docxmlater 10.4.1 → 11.0.4

package/src/elements/CLAUDE.md

@@ -1,142 +0,0 @@
-# Elements Module
-
-`src/elements/` — All document element classes (40 TypeScript files).
-
-## Element Hierarchy
-
-```
-Body Elements (added to document body)
-├── Paragraph
-├── Table
-├── TableOfContents / TableOfContentsElement
-├── StructuredDocumentTag
-├── AlternateContent          # mc:AlternateContent raw XML passthrough
-├── MathParagraph              # m:oMathPara raw XML passthrough
-├── CustomXmlBlock             # w:customXml raw XML passthrough
-└── PreservedElement           # Generic XML passthrough (block context)
-
-Inline Elements (within paragraphs)
-├── Run (formatted text span)
-├── ImageRun (inline image)
-├── Hyperlink
-├── Bookmark / RangeMarker
-├── Field
-├── Comment (anchor)
-├── Footnote/Endnote (reference)
-├── StructuredDocumentTag (inline)
-├── MathExpression             # m:oMath inline raw XML passthrough
-└── PreservedElement (inline)
-
-Header/Footer Elements
-├── Header / Footer
-└── Section (page setup, header/footer references)
-```
-
-## Core Elements
-
-### Paragraph (`Paragraph.ts`)
-
-Container for text and inline content. Key formatting: `alignment`, `spacing`, `indentation`, `pPrChange` (revision tracking), `keepNext`/`keepLines`/`pageBreakBefore` (pagination).
-
-**CT_OnOff properties:** `keepNext`, `keepLines`, `pageBreakBefore`, `suppressLineNumbers`, `suppressAutoHyphens`, `contextualSpacing`, `mirrorIndents`, `suppressOverlap` all use `!== undefined` semantics with `w:val` attribute, supporting explicit `false` to override style inheritance.
-
-**Rule:** `setKeepNext(true)` / `setKeepLines(true)` auto-clears `pageBreakBefore` to `undefined` (absent, not explicit false).
-
-**Alignment:** `ParagraphAlignment` includes bidi-aware values (`start`, `end`), CJK (`distribute`), Arabic kashida (`mediumKashida`, `highKashida`, `lowKashida`), and Thai (`thaiDistribute`).
-
-**w14 attributes:** `paraId` and `textId` (Word 2010+) are auto-generated for new paragraphs during `prepareSave()`. Values are 8-char hex, must be < `0x80000000`.
-
-CJK methods: `setKinsoku()`, `setWordWrap()`, `setOverflowPunct()`, `setTopLinePunct()`, `setAutoSpaceDE()`, `setAutoSpaceDN()`.
-
-### Run (`Run.ts`)
-
-Formatted text span. Character formatting (bold, italic, underline, strikethrough), font properties, color (hex or `'auto'`), highlight, special chars (tabs, breaks, symbols), complex script support.
-
-**CT_OnOff properties:** Boolean properties use `!== undefined` semantics: `undefined` = inherit from style, `true` = on, `false` = explicitly off (`w:val="0"`). This applies to: bold, bCs, italic, iCs, caps, smallCaps, strike, dstrike, snapToGrid.
-
-**Vertical alignment:** `setSubscript()`, `setSuperscript()`, and `vertAlignBaseline` field for explicit baseline reset per ST_VerticalAlignRun.
-
-Theme fonts: `setFontAsciiTheme()`, `setFontHAnsiTheme()`, `setFontEastAsiaTheme()`, `setFontCsTheme()`.
-Underline: `setUnderlineColor()`, `setUnderlineThemeColor()`.
-w14 effects: `addRawW14Property(xml)` for Word 2010+ ligatures, numForm, textOutline.
-Form fields: `FormFieldData`, `FormFieldTextInput`, `FormFieldCheckBox`, `FormFieldDropDownList` interfaces.
-
-**Static helper:** `Run.generateRunPropertiesXML(formatting)` — generates `w:rPr` with all 38+ CT_RPr elements in strict ECMA-376 order. Used by both Run and Style serialization.
-
-### Table (`Table.ts`, `TableRow.ts`, `TableCell.ts`)
-
-Row/column management, cell merging (horizontal + vertical), border styling, property change tracking (`w:tblPrChange`), inter-row content preservation (bookmarks, comments between rows).
-
-Navigation: `getFirstParagraph()`, `getLastParagraph()`.
-Layout: `getLayout()` returns `'fixed'` or `'auto'`.
