match[1].toLowerCase();
-    style.classes.push(cls);
-
-    // Background styles
-    if (['dark', 'light', 'accent', 'inverse'].includes(cls)) {
-      style.background = cls;
-    }
-    // Special slide types
-    else if (['cover', 'thanks', 'section', 'plain'].includes(cls)) {
-      style.type = cls;
-      // Cover, thanks, section slides default to no numbering
-      if (['cover', 'thanks', 'section'].includes(cls)) {
-        style.nonumber = true;
-        style.center = true;
-      }
-    }
-    // Explicit options
-    else if (cls === 'nonumber' || cls === 'unnumbered') {
-      style.nonumber = true;
-    } else if (cls === 'center' || cls === 'centered') {
-      style.center = true;
-    }
-  }
-
-  return { title, style };
-}
-
-/**
- * Parse a single slide's content into steps and notes
- */
-export function parseSlide(slideContent: string): Slide {
-  const lines = slideContent.split('\n');
-
-  // Extract title (first heading)
-  let title = '';
-  let titleLevel = 2;
-  let titleLineIndex = -1;
-  let style: SlideStyle = {
-    background: null,
-    type: null,
-    nonumber: false,
-    center: false,
-    classes: [],
-  };
-
-  for (let i = 0; i < lines.length; i++) {
-    const line = lines[i];
-    if (!line) continue;
-    const match = line.match(/^(#{1,6})\s+(.+)$/);
-    if (match && match[1] && match[2]) {
-      titleLevel = match[1].length;
-      // Parse style from heading
-      const parsed = parseSlideStyle(match[2]);
-      title = parsed.title;
-      style = parsed.style;
-      titleLineIndex = i;
-      break;
-    }
-  }
-
-  // Content after title
-  const contentStart = titleLineIndex >= 0 ? titleLineIndex + 1 : 0;
-  const bodyContent = lines.slice(contentStart).join('\n');
-
-  // Parse ::: step and ::: notes blocks
-  const steps: Step[] = [];
-  let notes: string | null = null;
-  let preamble = '';
-
-  // Regex to match fenced div blocks
-  const bodyLines = bodyContent.split('\n');
-  let inBlock = false;
-  let blockType: string | null = null;
-  let currentBlockContent: string[] = [];
-  let beforeFirstStep: string[] = [];
-  let foundFirstStep = false;
-
-  for (let i = 0; i < bodyLines.length; i++) {
-    const line = bodyLines[i];
-    if (!line && line !== '') continue;
-
-    if (!inBlock) {
-      // Check for block start
-      const stepMatch = line.match(/^:::\s*step\s*$/);
-      const notesMatch = line.match(/^:::\s*notes\s*$/);
-
-      if (stepMatch) {
-        inBlock = true;
-        blockType = 'step';
-        currentBlockContent = [];
-        foundFirstStep = true;
-      } else if (notesMatch) {
-        inBlock = true;
-        blockType = 'notes';
-        currentBlockContent = [];
-      } else if (!foundFirstStep) {
-        beforeFirstStep.push(line);
-      }
-    } else {
-      // Check for block end
-      if (line.match(/^:::\s*$/)) {
-        // End of block
-        if (blockType === 'step') {
-          steps.push({
-            index: steps.length + 1,
-            content: currentBlockContent.join('\n').trim(),
-          });
-        } else if (blockType === 'notes') {
-          notes = currentBlockContent.join('\n').trim();
-        }
-        inBlock = false;
-        blockType = null;
-        currentBlockContent = [];
-      } else {
-        currentBlockContent.push(line);
-      }
-    }
-  }
-
-  // Handle content before first step as preamble
-  preamble = beforeFirstStep.join('\n').trim();
-
-  // If no explicit steps, treat entire body as single step
-  if (steps.length === 0) {
-    // Remove notes from body if present
-    let bodyWithoutNotes = bodyContent;
-    const notesBlockMatch = bodyContent.match(/^:::\s*notes\s*$[\s\S]*?^:::\s*$/m);
-    if (notesBlockMatch) {
-      bodyWithoutNotes = bodyContent.replace(notesBlockMatch[0], '').trim();
-    }
-
-    steps.push({
-      index: 1,
-      content: bodyWithoutNotes.trim(),
-    });
-    preamble = '';
-  }
-
-  return {
