Fix E2E

.editorconfig

@@ -1,16 +0,0 @@
-# https://editorconfig.org
-
-root = true
-
-[*]
-end_of_line = lf
-insert_final_newline = true
-trim_trailing_whitespace = true
-charset = utf-8
-indent_style = space
-indent_size = 4
-tab_width = 4
-
-[*.{yml,yaml,md}]
-indent_size = 2
-tab_width = 2

.forgejo/workflows/check.yml

@@ -1,35 +0,0 @@
-name: Check
-
-on:
-  push:
-    branches: ["main"]
-  pull_request:
-    branches: ["main"]
-  workflow_dispatch:
-
-env:
-  CARGO_TERM_COLOR: always
-  RUSTFLAGS: "-Dwarnings"
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Setup Node.js environment
-        uses: actions/setup-node@v4
-        with:
-          node-version: "25.x"
-
-      - name: Setup Rust toolchain
-        uses: dtolnay/rust-toolchain@stable
-        with:
-          toolchain: "1.92.0"
-          components: clippy, rustfmt
-
-      - name: Lint & test
-        run: scripts/check.sh

.forgejo/workflows/deploy-docs.yml

@@ -1,38 +0,0 @@
-name: Deploy Documentation
-
-on:
-  push:
-    branches:
-      - main
-    paths:
-      - "docs/**"
-      - ".forgejo/workflows/deploy-docs.yml"
-  workflow_dispatch:
-
-concurrency:
-  group: pages
-  cancel-in-progress: false
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Setup Node.js environment
-        uses: actions/setup-node@v4
-        with:
-          node-version: "25.x"
-
-      - name: Build docs
-        run: scripts/build-docs.sh
-
-      - name: Upload artifact
-        uses: actions/upload-artifact@v4
-        with:
-          name: docs
-          path: docs/.vitepress/dist

.forgejo/workflows/e2e.yml

@@ -1,71 +0,0 @@
-name: E2E tests
-
-on:
-  push:
-    branches: ["main"]
-  pull_request:
-    branches: ["main"]
-  schedule:
-    - cron: "0 * * * *"
-  workflow_dispatch:
-
-concurrency:
-  group: e2e-tests
-  cancel-in-progress: false
-
-env:
-  RUSTFLAGS: "-Dwarnings"
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Setup Node.js environment
-        uses: actions/setup-node@v4
-        with:
-          node-version: "25.x"
-
-      - name: Setup Rust toolchain
-        uses: dtolnay/rust-toolchain@stable
-        with:
-          toolchain: "1.92.0"
-          components: clippy, rustfmt
-
-      - name: Setup rust
-        run: |
-          which sqlx || cargo install sqlx-cli
-          cd sync-server
-          sqlx database create --database-url sqlite://db.sqlite3
-          sqlx migrate run --source src/app_state/database/migrations --database-url sqlite://db.sqlite3
-
-      - name: E2E tests
-        run: |
-          cd sync-server
-          cargo run config-e2e.yml --color never &
-          SERVER_PID=$!
-          cd ..
-
-          scripts/e2e.sh 8
-          EXIT_CODE=$?
-
-          kill $SERVER_PID 2>/dev/null || true
-          wait $SERVER_PID 2>/dev/null || true
-
-          exit $EXIT_CODE
-
-      - name: Upload e2e logs
-        if: always()
-        uses: actions/upload-artifact@v4
-        with:
-          name: e2e-logs
-          path: logs/
-          retention-days: 30
-
-      - name: Cleanup
-        if: always()
-        run: scripts/clean-up.sh

.forgejo/workflows/publish-cli-docker.yml

@@ -1,51 +0,0 @@
-name: Publish CLI
-
-on:
-  push:
-    branches: ["main"]
-    tags: ["*"]
-  pull_request:
-    branches: ["main"]
-
-jobs:
-  publish-docker:
-    runs-on: ubuntu-docker
-
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Extract registry hostname
-        id: registry
-        run: echo "host=$(echo '${{ github.server_url }}' | sed 's|https\?://||')" >> "$GITHUB_OUTPUT"
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Log into container registry
-        uses: docker/login-action@v3
-        with:
-          registry: ${{ steps.registry.outputs.host }}
-          username: ${{ github.repository_owner }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Extract Docker metadata
-        id: meta
-        uses: docker/metadata-action@v5
-        with:
-          images: ${{ steps.registry.outputs.host }}/${{ github.repository }}-cli
-
-      - name: Build and push Docker image
-        id: build-and-push
-        uses: docker/build-push-action@v5
-        with:
-          context: frontend
-          file: frontend/local-client-cli/Dockerfile
-          platforms: linux/amd64
-          push: true
-          tags: ${{ steps.meta.outputs.tags }}
-          labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=registry,ref=${{ steps.registry.outputs.host }}/${{ github.repository }}-cli:buildcache
-          cache-to: type=registry,ref=${{ steps.registry.outputs.host }}/${{ github.repository }}-cli:buildcache,mode=max

.forgejo/workflows/publish-plugin.yml

@@ -1,71 +0,0 @@
-name: Publish Obsidian plugin
-
-on:
-  push:
-    tags: ["*"]
-
-env:
-  CARGO_TERM_COLOR: always
-
-jobs:
-  publish-plugin:
-    runs-on: ubuntu-latest
-
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Setup Node.js environment
-        uses: actions/setup-node@v4
-        with:
-          node-version: "25.x"
-
-      - name: Build plugin
-        run: |
-          cd frontend
-          npm ci
-          npm run build
-
-      - name: Setup Rust toolchain
-        uses: dtolnay/rust-toolchain@stable
-        with:
-          toolchain: "1.92.0"
-          components: clippy, rustfmt
-
-      - name: Install cross-compilation tools
-        run: |
-          apt update
-          apt install -y gcc-aarch64-linux-gnu musl-tools gcc-mingw-w64-x86-64 jq
-
-      - name: Build Linux and Windows binaries
-        run: ./scripts/build-sync-server-binaries.sh
-
-      - name: Create release
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          SERVER_URL: ${{ github.server_url }}
-          REPO: ${{ github.repository }}
-        run: |
-          tag="${GITHUB_REF#refs/tags/}"
-
-          mkdir -p release
-          cp frontend/obsidian-plugin/dist/* release/
-          cp sync-server/artifacts/sync-server-* release/
-
-          # Create draft release via Forgejo API
-          RELEASE_ID=$(curl -s -X POST \
-            "${SERVER_URL}/api/v1/repos/${REPO}/releases" \
-            -H "Authorization: token ${GITHUB_TOKEN}" \
-            -H "Content-Type: application/json" \
-            -d "{\"tag_name\": \"${tag}\", \"name\": \"${tag}\", \"draft\": true}" \
-            | jq -r '.id')
-
-          # Upload release assets
-          for file in release/*; do
-            filename=$(basename "$file")
-            curl -s -X POST \
-              "${SERVER_URL}/api/v1/repos/${REPO}/releases/${RELEASE_ID}/assets?name=${filename}" \
-              -H "Authorization: token ${GITHUB_TOKEN}" \
-              -F "attachment=@${file}"
-          done

.forgejo/workflows/publish-server-docker.yml

@@ -1,51 +0,0 @@
-name: Publish server Docker image
-
-on:
-  push:
-    branches: ["main"]
-    tags: ["*"]
-  pull_request:
-    branches: ["main"]
-
-jobs:
-  publish-docker:
-    runs-on: ubuntu-docker
-
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      - name: Extract registry hostname
-        id: registry
-        run: echo "host=$(echo '${{ github.server_url }}' | sed 's|https\?://||')" >> "$GITHUB_OUTPUT"
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Log into container registry
-        if: github.ref_type == 'tag'
-        uses: docker/login-action@v3
-        with:
-          registry: ${{ steps.registry.outputs.host }}
-          username: ${{ github.repository_owner }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Extract Docker metadata
-        id: meta
-        uses: docker/metadata-action@v5
-        with:
-          images: ${{ steps.registry.outputs.host }}/${{ github.repository }}
-
-      - name: Build and push Docker image
-        id: build-and-push
-        uses: docker/build-push-action@v5
-        with:
-          context: sync-server
-          platforms: linux/amd64,linux/arm64
-          push: ${{ github.ref_type == 'tag' }}
-          tags: ${{ steps.meta.outputs.tags }}
-          labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=registry,ref=${{ steps.registry.outputs.host }}/${{ github.repository }}:buildcache
-          cache-to: type=registry,ref=${{ steps.registry.outputs.host }}/${{ github.repository }}:buildcache,mode=max

.github/dependabot.yml

@@ -0,0 +1,27 @@
+# To get started with Dependabot version updates, you'll need to specify which
+# package ecosystems to update and where the package manifests are located.
+# Please see the documentation for all configuration options:
+# https://docs.github.com/code-security/dependabot/dependabot-version-updates/configuration-options-for-the-dependabot.yml-file
+
+version: 2
+updates:
+  - package-ecosystem: "npm"
+    directories: ["**"]
+    schedule:
+      interval: "daily"
+
+  - package-ecosystem: "docker"
+    directories: ["**"]
+    schedule:
+      interval: "daily"
+
+  - package-ecosystem: "cargo"
+    directories: ["**"]
+    schedule:
+      interval: "daily"
+
+  # Disable this for security reasons
+  # - package-ecosystem: "github-actions"
+  #   directories: ["**"]
+  #   schedule:
+  #     interval: "daily"

.github/workflows/check.yml

@@ -0,0 +1,66 @@
+name: Check
+
+on:
+  push:
+    branches: ["main"]
+  pull_request:
+    branches: ["main"]
+
+env:
+  CARGO_TERM_COLOR: always
+  RUSTFLAGS: "-Dwarnings"
+
+jobs:
+  build:
+    runs-on: self-hosted
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Node.js environment
+        uses: actions/setup-node@v4.2.0
+        with:
+          node-version: "22.x"
+          check-latest: true
+
+      - name: Setup rust
+        run: |
+          cargo install sqlx-cli wasm-pack
+          cd backend
+          sqlx database create --database-url sqlite://db.sqlite3
+          sqlx migrate run --source sync_server/src/app_state/database/migrations --database-url sqlite://db.sqlite3
+
+      - name: Build wasm
+        run: |
+          cd backend
+          wasm-pack build --target web sync_lib
+
+      - name: Lint backend
+        run: |
+          cd backend
+          cargo clippy --all-targets --all-features
+          cargo fmt --all -- --check
+
+      - name: Test backend
+        run: |
+          cd backend
+          cargo test --verbose
+          cd sync_lib
+          wasm-pack test --node
+
+      - name: Lint frontend
+        run: |
+          cd frontend
+          npm ci
+          npm run build
+          npm run lint
+          if [[ $(git status --porcelain) ]]; then
+              git status --porcelain
+              echo "Failing CI because the working directory is not clean after linting"
+              exit 1
+          fi
+
+      - name: Test frontend
+        run: |
+          cd frontend
+          npm run test

.github/workflows/e2e.yml

@@ -0,0 +1,56 @@
+name: E2E tests
+
+on:
+  push:
+    branches: ["main"]
+  pull_request:
+    branches: ["main"]
+
+env:
+  CARGO_TERM_COLOR: always
+  RUSTFLAGS: "-Dwarnings"
+
+jobs:
+  build:
+    runs-on: self-hosted
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Node.js environment
+        uses: actions/setup-node@v4.2.0
+        with:
+          node-version: "22.x"
+          check-latest: true
+
+      - name: Setup rust
+        run: |
+          cargo install sqlx-cli wasm-pack
+          cd backend
+          sqlx database create --database-url sqlite://db.sqlite3
+          sqlx migrate run --source sync_server/src/app_state/database/migrations --database-url sqlite://db.sqlite3
+
+      - name: Build wasm
+        run: |
+          cd backend
+          wasm-pack build --target web sync_lib
+
+      - name: E2E tests
+        run: |
+          cd backend
+          cargo run -p sync_server config-e2e.yml --color never &
+          cd ..
+
+          scripts/update-api-types.sh
+          cd frontend
+          npm ci
+          npm run build
+          npm run lint
+          if [[ $(git status --porcelain) ]]; then
+              git status --porcelain
+              echo "Failing CI because the working directory is not clean after updating the API types"
+              exit 1
+          fi
+
+          cd ..
+          scripts/e2e.sh 32

