chore: promote dev to main (top-down dep tree sweep)

.github/workflows/ci.yml

@@ -0,0 +1,80 @@
+name: CI
+
+on:
+  push:
+    branches: [dev, main]
+  pull_request:
+    branches: [dev, main]
+
+permissions:
+  contents: read
+
+env:
+  GOFLAGS: -buildvcs=false
+  GOWORK: "off"
+  GOPROXY: "direct"
+  GOSUMDB: "off"
+
+jobs:
+  test:
+    name: Test + Coverage
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          fetch-depth: 0
+      - uses: actions/setup-go@v6
+        with:
+          go-version: '1.26'
+      - name: Test with coverage
+        working-directory: go
+        run: go test -race -coverprofile=coverage.out -covermode=atomic -count=1 ./...
+      - name: Upload coverage to Codecov
+        uses: codecov/codecov-action@v5
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          files: go/coverage.out
+          flags: unittests
+          fail_ci_if_error: false
+
+  lint:
+    name: golangci-lint
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - uses: actions/setup-go@v6
+        with:
+          go-version: '1.26'
+      - uses: golangci/golangci-lint-action@v9
+        with:
+          version: latest
+          working-directory: go
+          args: --timeout=5m --tests=false
+
+  sonarcloud:
+    name: SonarCloud
+    runs-on: ubuntu-latest
+    needs: test
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          fetch-depth: 0
+      - uses: actions/setup-go@v6
+        with:
+          go-version: '1.26'
+      - name: Test for coverage
+        working-directory: go
+        run: go test -coverprofile=coverage.out -covermode=atomic -count=1 ./...
+      - name: SonarCloud Scan
+        uses: SonarSource/sonarqube-scan-action@v6
+        env:
+          SONAR_TOKEN: ${{ secrets.SONAR_TOKEN }}
+        with:
+          args: >
+            -Dsonar.organization=dappcore
+            -Dsonar.projectKey=dappcore_go-rag
+            -Dsonar.sources=go
+            -Dsonar.exclusions=**/vendor/**,**/third_party/**,**/.tmp/**,**/*_test.go
+            -Dsonar.tests=go
+            -Dsonar.test.inclusions=**/*_test.go
+            -Dsonar.go.coverage.reportPaths=go/coverage.out

.gitmodules

@@ -0,0 +1,8 @@
+[submodule "external/go"]
+	path = external/go
+	url = https://github.com/dappcore/go.git
+	branch = dev
+[submodule "external/log"]
+	path = external/log
+	url = https://github.com/dappcore/go-log.git
+	branch = dev

.woodpecker.yml

@@ -14,7 +14,7 @@ steps:
       GOFLAGS: -buildvcs=false
       GOWORK: "off"
     commands:
-      - golangci-lint run --timeout=5m ./...
+      - cd go && golangci-lint run --timeout=5m ./...
 
   - name: go-test
     image: golang:1.26-alpine
@@ -25,7 +25,7 @@ steps:
       CGO_ENABLED: "1"
     commands:
       - apk add --no-cache git build-base
-      - go test -race -coverprofile=coverage.out -covermode=atomic -count=1 ./...
+      - cd go && go test -race -coverprofile=coverage.out -covermode=atomic -count=1 ./...
   - name: sonar
     image: sonarsource/sonar-scanner-cli:latest
     depends_on: [go-test]

CLAUDE.md

@@ -2,49 +2,80 @@
 
 This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
 
-Module: `forge.lthn.ai/core/go-rag`
+Module: `dappco.re/go/rag`
+
+## Repo layout
+
+All Go sources, module files, and Go-only infra are now under `go/`.
+
+```text
+/                                  # Repo root
+├── go/
+│   ├── go.mod
+│   ├── go.sum
+│   ├── go.work
+│   ├── go.work.sum
+│   ├── cmd/
+│   ├── internal/
+│   ├── tests/
+│   ├── tools/
+│   ├── *.go
+│   ├── README.md      -> symlink to root README.md
+│   ├── CLAUDE.md      -> symlink to root CLAUDE.md
+│   ├── AGENTS.md      -> symlink to root AGENTS.md
+│   └── docs/          -> symlink to root docs/
+├── docs/             (RAG specs and docs)
+├── specs/            (RAG spec definitions)
+├── README.md
+├── CLAUDE.md
+├── AGENTS.md
+├── CONTRIBUTING.md
+├── sonar-project.properties
+├── LICENSE / LICENCE
+└── .woodpecker.yml
+```
 