-    title,
-    titleLevel,
-    steps,
-    notes,
-    preamble,
-    style,
-  };
-}
-
-/**
- * Parse markdown document into slides
- */
-export function parseSlides(markdown: string): Slide[] {
-  // Normalize line endings to \n
-  const normalized = markdown.replace(/\r\n/g, '\n');
-
-  // Split by --- (horizontal rule / slide delimiter)
-  // Handle YAML frontmatter by checking for --- at start
-  let content = normalized;
-  let frontmatter: string | null = null;
-
-  // Extract YAML frontmatter if present
-  if (normalized.startsWith('---')) {
-    const endMatch = normalized.slice(3).indexOf('\n---');
-    if (endMatch !== -1) {
-      frontmatter = normalized.slice(0, endMatch + 7); // Include both ---
-      content = normalized.slice(endMatch + 7).trim();
-    }
-  }
-
-  // Split remaining content by ---
-  const parts = content.split(/\n---\n/);
-
-  const slides: Slide[] = [];
-  for (const part of parts) {
-    if (!part) continue;
-    const trimmed = part.trim();
-    if (trimmed) {
-      slides.push(parseSlide(trimmed));
-    }
-  }
-
-  // Attach frontmatter to first slide's preamble if exists
-  if (frontmatter && slides.length > 0 && slides[0]) {
-    slides[0]._frontmatter = frontmatter;
-  }
-
-  return slides;
-}
-
-/**
- * Generate Beamer markdown using pandoc's native slide structure
- * Works WITH pandoc, not against it - pandoc creates frames, we add overlays
- */
-export function generateBeamerMarkdown(slides: Slide[]): string {
-  const output: string[] = [];
-
-  // Check if first slide is a cover slide - if so, skip pandoc's auto title page
-  const hasExplicitCover = slides.length > 0 && slides[0] && slides[0].style && slides[0].style.type === 'cover';
-
-  for (const slide of slides) {
-    // Include frontmatter if present
-    if (slide._frontmatter) {
-      let frontmatter = slide._frontmatter;
-
-      // If we have an explicit cover slide, remove title/author/date to prevent
-      // pandoc from generating a duplicate title frame
-      if (hasExplicitCover) {
-        // Remove title, author, date lines but keep other frontmatter
-        frontmatter = frontmatter
-          .replace(/^title:.*\n?/m, '')
-          .replace(/^author:.*\n?/m, '')
-          .replace(/^date:.*\n?/m, '')
-          .replace(/\n{2,}/g, '\n'); // Clean up extra blank lines
-
-        // Check if frontmatter is now empty (just --- and ---)
-        const content = frontmatter.replace(/---/g, '').trim();
-        if (!content) {
-          // Skip empty frontmatter entirely - don't output anything
-          // The slide content will follow directly
-        } else {
-          output.push(frontmatter);
-          output.push('');
-        }
-      } else {
-        output.push(frontmatter);
-        output.push('');
-      }
-    }
-
-    // Build pandoc heading with beamer attributes
-    let headingAttrs = '';
-    const attrList: string[] = [];
-
-    if (slide.style) {
-      if (slide.style.type === 'cover' || slide.style.type === 'thanks' || slide.style.type === 'section') {
-        attrList.push('.plain');
-        attrList.push('.noframenumbering');
-        attrList.push('.c');
-      } else if (slide.style.type === 'plain') {
-        attrList.push('.plain');
-      }
-      if (slide.style.nonumber && !attrList.includes('.noframenumbering')) {
-        attrList.push('.noframenumbering');
-      }
-      if (slide.style.center && !attrList.includes('.c')) {
-        attrList.push('.c');
-      }
-    }
-
-    if (attrList.length > 0) {
-      headingAttrs = ' {' + attrList.join(' ') + '}';
-    }
-
-    // Frame heading - pandoc creates the frame from this
-    const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + headingAttrs;
-    output.push(heading);
-    output.push('');
-
-    // Note: Per-frame background colors (.dark, .accent) are not yet supported
-    // in the pandoc-based beamer output. The classes are preserved for PPTX.