.github/workflows/publish-docker.yml

@@ -0,0 +1,89 @@
+name: Publish server Docker image
+
+on:
+  push:
+    branches: ["main"]
+    tags: ["*"]
+  pull_request:
+    branches: ["main"]
+
+env:
+  # Use docker.io for Docker Hub if empty
+  REGISTRY: ghcr.io
+  # github.repository as <account>/<repo>
+  IMAGE_NAME: ${{ github.repository }}
+
+jobs:
+  publish-docker:
+    runs-on: self-hosted
+
+    permissions:
+      contents: read
+      packages: write
+      # This is used to complete the identity challenge
+      # with sigstore/fulcio.
+      id-token: write
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      # Install the cosign tool
+      # https://github.com/sigstore/cosign-installer
+      - name: Install cosign
+        if: github.ref_type == 'tag'
+        uses: sigstore/cosign-installer@59acb6260d9c0ba8f4a2f9d9b48431a222b68e20 #v3.5.0
+        with:
+          cosign-release: "v2.2.4"
+
+      # Set up BuildKit Docker container builder to be able to build
+      # multi-platform images and export cache
+      # https://github.com/docker/setup-buildx-action
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@f95db51fddba0c2d1ec667646a06c2ce06100226 # v3.0.0
+
+      # Login against a Docker registry
+      # https://github.com/docker/login-action
+      - name: Log into registry ${{ env.REGISTRY }}
+        if: github.ref_type == 'tag'
+        uses: docker/login-action@343f7c4344506bcbf9b4de18042ae17996df046d # v3.0.0
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      # Extract metadata (tags, labels) for Docker
+      # https://github.com/docker/metadata-action
+      - name: Extract Docker metadata
+        id: meta
+        uses: docker/metadata-action@96383f45573cb7f253c731d3b3ab81c87ef81934 # v5.0.0
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
+
+      # Build and push Docker image with Buildx
+      # https://github.com/docker/build-push-action
+      - name: Build and push Docker image
+        id: build-and-push
+        uses: docker/build-push-action@0565240e2d4ab88bba5387d719585280857ece09 # v5.0.0
+        with:
+          context: backend
+          push: ${{ github.ref_type == 'tag' }}
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+          cache-from: type=gha
+          cache-to: type=gha,mode=max
+
+      # Sign the resulting Docker image digest.
+      # This will only write to the public Rekor transparency log when the Docker
+      # repository is public to avoid leaking data. If you would like to publish
+      # transparency data even for private images, pass --force to cosign below.
+      # https://github.com/sigstore/cosign
+      - name: Sign the published Docker image
+        if: ${{ github.ref_type == 'tag' }}
+        env:
+          # https://docs.github.com/en/actions/security-guides/security-hardening-for-github-actions#using-an-intermediate-environment-variable
+          TAGS: ${{ steps.meta.outputs.tags }}
+          DIGEST: ${{ steps.build-and-push.outputs.digest }}
+        # This step uses the identity token to provision an ephemeral certificate
+        # against the sigstore community Fulcio instance.
+        run: echo "${TAGS}" | xargs -I {} cosign sign --yes {}@${DIGEST}

.github/workflows/publish-plugin.yml

@@ -0,0 +1,46 @@
+name: Publish Obsidian plugin
+
+on:
+  push:
+    tags: ["*"]
+
+env:
+  CARGO_TERM_COLOR: always
+
+jobs:
+  publish-plugin:
+    runs-on: self-hosted
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Node.js environment
+        uses: actions/setup-node@v4.2.0
+        with:
+          node-version: "22.x"
+          check-latest: true
+
+      - name: Build wasm
+        run: |
+          cd backend
+          cargo install wasm-pack
+          wasm-pack build --target web sync_lib
+
+      - name: Build plugin
+        run: |
+          cd frontend
+          npm ci
+          npm run build
+
+      - name: Create release
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          tag="${GITHUB_REF#refs/tags/}"
+
+          cd frontend/obsidian-plugin/dist
+
+          gh release create "$tag" \
+            --title="$tag" \
+            --draft \
+            main.js manifest.json styles.css

.gitignore

@@ -1,24 +1,17 @@
-# npm
-node_modules
-
-# Exclude macOS Finder (System Explorer) View States
-.DS_Store
-
-# Frontend build folders
-frontend/*/dist
-
-# Rust build folders
-sync-server/target
-sync-server/artifacts
-sync-server/bindings/*.ts
-
-# build folders
-sync-server/db.sqlite3*
-**/databases
-
-*.log
-*.sqlx
-
-target
-
-.task
+# npm
+node_modules
+
+# Exclude macOS Finder (System Explorer) View States
+.DS_Store
+
+# Rust build folder
+backend/target
+
+frontend/*/dist
+
+backend/db.sqlite3*
+backend/databases
+
+*.log
+
+*.sqlx

.vscode/settings.json

@@ -3,8 +3,6 @@
     "jest.rootPath": "plugin",
     "files.exclude": {
         "**/dist": true,
-        "**/node_modules": true,
-        "**/.sqlx": true,
-        "**/target": true
+        "**/node_modules": true
     }
-}
+}

CLAUDE.md

@@ -1,155 +0,0 @@
-# CLAUDE.md
-
-This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
-
-## Project shape
-
-VaultLink is a self-hosted Obsidian file-sync system. Two halves of one repo:
-
-- `sync-server/` — Rust (axum + sqlx/SQLite). Source of truth for vault state, broadcasts changes via WebSocket.
-- `frontend/` — npm workspaces. The sync engine (`sync-client`) is consumed by an Obsidian plugin, a standalone CLI, a fuzz E2E harness, a scripted determinism harness, and a history UI.
-
-The HTTP/WS API types are generated from Rust (`ts-rs`) and mirrored into the TS workspaces. **Never hand-edit files in `frontend/sync-client/src/services/types/` or `frontend/history-ui/src/lib/types/`** — run `scripts/update-api-types.sh` after changing anything Serde-derived in the server.
-
-### Frontend workspaces
-
-- `sync-client` — the sync engine; published to consumers via `dist/`. All other TS workspaces depend on it via `file:../sync-client`.
-- `obsidian-plugin` — Obsidian plugin built from `sync-client`.
-- `local-client-cli` — same engine wrapped as a standalone CLI.
-- `history-ui` — vault-history web UI.
-- `test-client` — fuzz E2E harness (random ops across N processes).
-- `deterministic-tests` — scripted multi-client tests with an in-memory FS, run against a real server.
-
-## Common commands
-
-Pre-push hygiene (formats, lints, runs tests, requires clean git state):
-
-```sh
-scripts/check.sh --fix
-```
-
-Run the fuzz E2E (N parallel processes):
-
-```sh
-scripts/e2e.sh 12
-# Logs land in logs/log_<i>.log. Clean with scripts/clean-up.sh
-```
-
-Run deterministic tests (require a release-built server in `sync-server/target/release/sync_server` — they spawn it themselves):
-
-```sh
-cd sync-server && cargo build --release && cd ..
-cd frontend
-npm run build -w sync-client -w deterministic-tests
-node deterministic-tests/dist/cli.js                       # all
-node deterministic-tests/dist/cli.js --filter=rename       # subset
-node deterministic-tests/dist/cli.js --filter=… -j 4       # cap parallelism
-```
-
-Run a single sync-client unit test by file:
-
-```sh
-cd frontend/sync-client && npx tsx --test 'src/**/sync-event-queue.test.ts'
-```
-
-Server: dev runs from `sync-server/` against `config-e2e.yml`:
-
-```sh
-cd sync-server
-cargo run config-e2e.yml          # dev
-cargo build --release             # used by both e2e harnesses
-cargo test                        # unit + ts-rs binding export tests
-```
-
-Frontend dev (sync-client + obsidian-plugin watch in parallel):
-
-```sh
-cd frontend && npm install && npm run dev
-```
-
-Regenerate TS bindings from Rust types (touches `frontend/{sync-client,history-ui}/src/.../types/`):
-
-```sh
-scripts/update-api-types.sh
-```
-
-## SQLite / sqlx
-
-The server uses `sqlx::query!` macros that need a prepared `.sqlx` cache to compile offline. Touching any SQL means regenerating it:
-
-```sh
-cd sync-server
-sqlx database create --database-url sqlite://db.sqlite3
-sqlx migrate run --source src/app_state/database/migrations --database-url sqlite://db.sqlite3
-cargo sqlx prepare --workspace
-```
-
-New migrations: `sqlx migrate add --source src/app_state/database/migrations <name>`.
-
-## Sync engine architecture
-
-Read `frontend/sync-client/src/sync-operations/` to follow the sync engine; the rest of `sync-client` is plumbing (filesystem ops, persistence, services, telemetry).
-
-The engine is **two independent loops with separate invariants**:
-
-- **Wire loop** (`syncer.ts`) — drains the single-consumer FIFO queue. HTTP and WS handlers update record fields (`remoteRelativePath`, `parentVersionId`, `remoteHash`) and write content to the file at `record.localPath`. They never move files for path placement.
-- **Path reconciler** (`reconciler.ts`) — runs after every drained event. Best-effort pass that moves files to make `localPath === remoteRelativePath`. The move graph is topologically sorted; cycles are resolved by reading every file in the cycle into memory and writing each back to its new slot (no tmp files). Records with pending local events are skipped on each pass — the reconciler operates only on settled records. Failures (slot occupied by an untracked file, etc.) are silent skips; the next pass retries.
-
-**`SyncEventQueue`** (`sync-event-queue.ts`) holds:
-
-- `byDocId: Map<DocumentId, DocumentRecord>` — primary record store.
-- `byLocalPath: Map<RelativePath, DocumentRecord>` — derived index for path lookups, maintained at every mutation point.
-- `events: SyncEvent[]` — pending wire ops in FIFO drain order.
-
-```ts
-DocumentRecord = {
-    documentId,
-    parentVersionId,
-    remoteHash?,
-    remoteRelativePath,
-    localPath: RelativePath | undefined
-}
-```
-
-`localPath === undefined` means the doc has no local file yet — typically a remote create whose target slot was occupied at receive time; the reconciler will fetch and place when the slot frees (the bytes wait in `pendingPlacementContent`).
-
-Local FS events from the watcher update `localPath` synchronously at enqueue time via `setLocalPath` / `upsertRecord`. The wire loop never updates it for path placement; only the reconciler does. A user rename onto a tracked slot enqueues a `LocalDelete` for the displaced doc (the OS rename clobbered its content) and clears that doc's `localPath`.
-
-**Pending creates** use a `Promise<DocumentId>` chain to serialize dependent ops (`LocalUpdate`, `LocalDelete`) behind the still-in-flight `LocalCreate`. `resolveCreate` resolves the promise once the server returns a docId, and `replacePendingDocumentId` swaps the resolved id across already-queued events. `findLatestCreateForPath` is the lookup the watcher uses to attach dependents; `updatePendingCreatePath` rewrites a pending create's `event.path` in place when the user renames the file before its create has acked.
-
-**Watermark.** `lastSeenUpdateId` uses a `MinCovered` (a contiguous-prefix tracker over a stream of integers): we only advance the published min when the next consecutive id has been processed, so out-of-order RemoteChange ids don't fool the WebSocket handshake into requesting a too-recent catch-up.
-
-**Server catch-up.** The server's WS handshake replays events newer than the client's `last_seen_vault_update_id` from the `latest_document_versions` view (one row per doc, the latest). On those replayed rows `is_new_file` means _new to this client_ (`creation_vault_update_id > last_seen_vault_update_id`), not "this row is the doc's first version" — necessary because the catch-up only carries the latest version; if a doc was created and updated past the watermark, the client never sees its create otherwise.
-
-## Edge-case patterns the sync engine has to survive
-
-The two-loop split defuses most of the old race catalogue (slot-collision stashes, conflict-uuid divergence, `MoveOnConflict.NEW`/`EXISTING` policy choices) by separating wire transport from path placement. What's left:
-
-**Pending-create docId is a `Promise`, not a string, until the create acks.** Any `LocalUpdate` / `LocalDelete` queued behind a still-in-flight `LocalCreate` carries the create's `resolvers.promise` as its `documentId`. `replacePendingDocumentId` swaps the resolved id across queued events when the create resolves; `===` comparisons against the resolved string elsewhere will silently fail until that swap runs. Anything that walks `events[]` looking for a docId match must either run after the swap or be tolerant of `Promise`-typed ids.
-
-**`processCreate` reads `event.path` live, not `event.originalPath`.** The watcher rewrites `event.path` in place via `updatePendingCreatePath` when the user renames a pending-create file. `originalPath` was removed from `LocalCreate` events specifically because reading it would send the stale pre-rename path to the server.
-
-**`record.localPath` mutates in place across awaits.** When the watcher renames a doc while a drain handler is awaiting an HTTP roundtrip, the queue mutates the in-flight event's record so subsequent reads see the new path. Snapshotting `record.localPath` into a local at function entry and using it after an `await` reads/writes a now-vacated slot. Read `record.localPath` live; only snapshot for the deliberate "did it change while I was awaiting" comparison.
-
-**Reconciler-defer is the wire-loop's contract with the reconciler.** The reconciler skips records where `hasPendingLocalEventsForDocumentId` returns true. Wire-loop handlers can therefore freely write `remoteRelativePath` to whatever the server returned — even if it disagrees with `localPath` — knowing the reconciler won't move the file out from under a queued user rename.
-
-**Watermark advancement is load-bearing both ways.** Branches that _skip_ a remote event without advancing `lastSeenUpdateId` create permanent gaps that re-deliver forever. Branches that _advance_ without applying the content lose data: the server has no further event to re-deliver, the catch-up only carries the latest version, and any state in between is gone. Don't advance unless the event was actually applied (or deliberately discarded after weighing both halves).
-
-**`isNewFile` semantics differ between catch-up and real-time.** On WS handshake replay it means _new to this client_ (`creation_vault_update_id > last_seen_vault_update_id`); on real-time broadcasts it means _this version is the create_ (`creation_vault_update_id == vault_update_id`). A handler that decides based on one interpretation will be wrong on the other channel; reasoning about fetch-and-treat-as-new vs. ignore needs to know which channel delivered the event.
-
-**Pause / disable-sync mid-flight** is the one race the new model doesn't structurally fix. An HTTP that committed server-side but whose response was discarded leaves the server holding a doc the client has no record of. Resume → offline scan → server-side dedupe handles it (the server merges the duplicate create into the existing doc), but if the merge produces a deconflict, the client picks up an extra file. Out of scope for the two-loop split.
-
-**Cycle reconciliation uses in-memory content swap.** When the move graph contains a cycle, the reconciler reads every file in the cycle into memory and writes each back to its new slot, with no tmp files. A write-ahead marker at `.vaultlink/swap-<uuid>.json` lists each leg; on startup the reconciler reads the marker, hashes each `from` to determine which legs ran, and replays the rest. The `.vaultlink/**` glob is hard-coded as an internal ignore pattern so swap markers don't get sync'd.
-
-## Two complementary E2E harnesses
-
-- **`test-client` (fuzz):** random ops across N parallel processes for many minutes. Used by `scripts/e2e.sh`. Catches bugs nobody thought to write a test for, but reproductions are noisy.
-- **`deterministic-tests`:** scripted scenarios with an in-memory FS pinned to a real server. Used to _capture_ a fuzz-discovered bug as a minimal repro before fixing it. See `frontend/deterministic-tests/README.md` for the step grammar (`pause-server`, `pause-websocket`, `barrier`, `assert-consistent`, etc.).
-
-When a fuzz failure surfaces, the workflow is: root-cause from logs → write a deterministic test that fails on the bug → fix → confirm both the deterministic test and `e2e.sh` pass.
-
-## Style
-
-- TS: 4-space indent, no tabs, LF, prettier (`trailingComma: "none"`). YAML/MD use 2-space indent.
-- Rust: `rustfmt.toml` enforces 4-space spaces, LF.
-- Lint: ESLint for TS, Clippy for Rust, `cargo machete` for unused deps. All wired into `scripts/check.sh`.