-Legacy merge: `setHorizontalMerge()` / `getHorizontalMerge()` (hMerge).
-
-**Constraints:** `removeRow()` prevents removing the last row (ECMA-376 requires >= 1 row). `insertRow()`/`insertRows()` use grid span count for correct column count with merged cells. `tblPrEx` is serialized before `trPr` in `w:tr` per CT_Row.
-
-**TableRow:** `setHeightRule(rule)` sets height rule independently. When `w:hRule` is absent, the spec default is `"auto"` (not `"atLeast"`).
-
-**Nested Tables:** Raw XML passthrough via `TableCell.rawNestedContent[]`.
-```typescript
-cell.addRawNestedContent(position, xml, "table")
-cell.getRawNestedContent()
-cell.hasNestedTables()
-```
-
-**Inter-Row Content:** Per ECMA-376, elements like bookmarkEnd can appear between rows.
-```typescript
-table.addInterRowContent(afterRowIndex, xml, type)
-table.getInterRowContent()
-```
-
-### Image (`Image.ts`, `ImageRun.ts`)
-
-PNG/JPEG/GIF/SVG/EMF/WMF with buffer-based format detection. Inline and floating positioning. Full ECMA-376 attribute coverage: presetGeometry, compressionState, bwMode, picLocks, inline dist, transparency (`a:alphaModFix`), enhanced border model (`ImageBorder`), 9 raw XML passthrough slots, linked images (`r:link`), SVG dual-relationship (`asvg:svgBlip`).
-
-## Revision System
-
-### Revision (`Revision.ts`)
-
-Types: `insert`, `delete`, `moveFrom`, `moveTo`, `runPropertiesChange`, `paragraphPropertiesChange`, `tablePropertiesChange`, `tableCellInsert`, `tableCellDelete`, `tableCellMerge`, `numberingChange`, `sectionPropertiesChange`.
-
-### RevisionManager (`RevisionManager.ts`)
-
-Registration, retrieval, category/author/date filtering, location-aware queries, statistics, validation (duplicate IDs, orphaned moves).
-
-### MoveOperationHelper (`../utils/MoveOperationHelper.ts`)
-
-Creates complete move operations with all required ECMA-376 range markers (`moveFromRangeStart/End`, `moveToRangeStart/End`). Without range markers, Word displays moves as separate delete+insert pairs.
-
-```typescript
-MoveOperationHelper.addMoveOperation(sourcePara, destPara, {
-  author: 'John Doe',
-  content: new Run('moved text'),
-});
-```
-
-## Other Elements
-
-- **Hyperlink**: Internal/external URLs, tooltip, relationship management
-- **Bookmark / BookmarkManager**: Named locations, cross-references, unique ID assignment
-- **Field / FieldHelpers**: MERGEFIELD, DATE, TIME, PAGE, NUMPAGES, TOC, REF
-- **Comment / CommentManager**: Author/date, reply threading, anchor positioning
-- **Header / Footer / HeaderFooterManager**: Default, first page, even pages
-- **Section**: Page size, margins, orientation, columns, page numbering
-- **Footnote / Endnote / FootnoteManager / EndnoteManager**: Note content, numbering, separators
-- **StructuredDocumentTag**: Rich/plain text, date pickers, dropdowns, checkboxes
-- **Shape / TextBox**: Basic shapes, text boxes with content
-- **FontManager**: Font table and declarations
-- **ImageManager**: Image collection with relationship handling
-
-## Preserved Element Classes
-
-Store elements as raw XML for round-trip fidelity (internal structure not modified by framework):
-- **AlternateContent**: `mc:AlternateContent` — Word 2010+ drawing with VML fallback
-- **MathParagraph / MathExpression**: `m:oMathPara` / `m:oMath` — math content
-- **CustomXmlBlock**: `w:customXml` — embedded custom XML
-- **PreservedElement**: Generic wrapper for `w:proofErr`, `w:permStart`, `w:permEnd`, `w:altChunk`, `w:ruby`
-
-All share the same API pattern: `constructor(rawXml)`, `toXML()`, `getRawXml()`, `getType()`.
-
-## Serialization
-
-Every element has `toXML()` returning `XMLElement { name, attributes, children }`. Elements don't hold references to their parent document.