-
-    // Add preamble if present (visible on all overlays)
-    if (slide.preamble) {
-      output.push(slide.preamble);
-      output.push('');
-    }
-
-    // Generate content
-    if (slide.steps.length === 1 && slide.steps[0]) {
-      // Single step - just output content, no overlays needed
-      output.push(slide.steps[0].content);
-    } else {
-      // Multiple steps - use \pause between steps
-      // This is the pandoc-friendly way to do incremental reveals
-      for (let i = 0; i < slide.steps.length; i++) {
-        const step = slide.steps[i];
-        if (!step) continue;
-        output.push(step.content);
-        output.push('');
-        // Add pause after each step except the last
-        if (i < slide.steps.length - 1) {
-          output.push('. . .');
-          output.push('');
-        }
-      }
-    }
-
-    // Add speaker notes as LaTeX \note{} command
-    if (slide.notes) {
-      output.push('');
-      // Escape special LaTeX characters in notes
-      const notes = slide.notes || '';
-      const escapedNotes = notes
-        .replace(/\\/g, '\\textbackslash{}')
-        .replace(/[&%$#_{}]/g, '\\$&')
-        .replace(/\n/g, '\\\\');
-      output.push('\\note{' + escapedNotes + '}');
-    }
-
-    output.push('');
-    output.push('---');
-    output.push('');
-  }
-
-  // Remove trailing ---
-  while (output.length > 0) {
-    const last = output[output.length - 1];
-    if (!last || last.trim() !== '') break;
-    output.pop();
-  }
-  if (output.length > 0 && output[output.length - 1] === '---') {
-    output.pop();
-  }
-
-  return output.join('\n');
-}
-
-/**
- * Build PPTX slide class attribute string
- */
-function buildPptxSlideClasses(style: SlideStyle | null): string {
-  if (!style || !style.classes || style.classes.length === 0) {
-    return '';
-  }
-  return ' {.' + style.classes.join(' .') + '}';
-}
-
-/**
- * Generate PPTX markdown with duplicated slides for steps
- * Each step becomes a separate physical slide
- * Handles ::: buildup blocks by expanding them into multiple slides
- */
-export function generatePptxMarkdown(slides: Slide[]): string {
-  const output: string[] = [];
-
-  for (const slide of slides) {
-    // Include frontmatter if present (only on first slide)
-    if (slide._frontmatter) {
-      output.push(slide._frontmatter);
-      output.push('');
-    }
-
-    // Check if any step contains a buildup block
-    const stepsWithBuildup: StepWithBuildup[] = slide.steps.map((step) => {
-      const result = processBuildupBlocks(step.content, 'pptx');
-      return {
-        ...step,
-        beforeBuildup: result.content,
-        afterBuildup: result.afterContent || '',
-        buildupSteps: result.buildupSteps,
-      };
-    });
-
-    // Generate slides - handle buildup expansion
-    for (let i = 0; i < stepsWithBuildup.length; i++) {
-      const step = stepsWithBuildup[i];
-      if (!step) continue;
-
-      if (step.buildupSteps && step.buildupSteps.length > 0) {
-        // This step has a buildup block - generate one slide per buildup step
-        for (const buildupContent of step.buildupSteps) {
-          const classes = buildPptxSlideClasses(slide.style);
-          const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + classes;
-          output.push(heading);
-          output.push('');
-
-          // Add preamble if present
-          if (slide.preamble) {
-            output.push(slide.preamble);
-            output.push('');
-          }
-
-          // Add content before buildup block
-          if (step.beforeBuildup) {
-            output.push(step.beforeBuildup);
-            output.push('');
-          }
-
-          // Add this buildup step content
-          output.push(buildupContent);
-          output.push('');
-
-          // Add content after buildup block
-          if (step.afterBuildup) {
-            output.push(step.afterBuildup);
-            output.push('');
-          }
-
-          // Add speaker notes
-          if (slide.notes) {
-            output.push('::: notes');
-            output.push(slide.notes);
-            output.push(':::');
-          }
-
-          output.push('');
-          output.push('---');
-          output.push('');
-        }
-      } else {
-        // Regular step - no buildup
-        const classes = buildPptxSlideClasses(slide.style);
-        const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + classes;
-        output.push(heading);
-        output.push('');
-
-        // Add preamble if present
-        if (slide.preamble) {
-          output.push(slide.preamble);
-          output.push('');
-        }
-
-        // Add cumulative steps up to current
-        for (let j = 0; j <= i; j++) {
-          const s = stepsWithBuildup[j];
-          if (s) {
-            output.push(s.content);
-            output.push('');
-          }
-        }
-
-        // Add speaker notes
-        if (slide.notes) {
-          output.push('::: notes');
-          output.push(slide.notes);
-          output.push(':::');
-        }
-
-        output.push('');
-        output.push('---');
-        output.push('');
-      }
-    }
-  }
-
-  // Remove trailing ---
-  while (output.length > 0) {
-    const last = output[output.length - 1];
-    if (!last || last.trim() !== '') break;
-    output.pop();
-  }
-  if (output.length > 0 && output[output.length - 1] === '---') {
-    output.pop();
-  }
-
-  return output.join('\n');
-}
-
-/**
- * Process markdown for slide output format
- */
-export function processSlideMarkdown(markdown: string, format: 'beamer' | 'pptx'): string {
-  const slides = parseSlides(markdown);
-
-  if (format === 'beamer') {
-    return generateBeamerMarkdown(slides);
-  } else if (format === 'pptx') {
-    return generatePptxMarkdown(slides);
-  }
-
-  return markdown;
-}
-
-/**
- * Check if markdown contains slide syntax (steps, notes, buildup, or slide styles)
- */
-export function hasSlideSyntax(markdown: string): boolean {
-  // Check for ::: step, ::: notes, or ::: buildup
-  if (/^:::\s*(step|notes|buildup)\s*$/m.test(markdown)) {
-    return true;
-  }
-  // Check for slide style attributes on headings: ## Title {.dark}
-  if (/^#{1,6}\s+.+\{[^}]*\.(dark|light|accent|inverse|cover|thanks|section|plain|nonumber|center)/m.test(markdown)) {
-    return true;
-  }
-  return false;
-}
-
-// Export style parser for testing
-export { parseSlideStyle };
+/**
+ * Slide processing for Beamer and PPTX output
+ *
+ * Handles:
+ * - ::: step blocks for incremental reveals
+ * - ::: buildup blocks for progressive bullet reveals with greying
+ * - ::: notes blocks for speaker notes
+ * - Slide boundaries (---)
+ * - Slide styles: {.dark}, {.light}, {.accent}, {.inverse}
+ * - Special slides: {.cover}, {.thanks}, {.section}, {.plain}
+ *
+ * Syntax examples:
+ *   ## Title {.dark}           - Dark background slide
+ *   ## Welcome {.cover}        - Cover slide (no numbering, centered)
+ *   ## Thank You {.thanks}     - Thanks slide (no numbering)
+ *   # Part 1 {.section}        - Section divider slide
+ *   ## Image {.plain}          - No header/footer, full content
+ *   ## Highlight {.accent .nonumber}  - Accent color, no slide number
+ *
+ * Buildup syntax:
+ *   ::: buildup
+ *   - First point
+ *     - Sub A
+ *     - Sub B
+ *   - Second point
+ *   :::
+ *
+ * Generates slides where current point is colored, previous are greyed out.