README.md

@@ -2,24 +2,25 @@
 
 [![Check](https://github.com/schmelczer/vault-link/actions/workflows/check.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/check.yml)
 [![E2E tests](https://github.com/schmelczer/vault-link/actions/workflows/e2e.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/e2e.yml)
-[![Publish server Docker image](https://github.com/schmelczer/vault-link/actions/workflows/publish-server-docker.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/publish-server-docker.yml)
-[![Publish CLI](https://github.com/schmelczer/vault-link/actions/workflows/publish-cli-docker.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/publish-cli-docker.yml)
+[![Publish server Docker image](https://github.com/schmelczer/vault-link/actions/workflows/publish-docker.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/publish-docker.yml)
 [![Publish Obsidian plugin](https://github.com/schmelczer/vault-link/actions/workflows/publish-plugin.yml/badge.svg)](https://github.com/schmelczer/vault-link/actions/workflows/publish-plugin.yml)
 
+
 ## Develop
 
-### Set up Node.JS 25 with [nvm](https://github.com/nvm-sh/nvm)
+### Install [nvm](https://github.com/nvm-sh/nvm)
 
 - `curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.1/install.sh | bash`
-- `nvm install 25`
-- `nvm use 25`
-- Optionally, set the system-wide default: `nvm alias default 25`
+- `nvm install 22`
+- `nvm use 22`
+- Optionally set the system-wide default: `nvm alias default 22`
+
 
 ### Set up Rust
 
 - Install [`rustup`](https://rustup.rs): `curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh`
 - Install [`wasm-pack`](https://rustwasm.github.io/wasm-pack/installer): `curl https://rustwasm.github.io/wasm-pack/installer/init.sh -sSf | sh`
-- `cargo install cargo-insta sqlx-cli`
+- `cargo install cargo-insta sqlx-cli cargo-edit`
 
 ### Install Obsidian on Linux
 
@@ -30,50 +31,25 @@ flatpak install flathub md.obsidian.Obsidian
 flatpak run md.obsidian.Obsidian
 ```
 
-#### Run in development mode
-
-Start the server:
-
-```sh
-cargo install sqlx-cli
-cd sync-server
-cargo run config-e2e.yml
-```
-
-```sh
-cd frontend
-npm install
-npm run dev
-```
-
 ### Scripts
 
-#### Before pushing
-
-```sh
-scripts/check.sh --fix
-```
-
 #### Update HTTP API TS bindings
 
-```sh
+```sh 
 scripts/update-api-types.sh
 ```
 
-#### Publish new version
+#### Publish new version 
 
 ```sh
 scripts/bump-version.sh patch
 ```
 