 Retrieval-Augmented Generation library for Go — document chunking, Ollama embeddings, Qdrant vector storage and cosine-similarity search.
 
 ## Commands
 
 ```bash
-go build ./...                       # Build library + CLI
-go test ./...                        # Unit + mock tests (no services needed)
-go test -tags rag ./...              # Full suite including live Qdrant + Ollama
-go test -v -run TestName ./...       # Single test
-go test -race ./...                  # Race detector
-go test -bench=. -benchmem ./...     # Benchmarks (mock-only)
-go test -tags rag -bench=. ./...     # Benchmarks with live services
-gofmt -w .                           # Format
-go vet ./...                         # Vet
-golangci-lint run ./...              # Lint
+cd go && go build ./...                       # Build library + CLI
+cd go && go test ./...                        # Unit + mock tests (no services needed)
+cd go && go test -tags rag ./...              # Full suite including live Qdrant + Ollama
+cd go && go test -v -run TestName ./...       # Single test
+cd go && go test -race ./...                  # Race detector
+cd go && go test -bench=. -benchmem ./...     # Benchmarks (mock-only)
+cd go && go test -tags rag -bench=. ./...     # Benchmarks with live services
+cd go && gofmt -w .                           # Format
+cd go && go vet ./...                         # Vet
+cd go && golangci-lint run ./...              # Lint
 ```
 
 ## Architecture
 
-The library is built around two core interfaces (`Embedder` in `embedder.go`, `VectorStore` in `vectorstore.go`) that decouple business logic from service implementations. All pipeline code operates against these interfaces; concrete clients (`OllamaClient`, `QdrantClient`) and test mocks (`mockEmbedder`, `mockVectorStore` in `mock_test.go`) satisfy them.
+The library is built around two core interfaces (`Embedder` in `go/embedder.go`, `VectorStore` in `go/vectorstore.go`) that decouple business logic from service implementations. All pipeline code operates against these interfaces; concrete clients (`OllamaClient`, `QdrantClient`) and test mocks (`mockEmbedder`, `mockVectorStore` in `go/mock_test.go`) satisfy them.
 
-**Two pipeline flows** drive everything — see `docs/architecture.md` for the full data-flow diagrams:
+**Two pipeline flows** drive everything:
 
-1. **Ingestion** (`ingest.go`): directory walk -> `ChunkMarkdown` (three-level split in `chunk.go`) -> embed per chunk -> batch upsert to vector store
-2. **Query** (`query.go`): embed question -> vector search -> threshold filter -> optional keyword boosting (`keyword.go`) -> format results (text/XML/JSON)
+1. **Ingestion** (`go/ingest.go`): directory walk -> `ChunkMarkdown` (three-level split in `go/chunk.go`) -> embed per chunk -> batch upsert to vector store
+2. **Query** (`go/query.go`): embed question -> vector search -> threshold filter -> optional keyword boosting (`go/keyword.go`) -> format results (text/XML/JSON)
 
-**Two-tier helper pattern** in `helpers.go`:
+**Two-tier helper pattern** in `go/helpers.go`:
 - `*With` variants (e.g. `QueryWith`, `IngestDirWith`) accept pre-constructed `VectorStore`+`Embedder` — use for long-lived processes
 - Default-client wrappers (e.g. `QueryDocs`, `IngestDirectory`) create fresh connections per call with health checks — use for CLI/one-shot operations
 
-**CLI** (`cmd/rag/`): CLI subcommands registered via `AddRAGSubcommands()`, mounted under `core ai rag`. Uses `forge.lthn.ai/core/cli` and `forge.lthn.ai/core/go-i18n` for i18n'd flag descriptions.
+**CLI** (`go/cmd/rag/`): CLI subcommands registered via `AddRAGSubcommands()`, mounted under `core ai rag`.
 