package/src/formatting/CLAUDE.md

@@ -1,78 +0,0 @@
-# Formatting Module
-
-`src/formatting/` — Styles and numbering systems.
-
-## Style System
-
-### Style (`Style.ts`)
-
-Types: `paragraph`, `character`, `table`, `numbering`.
-
-Properties: `styleId`, `type`, `name`, `basedOn` (inheritance), `next`, `link`, `runFormatting`, `paragraphFormatting`, `tableFormatting`, `isDefault`.
-
-**Inheritance:** Styles inherit via `basedOn`. Resolution: default → basedOn chain → own formatting.
-
-**Run properties:** Style.ts delegates rPr generation to `Run.generateRunPropertiesXML()` for full ECMA-376 CT_RPr coverage (38+ elements in correct order).
-
-**Conditional formatting:** `tblStylePr` elements follow CT_Style order: pPr → rPr → tblPr → trPr → tcPr.
-
-Built-in styles: `Normal`, `Heading1`-`Heading6`, `Title`, `Subtitle`, `DefaultParagraphFont`, `TableGrid`.
-
-### StylesManager (`StylesManager.ts`)
-
-Stores, validates, generates `word/styles.xml`, parses from existing documents. Validates unique IDs, basedOn existence, no circular inheritance.
-
-## Numbering System
-
-### Concepts
-
-- **AbstractNumbering**: Reusable template defining structure (up to 9 levels) with format/indentation per level
-- **NumberingInstance**: Links abstract numbering to usage (referenced by paragraphs via numId). Can override levels.
-- **NumberingLevel**: Format for one level — format type, text pattern, alignment, indentation, font
-
-### NumberingLevel (`NumberingLevel.ts`)
-
-**Formats:** `decimal`, `upperRoman`, `lowerRoman`, `upperLetter`, `lowerLetter`, `bullet`.
-
-**Level text:** `%1.` (level 1 number), `%1.%2.` (two levels), `•` (bullet).
-
-**Standard indentation:** `leftIndent = 720 + (level * 360)` twips, `hangingIndent = 360` twips.
-
-**Factory methods:**
-```typescript
-NumberingLevel.createBulletLevel(level, symbol)
-NumberingLevel.createNumberedLevel(level, format, text)
-NumberingLevel.calculateStandardIndentation(level)
-NumberingLevel.getBulletSymbolWithFont(level, style)
-```
-
-**Text formatting:** `setItalic()`, `setUnderline(style)`, `setBold()`, `setColor()`.
-
-### AbstractNumbering (`AbstractNumbering.ts`)
-
-```typescript
-AbstractNumbering.createNumberedList(id, maxLevels)  // 1., 2., 3.
-AbstractNumbering.createBulletedList(id, maxLevels)   // bullet symbols
-```
-
-### NumberingInstance (`NumberingInstance.ts`)
-
-Links to abstract numbering. Level overrides:
-```typescript
-const instance = new NumberingInstance(numId, abstractNumId);
-instance.addLevelOverride(0, customLevel);
-```
-
-### NumberingManager (`NumberingManager.ts`)
-
-Manages abstract numberings + instances. Generates `word/numbering.xml`. Assigns unique IDs.
-
-```typescript
-manager.addAbstractNumbering(abstractNum);
-manager.addNumberingInstance(instance);
-paragraph.setNumbering(numId, level);
-```
-
-## Key Behavior: Indentation vs Numbering
-
-Per ECMA-376 §17.3.1.12, paragraph-level indentation overrides numbering-level indentation. The framework auto-clears conflicting paragraph indentation when numbering is applied (left, firstLine, hanging cleared; right preserved). To change list indentation, change the numbering **level** instead of setting indent directly.