+
 #### Run E2E tests
 
-```sh
-scripts/e2e.sh 8
+```sh 
+scripts/e2e.sh
 ```
 
 And to clean up the logs & database files, run `scripts/clean-up.sh`
-
-## Projects
-
-- [Sync server](./sync-server/README.md)

backend/.dockerignore

@@ -2,4 +2,5 @@ target
 Dockerfile
 .dockerignore
 databases
+sync_lib/pkg
 *.yml

backend/Cargo.toml

@@ -0,0 +1,75 @@
+[workspace]
+resolver = "2"
+members = [
+    "reconcile",
+    "sync_server", 
+    "sync_lib"
+]
+
+[workspace.package]
+rust-version = "1.83"
+authors = ["Andras Schmelczer <andras@schmelczer.dev>"]
+edition = "2024"
+license = "MIT"
+repository = "https://github.com/schmelczer/vault-link"
+version = "0.2.2"
+
+[workspace.dependencies]  
+serde = { version = "1.0.214", default-features = false, features = ["derive"] }
+thiserror = { version = "1.0.66", default-features = false  } 
+
+[profile.release]
+codegen-units = 1
+lto = true
+opt-level = 3
+strip="debuginfo"  # Keep some info for better panics
+
+[workspace.lints.rust]
+unsafe_code = "forbid"
+rust_2018_idioms = { level = "warn", priority = -1 }
+missing_debug_implementations = "warn"
+
+[workspace.lints.clippy]
+await_holding_lock = "warn"
+dbg_macro = "warn"
+empty_enum = "warn"
+enum_glob_use = "warn"
+exit = "warn"
+filter_map_next = "warn"
+fn_params_excessive_bools = "warn"
+if_let_mutex = "warn"
+imprecise_flops = "warn"
+inefficient_to_string = "warn"
+linkedlist = "warn"
+lossy_float_literal = "warn"
+macro_use_imports = "warn"
+match_on_vec_items = "warn"
+match_wildcard_for_single_variants = "warn"
+mem_forget = "warn"
+needless_borrow = "warn"
+needless_continue = "warn"
+option_option = "warn"
+rest_pat_in_fully_bound_structs = "warn"
+str_to_string = "warn"
+suboptimal_flops = "warn"
+todo = "warn"
+uninlined_format_args = "warn"
+unnested_or_patterns = "warn"
+unused_self = "warn"
+verbose_file_reads = "warn"
+
+cast_possible_truncation = { level = "allow", priority = 1 }
+doc_link_with_quotes = { level = "allow", priority = 1 }
+cast_sign_loss = { level = "allow", priority = 1 }
+cast_possible_wrap = { level = "allow", priority = 1 }
+struct_field_names = { level = "allow", priority = 1 }
+single_call_fn = { level = "allow", priority = 1 }
+absolute_paths = { level = "allow", priority = 1 }
+arithmetic_side_effects = { level = "allow", priority = 1 }
+similar_names = { level = "allow", priority = 1 }
+self_named_module_files = { level = "allow", priority = 1 }
+single_char_lifetime_names = { level = "allow", priority = 1 }
+missing_docs_in_private_items = { level = "allow", priority = 1 }
+question_mark_used =  { level = "allow", priority = 1 }
+implicit_return = { level = "allow", priority = 1 }
+pedantic = { level = "warn", priority = 0 }

backend/Dockerfile

@@ -0,0 +1,33 @@
+FROM rust:1.83 AS builder
+
+WORKDIR /usr/src/backend
+
+RUN apt update && apt install -y musl-tools
+RUN cargo install sqlx-cli
+
+COPY . .
+
+RUN sqlx database create --database-url sqlite://db.sqlite3
+RUN sqlx migrate run --source sync_server/src/app_state/database/migrations --database-url sqlite://db.sqlite3
+
+RUN cargo build --package sync_server --release --target x86_64-unknown-linux-musl
+
+# Runtime image
+FROM alpine:3.21.3
+
+LABEL org.opencontainers.image.authors="andras@schmelczer.dev"
+
+RUN apk add --no-cache curl
+
+COPY --from=builder /usr/src/backend/target/x86_64-unknown-linux-musl/release/sync_server /app/sync_server
+
+VOLUME /data
+EXPOSE 3000/tcp
+WORKDIR /data
+
+HEALTHCHECK \
+    --interval=30s \
+    --timeout=5s \
+    CMD curl -f http://localhost:3000/ping || exit 1
+
+ENTRYPOINT ["/app/sync_server"]

backend/config-e2e.yml

@@ -0,0 +1,22 @@
+database:
+  databases_directory_path: databases
+  max_connections: 12
+
+server:
+  host: 0.0.0.0
+  port: 3000
+  max_body_size_mb: 512
+  max_clients_per_vault: 256
+
+users:
+  user_tokens:
+  - name: admin
+    token: test-token-change-me
+    vaults: 
+      all: true
+
+  - name: test
+    token: other-test-token
+    vaults: 
+      allowed:
+       - default

backend/reconcile/Cargo.toml

@@ -0,0 +1,21 @@
+[package]
+name = "reconcile"
+version.workspace = true
+edition.workspace = true
+authors.workspace = true
+license.workspace = true
+repository.workspace = true
+
+[dependencies]
+serde = { version = "1.0.215", optional = true }
+
+[features]
+serde = [ "dep:serde" ]
+
+[dev-dependencies]
+insta = "1.41.1"
+pretty_assertions = "1.4.1"
+test-case = "3.3.1"
+
+[lints]
+workspace = true

backend/reconcile/src/diffs.rs

@@ -0,0 +1,2 @@
+pub mod myers;
+pub mod raw_operation;

backend/reconcile/src/diffs/myers.rs

@@ -0,0 +1,284 @@
+//! Taken from <https://github.com/mitsuhiko/similar/blob/7e15c44de11a1cd61e1149189929e189ef977fd8/src/algorithms/myers.rs>
+//! Myers' diff algorithm.
+//!
+//! * time: `O((N+M)D)`
+//! * space `O(N+M)`
+//!
+//! See [the original article by Eugene W. Myers](http://www.xmailserver.org/diff2.pdf)
+//! describing it.
+//!
+//! The implementation of this algorithm is based on the implementation by
+//! Brandon Williams.
+//!
+//! # Heuristics
+//!
+//! At present this implementation of Myers' does not implement any more
+//! advanced heuristics that would solve some pathological cases.  For instance
+//! passing two large and completely distinct sequences to the algorithm will
+//! make it spin without making reasonable progress.
+//! For potential improvements here see [similar#15](https://github.com/mitsuhiko/similar/issues/15).
+
+use std::{
+    ops::{Index, IndexMut, Range},
+    vec,
+};
+
+use super::raw_operation::RawOperation;
+use crate::{
+    tokenizer::token::Token,
+    utils::{common_prefix_len::common_prefix_len, common_suffix_len::common_suffix_len},
+};
+
+/// Myers' diff algorithm with deadline.
+///
+/// Diff `old`, between indices `old_range` and `new` between indices
+/// `new_range`.
+///
+/// This diff is done with an optional deadline that defines the maximal
+/// execution time permitted before it bails and falls back to an approximation.
+pub fn diff<T>(old: &[Token<T>], new: &[Token<T>]) -> Vec<RawOperation<T>>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    let max_d = (old.len() + new.len()).div_ceil(2) + 1;
+    let mut vb = V::new(max_d);
+    let mut vf = V::new(max_d);
+    let mut result: Vec<RawOperation<T>> = vec![];
+    conquer(
+        old,
+        0..old.len(),
+        new,
+        0..new.len(),
+        &mut vf,
+        &mut vb,
+        &mut result,
+    );
+    result
+}
+
+// A D-path is a path which starts at (0,0) that has exactly D non-diagonal
+// edges. All D-paths consist of a (D - 1)-path followed by a non-diagonal edge
+// and then a possibly empty sequence of diagonal edges called a snake.
+
+/// `V` contains the endpoints of the furthest reaching `D-paths`. For each
+/// recorded endpoint `(x,y)` in diagonal `k`, we only need to retain `x`
+/// because `y` can be computed from `x - k`. In other words, `V` is an array of
+/// integers where `V[k]` contains the row index of the endpoint of the furthest
+/// reaching path in diagonal `k`.
+///
+/// We can't use a traditional Vec to represent `V` since we use `k` as an index
+/// and it can take on negative values. So instead `V` is represented as a
+/// light-weight wrapper around a Vec plus an `offset` which is the maximum
+/// value `k` can take on in order to map negative `k`'s back to a value >= 0.
+#[derive(Debug)]
+struct V {
+    offset: isize,
+    v: Vec<usize>, // Look into initializing this to -1 and storing isize
+}
+
+impl V {
+    fn new(max_d: usize) -> Self {
+        Self {
+            offset: max_d as isize,
+            v: vec![0; 2 * max_d],
+        }
+    }
+
+    fn len(&self) -> usize { self.v.len() }
+}
+
+impl Index<isize> for V {
+    type Output = usize;
+
+    fn index(&self, index: isize) -> &Self::Output { &self.v[(index + self.offset) as usize] }
+}
+
+impl IndexMut<isize> for V {
+    fn index_mut(&mut self, index: isize) -> &mut Self::Output {
+        &mut self.v[(index + self.offset) as usize]
+    }
+}
+
+fn split_at(range: Range<usize>, at: usize) -> (Range<usize>, Range<usize>) {
+    (range.start..at, at..range.end)
+}
+
+/// A `Snake` is a sequence of diagonal edges in the edit graph.  Normally
+/// a snake has a start end end point (and it is possible for a snake to have
+/// a length of zero, meaning the start and end points are the same) however
+/// we do not need the end point which is why it's not implemented here.
+///
+/// The divide part of a divide-and-conquer strategy. A D-path has D+1 snakes
+/// some of which may be empty. The divide step requires finding the ceil(D/2) +
+/// 1 or middle snake of an optimal D-path. The idea for doing so is to
+/// simultaneously run the basic algorithm in both the forward and reverse
+/// directions until furthest reaching forward and reverse paths starting at
+/// opposing corners 'overlap'.
+fn find_middle_snake<T>(
+    old: &[Token<T>],
+    old_range: Range<usize>,
+    new: &[Token<T>],
+    new_range: Range<usize>,
+    vf: &mut V,
+    vb: &mut V,
+) -> Option<(usize, usize)>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    let n = old_range.len();
+    let m = new_range.len();
+
+    // By Lemma 1 in the paper, the optimal edit script length is odd or even as
+    // `delta` is odd or even.
+    let delta = n as isize - m as isize;
+    let odd = delta & 1 == 1;
+
+    // The initial point at (0, -1)
+    vf[1] = 0;
+    // The initial point at (N, M+1)
+    vb[1] = 0;
+
+    let d_max = (n + m).div_ceil(2) + 1;
+    assert!(vf.len() >= d_max);
+    assert!(vb.len() >= d_max);
+
+    for d in 0..d_max as isize {
+        // Forward path
+        for k in (-d..=d).rev().step_by(2) {
+            let mut x = if k == -d || (k != d && vf[k - 1] < vf[k + 1]) {
+                vf[k + 1]
+            } else {
+                vf[k - 1] + 1
+            };
+            let y = (x as isize - k) as usize;
+
+            // The coordinate of the start of a snake
+            let (x0, y0) = (x, y);
+            //  While these sequences are identical, keep moving through the
+            //  graph with no cost
+            if x < old_range.len() && y < new_range.len() {
+                let advance = common_prefix_len(
+                    old,
+                    old_range.start + x..old_range.end,
+                    new,
+                    new_range.start + y..new_range.end,
+                );
+                x += advance;
+            }
+
+            // This is the new best x value
+            vf[k] = x;
+
+            // Only check for connections from the forward search when N - M is
+            // odd and when there is a reciprocal k line coming from the other
+            // direction.
+            if odd && (k - delta).abs() <= (d - 1) {
+                // TODO optimize this so we don't have to compare against n
+                if vf[k] + vb[-(k - delta)] >= n {
+                    // Return the snake
+                    return Some((x0 + old_range.start, y0 + new_range.start));
+                }
+            }
+        }
+
+        // Backward path
+        for k in (-d..=d).rev().step_by(2) {
+            let mut x = if k == -d || (k != d && vb[k - 1] < vb[k + 1]) {
+                vb[k + 1]
+            } else {
+                vb[k - 1] + 1
+            };
+            let mut y = (x as isize - k) as usize;
+
+            // The coordinate of the start of a snake
+            if x < n && y < m {
+                let advance = common_suffix_len(
+                    old,
+                    old_range.start..old_range.start + n - x,
+                    new,
+                    new_range.start..new_range.start + m - y,
+                );
+                x += advance;
+                y += advance;
+            }
+
+            // This is the new best x value
+            vb[k] = x;
+
+            if !odd && (k - delta).abs() <= d {
+                // TODO optimize this so we don't have to compare against n
+                if vb[k] + vf[-(k - delta)] >= n {
+                    // Return the snake
+                    return Some((n - x + old_range.start, m - y + new_range.start));
+                }
+            }
+        }
+
+        // TODO: Maybe there's an opportunity to optimize and bail early?
+    }
+
+    None
+}
+
+fn conquer<T>(
+    old: &[Token<T>],
+    mut old_range: Range<usize>,
+    new: &[Token<T>],
+    mut new_range: Range<usize>,
+    vf: &mut V,
+    vb: &mut V,
+    result: &mut Vec<RawOperation<T>>,
+) where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    // Check for common prefix
+    let common_prefix_len = common_prefix_len(old, old_range.clone(), new, new_range.clone());
+    if common_prefix_len > 0 {
+        result.push(RawOperation::Equal(
+            old[old_range.start..old_range.start + common_prefix_len].to_vec(),
+        ));
+    }
+    old_range.start += common_prefix_len;
+    new_range.start += common_prefix_len;
+
+    // Check for common suffix
+    let common_suffix_len = common_suffix_len(old, old_range.clone(), new, new_range.clone());
+    let common_suffix = (
+        old_range.end - common_suffix_len,
+        new_range.end - common_suffix_len,
+    );
+    old_range.end -= common_suffix_len;
+    new_range.end -= common_suffix_len;
+
+    if old_range.is_empty() && new_range.is_empty() {
+        // Do nothing
+    } else if new_range.is_empty() {
+        result.push(RawOperation::Delete(
+            old[old_range.start..old_range.start + old_range.len()].to_vec(),
+        ));
+    } else if old_range.is_empty() {
+        result.push(RawOperation::Insert(
+            new[new_range.start..new_range.start + new_range.len()].to_vec(),
+        ));
+    } else if let Some((x_start, y_start)) =
+        find_middle_snake(old, old_range.clone(), new, new_range.clone(), vf, vb)
+    {
+        let (old_a, old_b) = split_at(old_range, x_start);
+        let (new_a, new_b) = split_at(new_range, y_start);
+        conquer(old, old_a, new, new_a, vf, vb, result);
+        conquer(old, old_b, new, new_b, vf, vb, result);
+    } else {
+        result.push(RawOperation::Delete(
+            old[old_range.start..old_range.end].to_vec(),
+        ));
+        result.push(RawOperation::Insert(
+            new[new_range.start..new_range.end].to_vec(),
+        ));
+    }
+
+    if common_suffix_len > 0 {
+        result.push(RawOperation::Equal(
+            old[common_suffix.0..common_suffix.0 + common_suffix_len].to_vec(),
+        ));
+    }
+}

backend/reconcile/src/diffs/raw_operation.rs