+ * Subpoints appear sequentially within their parent.
+ */
+
+interface Step {
+  index: number;
+  content: string;
+}
+
+interface SlideStyle {
+  background: string | null;
+  type: string | null;
+  nonumber: boolean;
+  center: boolean;
+  classes: string[];
+}
+
+interface Slide {
+  title: string;
+  titleLevel: number;
+  steps: Step[];
+  notes: string | null;
+  preamble: string;
+  style: SlideStyle;
+  _frontmatter?: string;
+}
+
+interface BulletItem {
+  text: string;
+  indent: number;
+  children: BulletItem[];
+}
+
+interface BuildupStep {
+  itemIndex: number;
+  subIndex: number | null;
+}
+
+interface BuildupResult {
+  content: string;
+  afterContent?: string;
+  buildupSteps: string[] | null;
+}
+
+interface StepWithBuildup extends Step {
+  beforeBuildup?: string;
+  afterBuildup?: string;
+  buildupSteps?: string[] | null;
+}
+
+/**
+ * Grey color for "completed" buildup items
+ */
+const GREY_COLOR = '#888888';
+
+/**
+ * Background styles mapped to Beamer options
+ */
+const BEAMER_BACKGROUNDS: Record<string, string> = {
+  dark: '\\setbeamercolor{background canvas}{bg=black}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
+  light: '\\setbeamercolor{background canvas}{bg=white}\\setbeamercolor{normal text}{fg=black}\\usebeamercolor[fg]{normal text}',
+  accent: '\\setbeamercolor{background canvas}{bg=structure.fg}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
+  inverse: '\\setbeamercolor{background canvas}{bg=structure.fg!90!black}\\setbeamercolor{normal text}{fg=white}\\usebeamercolor[fg]{normal text}',
+};
+
+/**
+ * Parse a bullet list into a tree structure
+ */
+function parseBulletList(content: string): BulletItem[] {
+  const lines = content.split('\n');
+  const items: BulletItem[] = [];
+  const stack: Array<BulletItem & { children: BulletItem[] }> = [{ children: items, indent: -1 } as any];
+
+  for (const line of lines) {
+    // Match bullet lines: "- text" or "  - text" etc.
