Slide2Doc

Slide2Doc is a Blazor Server application that compiles a DOCX document from a PPTX outline and one or more DOCX source documents. It extracts topics from slides, indexes headings from sources, matches topics to sections, and produces a compiled DOCX without writing intermediate files to disk.

Project Structure

/src
  /Slide2Doc.Domain        // Domain entities and enums
  /Slide2Doc.Application   // Use cases and ports
  /Slide2Doc.Infrastructure// OpenXML implementations
  /Slide2Doc.Web           // Blazor Server UI
/tests
  /Slide2Doc.Application.Tests
  /Slide2Doc.Infrastructure.Tests
.github/workflows/ci.yml

How it works

1. Upload: The PPTX (outline) and one or more DOCX sources are uploaded. Inputs are validated (extension, content type, size, max file count) and kept in memory.
2. Extract: OpenXmlPptxTopicExtractor reads slide titles (title placeholder fallback to first text run).
3. Index: OpenXmlDocxSectionIndexer walks DOCX body paragraphs, capturing Heading1/2/3 sections with their element ranges and source file name.
4. Match: TopicMatcher normalizes text and performs exact/contains matching to pair topics with document sections.
5. Compose: OpenXmlDocxComposer builds a new DOCX, adds each topic as Heading1, and copies matched section content (paragraphs/tables). Content stays in streams—no disk writes.
6. Download: The composed DOCX is streamed to the browser via a Blazor page and JavaScript download helper.

Run locally

1. Install .NET 8 SDK.
2. Restore and build:

   dotnet restore Slide2Doc.sln
   dotnet build Slide2Doc.sln

3. Run the Blazor Server app:

   dotnet run --project src/Slide2Doc.Web/Slide2Doc.Web.csproj

4. Open https://localhost:5001 (or shown URL). Navigate through Upload → Mapping → Generate.

Tests

Run unit and integration tests:

dotnet test Slide2Doc.sln

CI

GitHub Actions workflow (.github/workflows/ci.yml) restores, builds, and tests the solution.

Architecture Notes

• Domain: Topic, DocumentSection, TopicMatch, CompileOptions plus enums for heading and match status.
• Application: Interfaces for PPTX extraction, DOCX indexing, matching, composing, and a CompileDocumentService orchestrating prepare/compose steps.
• Infrastructure: OpenXML-based implementations for PPTX topic extraction, DOCX heading indexing, and DOCX composing (paragraph/table level only).
• Web: Blazor Server UI with in-memory state per session, FluentValidation for uploads, logging, and centralized error messaging.

Limitations

• Images and advanced styling are not copied into the output (paragraphs and tables only).
• Fuzzy matching beyond exact/contains is not implemented in the MVP.
• Table of contents rendering is a placeholder.
• Upload size limits are enforced but can be tuned in UploadRequestValidator.

Roadmap

• Preserve styles and media during composition.
• Add fuzzy matching and manual drag/drop mapping UX.
• Persist sessions or allow download of mapping configuration.
• Add timeout/streaming safeguards for very large files.