@@ -0,0 +1,48 @@
+use crate::tokenizer::token::Token;
+
+#[derive(Debug, Clone, PartialEq)]
+pub enum RawOperation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    Insert(Vec<Token<T>>),
+    Delete(Vec<Token<T>>),
+    Equal(Vec<Token<T>>),
+}
+
+impl<T> RawOperation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    pub fn tokens(&self) -> &Vec<Token<T>> {
+        match self {
+            RawOperation::Insert(tokens)
+            | RawOperation::Delete(tokens)
+            | RawOperation::Equal(tokens) => tokens,
+        }
+    }
+
+    pub fn original_text_length(&self) -> usize {
+        self.tokens().iter().map(Token::get_original_length).sum()
+    }
+
+    pub fn get_original_text(self) -> String { self.tokens().iter().map(Token::original).collect() }
+
+    /// Extends the operation with another operation if returning the new
+    /// operation. Only operations of the same type can be used to extend.
+    /// If the operations are of different types, returns None.
+    pub fn extend(self, other: RawOperation<T>) -> Option<RawOperation<T>> {
+        match (self, other) {
+            (RawOperation::Insert(tokens1), RawOperation::Insert(tokens2)) => Some(
+                RawOperation::Insert(tokens1.into_iter().chain(tokens2).collect()),
+            ),
+            (RawOperation::Delete(tokens1), RawOperation::Delete(tokens2)) => Some(
+                RawOperation::Delete(tokens1.into_iter().chain(tokens2).collect()),
+            ),
+            (RawOperation::Equal(tokens1), RawOperation::Equal(tokens2)) => Some(
+                RawOperation::Equal(tokens1.into_iter().chain(tokens2).collect()),
+            ),
+            _ => None,
+        }
+    }
+}

backend/reconcile/src/lib.rs

@@ -0,0 +1,7 @@
+mod diffs;
+mod operation_transformation;
+mod tokenizer;
+mod utils;
+
+pub use operation_transformation::{EditedText, reconcile, reconcile_with_tokenizer};
+pub use tokenizer::token::Token;

backend/reconcile/src/operation_transformation.rs

@@ -0,0 +1,226 @@
+mod edited_text;
+mod merge_context;
+mod operation;
+
+pub use edited_text::EditedText;
+pub use operation::Operation;
+
+use crate::tokenizer::Tokenizer;
+
+#[must_use]
+pub fn reconcile(original: &str, left: &str, right: &str) -> String {
+    // Common trivial cases
+    if left == right {
+        return left.to_owned();
+    }
+
+    if original == left {
+        return right.to_owned();
+    }
+
+    if original == right {
+        return left.to_owned();
+    }
+
+    // 3-way merge
+    let left_operations = EditedText::from_strings(original, left);
+    let right_operations = EditedText::from_strings(original, right);
+
+    let merged_operations = left_operations.merge(right_operations);
+    merged_operations.apply()
+}
+
+pub fn reconcile_with_tokenizer<F, T>(
+    original: &str,
+    left: &str,
+    right: &str,
+    tokenizer: &Tokenizer<T>,
+) -> String
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    let left_operations = EditedText::from_strings_with_tokenizer(original, left, tokenizer);
+    let right_operations = EditedText::from_strings_with_tokenizer(original, right, tokenizer);
+
+    let merged_operations = left_operations.merge(right_operations);
+    merged_operations.apply()
+}
+
+#[cfg(test)]
+mod test {
+    use std::{fs, ops::Range, path::Path};
+
+    use pretty_assertions::assert_eq;
+    use test_case::test_matrix;
+
+    use super::*;
+
+    #[test]
+    fn test_merges() {
+        // Both replaced one token but different
+        test_merge_both_ways(
+            "original_1 original_2 original_3",
+            "original_1 edit_1 original_3",
+            "original_1 original_2 edit_2",
+            "original_1 edit_1 edit_2",
+        );
+
+        // Both replaced the same one token
+        test_merge_both_ways(
+            "original_1 original_2 original_3",
+            "original_1 edit_1 original_3",
+            "original_1 edit_1 original_3",
+            "original_1 edit_1 original_3",
+        );
+
+        // One deleted a large range, the other deleted subranges and inserted as
+        // well
+        test_merge_both_ways(
+            "original_1 original_2 original_3 original_4 original_5",
+            "original_1 original_5",
+            "original_1 edit_1 original_3 edit_2 original_5",
+            "original_1 edit_1 edit_2 original_5",
+        );
+
+        // One deleted a large range, the other inserted and deleted a partially
+        // overlapping range
+        test_merge_both_ways(
+            "original_1 original_2 original_3 original_4 original_5",
+            "original_1 original_5",
+            "original_1 edit_1 original_3 edit_2",
+            "original_1 edit_1 edit_2",
+        );
+
+        // Merge a replace and an append
+        test_merge_both_ways("a b ", "c d ", "a b c d ", "c d c d ");
+
+        test_merge_both_ways("a b c d e", "a e", "a c e", "a e");
+
+        test_merge_both_ways("a 0 1 2 b", "a b", "a E 1 F b", "a E F b");
+
+        test_merge_both_ways(
+            "a this one delete b",
+            "a b",
+            "a my one change b",
+            "a my change b",
+        );
+
+        test_merge_both_ways(
+            "this stays, this is one big delete, don't touch this",
+            "this stays, don't touch this",
+            "this stays, my one change, don't touch this",
+            "this stays, my change, don't touch this",
+        );
+
+        test_merge_both_ways("1 2 3 4 5 6", "1 6", "1 2 4 ", "1 ");
+
+        test_merge_both_ways(
+            "hello world",
+            "hi, world",
+            "hello my friend!",
+            "hi, my friend!",
+        );
+
+        test_merge_both_ways(
+            "both delete the same word",
+            "both the same word",
+            "both the same word",
+            "both the same word",
+        );
+
+        test_merge_both_ways("    ", "it’s utf-8!", "   ", "it’s utf-8!");
+
+        test_merge_both_ways(
+            "both delete the same word but one a bit more",
+            "both the same word",
+            "both same word",
+            "both same word",
+        );
+
+        test_merge_both_ways(
+            "long text with one big delete and many small",
+            "long small",
+            "long with big and small",
+            "long small",
+        );
+    }
+
+    #[test]
+    fn test_reconcile_idempotent_inserts() {
+        // Both inserted the same prefix; this should get deduped
+        test_merge_both_ways(
+            "hi ",
+            "hi there ",
+            "hi there my friend ",
+            "hi there my friend ",
+        );
+
+        // The prefix of the 2nd appears on the 1st so it shouldn't get duplicated
+        test_merge_both_ways(
+            "hi ",
+            "hi there you ",
+            "hi there my friend ",
+            "hi there my friend you ",
+        );
+
+        test_merge_both_ways("a", "a b c", "a b c d", "a b c d");
+
+        test_merge_both_ways(
+            "      |7ca2b36d-6ee7-49eb-8eb1-d77e4cc1a001|   ",
+              "      |7ca2b36d-6ee7-49eb-8eb1-d77e4cc1a001|      |cd9195cc-103a-4f13-90c8-4fba0ba421ee|      |d39156cc-cfd6-42a8-b70a-75020896069d|      |fbad794c-9c47-41f2-a343-490284ecb5a0|      |dup|   ",
+             "       |7ca2b36d-6ee7-49eb-8eb1-d77e4cc1a001|      |cd9195cc-103a-4f13-90c8-4fba0ba421ee|      |dup|   ",
+            "      |7ca2b36d-6ee7-49eb-8eb1-d77e4cc1a001|      |cd9195cc-103a-4f13-90c8-4fba0ba421ee|      |d39156cc-cfd6-42a8-b70a-75020896069d|      |fbad794c-9c47-41f2-a343-490284ecb5a0|      |dup|      |dup|   ");
+    }
+
+    #[test_matrix( [
+        "pride_and_prejudice.txt",
+        "romeo_and_juliet.txt",
+        "room_with_a_view.txt",
+        "kun_lu.txt",
+
+    ],  [
+        "pride_and_prejudice.txt",
+        "romeo_and_juliet.txt",
+        "room_with_a_view.txt",
+        "kun_lu.txt"
+    ],  [
+        "pride_and_prejudice.txt",
+        "romeo_and_juliet.txt",
+        "room_with_a_view.txt",
+        "kun_lu.txt"
+    ], [0..10000, 10000..20000], [0..10000, 10000..20000], [0..10000, 10000..20000])]
+    fn test_merge_files_without_panic(
+        file_name_1: &str,
+        file_name_2: &str,
+        file_name_3: &str,
+        range_1: Range<usize>,
+        range_2: Range<usize>,
+        range_3: Range<usize>,
+    ) {
+        let files = [file_name_1, file_name_2, file_name_3];
+        let permutations = [range_1, range_2, range_3];
+
+        let root = Path::new("test/resources/");
+
+        let contents = files
+            .iter()
+            .zip(permutations.iter())
+            .map(|(file, range)| {
+                let path = root.join(file);
+                fs::read_to_string(&path)
+                    .unwrap()
+                    .chars()
+                    .skip(range.start)
+                    .take(range.end)
+                    .collect::<String>()
+            })
+            .collect::<Vec<_>>();
+
+        let _ = reconcile(&contents[0], &contents[1], &contents[2]);
+    }
+
+    fn test_merge_both_ways(original: &str, edit_1: &str, edit_2: &str, expected: &str) {
+        assert_eq!(reconcile(original, edit_1, edit_2), expected);
+        assert_eq!(reconcile(original, edit_2, edit_1), expected);
+    }
+}

backend/reconcile/src/operation_transformation/edited_text.rs

@@ -0,0 +1,305 @@
+use core::iter;
+
+#[cfg(feature = "serde")]
+use serde::{Deserialize, Serialize};
+
+use super::Operation;
+use crate::{
+    diffs::{myers::diff, raw_operation::RawOperation},
+    operation_transformation::merge_context::MergeContext,
+    tokenizer::{Tokenizer, word_tokenizer::word_tokenizer},
+    utils::{
+        merge_iters::MergeSorted as _, ordered_operation::OrderedOperation, side::Side,
+        string_builder::StringBuilder,
+    },
+};
+
+/// A sequence of operations that can be applied to a text document.
+/// `EditedText` supports merging two sequences of operations using the
+/// principle of Operational Transformation.
+///
+/// It's mainly created through the `from_strings` method, then merged with
+/// another `EditedText` derived from the same original text and then applied to
+/// the original text to get the reconciled text of concurrent edits.
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone, PartialEq, Default)]
+pub struct EditedText<'a, T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    text: &'a str,
+    operations: Vec<OrderedOperation<T>>,
+}
+
+impl<'a> EditedText<'a, String> {
+    /// Create an `EditedText` from the given original (old) and updated (new)
+    /// strings. The returned `EditedText` represents the changes from the
+    /// original to the updated text. When the return value is applied to
+    /// the original text, it will result in the updated text. The default
+    /// word tokenizer is used to tokenize the text which splits the text on
+    /// whitespaces.
+    #[must_use]
+    pub fn from_strings(original: &'a str, updated: &str) -> Self {
+        Self::from_strings_with_tokenizer(original, updated, &word_tokenizer)
+    }
+}
+
+impl<'a, T> EditedText<'a, T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    /// Create an `EditedText` from the given original (old) and updated (new)
+    /// strings. The returned `EditedText` represents the changes from the
+    /// original to the updated text. When the return value is applied to
+    /// the original text, it will result in the updated text. The tokenizer
+    /// function is used to tokenize the text.
+    pub fn from_strings_with_tokenizer(
+        original: &'a str,
+        updated: &str,
+        tokenizer: &Tokenizer<T>,
+    ) -> Self {
+        let original_tokens = (tokenizer)(original);
+        let updated_tokens = (tokenizer)(updated);
+
+        let diff: Vec<RawOperation<T>> = diff(&original_tokens, &updated_tokens);
+
+        Self::new(
+            original,
+            Self::cook_operations(Self::elongate_operations(diff)).collect(),
+        )
+    }
+
+    // Turn raw operations into ordered operations while keeping track of old & new
+    // indexes.
+    fn cook_operations<I>(raw_operations: I) -> impl Iterator<Item = OrderedOperation<T>>
+    where
+        I: IntoIterator<Item = RawOperation<T>>,
+    {
+        let mut new_index = 0; // this is the start index of the operation on the new text
+        let mut order = 0; // this is the start index of the operation on the original text
+
+        raw_operations.into_iter().flat_map(move |raw_operation| {
+            let length = raw_operation.original_text_length();
+
+            let operation = match raw_operation {
+                RawOperation::Equal(..) => {
+                    new_index += length;
+                    order += length;
+
+                    None
+                }
+                RawOperation::Insert(tokens) => {
+                    let op = Operation::create_insert(new_index, tokens)
+                        .map(|operation| OrderedOperation { order, operation });
+
+                    new_index += length;
+
+                    op
+                }
+                RawOperation::Delete(..) => {
+                    let op = if cfg!(debug_assertions) {
+                        Operation::create_delete_with_text(
+                            new_index,
+                            raw_operation.get_original_text(),
+                        )
+                    } else {
+                        Operation::create_delete(new_index, length)
+                    }
+                    .map(|operation| OrderedOperation { order, operation });
+
+                    order += length;
+
+                    op
+                }
+            };
+
+            operation.into_iter()
+        })
+    }
+
+    fn elongate_operations<I>(raw_operations: I) -> Vec<RawOperation<T>>
+    where
+        I: IntoIterator<Item = RawOperation<T>>,
+    {
+        let mut maybe_previous_insert: Option<RawOperation<T>> = None;
+        let mut maybe_previous_delete: Option<RawOperation<T>> = None;
+
+        let mut result: Vec<RawOperation<T>> = raw_operations
+            .into_iter()
+            .flat_map(|next| match next {
+                RawOperation::Insert(..) => {
+                    if let Some(prev) = maybe_previous_insert.take() {
+                        maybe_previous_insert = prev.extend(next);
+                    } else {
+                        maybe_previous_insert = Some(next);
+                    }
+