+    const match = line.match(/^(\s*)[-*]\s+(.+)$/);
+    if (!match) continue;
+
+    const indent = (match[1] || '').length;
+    const text = (match[2] || '').trim();
+
+    const item: BulletItem = { text, indent, children: [] };
+
+    // Find parent based on indentation
+    while (stack.length > 1) {
+      const top = stack[stack.length - 1];
+      if (!top || top.indent < indent) break;
+      stack.pop();
+    }
+
+    const parent = stack[stack.length - 1];
+    if (parent) {
+      parent.children.push(item);
+    }
+    stack.push(item as any);
+  }
+
+  return items;
+}
+
+/**
+ * Flatten bullet tree into sequential reveal steps
+ * Each step is: { itemIndex, subIndex, isSubItem }
+ */
+function flattenBuildupSteps(items: BulletItem[]): BuildupStep[] {
+  const steps: BuildupStep[] = [];
+
+  for (let i = 0; i < items.length; i++) {
+    const item = items[i];
+    if (!item) continue;
+
+    if (item.children.length === 0) {
+      // No children - single step for this item
+      steps.push({ itemIndex: i, subIndex: null });
+    } else {
+      // Has children - first show parent, then each child
+      steps.push({ itemIndex: i, subIndex: -1 }); // Parent only
+
+      for (let j = 0; j < item.children.length; j++) {
+        steps.push({ itemIndex: i, subIndex: j });
+      }
+    }
+  }
+
+  return steps;
+}
+
+/**
+ * Render a bullet item with optional color
+ */
+function renderBulletItem(
+  item: BulletItem,
+  color: string | null,
+  indentLevel: number = 0,
+  showChildrenUpTo: number | null = null
+): string {
+  const indent = '  '.repeat(indentLevel);
+  const lines: string[] = [];
+
+  // Render the main item
+  const text = color ? `[${item.text}]{color=${color}}` : item.text;
+  lines.push(`${indent}- ${text}`);
+
+  // Render children if any should be shown
+  if (showChildrenUpTo !== null && showChildrenUpTo >= 0 && item.children) {
+    for (let i = 0; i <= showChildrenUpTo && i < item.children.length; i++) {
+      const child = item.children[i];
+      if (!child) continue;
+      const childColor = color; // Children inherit parent's color state
+      const childText = childColor ? `[${child.text}]{color=${childColor}}` : child.text;
+      lines.push(`${indent}  - ${childText}`);
+    }
+  }
+
+  return lines.join('\n');
+}
+
+/**
+ * Generate buildup slides from a bullet list
+ */
+function expandBuildup(content: string, format: string): string[] {
+  const items = parseBulletList(content);
+  if (items.length === 0) return [content];
+
+  const steps = flattenBuildupSteps(items);
+  const slideContents: string[] = [];
+
+  for (const step of steps) {
+    const lines: string[] = [];
+
+    for (let i = 0; i < items.length; i++) {
+      const item = items[i];
+      if (!item) continue;
+
+      if (i < step.itemIndex) {
+        // Previous top-level item - greyed out with all children
+        const allChildrenIndex = item.children.length > 0 ? item.children.length - 1 : null;
+        lines.push(renderBulletItem(item, GREY_COLOR, 0, allChildrenIndex));
+      } else if (i === step.itemIndex) {
+        // Current top-level item - colored
+        if (step.subIndex === null) {
+          // No children case - just show item
+          lines.push(renderBulletItem(item, null, 0, null));
+        } else if (step.subIndex === -1) {
+          // Has children but showing parent only first
+          lines.push(renderBulletItem(item, null, 0, null));
+        } else {
+          // Showing parent + children up to subIndex
+          lines.push(renderBulletItem(item, null, 0, step.subIndex));
+        }
+      }
+      // Future items (i > step.itemIndex) - not shown yet
+    }
+
+    slideContents.push(lines.join('\n'));
+  }
+
+  return slideContents;
+}
+
+/**
+ * Process ::: buildup blocks in content and expand to steps
+ */
+function processBuildupBlocks(content: string, format: string): BuildupResult {
+  const buildupMatch = content.match(/^:::\s*buildup\s*\n([\s\S]*?)\n:::\s*$/m);
+
+  if (!buildupMatch || buildupMatch.index === undefined) {
+    return { content, buildupSteps: null };
+  }
+
+  const buildupContent = buildupMatch[1] || '';
+  const buildupSteps = expandBuildup(buildupContent, format);
+
+  // Remove the buildup block from content (will be replaced by steps)
+  const beforeBuildup = content.slice(0, buildupMatch.index).trim();
+  const afterBuildup = content.slice(buildupMatch.index + buildupMatch[0].length).trim();
+
+  return {
+    content: beforeBuildup,
+    afterContent: afterBuildup,
+    buildupSteps,
+  };
+}
+
+/**
+ * Parse slide style attributes from heading
+ */
+function parseSlideStyle(heading: string): { title: string; style: SlideStyle } {
+  const style: SlideStyle = {
+    background: null,
+    type: null,
+    nonumber: false,
+    center: false,
+    classes: [],
+  };
+
+  // Match {.class1 .class2} at end of heading