package/src/managers/CLAUDE.md

@@ -1,47 +0,0 @@
-# Managers Module
-
-`src/managers/` — Specialized manager classes for document features.
-
-## Manager Pattern
-
-Managers encapsulate complex operations for specific features. Common interface: `add`, `get`, `getAll`, `remove`, `dispose`. All managers are created by Document and accessed via getters (e.g., `doc.getImageManager()`). Always access through Document — never instantiate directly.
-
-## DrawingManager (`DrawingManager.ts`)
-
-Manages shapes, text boxes, and preserved drawings. Drawing types: `shape`, `picture`, `textBox`, `diagram`, `chart`, `preserved`.
-
-**PreservedDrawing:** Unknown drawing elements stored as raw XML during parsing for round-trip fidelity.
-
-```typescript
-const drawingManager = doc.getDrawingManager();
-drawingManager.addDrawing(shape);
-drawingManager.getAllDrawings();
-```
-
-## Related Managers (in other modules)
-
-| Manager | Location | Purpose |
-|---------|----------|---------|
-| ImageManager | `src/elements/` | Image buffering, relationship management, format detection |
-| HeaderFooterManager | `src/elements/` | Default/first/even page headers and footers |
-| BookmarkManager | `src/elements/` | Bookmark registration, unique IDs, start/end pairing |
-| CommentManager | `src/elements/` | Comment storage, author tracking, reply threading |
-| RevisionManager | `src/elements/` | Track changes, RSID management |
-| FootnoteManager | `src/elements/` | Footnote numbering, separators |
-| EndnoteManager | `src/elements/` | Endnote numbering, separators |
-| FontManager | `src/elements/` | Font table declarations |
-| StylesManager | `src/formatting/` | Style definitions, validation, inheritance |
-| NumberingManager | `src/formatting/` | List numbering definitions |
-| RelationshipManager | `src/core/` | Document part relationships |
-
-## Lifecycle
-
-1. **Init:** Created when Document is instantiated
-2. **Use:** Accessed via `doc.get*Manager()` getters
-3. **Cleanup:** All disposed when `doc.dispose()` is called
-
-## Best Practices
-
-- Check existence before adding: `if (!manager.hasBookmark('name')) { ... }`
-- Use manager methods over manual calculation (e.g., `imageManager.getTotalImageSize()`)
-- Always call `doc.dispose()` — cleans up all managers

package/src/tracking/CLAUDE.md

@@ -1,30 +0,0 @@
-# Tracking Module
-
-`src/tracking/` — Change tracking context for revision-aware document editing.
-
-## Interfaces & Classes
-
-### TrackingContext (interface)
-
-Contract for tracking changes on document elements. Methods:
-
-- `trackRunPropertyChange(run, previousProps)` — creates `w:rPrChange`
-- `trackParagraphPropertyChange(para, previousProps)` — creates `w:pPrChange`
-- `trackTableChange(table, previousProps)` — creates `w:tblPrChange`
-- `trackInsertion(element)` / `trackDeletion(element)` — wraps in `w:ins`/`w:del`
-- `flushPendingChanges()` — converts queued changes to Revision objects
-- `isEnabled()` — check if tracking is active
-
-### DocumentTrackingContext (`DocumentTrackingContext.ts`)
-
-Implementation of `TrackingContext`. Key behaviors:
-
-- **State:** enabled flag, author name, `trackFormatting` toggle, pending changes map
-- **Consolidation:** Groups changes by element with time-window keys to batch similar edits (prevents one revision per keystroke)
-- **PrChange application:** Creates/merges `*PrChange` markers (`rPrChange`, `pPrChange`, `tblPrChange`, `tcPrChange`, `sectPrChange`) per ECMA-376
-- **Full snapshot tracking:** Stores complete previous properties (not deltas) for table/cell/section changes
-- **WeakMap for element IDs:** Prevents GC issues with stable element references
-
-## Key Pattern
-
-Call `doc.flushPendingChanges()` before inspecting `pPrChange` or `rPrChange` in tests — pending changes are batched and only materialized on flush or save.

