Enhance refactor commands with controller-aware Route() updates and fix code quality violations

Add semantic token highlighting for 'that' variable and comment file references in VS Code extension
Add Phone_Text_Input and Currency_Input components with formatting utilities
Implement client widgets, form standardization, and soft delete functionality
Add modal scroll lock and update documentation
Implement comprehensive modal system with form integration and validation
Fix modal component instantiation using jQuery plugin API
Implement modal system with responsive sizing, queuing, and validation support
Implement form submission with validation, error handling, and loading states
Implement country/state selectors with dynamic data loading and Bootstrap styling
Revert Rsx::Route() highlighting in Blade/PHP files
Target specific PHP scopes for Rsx::Route() highlighting in Blade
Expand injection selector for Rsx::Route() highlighting
Add custom syntax highlighting for Rsx::Route() and Rsx.Route() calls
Update jqhtml packages to v2.2.165
Add bundle path validation for common mistakes (development mode only)
Create Ajax_Select_Input widget and Rsx_Reference_Data controller
Create Country_Select_Input widget with default country support
Initialize Tom Select on Select_Input widgets
Add Tom Select bundle for enhanced select dropdowns
Implement ISO 3166 geographic data system for country/region selection
Implement widget-based form system with disabled state support

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
root
2025-10-30 06:21:56 +00:00
parent e678b987c2
commit f6ac36c632
5683 changed files with 5854736 additions and 22329 deletions

74
node_modules/resolve/.github/THREAT_MODEL.md generated vendored Normal file
View File

@@ -0,0 +1,74 @@
## Threat Model for resolve (module path resolution library)
### 1. Library Overview
- **Library Name:** resolve
- **Brief Description:** Implements Node.js `require.resolve()` algorithm for synchronous and asynchronous file path resolution. Used to locate modules and files in Node.js projects.
- **Key Public APIs/Functions:** `resolve.sync()` / `resolve/sync`, `resolve()` / `resolve/async`
### 2. Define Scope
This threat model focuses on the core path resolution algorithm, including filesystem interaction, option handling, and cache management.
### 3. Conceptual System Diagram
```
Caller Application → resolve(id, options) → Resolution Algorithm → File System
└→ Options Handling
└→ Cache System
```
**Trust Boundaries:**
- **Input module IDs:** May come from untrusted sources (user input, configuration)
- **Filesystem access:** The library interacts with the filesystem to resolve paths
- **Options:** Provided by the caller
- **Cache:** Used to improve performance, but could be a vector for tampering or information disclosure if not handled securely
### 4. Identify Assets
- **Integrity of resolution output:** Ensure correct and safe file path matching.
- **Confidentiality of configuration:** Prevent sensitive path information from being leaked.
- **Availability/performance for host application:** Prevent crashes or resource exhaustion.
- **Security of host application:** Prevent path traversal or unintended filesystem access.
- **Reputation of library:** Maintain trust by avoiding supply chain attacks and vulnerabilities[1][3][4].
### 5. Identify Threats
| Component / API / Interaction | S | T | R | I | D | E |
|-----------------------------------------------------|----|----|----|----|----|----|
| Public API Call (`resolve/async`, `resolve/sync`) | ✓ | ✓ | | ✓ | | |
| Filesystem Access | | ✓ | | ✓ | ✓ | |
| Options Handling | ✓ | ✓ | | ✓ | | |
| Cache System | | ✓ | | ✓ | | |
**Key Threats:**
- **Spoofing:** Malicious module IDs mimicking legitimate packages, or spoofing configuration options[1].
- **Tampering:** Caller-provided paths altering resolution order, or cache tampering leading to incorrect results[1][4].
- **Information Disclosure:** Error messages revealing filesystem structure or sensitive paths[1].
- **Denial of Service:** Recursive or excessive resolution exhausting filesystem handles or causing application crashes[1].
- **Path Traversal:** Malicious input allowing access to files outside the intended directory[4].
### 6. Mitigation/Countermeasures
| Threat Identified | Proposed Mitigation |
|--------------------------------------------|---------------------|
| Spoofing (malicious module IDs/config) | Sanitize input IDs; validate against known patterns; restrict `basedir` to app-controlled paths[1][4]. |
| Tampering (path traversal, cache) | Validate input IDs for directory escapes; secure cache reads/writes; restrict cache to trusted sources[1][4]. |
| Information Disclosure (error messages) | Generic "not found" errors without internal paths; avoid exposing sensitive configuration in errors[1]. |
| Denial of Service (resource exhaustion) | Limit recursive resolution depth; implement timeout; monitor for excessive filesystem operations[1]. |
### 7. Risk Ranking
- **High:** Path traversal via malicious IDs (if not properly mitigated)
- **Medium:** Cache tampering or spoofing (if cache is not secured)
- **Low:** Information disclosure in errors (if error handling is generic)
### 8. Next Steps & Review
1. **Implement input sanitization for module IDs and configuration.**
2. **Add resolution depth limiting and timeout.**
3. **Audit cache handling for race conditions and tampering.**
4. **Regularly review dependencies for vulnerabilities.**
5. **Keep documentation and threat model up to date.**
6. **Monitor for new threats as the ecosystem and library evolve[1][3].**