+  const attrMatch = heading.match(/\s*\{([^}]+)\}\s*$/);
+  if (!attrMatch || attrMatch.index === undefined) {
+    return { title: heading.trim(), style };
+  }
+
+  const title = heading.slice(0, attrMatch.index).trim();
+  const attrs = attrMatch[1] || '';
+
+  // Parse each .class
+  const classMatches = attrs.matchAll(/\.(\w+)/g);
+  for (const match of classMatches) {
+    if (!match[1]) continue;
+    const cls = match[1].toLowerCase();
+    style.classes.push(cls);
+
+    // Background styles
+    if (['dark', 'light', 'accent', 'inverse'].includes(cls)) {
+      style.background = cls;
+    }
+    // Special slide types
+    else if (['cover', 'thanks', 'section', 'plain'].includes(cls)) {
+      style.type = cls;
+      // Cover, thanks, section slides default to no numbering
+      if (['cover', 'thanks', 'section'].includes(cls)) {
+        style.nonumber = true;
+        style.center = true;
+      }
+    }
+    // Explicit options
+    else if (cls === 'nonumber' || cls === 'unnumbered') {
+      style.nonumber = true;
+    } else if (cls === 'center' || cls === 'centered') {
+      style.center = true;
+    }
+  }
+
+  return { title, style };
+}
+
+/**
+ * Parse a single slide's content into steps and notes
+ */
+export function parseSlide(slideContent: string): Slide {
+  const lines = slideContent.split('\n');
+
+  // Extract title (first heading)
+  let title = '';
+  let titleLevel = 2;
+  let titleLineIndex = -1;
+  let style: SlideStyle = {
+    background: null,
+    type: null,
+    nonumber: false,
+    center: false,
+    classes: [],
+  };
+
+  for (let i = 0; i < lines.length; i++) {
+    const line = lines[i];
+    if (!line) continue;
+    const match = line.match(/^(#{1,6})\s+(.+)$/);
+    if (match && match[1] && match[2]) {
+      titleLevel = match[1].length;
+      // Parse style from heading
+      const parsed = parseSlideStyle(match[2]);
+      title = parsed.title;
+      style = parsed.style;
+      titleLineIndex = i;
+      break;
+    }
+  }
+
+  // Content after title
+  const contentStart = titleLineIndex >= 0 ? titleLineIndex + 1 : 0;
+  const bodyContent = lines.slice(contentStart).join('\n');
+
+  // Parse ::: step and ::: notes blocks
+  const steps: Step[] = [];
+  let notes: string | null = null;
+  let preamble = '';
+
+  // Regex to match fenced div blocks
+  const bodyLines = bodyContent.split('\n');
+  let inBlock = false;
+  let blockType: string | null = null;
+  let currentBlockContent: string[] = [];
+  let beforeFirstStep: string[] = [];
+  let foundFirstStep = false;
+
+  for (let i = 0; i < bodyLines.length; i++) {
+    const line = bodyLines[i];
+    if (!line && line !== '') continue;
+
+    if (!inBlock) {
+      // Check for block start
+      const stepMatch = line.match(/^:::\s*step\s*$/);
+      const notesMatch = line.match(/^:::\s*notes\s*$/);
+
+      if (stepMatch) {
+        inBlock = true;
+        blockType = 'step';
+        currentBlockContent = [];
+        foundFirstStep = true;
+      } else if (notesMatch) {
+        inBlock = true;
+        blockType = 'notes';
+        currentBlockContent = [];
+      } else if (!foundFirstStep) {
+        beforeFirstStep.push(line);
+      }
+    } else {
+      // Check for block end
+      if (line.match(/^:::\s*$/)) {
+        // End of block
+        if (blockType === 'step') {
+          steps.push({
+            index: steps.length + 1,
+            content: currentBlockContent.join('\n').trim(),
+          });
+        } else if (blockType === 'notes') {
+          notes = currentBlockContent.join('\n').trim();
+        }
+        inBlock = false;
+        blockType = null;
+        currentBlockContent = [];
+      } else {
+        currentBlockContent.push(line);
+      }
+    }
+  }
+
+  // Handle content before first step as preamble
+  preamble = beforeFirstStep.join('\n').trim();
+
+  // If no explicit steps, treat entire body as single step
+  if (steps.length === 0) {
+    // Remove notes from body if present
+    let bodyWithoutNotes = bodyContent;
+    const notesBlockMatch = bodyContent.match(/^:::\s*notes\s*$[\s\S]*?^:::\s*$/m);
+    if (notesBlockMatch) {
+      bodyWithoutNotes = bodyContent.replace(notesBlockMatch[0], '').trim();
+    }
+
+    steps.push({
+      index: 1,
+      content: bodyWithoutNotes.trim(),
+    });
+    preamble = '';
+  }
+
+  return {
+    title,
+    titleLevel,
+    steps,
+    notes,
+    preamble,
+    style,
+  };
+}
+
+/**
+ * Parse markdown document into slides
+ */
+export function parseSlides(markdown: string): Slide[] {
+  // Normalize line endings to \n
+  const normalized = markdown.replace(/\r\n/g, '\n');
+
+  // Split by --- (horizontal rule / slide delimiter)
+  // Handle YAML frontmatter by checking for --- at start
+  let content = normalized;
+  let frontmatter: string | null = null;
+
+  // Extract YAML frontmatter if present
+  if (normalized.startsWith('---')) {