package/src/types/CLAUDE.md

@@ -1,39 +0,0 @@
-# Types Module
-
-`src/types/` — TypeScript type definitions shared across modules.
-
-## Files
-
-### `formatting.ts`
-Formatting-related types used by Document search/replace and paragraph formatting:
-- `BorderStyleType`, `BorderStyle`, `ParagraphBorder` — border definitions
-- `ShadingPattern`, `ParagraphShading` — re-exported from CommonTypes
-- `TabAlignment`, `TabLeader`, `TabStop` — tab stop configuration
-- `FindOptions`, `ReplaceOptions`, `SearchResult` — text search/replace options
-- `EmphasisType`, `ListPrefix`, `FormatOptions`, `StyleApplyOptions` — style application
-
-### `styleConfig.ts`
-Style configuration interfaces for Document's style application methods:
-- `StyleRunFormatting` — character-level formatting (font, size, bold, etc.)
-- `StyleParagraphFormatting` — paragraph-level formatting (alignment, spacing, indentation)
-- `StyleConfig`, `Heading2Config`, `NormalConfig` — predefined style configurations
-- `ApplyStylesOptions` — options for `doc.applyStyles()` / `doc.applyStylesFromObjects()`
-- `ApplyCustomFormattingOptions` — deprecated alias for `ApplyStylesOptions`
-
-### `settings-types.ts`
-Document settings interfaces:
-- `DocumentProtection` — password protection settings
-- `RevisionViewSettings` — revision display mode
-- `TrackChangesSettings` — track changes configuration
-- `WebSettingsInfo` — web settings (div cleanup, encoding)
-
-### `list-types.ts`
-List detection and normalization:
-- `ListCategory`, `NumberFormat`, `BulletFormat` — list type classification
-- `ListDetectionResult`, `ListAnalysis` — detection output
-- `ListNormalizationOptions`, `ListNormalizationReport` — normalization config/results
-
-### `compatibility-types.ts`
-Word compatibility mode types:
-- `CompatSetting` — individual compatibility setting (name, uri, val)
-- `CompatibilityInfo` — full compatibility state (mode value, legacy flags, modern settings)