 ## Coding Standards
 
 - UK English (colour, organisation, initialise, behaviour)
 - Conventional commits: `type(scope): description`
 - Co-Author: `Co-Authored-By: Virgil <virgil@lethean.io>`
 - Licence: EUPL-1.2
-- Tests: stdlib `testing` with the local helpers in `test_helpers_test.go`; do not add testify
+- Tests: stdlib `testing` with the local helpers in `go/test_helpers_test.go`; do not add testify
 - Integration tests: `//go:build rag` build tag — requires live Qdrant + Ollama
-- Mocks: `mockEmbedder` and `mockVectorStore` in `mock_test.go` — error injection via fields, call tracking for verification
+- Mocks: `mockEmbedder` and `mockVectorStore` in `go/mock_test.go` — error injection via fields, call tracking for verification
 
 ## Environment Variables
 

README.md

@@ -1,8 +1,20 @@
-[![Go Reference](https://pkg.go.dev/badge/forge.lthn.ai/core/go-rag.svg)](https://pkg.go.dev/forge.lthn.ai/core/go-rag)
-[![License: EUPL-1.2](https://img.shields.io/badge/License-EUPL--1.2-blue.svg)](LICENSE.md)
-[![Go Version](https://img.shields.io/badge/Go-1.26-00ADD8?style=flat&logo=go)](go.mod)
+<!-- SPDX-License-Identifier: EUPL-1.2 -->
+
+
+
+> Retrieval-augmented generation — chunking, embeddings, qdrant + ollama
+
+[![CI](https://github.com/dappcore/go-rag/actions/workflows/ci.yml/badge.svg?branch=dev)](https://github.com/dappcore/go-rag/actions/workflows/ci.yml)
+[![Quality Gate](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=alert_status)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Coverage](https://codecov.io/gh/dappcore/go-rag/branch/dev/graph/badge.svg)](https://codecov.io/gh/dappcore/go-rag)
+[![Security Rating](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=security_rating)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Maintainability Rating](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=sqale_rating)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Reliability Rating](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=reliability_rating)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Code Smells](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=code_smells)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Lines of Code](https://sonarcloud.io/api/project_badges/measure?project=dappcore_go-rag&metric=ncloc)](https://sonarcloud.io/dashboard?id=dappcore_go-rag)
+[![Go Reference](https://pkg.go.dev/badge/dappco.re/go/go-rag.svg)](https://pkg.go.dev/dappco.re/go/go-rag)
+[![License: EUPL-1.2](https://img.shields.io/badge/License-EUPL--1.2-blue.svg)](https://eupl.eu/1.2/en/)
 
-# go-rag
 
 Retrieval-Augmented Generation library for Go. Provides document chunking with three-level Markdown splitting plus sentence- and paragraph-based chunkers, configurable overlap, embedding generation via Ollama, vector storage and cosine-similarity search via Qdrant (gRPC), TF-IDF keyword fallback and keyword boosting post-filter, and result formatting in plain text, XML (for LLM prompt injection), or JSON. Ingestion accepts Markdown, text, PDF, and `.markdown` documents via `ShouldProcess()` / `FileExtensions()`. Designed around `Embedder` and `VectorStore` interfaces that decouple business logic from service implementations and enable mock-based testing.
 