+    const endMatch = normalized.slice(3).indexOf('\n---');
+    if (endMatch !== -1) {
+      frontmatter = normalized.slice(0, endMatch + 7); // Include both ---
+      content = normalized.slice(endMatch + 7).trim();
+    }
+  }
+
+  // Split remaining content by ---
+  const parts = content.split(/\n---\n/);
+
+  const slides: Slide[] = [];
+  for (const part of parts) {
+    if (!part) continue;
+    const trimmed = part.trim();
+    if (trimmed) {
+      slides.push(parseSlide(trimmed));
+    }
+  }
+
+  // Attach frontmatter to first slide's preamble if exists
+  if (frontmatter && slides.length > 0 && slides[0]) {
+    slides[0]._frontmatter = frontmatter;
+  }
+
+  return slides;
+}
+
+/**
+ * Generate Beamer markdown using pandoc's native slide structure
+ * Works WITH pandoc, not against it - pandoc creates frames, we add overlays
+ */
+export function generateBeamerMarkdown(slides: Slide[]): string {
+  const output: string[] = [];
+
+  // Check if first slide is a cover slide - if so, skip pandoc's auto title page
+  const hasExplicitCover = slides.length > 0 && slides[0] && slides[0].style && slides[0].style.type === 'cover';
+
+  for (const slide of slides) {
+    // Include frontmatter if present
+    if (slide._frontmatter) {
+      let frontmatter = slide._frontmatter;
+
+      // If we have an explicit cover slide, remove title/author/date to prevent
+      // pandoc from generating a duplicate title frame
+      if (hasExplicitCover) {
+        // Remove title, author, date lines but keep other frontmatter
+        frontmatter = frontmatter
+          .replace(/^title:.*\n?/m, '')
+          .replace(/^author:.*\n?/m, '')
+          .replace(/^date:.*\n?/m, '')
+          .replace(/\n{2,}/g, '\n'); // Clean up extra blank lines
+
+        // Check if frontmatter is now empty (just --- and ---)
+        const content = frontmatter.replace(/---/g, '').trim();
+        if (!content) {
+          // Skip empty frontmatter entirely - don't output anything
+          // The slide content will follow directly
+        } else {
+          output.push(frontmatter);
+          output.push('');
+        }
+      } else {
+        output.push(frontmatter);
+        output.push('');
+      }
+    }
+
+    // Build pandoc heading with beamer attributes
+    let headingAttrs = '';
+    const attrList: string[] = [];
+
+    if (slide.style) {
+      if (slide.style.type === 'cover' || slide.style.type === 'thanks' || slide.style.type === 'section') {
+        attrList.push('.plain');
+        attrList.push('.noframenumbering');
+        attrList.push('.c');
+      } else if (slide.style.type === 'plain') {
+        attrList.push('.plain');
+      }
+      if (slide.style.nonumber && !attrList.includes('.noframenumbering')) {
+        attrList.push('.noframenumbering');
+      }
+      if (slide.style.center && !attrList.includes('.c')) {
+        attrList.push('.c');
+      }
+    }
+
+    if (attrList.length > 0) {
+      headingAttrs = ' {' + attrList.join(' ') + '}';
+    }
+
+    // Frame heading - pandoc creates the frame from this
+    const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + headingAttrs;
+    output.push(heading);
+    output.push('');
+
+    // Note: Per-frame background colors (.dark, .accent) are not yet supported
+    // in the pandoc-based beamer output. The classes are preserved for PPTX.
+
+    // Add preamble if present (visible on all overlays)
+    if (slide.preamble) {
+      output.push(slide.preamble);
+      output.push('');
+    }
+
+    // Generate content
+    if (slide.steps.length === 1 && slide.steps[0]) {
+      // Single step - just output content, no overlays needed
+      output.push(slide.steps[0].content);
+    } else {
+      // Multiple steps - use \pause between steps
+      // This is the pandoc-friendly way to do incremental reveals
+      for (let i = 0; i < slide.steps.length; i++) {
+        const step = slide.steps[i];
+        if (!step) continue;
+        output.push(step.content);
+        output.push('');
+        // Add pause after each step except the last
+        if (i < slide.steps.length - 1) {
+          output.push('. . .');
+          output.push('');
+        }
+      }
+    }
+
+    // Add speaker notes as LaTeX \note{} command
+    if (slide.notes) {
+      output.push('');
+      // Escape special LaTeX characters in notes
+      const notes = slide.notes || '';
+      const escapedNotes = notes
+        .replace(/\\/g, '\\textbackslash{}')
+        .replace(/[&%$#_{}]/g, '\\$&')
+        .replace(/\n/g, '\\\\');
+      output.push('\\note{' + escapedNotes + '}');
+    }
+
+    output.push('');
+    output.push('---');
+    output.push('');
+  }
+
+  // Remove trailing ---
+  while (output.length > 0) {
+    const last = output[output.length - 1];
+    if (!last || last.trim() !== '') break;
+    output.pop();
+  }
