docrev 0.5.0 → 0.5.1

package/lib/annotations.js

@@ -173,6 +173,7 @@ export function parseAnnotations(text) {
 
 /**
  * Strip annotations from text, applying changes
+ * Handles nested annotations by iterating until stable
  * @param {string} text
  * @param {{keepComments?: boolean}} options
  * @returns {string}
@@ -180,26 +181,88 @@ export function parseAnnotations(text) {
 export function stripAnnotations(text, options = {}) {
   const { keepComments = false } = options;
 
-  // Apply substitutions: {~~old~>new~~} → new
-  text = text.replace(PATTERNS.substitute, '$2');
+  // Iterate until no more changes (handles nested annotations)
+  let prev;
+  let iterations = 0;
+  const maxIterations = 20; // Safety limit
 
-  // Apply insertions: {++text++} → text
-  text = text.replace(PATTERNS.insert, '$1');
+  do {
+    prev = text;
 
-  // Apply deletions: {--text--} → nothing
-  text = text.replace(PATTERNS.delete, '');
+    // Apply substitutions: {~~old~>new~~} → new
+    text = text.replace(PATTERNS.substitute, '$2');
 
-  // Remove highlights: {==text==} → text
-  text = text.replace(PATTERNS.highlight, '$1');
+    // Apply insertions: {++text++} → text
+    text = text.replace(PATTERNS.insert, '$1');
 
-  // Remove comments unless keeping
-  if (!keepComments) {
-    text = text.replace(PATTERNS.comment, '');
-  }
+    // Apply deletions: {--text--} → nothing
+    text = text.replace(PATTERNS.delete, '');
+
+    // Remove highlights: {==text==} → text
+    text = text.replace(PATTERNS.highlight, '$1');
+
+    // Remove comments unless keeping
+    if (!keepComments) {
+      text = text.replace(PATTERNS.comment, '');
+    }
+
+    // Clean up partial/orphaned markers within the loop
+    // This handles cases where nested annotations leave behind fragments
+
+    // Empty annotations (from nested stripping)
+    text = text.replace(/\{----\}/g, '');
+    text = text.replace(/\{\+\+\+\+\}/g, '');
+    text = text.replace(/\{--\s*--\}/g, '');
+    text = text.replace(/\{\+\+\s*\+\+\}/g, '');
+
+    // Orphaned substitution fragments: ~>text~~} or {~~text (no proper pairs)
+    text = text.replace(/~>[^{]*?~~\}/g, '');
+    text = text.replace(/\{~~[^~}]*$/gm, '');
+
+    // Handle malformed substitution from nested: {~~{~~old → just strip the {~~
+    text = text.replace(/\{~~\{~~/g, '{~~');
+    text = text.replace(/~~\}~~\}/g, '~~}');
+
+    iterations++;
+  } while (text !== prev && iterations < maxIterations);
+
+  // Final cleanup of any remaining orphaned markers
+  // Orphaned closing markers
+  text = text.replace(/--\}(?:--\})+/g, '');
+  text = text.replace(/\+\+\}(?:\+\+\})+/g, '');
+  text = text.replace(/~~\}(?:~~\})+/g, '');
+  text = text.replace(/--\}/g, '');
+  text = text.replace(/\+\+\}/g, '');
+  text = text.replace(/~~\}/g, '');
+
+  // Orphaned opening markers
+  text = text.replace(/\{--(?:\{--)+/g, '');
+  text = text.replace(/\{\+\+(?:\{\+\+)+/g, '');
+  text = text.replace(/\{~~(?:\{~~)+/g, '');
+  text = text.replace(/\{--/g, '');
+  text = text.replace(/\{\+\+/g, '');
+  text = text.replace(/\{~~/g, '');
+  text = text.replace(/~>/g, '');
+
+  // Clean up multiple spaces (but preserve structure like newlines)
+  text = text.replace(/  +/g, ' ');
 
   return text;
 }
 
+/**
+ * Check if text contains any CriticMarkup annotations
+ * @param {string} text
+ * @returns {boolean}
+ */
+export function hasAnnotations(text) {
+  return PATTERNS.insert.test(text) ||
+         PATTERNS.delete.test(text) ||
+         PATTERNS.substitute.test(text) ||
+         PATTERNS.comment.test(text) ||
+         PATTERNS.highlight.test(text);
+}
+
 /**
  * Apply a decision to a single annotation
  * @param {string} text

package/lib/import.js

@@ -5,6 +5,7 @@
 import * as fs from 'fs';
 import * as path from 'path';
 import { diffWords } from 'diff';
+import { stripAnnotations } from './annotations.js';
 
 /**
  * Extract comments directly from Word docx comments.xml
@@ -985,7 +986,11 @@ export async function importFromWord(docxPath, originalMdPath, options = {}) {
   }
 
   // Read original markdown
-  const originalMd = fs.readFileSync(originalMdPath, 'utf-8');
+  let originalMd = fs.readFileSync(originalMdPath, 'utf-8');
+
+  // IMPORTANT: Strip any existing annotations to prevent nested annotations
+  // This ensures we always diff clean text against Word text
+  originalMd = stripAnnotations(originalMd, { keepComments: false });
 
   // Generate diff
   let annotated = generateSmartDiff(originalMd, wordText, author);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "docrev",
-  "version": "0.5.0",
+  "version": "0.5.1",
   "description": "Academic paper revision workflow: Word ↔ Markdown round-trips, DOI validation, reviewer comments",
   "type": "module",
   "types": "types/index.d.ts",