package/src/utils/CLAUDE.md

@@ -1,168 +0,0 @@
-# Utils Module
-
-`src/utils/` — Standalone utilities (22 TypeScript files) for unit conversions, revision processing, validation, logging, and helpers.
-
-## Architecture Classification
-
-**Pure utilities** (no element/core imports — safe to use anywhere):
-- `units.ts`, `validation.ts`, `deepClone.ts`, `dateFormatting.ts`, `errorHandling.ts`
-- `xmlSanitization.ts`, `formatting.ts`, `parsingHelpers.ts`, `logger.ts`, `diagnostics.ts`
-- `textDiff.ts`, `list-detection.ts`, `corruptionDetection.ts`
-
-**Domain-specific processors** (import from elements/core — should arguably be in `src/processors/`):
-- `acceptRevisions.ts`, `RevisionWalker.ts`, `InMemoryRevisionAcceptor.ts`
-- `SelectiveRevisionAcceptor.ts`, `RevisionAwareProcessor.ts`, `MoveOperationHelper.ts`
-- `ChangelogGenerator.ts`, `ShadingResolver.ts`, `stripTrackedChanges.ts`
-- `CompatibilityUpgrader.ts`, `cnfStyleDecoder.ts`
-
-This split explains the 36+ dependency-cruiser warnings about utils importing elements — the domain-specific files are processors, not pure utilities.
-
-## Unit Conversions (`units.ts`)
-
-40+ conversion functions between twips, EMUs, points, half-points, pixels, inches, centimeters.
-
-Key constants: 1 inch = 1440 twips = 914400 EMUs = 72 points. 1 point = 20 twips. Font sizes use half-points.
-
-## Validation (`validation.ts`)
-
-- `validateRunText(text)` — Detects XML patterns accidentally passed as text
-- `cleanXmlFromText(text)` — Removes XML patterns from text
-- `isValidDocxStructure(files)` — Validates DOCX file structure
-
-## Corruption Detection (`corruptionDetection.ts`)
-
-```typescript
-const report = detectCorruptionInDocument(doc);
-// report: { isCorrupted, summary, locations: [{ suggestedFix }] }
-```
-
-Detects XML markup in text, invalid characters, malformed structures.
-
-## XML Sanitization (`xmlSanitization.ts`)
-
-XML 1.0 control character validation. Auto-applied by `XMLBuilder.escapeXmlText()`.
-
-- `removeInvalidXmlChars(text)` — Strips 0x00-0x08, 0x0B-0x0C, 0x0E-0x1F, 0x7F
-- `hasInvalidXmlChars(text)` / `findInvalidXmlChars(text)`
-- Tab (0x09), newline (0x0A), CR (0x0D) are preserved
-
-## Parsing Helpers (`parsingHelpers.ts`)
-
-Safe OOXML attribute parsing with zero-value edge case handling:
-
-- `safeParseInt(value, default?)` — NaN-safe integer parsing
-- `parseOoxmlBoolean(prop)` — Handles self-closing tags, `val="1"`, `val="true"`, `val="on"`
-- `isExplicitlySet(value)` — Handles 0 correctly (common bug: treating 0 as falsy)
-- `parseNumericAttribute(value, default?)` — Zero-safe numeric parsing
-
-## Formatting Utilities (`formatting.ts`)
-
-- `normalizeColor(color)` — Uppercase 6-char hex
-
-## Deep Clone (`deepClone.ts`)
-
-- `deepClone<T>(obj)` — Safe deep cloning
-
-## Date Formatting (`dateFormatting.ts`)
-
-- `formatDateForXml(date)` — ISO 8601 without milliseconds (Word rejects dates with ms in `w:date`)
-
-## Revision Processing
-
-### acceptRevisions (`acceptRevisions.ts`)
-
-Main entry point for accepting tracked changes via raw XML:
-```typescript
-await acceptAllRevisions(zipHandler, options?)
-// Options: acceptInsertions, acceptDeletions, acceptMoves, acceptPropertyChanges, useDomParser
-```
-Processes document.xml + all headers/footers, remaps image relationship IDs, cleans metadata.
-
-### RevisionWalker (`RevisionWalker.ts`)
-
-DOM-based tree walker for revision acceptance. Handles nested revisions, preserves element order via `_orderedChildren`. ReDoS-safe.
-
-### ChangelogGenerator (`ChangelogGenerator.ts`)
-
-Converts revisions to structured changelog data:
-```typescript
-const entries = ChangelogGenerator.fromDocument(doc, options?);
-ChangelogGenerator.toMarkdown(entries)  // Also: toJSON, toHTML, toCSV
-ChangelogGenerator.getSummary(entries)
-ChangelogGenerator.consolidate(entries)
-```
-Options: `includeFormattingChanges`, `consolidate`, `filterAuthors`, `filterCategories`, `sortBy`, `sortOrder`.
-
-### SelectiveRevisionAcceptor (`SelectiveRevisionAcceptor.ts`)
-
-Accept/reject revisions by criteria (authors, types, date range, categories).
-
-### MoveOperationHelper (`MoveOperationHelper.ts`)
-
-Creates complete move operations with ECMA-376 range markers. See `src/elements/CLAUDE.md` for usage.
-
-### stripTrackedChanges (`stripTrackedChanges.ts`)
-
-Removes all revision markup: `await stripTrackedChanges(zipHandler)`.
-
-## Text Diff (`textDiff.ts`)
-
-Character-level diff for granular tracking in `Run.setText()`:
-```typescript
-const segments = diffText('Hello World', 'Hello Earth');
-// [{ type: 'equal', text: 'Hello ' }, { type: 'delete', text: 'World' }, { type: 'insert', text: 'Earth' }]
-if (diffHasUnchangedParts(segments)) { /* granular tracking */ }
-```
-
-## Compatibility Utilities
-
-### CompatibilityUpgrader (`CompatibilityUpgrader.ts`)
-
-Upgrades `w:compat` blocks to modern Word 2013+ format. Used by `Document.upgradeToModernFormat()`.
-
-Related constants in `src/constants/legacyCompatFlags.ts`:
-- `LEGACY_COMPAT_ELEMENTS` — 65 legacy boolean flags
-- `MODERN_COMPAT_SETTINGS` — 5 default modern settings
-- `MS_WORD_COMPAT_URI` — Microsoft namespace URI
-
-## Logging (`logger.ts`)
-
-**Levels:** `DEBUG`, `INFO`, `WARN`, `ERROR`
-
-**Implementations:** `ConsoleLogger` (timestamps), `SilentLogger` (no output), `CollectingLogger` (in-memory).
-
-**Environment:**
-```bash
-DOCXMLATER_LOG_LEVEL=info npm test    # info/debug/silent
-DEBUG=docxmlater npm test
-```
-
-**Programmatic:**
-```typescript
-import { setGlobalLogger, ConsoleLogger, LogLevel } from 'docxmlater';
-setGlobalLogger(new ConsoleLogger(LogLevel.INFO));
-```
-
-**Scoped:**
-```typescript
-const logger = createScopedLogger(getGlobalLogger(), 'MyComponent');
-logger.info('Done', { count: 10 });
-// Output: 12:34:56.789 [INFO ] [MyComponent] Done count=10
-```
-
-| Component | INFO | DEBUG |
-|-----------|------|-------|
-| Document | Load/save ops | + details |
-| ZipHandler | File ops, validation | + counts, sizes |
-| DocumentParser | Completion, warnings | + element counts |
-| DocumentGenerator | Generation steps | + XML sizes |
-| RevisionManager | Clear, summary | + each registration |
-
-## Error Handling (`errorHandling.ts`)
-
-`DocXMLaterError` → `InvalidDocxError`, `ParseError`, `ValidationError`.
-
-## Diagnostics (`diagnostics.ts`)
-
-- `debugXmlStructure(xml)` — Debug XML structure
-- `inspectElement(element)` — Element inspection