+                    Box::new(iter::empty()) as Box<dyn Iterator<Item = RawOperation<T>>>
+                }
+                RawOperation::Delete(..) => {
+                    if let Some(prev) = maybe_previous_delete.take() {
+                        maybe_previous_delete = prev.extend(next);
+                    } else {
+                        maybe_previous_delete = Some(next);
+                    }
+
+                    Box::new(iter::empty()) as Box<dyn Iterator<Item = RawOperation<T>>>
+                }
+                RawOperation::Equal(..) => Box::new(
+                    maybe_previous_insert
+                        .take()
+                        .into_iter()
+                        .chain(maybe_previous_delete.take())
+                        .chain(iter::once(next)),
+                )
+                    as Box<dyn Iterator<Item = RawOperation<T>>>,
+            })
+            .collect();
+
+        if let Some(prev) = maybe_previous_insert {
+            result.push(prev);
+        }
+
+        if let Some(prev) = maybe_previous_delete {
+            result.push(prev);
+        }
+
+        result
+    }
+
+    /// Create a new `EditedText` with the given operations.
+    /// The operations must be in the order in which they are meant to be
+    /// applied. The operations must not overlap.
+    fn new(text: &'a str, operations: Vec<OrderedOperation<T>>) -> Self {
+        operations
+            .iter()
+            .zip(operations.iter().skip(1))
+            .for_each(|(previous, next)| {
+                debug_assert!(
+                    previous.operation.start_index() <= next.operation.start_index(),
+                    "{} must not come before {} yet it does",
+                    previous.operation,
+                    next.operation
+                );
+            });
+
+        Self { text, operations }
+    }
+
+    #[must_use]
+    pub fn merge(self, other: Self) -> Self {
+        debug_assert_eq!(
+            self.text, other.text,
+            "`EditedText`-s must be derived from the same text to be mergable"
+        );
+
+        let mut left_merge_context = MergeContext::default();
+        let mut right_merge_context = MergeContext::default();
+
+        Self::new(
+            self.text,
+            self.operations
+                .into_iter()
+                .map(|op| (op, Side::Left))
+                .merge_sorted_by_key(
+                    other.operations.into_iter().map(|op| (op, Side::Right)),
+                    |(operation, _)| {
+                        (
+                            operation.order,
+                            // Operations on the left and right must come in the same order so that
+                            // inserts can be merged with other inserts and deletes with deletes.
+                            usize::from(matches!(operation.operation, Operation::Delete { .. })),
+                            // Make sure that the ordering is deterministic regardless which text
+                            // is left or right.
+                            match &operation.operation {
+                                Operation::Insert { text, .. } => text
+                                    .iter()
+                                    .map(super::super::tokenizer::token::Token::original)
+                                    .collect::<String>(),
+                                Operation::Delete {
+                                    deleted_character_count,
+                                    ..
+                                } => deleted_character_count.to_string(),
+                            },
+                        )
+                    },
+                )
+                .flat_map(|(OrderedOperation { order, operation }, side)| {
+                    match side {
+                        Side::Left => operation.merge_operations_with_context(
+                            &mut right_merge_context,
+                            &mut left_merge_context,
+                        ),
+                        Side::Right => operation.merge_operations_with_context(
+                            &mut left_merge_context,
+                            &mut right_merge_context,
+                        ),
+                    }
+                    .map(|operation| OrderedOperation { order, operation })
+                    .into_iter()
+                })
+                .collect(),
+        )
+    }
+
+    /// Apply the operations to the text and return the resulting text.
+    #[must_use]
+    pub fn apply(&self) -> String {
+        let mut builder: StringBuilder<'_> = StringBuilder::new(self.text);
+
+        for OrderedOperation { operation, .. } in &self.operations {
+            builder = operation.apply(builder);
+        }
+
+        builder.build()
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use std::env;
+
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    fn test_calculate_operations() {
+        let left = "hello world! How are you?  Adam";
+        let right = "Hello, my friend! How are you doing? Albert";
+
+        let operations = EditedText::from_strings(left, right);
+
+        insta::assert_debug_snapshot!(operations);
+
+        let new_right = operations.apply();
+        assert_eq!(new_right.to_string(), right);
+    }
+
+    #[test]
+    fn test_calculate_operations_with_no_diff() {
+        let text = "hello world!";
+
+        let operations = EditedText::from_strings(text, text);
+
+        assert_eq!(operations.operations.len(), 0);
+
+        let new_right = operations.apply();
+
+        assert_eq!(new_right.to_string(), text);
+    }
+
+    #[test]
+    fn test_calculate_operations_with_insert() {
+        let original = "hello world! ...";
+        let left = "Hello world! I'm Andras.";
+        let right = "Hello world! How are you?";
+        let expected = "Hello world! How are you? I'm Andras.";
+
+        let operations_1 = EditedText::from_strings(original, left);
+        let operations_2 = EditedText::from_strings(original, right);
+
+        let operations = operations_1.merge(operations_2);
+        assert_eq!(operations.apply(), expected);
+    }
+}

backend/reconcile/src/operation_transformation/merge_context.rs

@@ -0,0 +1,91 @@
+use core::fmt::Debug;
+
+use crate::operation_transformation::Operation;
+
+#[derive(Clone)]
+pub struct MergeContext<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    last_operation: Option<Operation<T>>,
+    pub shift: i64,
+}
+
+impl<T> Default for MergeContext<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    fn default() -> Self {
+        MergeContext {
+            last_operation: None,
+            shift: 0,
+        }
+    }
+}
+
+impl<T> Debug for MergeContext<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
+        f.debug_struct("MergeContext")
+            .field("last_operation", &self.last_operation)
+            .field("shift", &self.shift)
+            .finish()
+    }
+}
+
+impl<T> MergeContext<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    pub fn last_operation(&self) -> Option<&Operation<T>> { self.last_operation.as_ref() }
+
+    /// Replace the last delete operation (if there was one) with a new one
+    /// while applying it to the shift.
+    pub fn consume_and_replace_last_operation(&mut self, operation: Option<Operation<T>>) {
+        if let Some(Operation::Delete {
+            deleted_character_count,
+            ..
+        }) = self.last_operation.take()
+        {
+            self.shift -= deleted_character_count as i64;
+        }
+
+        self.last_operation = operation;
+    }
+
+    pub fn replace_last_operation(&mut self, operation: Option<Operation<T>>) {
+        self.last_operation = operation;
+    }
+
+    /// Remove the last operation (if there was one) in case it is behind the
+    /// threshold operation. This changes the shift in case the last operation
+    /// was a delete.
+    pub fn consume_last_operation_if_it_is_too_behind(
+        &mut self,
+        threshold_operation: &Operation<T>,
+    ) {
+        if let Some(last_operation) = self.last_operation.as_ref() {
+            if let Operation::Delete {
+                deleted_character_count,
+                ..
+            } = last_operation
+            {
+                if threshold_operation.start_index() as i64 + self.shift
+                    > last_operation.end_index() as i64
+                {
+                    self.shift -= *deleted_character_count as i64;
+                    self.last_operation = None;
+                }
+            } else if let Operation::Insert { .. } = last_operation {
+                if threshold_operation.start_index() as i64 + self.shift
+                    - last_operation.len() as i64
+                    > last_operation.end_index() as i64
+                {
+                    self.last_operation = None;
+                }
+            }
+        }
+    }
+}

backend/reconcile/src/operation_transformation/operation.rs

@@ -0,0 +1,377 @@
+use core::fmt::{Debug, Display};
+use std::ops::Range;
+
+#[cfg(feature = "serde")]
+use serde::{Deserialize, Serialize};
+
+use super::merge_context::MergeContext;
+use crate::{
+    Token,
+    utils::{
+        find_longest_prefix_contained_within::find_longest_prefix_contained_within,
+        string_builder::StringBuilder,
+    },
+};
+
+/// Represents a change that can be applied to a text document.
+/// Operation is tied to a `ropey::Rope` and is mainly expected to be
+/// created by `EditedText`.
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Clone, PartialEq)]
+pub enum Operation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    Insert {
+        index: usize,
+        text: Vec<Token<T>>,
+    },
+
+    Delete {
+        index: usize,
+        deleted_character_count: usize,
+
+        #[cfg(debug_assertions)]
+        deleted_text: Option<String>,
+    },
+}
+
+impl<T> Operation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    /// Creates an insert operation with the given index and text.
+    /// If the text is empty (meaning that the operation would be a no-op),
+    /// returns None.
+    pub fn create_insert(index: usize, text: Vec<Token<T>>) -> Option<Self> {
+        if text.is_empty() {
+            return None;
+        }
+
+        Some(Operation::Insert { index, text })
+    }
+
+    /// Creates a delete operation with the given index and number of
+    /// to-be-deleted characters. If the operation would delete 0 (meaning
+    /// that the operation would be a no-op), returns None.
+    pub fn create_delete(index: usize, deleted_character_count: usize) -> Option<Self> {
+        if deleted_character_count == 0 {
+            return None;
+        }
+
+        Some(Operation::Delete {
+            index,
+            deleted_character_count,
+
+            #[cfg(debug_assertions)]
+            deleted_text: None,
+        })
+    }
+
+    pub fn create_delete_with_text(index: usize, text: String) -> Option<Self> {
+        if text.is_empty() {
+            return None;
+        }
+
+        Some(Operation::Delete {
+            index,
+            deleted_character_count: text.chars().count(),
+
+            #[cfg(debug_assertions)]
+            deleted_text: Some(text),
+        })
+    }
+
+    /// Applies the operation to the given `StringBuilder`, returning the
+    /// modified `StringBuilder`.
+    ///
+    /// When compiled in debug mode, panics if a delete operation is attempted
+    /// on a range of text that does not match the text to be deleted.
+    pub fn apply<'a>(&self, mut builder: StringBuilder<'a>) -> StringBuilder<'a> {
+        match self {
+            Operation::Insert { text, .. } => builder.insert(
+                self.start_index(),
+                &text.iter().map(Token::original).collect::<String>(),
+            ),
+            Operation::Delete {
+                #[cfg(debug_assertions)]
+                deleted_text,
+                ..
+            } => {
+                #[cfg(debug_assertions)]
+                debug_assert!(
+                    deleted_text
+                        .as_ref()
+                        .is_none_or(|text| builder.get_slice(self.range()) == *text),
+                    "Text to delete does not match the text in the range"
+                );
+
+                builder.delete(self.range());
+            }
+        }
+
+        builder
+    }
+
+    /// Returns the index of the first character that the operation affects.
+    pub fn start_index(&self) -> usize {
+        match self {
+            Operation::Insert { index, .. } | Operation::Delete { index, .. } => *index,
+        }
+    }
+
+    /// Returns the index of the last character that the operation affects.
+    pub fn end_index(&self) -> usize {
+        debug_assert!(
+            self.len() > 0,
+            " len() must be greater than 0 because operations must be non-empty"
+        );
+        self.start_index() + self.len() - 1
+    }
+
+    /// Returns the range of indices of characters that the operation affects.
+    #[allow(clippy::range_plus_one)]
+    pub fn range(&self) -> Range<usize> { self.start_index()..self.end_index() + 1 }
+
+    /// Returns the number of affected characters. It is always greater than 0
+    /// because empty operations cannot be created.
+    pub fn len(&self) -> usize {
+        match self {
+            Operation::Insert { text, .. } => text.iter().map(Token::get_original_length).sum(),
+            Operation::Delete {
+                deleted_character_count,
+                ..
+            } => *deleted_character_count,
+        }
+    }
+
+    /// Creates a new operation with the same type and text but with the given
+    /// index.
+    pub fn with_index(self, index: usize) -> Self {
+        match self {
+            Operation::Insert { text, .. } => Operation::Insert { index, text },
+            Operation::Delete {
+                deleted_character_count,
+
+                #[cfg(debug_assertions)]
+                deleted_text,
+                ..
+            } => Operation::Delete {
+                index,
+                deleted_character_count,
+
+                #[cfg(debug_assertions)]
+                deleted_text,
+            },
+        }
+    }
+
+    /// Creates a new operation with the same type and text but with the index
+    /// shifted by the given offset. The offset can be negative but the
+    /// resulting index must be non-negative.
+    ///
+    /// # Panics
