updating generation code to straight JS

CLAUDE.md

@@ -1,49 +1,50 @@
-# CLAUDE.md
+# Resume Generator with docx-js
 
-This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+## Overview
+This project generates professional Word documents (.docx) from JSON Resume format using the `docx` npm package.
 
-## Repository Purpose
-
-This repository generates resumes based on existing resume documents and recent work history. Resumes can be general-purpose or tailored for specific job roles/titles.
-
-## Output Structure
-
-All generated resumes must follow this folder structure:
-```
-YEAR/ROLE/Resume.md
-YEAR/ROLE/Resume.pdf (or Resume.docx if PDF not possible)
+## Setup
+```bash
+npm install docx
 ```
 
-Example: `2025/Security-Director/Resume.md`
+## Usage
+1. Place your resume data in `resume.json` following the JSON Resume schema (https://jsonresume.org/schema/)
+2. Run: `node generate_resume.js`
+3. Output: `Phillip_Tarrant_Resume_{YEAR}.docx`
 
-## Output Formats
+## Key docx-js Patterns
 
-Every resume generation requires two outputs:
-1. **Markdown version** - Primary editable format
-2. **PDF version** - Preferred final format (use MS Word .docx if PDF generation is not possible)
+### Page Setup
+- Always set explicit page size (US Letter: 12240 x 15840 DXA, where 1440 DXA = 1 inch)
+- Set margins in DXA units
 
-## Source Materials
+### Text Formatting
+- Use `TextRun` for text with formatting (bold, italic, size, font)
+- Size is in half-points (22 = 11pt, 24 = 12pt, 32 = 16pt)
+- Always specify font explicitly for consistency
 
-Reference these files in `sources/` when generating resumes:
+### Lists/Bullets
+- Never use unicode bullets manually (like "•" in text)
+- Use `numbering` config with `LevelFormat.BULLET`
+- Reference the numbering config in paragraphs
 
-**Resume Documents:**
-- `sources/Phillip_Tarrant_Resume_2024.docx` - Most recent general resume
-- `sources/Phillip_Tarrant_Resume_2024_c_level.docx` - Executive/C-level focused version
-- `sources/Phillip_Tarrant_Resume_2023.docx` - Previous year resume
+### Critical Rules
+1. Never use `\n` for line breaks - use separate Paragraph elements
+2. Always set table `width` when using tables
+3. PageBreak must be inside a Paragraph
+4. Use Arial as default font (universally supported)
 
-**Work History Details:**
-- `sources/compuquip.md` - Compuquip role accomplishments
-- `sources/contractor.md` - Recent contract roles
-- `sources/briarwood.md` - Briarwood experience
-- `sources/ice.md` - ICE experience
-- `sources/morgan.md` - Morgan experience
-- `sources/nwtf.md` - NWTF experience
-- `sources/tst.md` - TST experience
-- `sources/older_experience.md` - Earlier career history
+## Customization
+- Modify margins in `sections[0].properties.page.margin`
+- Change fonts by updating the `font` property in TextRun objects
+- Adjust spacing with `spacing: { before: X, after: Y }` on Paragraphs
 
-## Resume Generation Workflow
-
-1. **General resume**: Use 2024 resume as base, incorporate recent history from markdown files
-2. **Role-specific resume**: Tailor content to emphasize relevant experience for the target role/title
-3. Create the YEAR/ROLE directory structure before generating files
-4. Generate both Markdown and PDF/Word outputs
+## JSON Resume Schema
+The input JSON should follow the standard JSON Resume format with these sections:
+- `basics`: name, label, email, phone, url, summary, location, profiles
+- `work`: array of positions with name, position, startDate, endDate, summary, highlights
+- `education`: institution, area, studyType, startDate, endDate, score
+- `certificates`: name, issuer, date
+- `awards`: title, awarder, date, summary
+- `skills`: array of skill categories with name, level, keywords