package/src/validation/CLAUDE.md

@@ -1,40 +0,0 @@
-# Validation Module
-
-`src/validation/` — Revision validation, rule definitions, and auto-fix.
-
-## Classes
-
-### RevisionValidator (`RevisionValidator.ts`)
-
-Static validation class. Entry point: `validate(doc, options?)`.
-
-Runs all rules with configurable strictness (`strict` / `normal` / `lenient`). Returns `ValidationResult` with categorized issues (errors, warnings, infos) and summary stats.
-
-Individual validators: `validateRevisionIds()`, `validateMovePairs()`, `validateAuthors()`, `validateDates()`, `validateContent()`, `validateSequentialIds()`.
-
-### ValidationRules (`ValidationRules.ts`)
-
-10 rules organized by severity:
-
-| Rule   | Severity | Description                | Auto-fixable |
-|--------|----------|----------------------------|-------------|
-| REV001 | Error    | Duplicate revision IDs     | Yes         |
-| REV002 | Error    | Missing authors            | Yes         |
-| REV003 | Error    | Orphaned moveFrom          | Yes         |
-| REV004 | Error    | Orphaned moveTo            | Yes         |
-| REV101 | Warning  | Missing dates              | Yes         |
-| REV102 | Warning  | Invalid date format        | Yes         |
-| REV103 | Warning  | Empty revisions            | Yes         |
-| REV104 | Warning  | Non-sequential IDs         | Yes         |
-| REV201 | Info     | Large revision count       | No          |
-| REV202 | Info     | Old revision dates         | No          |
-
-**Auto-fix dependency order:** REV001 -> REV002 -> REV003/004 -> REV103 -> REV104
-
-### RevisionAutoFixer (`RevisionAutoFixer.ts`)
-
-Applies fixes in correct dependency order. Supports dry-run mode via `preview()`.
-
-Entry point: `fix(doc, options?)`. Returns `AutoFixResult` with detailed actions taken.
-
-Individual fixers: `fixDuplicateIds()`, `fixMissingAuthors()`, `fixOrphanedMoveMarkers()`, `fixEmptyRevisions()`, `fixNonSequentialIds()`.