+    ///
+    /// In debug mode, panics if the resulting index is negative.
+    pub fn with_shifted_index(self, offset: i64) -> Self {
+        let index = self.start_index() as i64 + offset;
+        debug_assert!(index >= 0, "Shifted index must be non-negative");
+
+        self.with_index(index as usize)
+    }
+
+    /// Merges the operation with the given context, producing a new operation
+    /// and updating the context. This implements a comples FSM that handles
+    /// the merging of operations in a way that is consistent with the text.
+    /// The contexts are updated in-place.
+    pub fn merge_operations_with_context(
+        self,
+        affecting_context: &mut MergeContext<T>,
+        produced_context: &mut MergeContext<T>,
+    ) -> Option<Operation<T>> {
+        affecting_context.consume_last_operation_if_it_is_too_behind(&self);
+        let operation = self.with_shifted_index(affecting_context.shift);
+
+        match (operation, affecting_context.last_operation()) {
+            (operation @ Operation::Insert { .. }, None) => {
+                produced_context.shift += operation.len() as i64;
+                produced_context.consume_and_replace_last_operation(Some(operation.clone()));
+                Some(operation)
+            }
+
+            (
+                Operation::Insert { text, index },
+                Some(Operation::Insert {
+                    text: previous_inserted_text,
+                    ..
+                }),
+            ) => {
+                // In case the current insert's prefix appears in the previously inserted text,
+                // we can trim the current insert to only include the non-overlapping part.
+                // This way, we don't end up duplicating text.
+                let offset_in_tokens =
+                    find_longest_prefix_contained_within(previous_inserted_text, &text);
+                let offset_in_length = text
+                    .iter()
+                    .take(offset_in_tokens)
+                    .map(Token::get_original_length)
+                    .sum::<usize>();
+                let trimmed_operation =
+                    Operation::create_insert(index, text[offset_in_tokens..].to_vec());
+
+                affecting_context.shift -= offset_in_length as i64;
+                produced_context.shift += trimmed_operation
+                    .as_ref()
+                    .map(Operation::len)
+                    .unwrap_or_default() as i64;
+                produced_context.consume_and_replace_last_operation(trimmed_operation.clone());
+
+                trimmed_operation
+            }
+
+            (operation @ Operation::Delete { .. }, None | Some(Operation::Insert { .. })) => {
+                produced_context.consume_and_replace_last_operation(Some(operation.clone()));
+                Some(operation)
+            }
+
+            (
+                operation @ Operation::Insert { .. },
+                Some(last_delete @ Operation::Delete { .. }),
+            ) => {
+                produced_context.shift += operation.len() as i64;
+
+                debug_assert!(
+                    last_delete.range().contains(&operation.start_index()),
+                    "There is a last delete ({last_delete}) but the operation ({operation}) is \
+                     not contained in it"
+                );
+
+                let difference = operation.start_index() as i64 - last_delete.start_index() as i64;
+
+                let moved_operation = operation.with_index(last_delete.start_index());
+
+                affecting_context.replace_last_operation(Operation::create_delete(
+                    moved_operation.end_index() + 1,
+                    (last_delete.len() as i64 - difference) as usize,
+                ));
+                affecting_context.shift -= difference;
+
+                produced_context.consume_and_replace_last_operation(Some(moved_operation.clone()));
+
+                Some(moved_operation)
+            }
+
+            (
+                operation @ Operation::Delete { .. },
+                Some(last_delete @ Operation::Delete { .. }),
+            ) => {
+                debug_assert!(
+                    last_delete.range().contains(&operation.start_index()),
+                    "There is a last delete ({last_delete}) but the operation ({operation}) is \
+                     not contained in it"
+                );
+
+                let difference = operation.start_index() as i64 - last_delete.start_index() as i64;
+
+                let updated_delete = Operation::create_delete(
+                    last_delete.start_index(),
+                    0.max(operation.end_index() as i64 - last_delete.end_index() as i64) as usize,
+                );
+
+                affecting_context.replace_last_operation(Operation::create_delete(
+                    last_delete.start_index(),
+                    0.max(last_delete.end_index() as i64 - operation.end_index() as i64) as usize,
+                ));
+                affecting_context.shift -= difference;
+
+                produced_context.consume_and_replace_last_operation(updated_delete.clone());
+
+                updated_delete
+            }
+        }
+    }
+}
+
+impl<T> Display for Operation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
+        match self {
+            Operation::Insert { index, text } => {
+                write!(
+                    f,
+                    "<insert '{}' from index {}>",
+                    text.iter().map(Token::original).collect::<String>(),
+                    index
+                )
+            }
+            Operation::Delete {
+                index,
+                deleted_character_count,
+
+                #[cfg(debug_assertions)]
+                deleted_text,
+            } => {
+                #[cfg(debug_assertions)]
+                write!(
+                    f,
+                    "<delete {} from index {}>",
+                    deleted_text
+                        .as_ref()
+                        .map(|text| format!("'{text}'"))
+                        .unwrap_or(format!("{deleted_character_count} characters")),
+                    index
+                )?;
+
+                #[cfg(not(debug_assertions))]
+                write!(
+                    f,
+                    "<delete {deleted_character_count} characters from index {index}>",
+                )?;
+
+                Ok(())
+            }
+        }
+    }
+}
+
+impl<T> Debug for Operation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result { write!(f, "{self}") }
+}
+
+#[cfg(test)]
+mod tests {
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    #[should_panic(expected = "Shifted index must be non-negative")]
+    fn test_shifting_error() {
+        insta::assert_debug_snapshot!(
+            Operation::create_insert(1, vec!["hi".into()])
+                .unwrap()
+                .with_shifted_index(-2)
+        );
+    }
+
+    #[test]
+    fn test_apply_delete_with_create() {
+        let builder = StringBuilder::new("hello world");
+        let operation = Operation::<()>::create_delete_with_text(5, " world".to_owned()).unwrap();
+
+        assert_eq!(operation.apply(builder).build(), "hello");
+    }
+
+    #[test]
+    fn test_apply_insert() {
+        let builder = StringBuilder::new("hello");
+        let operation = Operation::create_insert(5, vec![" my friend".into()]).unwrap();
+
+        assert_eq!(operation.apply(builder).build(), "hello my friend");
+    }
+}

backend/reconcile/src/operation_transformation/snapshots/reconcile__operation_transformation__edited_text__tests__calculate_operations.snap

@@ -0,0 +1,26 @@
+---
+source: reconcile/src/operation_transformation/edited_text.rs
+expression: operations
+snapshot_kind: text
+---
+EditedText {
+    text: "hello world! How are you?  Adam",
+    operations: [
+        OrderedOperation {
+            order: 0,
+            operation: <insert 'Hello, my friend!' from index 0>,
+        },
+        OrderedOperation {
+            order: 0,
+            operation: <delete 'hello world!' from index 17>,
+        },
+        OrderedOperation {
+            order: 20,
+            operation: <insert ' you doing? Albert' from index 25>,
+        },
+        OrderedOperation {
+            order: 20,
+            operation: <delete ' you?  Adam' from index 43>,
+        },
+    ],
+}

backend/reconcile/src/operation_transformation/snapshots/reconcile__operations__edited_text__tests__calculate_operations.snap

@@ -0,0 +1,61 @@
+---
+source: reconcile/src/operations/edited_text.rs
+expression: operations
+snapshot_kind: text
+---
+EditedText {
+    text: "hello world! How are you?  Adam",
+    operations: [
+        OrderedOperation {
+            order: 0,
+            operation: Insert {
+                index: 0,
+                text: "Hello, my friend! ",
+            },
+        },
+        OrderedOperation {
+            order: 0,
+            operation: Delete {
+                index: 18,
+                deleted_character_count: 13,
+                deleted_text: Some(
+                    "hello world! ",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 21,
+            operation: Delete {
+                index: 26,
+                deleted_character_count: 5,
+                deleted_text: Some(
+                    "you? ",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 26,
+            operation: Delete {
+                index: 26,
+                deleted_character_count: 5,
+                deleted_text: Some(
+                    " Adam",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 31,
+            operation: Insert {
+                index: 26,
+                text: "you ",
+            },
+        },
+        OrderedOperation {
+            order: 31,
+            operation: Insert {
+                index: 30,
+                text: "doing? Albert",
+            },
+        },
+    ],
+}

backend/reconcile/src/operation_transformation/snapshots/reconcile__operations__operation_sequence__tests__calculate_operations.snap

@@ -0,0 +1,60 @@
+---
+source: reconcile/src/operations/operation_sequence.rs
+expression: operations
+snapshot_kind: text
+---
+EditedText {
+    operations: [
+        OrderedOperation {
+            order: 0,
+            operation: Insert {
+                index: 0,
+                text: "Hello, my friend! ",
+            },
+        },
+        OrderedOperation {
+            order: 0,
+            operation: Delete {
+                index: 18,
+                deleted_character_count: 13,
+                deleted_text: Some(
+                    "hello world! ",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 21,
+            operation: Delete {
+                index: 26,
+                deleted_character_count: 5,
+                deleted_text: Some(
+                    "you? ",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 26,
+            operation: Delete {
+                index: 26,
+                deleted_character_count: 5,
+                deleted_text: Some(
+                    " Adam",
+                ),
+            },
+        },
+        OrderedOperation {
+            order: 31,
+            operation: Insert {
+                index: 26,
+                text: "you ",
+            },
+        },
+        OrderedOperation {
+            order: 31,
+            operation: Insert {
+                index: 30,
+                text: "doing? Albert",
+            },
+        },
+    ],
+}

backend/reconcile/src/tokenizer.rs

@@ -0,0 +1,6 @@
+use token::Token;
+
+pub mod token;
+pub mod word_tokenizer;
+
+pub type Tokenizer<T> = dyn Fn(&str) -> Vec<Token<T>>;

backend/reconcile/src/tokenizer/snapshots/reconcile__tokenizer__word_tokenizer__tests__with_snapshots-2.snap

@@ -0,0 +1,6 @@
+---
+source: reconcile/src/tokenizer/word_tokenizer.rs
+expression: "word_tokenizer(\"\")"
+snapshot_kind: text
+---
+[]

backend/reconcile/src/tokenizer/snapshots/reconcile__tokenizer__word_tokenizer__tests__with_snapshots-3.snap

@@ -0,0 +1,15 @@
+---
+source: reconcile/src/tokenizer/word_tokenizer.rs
+expression: "word_tokenizer(\" what? \")"
+snapshot_kind: text
+---
+[
+    Token {
+        normalised: "what?",
+        original: " what?",
+    },
+    Token {
+        normalised: "",
+        original: " ",
+    },
+]

backend/reconcile/src/tokenizer/snapshots/reconcile__tokenizer__word_tokenizer__tests__with_snapshots-4.snap

@@ -0,0 +1,23 @@
+---
+source: reconcile/src/tokenizer/word_tokenizer.rs
+expression: "word_tokenizer(\" hello, \\nwhere are you?\")"
+snapshot_kind: text
+---
+[
+    Token {
+        normalised: "hello,",
+        original: " hello,",
+    },
+    Token {
+        normalised: "where",
+        original: " \nwhere",
+    },
+    Token {
+        normalised: "are",
+        original: " are",
+    },
+    Token {
+        normalised: "you?",
+        original: " you?",
+    },
+]

backend/reconcile/src/tokenizer/snapshots/reconcile__tokenizer__word_tokenizer__tests__with_snapshots.snap

@@ -0,0 +1,15 @@
+---
+source: reconcile/src/tokenizer/word_tokenizer.rs
+expression: "word_tokenizer(\"Hi there!\")"
+snapshot_kind: text
+---
+[
+    Token {
+        normalised: "Hi",
+        original: "Hi",
+    },
+    Token {
+        normalised: "there!",
+        original: " there!",
+    },
+]

backend/reconcile/src/tokenizer/token.rs

@@ -0,0 +1,44 @@
+#[cfg(feature = "serde")]
+use serde::{Deserialize, Serialize};
+
+/// A token is a string that has been normalised in some way.
+/// The normalised form is used for comparison, while the original form is used
+/// for applying Operations.
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
+pub struct Token<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    normalised: T,
+    original: String,
+}
+
+impl From<&str> for Token<String> {
+    fn from(s: &str) -> Self { Token::new(s.trim().to_owned(), s.to_owned()) }
+}
+
+impl<T> Token<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    pub fn new(normalised: T, original: String) -> Self {
+        Token {
+            normalised,
+            original,
+        }
+    }
+
+    pub fn original(&self) -> &str { &self.original }
+
+    pub fn normalised(&self) -> &T { &self.normalised }
+
+    pub fn get_original_length(&self) -> usize { self.original.chars().count() }
+}
+
+impl<T> PartialEq for Token<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    fn eq(&self, other: &Self) -> bool { self.normalised == other.normalised }