go.work

@@ -1,3 +1,10 @@
-go 1.26.0
+go 1.26.2
 
-use .
+// Workspace mode for development: pulls fresh code from external/ submodules.
+// CI uses GOWORK=off to fall back to go/go.mod tags (reproducible).
+
+use (
+	./go
+	./external/go
+	./external/log
+)

go/AGENTS.md

@@ -0,0 +1 @@
+../AGENTS.md

go/CLAUDE.md

@@ -0,0 +1 @@
+../CLAUDE.md

go/README.md

@@ -0,0 +1 @@
+../README.md

chunk.go → go/chunk.go

@@ -556,14 +556,13 @@ func splitBySentencesSeq(text string) iter.Seq[string] {
 
 		for len(remaining) > 0 {
 			boundary := -1
+
+		BoundarySearch:
 			for i, r := range remaining {
 				switch r {
 				case '.', '!', '?', '\n':
 					boundary = i + len(string(r))
-					break
-				}
-				if boundary >= 0 {
-					break
+					break BoundarySearch
 				}
 			}
 
@@ -587,11 +586,6 @@ func splitBySentencesSeq(text string) iter.Seq[string] {
 	}
 }
 
-// splitBySections splits text by ## headers while preserving the header with its content.
-func splitBySections(text string) []string {
-	return slices.Collect(splitBySectionsSeq(text))
-}
-
 // splitBySectionsSeq returns an iterator that yields text sections split by ## headers.
 func splitBySectionsSeq(text string) iter.Seq[string] {
 	return func(yield func(string) bool) {
@@ -618,11 +612,6 @@ func splitBySectionsSeq(text string) iter.Seq[string] {
 	}
 }
 
-// splitByParagraphs splits text by double newlines.
-func splitByParagraphs(text string) []string {
-	return slices.Collect(splitByParagraphsSeq(text))
-}
-
 // splitByParagraphsSeq returns an iterator that yields paragraphs split by double newlines.
 func splitByParagraphsSeq(text string) iter.Seq[string] {
 	return func(yield func(string) bool) {
@@ -746,6 +735,11 @@ func markdownSectionTitle(section string) (string, string) {
 	return parentTitle, title
 }
 
+var (
+	_ = splitBySentences
+	_ = indexOf
+)
+
 func indexOf(s, substr string) int {
 	if substr == "" || len(substr) > len(s) {
 		return -1

go/docs

@@ -0,0 +1 @@
+../docs

go/go.work

@@ -0,0 +1,3 @@
+go 1.26.0
+
+use .

keyword.go → go/keyword.go

@@ -221,10 +221,10 @@ func tokenise(text string) []string {
 	}
 
 	for _, r := range text {
-		switch {
-		case r == ' ' || r == '\t' || r == '\n' || r == '\r':
+		switch r {
+		case ' ', '\t', '\n', '\r':
 			flush()
-		case r == '.' || r == ',' || r == ';' || r == ':' || r == '!' || r == '?' || r == '"' || r == '\'' || r == '(' || r == ')' || r == '[' || r == ']' || r == '{' || r == '}' || r == '<' || r == '>':
+		case '.', ',', ';', ':', '!', '?', '"', '\'', '(', ')', '[', ']', '{', '}', '<', '>':
 			flush()
 		default:
 			current.WriteRune(r)
@@ -340,6 +340,8 @@ func extractKeywordsSeq(query string) iter.Seq[string] {
 	}
 }
 
+var _ = fields
+
 func fields(text string) []string {
 	var words []string
 	current := core.NewBuilder()

sonar-project.properties

@@ -5,4 +5,4 @@ sonar.exclusions=**/vendor/**,**/third_party/**,**/.tmp/**,**/gomodcache/**,**/n
 sonar.tests=.
 sonar.test.inclusions=**/*_test.go,**/*.test.ts,**/*.test.js,**/*.spec.ts,**/*.spec.js
 sonar.test.exclusions=**/vendor/**,**/third_party/**,**/.tmp/**,**/gomodcache/**,**/node_modules/**,**/dist/**,**/build/**
-sonar.go.coverage.reportPaths=coverage.out
+sonar.go.coverage.reportPaths=go/coverage.out