package/src/xml/CLAUDE.md

@@ -1,65 +0,0 @@
-# XML Module
-
-`src/xml/` — XML generation and parsing (2 files).
-
-## XMLBuilder (`XMLBuilder.ts`)
-
-Generates well-formed XML using DOM-like approach (not string concatenation) to prevent injection.
-
-**Key methods:**
-- `createDocument(rootTag, namespaces, children)` — Root element with namespaces. Auto-generates `mc:Ignorable` when extended namespaces (w14, w15, wp14, w16se, etc.) are present.
-- `wElement(name, attributes, children)` — Element with `w:` namespace
-- `wSelf(name, attributes)` — Self-closing `w:` element
-- `escapeXmlText(text)` / `escapeXmlAttribute(value)` — Includes XML 1.0 character sanitization
-
-## XMLParser (`XMLParser.ts`)
-
-Position-based parser (ReDoS-safe). Maximum nesting depth: 256.
-
-**`parseToObject(xml, options?)`** — Returns fast-xml-parser compatible format:
-
-| Feature | Format |
-|---------|--------|
-| Attributes | `@_` prefix (`@_Id`, `@_Type`) |
-| Text content | `#text` property |
-| Multiple same-name children | Array `[]` |
-| Single child | Object `{}` |
-| Self-closing tags | Empty object `{}` |
-| Namespaces | Preserved in keys (`w:p`, `w:r`) |
-| Element order | `_orderedChildren` array |
-
-**Options:**
-```typescript
-interface ParseToObjectOptions {
-  ignoreAttributes?: boolean;      // default: false
-  attributeNamePrefix?: string;    // default: '@_'
-  textNodeName?: string;           // default: '#text'
-  parseAttributeValue?: boolean;   // default: true (numbers/booleans)
-  trimValues?: boolean;            // default: true
-  alwaysArray?: boolean;           // default: false
-  maxNestingDepth?: number;        // default: 256
-  ignoreNamespace?: boolean;       // default: false
-}
-```
-
-**Element Order (`_orderedChildren`):**
-
-Critical for parsing mixed content like runs with interleaved tabs and breaks:
-
-```typescript
-const xml = `<w:r><w:t>Hello</w:t><w:tab/><w:t>World</w:t></w:r>`;
-const result = XMLParser.parseToObject(xml);
-// result["w:r"]._orderedChildren = [
-//   { type: "w:t", index: 0 },
-//   { type: "w:tab", index: 0 },
-//   { type: "w:t", index: 1 }
-// ]
-```
-
-`type` = element tag name, `index` = position within elements of that type.
-
-## Testing
-
-- `tests/xml/XMLBuilder.test.ts` — Builder tests
-- `tests/xml/XMLBuilder-sanitization.test.ts` — 41 sanitization tests
-- `tests/xml/XMLParser.test.ts` — 39 parser tests

package/src/zip/CLAUDE.md

@@ -1,55 +0,0 @@
-# ZIP Module
-
-`src/zip/` — DOCX/ZIP archive handling (5 files). Dependency: JSZip.
-
-## Classes
-
-### ZipHandler (`ZipHandler.ts`)
-
-High-level facade combining ZipReader/ZipWriter. Used by Document for all file I/O.
-
-```typescript
-await zipHandler.loadFromFile(filepath) / loadFromBuffer(buffer)
-await zipHandler.saveToFile(filepath) / toBuffer()
-zipHandler.addFile(path, content)  // string or Buffer
-zipHandler.getFile(path)           // returns string | undefined
-zipHandler.hasFile(path) / getFileNames()
-```
-
-### ZipReader (`ZipReader.ts`)
-
-Reads/extracts from DOCX/ZIP. UTF-8 text extraction, binary file extraction, size validation.
-
-### ZipWriter (`ZipWriter.ts`)
-
-Creates DOCX/ZIP archives. UTF-8 string → Buffer conversion, compression control.
-
-## Types and Constants (`types.ts`)
-
-`DOCX_PATHS` — Standard file paths (`CONTENT_TYPES`, `RELS`, `DOCUMENT`, `STYLES`, `NUMBERING`, `SETTINGS`, `FONT_TABLE`, `THEME`, `CORE_PROPS`, `APP_PROPS`, `FOOTNOTES`, `ENDNOTES`, etc.).
-
-`REQUIRED_DOCX_FILES` — Files required for valid DOCX.
-
-`SizeLimitOptions` — Warning (50MB) and max (150MB) size limits.
-
-## Error Types (`errors.ts`)
-
-`DocxError` (base) → `DocxNotFoundError`, `InvalidDocxError`, `CorruptedArchiveError`, `MissingRequiredFileError`, `FileOperationError`.
-
-## UTF-8 Encoding
-
-All text is UTF-8 per ECMA-376:
-- `ZipWriter.addFile()` converts strings to UTF-8 Buffers
-- `ZipReader` extracts text as UTF-8 strings
-- Binary files (images) preserved as-is
-- XML declarations include `encoding="UTF-8"`
-
-Supports: ASCII, Latin diacritics, Greek, Cyrillic, Arabic, Hebrew, Devanagari, CJK, emoji, RTL.
-
-## Security
-
-File paths validated against path traversal (`../`, absolute paths, URL-encoded sequences).
-
-## Testing
-
-`tests/zip/ZipHandler.test.ts` — 62 tests including 11 UTF-8 encoding tests.