+  if (output.length > 0 && output[output.length - 1] === '---') {
+    output.pop();
+  }
+
+  return output.join('\n');
+}
+
+/**
+ * Build PPTX slide class attribute string
+ */
+function buildPptxSlideClasses(style: SlideStyle | null): string {
+  if (!style || !style.classes || style.classes.length === 0) {
+    return '';
+  }
+  return ' {.' + style.classes.join(' .') + '}';
+}
+
+/**
+ * Generate PPTX markdown with duplicated slides for steps
+ * Each step becomes a separate physical slide
+ * Handles ::: buildup blocks by expanding them into multiple slides
+ */
+export function generatePptxMarkdown(slides: Slide[]): string {
+  const output: string[] = [];
+
+  for (const slide of slides) {
+    // Include frontmatter if present (only on first slide)
+    if (slide._frontmatter) {
+      output.push(slide._frontmatter);
+      output.push('');
+    }
+
+    // Check if any step contains a buildup block
+    const stepsWithBuildup: StepWithBuildup[] = slide.steps.map((step) => {
+      const result = processBuildupBlocks(step.content, 'pptx');
+      return {
+        ...step,
+        beforeBuildup: result.content,
+        afterBuildup: result.afterContent || '',
+        buildupSteps: result.buildupSteps,
+      };
+    });
+
+    // Generate slides - handle buildup expansion
+    for (let i = 0; i < stepsWithBuildup.length; i++) {
+      const step = stepsWithBuildup[i];
+      if (!step) continue;
+
+      if (step.buildupSteps && step.buildupSteps.length > 0) {
+        // This step has a buildup block - generate one slide per buildup step
+        for (const buildupContent of step.buildupSteps) {
+          const classes = buildPptxSlideClasses(slide.style);
+          const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + classes;
+          output.push(heading);
+          output.push('');
+
+          // Add preamble if present
+          if (slide.preamble) {
+            output.push(slide.preamble);
+            output.push('');
+          }
+
+          // Add content before buildup block
+          if (step.beforeBuildup) {
+            output.push(step.beforeBuildup);
+            output.push('');
+          }
+
+          // Add this buildup step content
+          output.push(buildupContent);
+          output.push('');
+
+          // Add content after buildup block
+          if (step.afterBuildup) {
+            output.push(step.afterBuildup);
+            output.push('');
+          }
+
+          // Add speaker notes
+          if (slide.notes) {
+            output.push('::: notes');
+            output.push(slide.notes);
+            output.push(':::');
+          }
+
+          output.push('');
+          output.push('---');
+          output.push('');
+        }
+      } else {
+        // Regular step - no buildup
+        const classes = buildPptxSlideClasses(slide.style);
+        const heading = '#'.repeat(slide.titleLevel) + ' ' + slide.title + classes;
+        output.push(heading);
+        output.push('');
+
+        // Add preamble if present
+        if (slide.preamble) {
+          output.push(slide.preamble);
+          output.push('');
+        }
+
+        // Add cumulative steps up to current
+        for (let j = 0; j <= i; j++) {
+          const s = stepsWithBuildup[j];
+          if (s) {
+            output.push(s.content);
+            output.push('');
+          }
+        }
+
+        // Add speaker notes
+        if (slide.notes) {
+          output.push('::: notes');
+          output.push(slide.notes);
+          output.push(':::');
+        }
+
+        output.push('');
+        output.push('---');
+        output.push('');
+      }
+    }
+  }
+
+  // Remove trailing ---
+  while (output.length > 0) {
+    const last = output[output.length - 1];
+    if (!last || last.trim() !== '') break;
+    output.pop();
+  }
+  if (output.length > 0 && output[output.length - 1] === '---') {
+    output.pop();
+  }
+
+  return output.join('\n');
+}
+
+/**
+ * Process markdown for slide output format
+ */
+export function processSlideMarkdown(markdown: string, format: 'beamer' | 'pptx'): string {
+  const slides = parseSlides(markdown);
+
+  if (format === 'beamer') {
+    return generateBeamerMarkdown(slides);
+  } else if (format === 'pptx') {
+    return generatePptxMarkdown(slides);
+  }
+
+  return markdown;
+}
+
+/**
+ * Check if markdown contains slide syntax (steps, notes, buildup, or slide styles)
+ */
+export function hasSlideSyntax(markdown: string): boolean {
+  // Check for ::: step, ::: notes, or ::: buildup
+  if (/^:::\s*(step|notes|buildup)\s*$/m.test(markdown)) {
+    return true;
+  }
+  // Check for slide style attributes on headings: ## Title {.dark}
+  if (/^#{1,6}\s+.+\{[^}]*\.(dark|light|accent|inverse|cover|thanks|section|plain|nonumber|center)/m.test(markdown)) {
+    return true;
+  }
+  return false;
+}
+
+// Export style parser for testing
+export { parseSlideStyle };