+}

backend/reconcile/src/tokenizer/word_tokenizer.rs

@@ -0,0 +1,48 @@
+use super::token::Token;
+
+/// Splits on whitespace keeping the leading whitespace.
+///
+///     
+/// ## Example
+///
+/// "Hi there!" -> ["Hi", " there!"]
+pub fn word_tokenizer(text: &str) -> Vec<Token<String>> {
+    let mut result: Vec<Token<String>> = Vec::new();
+
+    let mut last_whitespace = 0;
+    let mut previous_char_is_whitespace = true;
+
+    for (i, c) in text.char_indices() {
+        let is_current_char_whitespace = c.is_whitespace();
+        if !previous_char_is_whitespace && is_current_char_whitespace {
+            result.push(text[last_whitespace..i].into());
+            last_whitespace = i;
+        }
+
+        previous_char_is_whitespace = is_current_char_whitespace;
+    }
+
+    if last_whitespace < text.len() {
+        result.push(text[last_whitespace..].into());
+    }
+
+    result
+}
+
+#[cfg(test)]
+mod tests {
+    use insta::assert_debug_snapshot;
+
+    use super::*;
+
+    #[test]
+    fn test_with_snapshots() {
+        assert_debug_snapshot!(word_tokenizer("Hi there!"));
+
+        assert_debug_snapshot!(word_tokenizer(""));
+
+        assert_debug_snapshot!(word_tokenizer(" what? "));
+
+        assert_debug_snapshot!(word_tokenizer(" hello, \nwhere are you?"));
+    }
+}

backend/reconcile/src/utils.rs

@@ -0,0 +1,7 @@
+pub mod common_prefix_len;
+pub mod common_suffix_len;
+pub mod find_longest_prefix_contained_within;
+pub mod merge_iters;
+pub mod ordered_operation;
+pub mod side;
+pub mod string_builder;

backend/reconcile/src/utils/common_prefix_len.rs

@@ -0,0 +1,47 @@
+use core::ops::{Index, Range};
+
+/// Given two lookups and ranges calculates the length of the common prefix.
+/// Copied from <https://github.com/mitsuhiko/similar/blob/7e15c44de11a1cd61e1149189929e189ef977fd8/src/algorithms/utils.rs>
+pub fn common_prefix_len<Old, New>(
+    old: &Old,
+    old_range: Range<usize>,
+    new: &New,
+    new_range: Range<usize>,
+) -> usize
+where
+    Old: Index<usize> + ?Sized,
+    New: Index<usize> + ?Sized,
+    New::Output: PartialEq<Old::Output>,
+{
+    new_range
+        .zip(old_range)
+        .take_while(|x| new[x.0] == old[x.1])
+        .count()
+}
+
+#[cfg(test)]
+mod tests {
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    fn test_common_prefix_len() {
+        assert_eq!(
+            common_prefix_len("".as_bytes(), 0..0, "".as_bytes(), 0..0),
+            0
+        );
+        assert_eq!(
+            common_prefix_len("foobarbaz".as_bytes(), 0..9, "foobarblah".as_bytes(), 0..10),
+            7
+        );
+        assert_eq!(
+            common_prefix_len("foobarbaz".as_bytes(), 0..9, "blablabla".as_bytes(), 0..9),
+            0
+        );
+        assert_eq!(
+            common_prefix_len("foobarbaz".as_bytes(), 3..9, "foobarblah".as_bytes(), 3..10),
+            4
+        );
+    }
+}

backend/reconcile/src/utils/common_suffix_len.rs

@@ -0,0 +1,48 @@
+use core::ops::{Index, Range};
+
+/// Given two lookups and ranges calculates the length of common suffix.
+/// Copied from <https://github.com/mitsuhiko/similar/blob/7e15c44de11a1cd61e1149189929e189ef977fd8/src/algorithms/utils.rs>
+pub fn common_suffix_len<Old, New>(
+    old: &Old,
+    old_range: Range<usize>,
+    new: &New,
+    new_range: Range<usize>,
+) -> usize
+where
+    Old: Index<usize> + ?Sized,
+    New: Index<usize> + ?Sized,
+    New::Output: PartialEq<Old::Output>,
+{
+    new_range
+        .rev()
+        .zip(old_range.rev())
+        .take_while(|x| new[x.0] == old[x.1])
+        .count()
+}
+
+#[cfg(test)]
+mod tests {
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    fn test_common_suffix_len() {
+        assert_eq!(
+            common_suffix_len("".as_bytes(), 0..0, "".as_bytes(), 0..0),
+            0
+        );
+        assert_eq!(
+            common_suffix_len("1234".as_bytes(), 0..4, "X0001234".as_bytes(), 0..8),
+            4
+        );
+        assert_eq!(
+            common_suffix_len("1234".as_bytes(), 0..4, "Xxxx".as_bytes(), 0..4),
+            0
+        );
+        assert_eq!(
+            common_suffix_len("1234".as_bytes(), 2..4, "01234".as_bytes(), 2..5),
+            2
+        );
+    }
+}

backend/reconcile/src/utils/find_longest_prefix_contained_within.rs

@@ -0,0 +1,103 @@
+use crate::Token;
+
+/// Given two lists of tokens, returns `length` where `old` list somewhere
+/// within contains the `length` prefix of the `new` list.
+///
+/// ## Example
+///
+/// ```not_rust
+/// old: [0, 1, 9, 0, 2, 5]
+/// new:       [9, 0, 2, 5, 1]
+/// ```
+/// > results in an length of 4
+///
+///
+/// ```not_rust
+/// old: [0, 1, 9, 0, 2, 5]
+/// new:          [0, 2]
+/// ```
+/// > results in an length of 2
+///
+/// ```not_rust
+/// old: [0, 1, 9, 0, 2, 5]
+/// new:          [0, 4]
+/// ```
+/// > results in an length of 1
+pub fn find_longest_prefix_contained_within<T>(old: &[Token<T>], new: &[Token<T>]) -> usize
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    let max_possible = new.len().min(old.len());
+
+    for len in (1..=max_possible).rev() {
+        let prefix = &new[..len];
+        if old.windows(len).any(|window| window == prefix) {
+            return len;
+        }
+    }
+
+    0
+}
+
+#[cfg(test)]
+mod tests {
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    fn test_common_overlap() {
+        assert_eq!(
+            find_longest_prefix_contained_within(&["".into()], &["".into()]),
+            1
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "b".into(), "c".into()],
+                &["b".into(), "c".into(), "a".into()]
+            ),
+            2
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "b".into(), "c".into()],
+                &["b".into(), "c".into()]
+            ),
+            2
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "b".into(), "c".into()],
+                &["b".into()]
+            ),
+            1
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "b".into(), "c".into(), "b".into(), "a".into()],
+                &["b".into(), "a".into()]
+            ),
+            2
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "a".into(), "a".into()],
+                &["a".into(), "b".into(), "c".into()]
+            ),
+            1
+        );
+
+        assert_eq!(
+            find_longest_prefix_contained_within(
+                &["a".into(), "b".into(), "c".into()],
+                &["d".into(), "e".into(), "a".into()]
+            ),
+            0
+        );
+    }
+}

backend/reconcile/src/utils/merge_iters.rs

@@ -0,0 +1,86 @@
+use core::{cmp::Ordering, iter::Peekable};
+
+pub struct MergeAscending<L, R, F, O>
+where
+    L: Iterator<Item = R::Item>,
+    R: Iterator,
+    F: Fn(&R::Item) -> O,
+    O: PartialOrd,
+{
+    left: Peekable<L>,
+    right: Peekable<R>,
+    get_key: F,
+}
+
+impl<L, R, F, O> MergeAscending<L, R, F, O>
+where
+    L: Iterator<Item = R::Item>,
+    R: Iterator,
+    F: Fn(&R::Item) -> O,
+    O: PartialOrd,
+{
+    fn new(left: L, right: R, get_key: F) -> Self {
+        MergeAscending {
+            left: left.peekable(),
+            right: right.peekable(),
+            get_key,
+        }
+    }
+}
+
+impl<L, R, F, O> Iterator for MergeAscending<L, R, F, O>
+where
+    L: Iterator<Item = R::Item>,
+    R: Iterator,
+    F: Fn(&R::Item) -> O,
+    O: PartialOrd,
+{
+    type Item = L::Item;
+
+    fn next(&mut self) -> Option<L::Item> {
+        let order = match (self.left.peek(), self.right.peek()) {
+            (Some(l), Some(r)) => (self.get_key)(l).partial_cmp(&(self.get_key)(r)),
+            (Some(_), None) => Some(Ordering::Less),
+            (None, Some(_)) => Some(Ordering::Greater),
+            (None, None) => return None,
+        };
+
+        match order {
+            Some(Ordering::Less | Ordering::Equal) | None => self.left.next(),
+            Some(Ordering::Greater) => self.right.next(),
+        }
+    }
+}
+
+pub trait MergeSorted: Iterator {
+    fn merge_sorted_by_key<R, F, O>(self, other: R, get_key: F) -> MergeAscending<Self, R, F, O>
+    where
+        Self: Sized,
+        R: Iterator<Item = Self::Item>,
+        F: Fn(&Self::Item) -> O,
+        O: PartialOrd,
+    {
+        MergeAscending::new(self, other, get_key)
+    }
+}
+
+impl<T: ?Sized> MergeSorted for T where T: Iterator {}
+
+#[cfg(test)]
+mod tests {
+    use pretty_assertions::assert_eq;
+
+    use super::*;
+
+    #[test]
+    fn test_merge_sorted_by_key() {
+        let left = [9, 7, 5, 3, 1];
+        let right = [7, 6, 5, 4, 3];
+
+        let result: Vec<i32> = left
+            .into_iter()
+            .merge_sorted_by_key(right.into_iter(), |x| -1 * x)
+            .collect();
+        assert_eq!(result, vec![9, 7, 7, 6, 5, 5, 4, 3, 3, 1]);
+    }
+}

backend/reconcile/src/utils/ordered_operation.rs

@@ -0,0 +1,14 @@
+#[cfg(feature = "serde")]
+use serde::{Deserialize, Serialize};
+
+use crate::operation_transformation::Operation;
+
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone, PartialEq)]
+pub struct OrderedOperation<T>
+where
+    T: PartialEq + Clone + std::fmt::Debug,
+{
+    pub order: usize,
+    pub operation: Operation<T>,
+}

backend/reconcile/src/utils/side.rs

@@ -0,0 +1,4 @@
+pub enum Side {
+    Left,
+    Right,
+}

backend/reconcile/src/utils/string_builder.rs

@@ -0,0 +1,111 @@
+use core::ops::Range;
+
+/// A helper for building a string in order based on an original string and a
+/// series of insertions and deletions applied to it. It is safe to use with
+/// UTF-8 strings as all operations are based on character indices.
+#[derive(Debug, Clone)]
+pub struct StringBuilder<'a> {
+    original: &'a str,
+    last_old_char_index: usize,
+    buffer: String,
+}
+
+impl StringBuilder<'_> {
+    pub fn new(original: &str) -> StringBuilder<'_> {
+        StringBuilder {
+            original,
+            last_old_char_index: 0,
+            buffer: String::with_capacity(original.len()),
+        }
+    }
+
+    /// Insert a string at the given index after copying the original string up
+    /// to that index from the last insertion or deletion.
+    pub fn insert(&mut self, from: usize, text: &str) {
+        self.copy_until(from);
+        self.buffer.push_str(text);
+    }
+
+    /// Delete a string at the given index after copying the original string up
+    /// to that index from the last insertion or deletion.
+    pub fn delete(&mut self, range: core::ops::Range<usize>) {
+        self.copy_until(range.start);
+        self.last_old_char_index += range.len();
+    }
+
+    fn copy_until(&mut self, index: usize) {
+        let current_char_count = self.buffer.chars().count();
+        debug_assert!(
+            index >= current_char_count,
+            "String builder only support building in order"
+        );
+
+        let jump = index - current_char_count;
+
+        self.buffer.push_str(
+            &self
+                .original
+                .chars()
+                .skip(self.last_old_char_index)
+                .take(jump)
+                .collect::<String>(),
+        );
+        self.last_old_char_index += jump;
+    }
+
+    /// Finish building the string after copying the remaining original string
+    /// since the last insertion or deletion.
+    pub fn build(mut self) -> String {
+        self.buffer.push_str(
+            &self
+                .original
+                .chars()
+                .skip(self.last_old_char_index)
+                .collect::<String>(),
+        );
+
+        self.buffer
+    }
+
+    #[allow(dead_code)]
+    pub fn get_slice(&self, range: Range<usize>) -> String {
+        let result = self
+            .buffer
+            .chars()
+            .chain(self.original.chars().skip(self.last_old_char_index))
+            .skip(range.start)
+            .take(range.end - range.start)
+            .collect::<String>();
+
+        debug_assert_eq!(result.chars().count(), range.len(), "Range out of bounds",);
+
+        result
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_string_builder() {
+        let original = "aaa bbb ccc";
+        let mut builder = StringBuilder::new(original);
+
+        builder.insert(0, "ddd ");
+        builder.delete(4..8);
+        builder.insert(11, " eee");
+
+        assert_eq!(builder.build(), "ddd bbb ccc eee");
+    }
+
+    #[test]
+    fn test_string_builder2() {
+        let original = "abcde";
+        let mut builder = StringBuilder::new(original);
+
+        builder.delete(1..4);
+
+        assert_eq!(builder.build(), "ae");
+    }
+}

backend/rust-toolchain.toml

@@ -0,0 +1,4 @@
+[toolchain]
+channel = "nightly-2025-03-14"
+targets = [ "x86_64-unknown-linux-gnu", "x86_64-unknown-linux-musl" ]
+profile = "default"

backend/rustfmt.toml

@@ -0,0 +1,8 @@
+imports_granularity = "crate"
+condense_wildcard_suffixes = true
+fn_single_line = true
+format_strings = true
+reorder_impl_items = true
+group_imports = "StdExternalCrate"
+use_field_init_shorthand = true
+wrap_comments=true

backend/sync_lib/Cargo.toml

@@ -0,0 +1,32 @@
+[package]
+name = "sync_lib"
+version.workspace = true
+edition.workspace = true
+authors.workspace = true
+license.workspace = true
+repository.workspace = true
+
+[lib]
+crate-type = ["cdylib", "rlib"]
+
+[dependencies]
+base64 = "0.22.1"
+reconcile = { path = "../reconcile" }
+wasm-bindgen = "0.2.99"
+thiserror = { workspace = true }
+
+# The `console_error_panic_hook` crate provides better debugging of panics by
+# logging them with `console.error`. This is great for development, but requires
+# all the `std::fmt` and `std::panicking` infrastructure, so isn't great for
+# code size when deploying.
+console_error_panic_hook = { version = "0.1.7", optional = true }
+
+[dev-dependencies]
+wasm-bindgen-test = "0.3.49"
+insta = "1.41.1"
+
+[features]
+default = ["console_error_panic_hook"]
+
+[lints]
+workspace = true

backend/sync_lib/src/errors.rs

@@ -0,0 +1,29 @@
+use base64::DecodeError;
+use thiserror::Error;
+use wasm_bindgen::JsValue;
+
+#[derive(Error, Debug)]
+pub enum SyncLibError {
+    #[error("Base64 decoding error because of {}", .reason)]
+    Base64DecodingError { reason: String },
+}
+
+impl From<DecodeError> for SyncLibError {
+    fn from(e: DecodeError) -> Self {
+        SyncLibError::Base64DecodingError {
+            reason: e.to_string(),
+        }
+    }
+}
+
+impl From<std::string::FromUtf8Error> for SyncLibError {
+    fn from(e: std::string::FromUtf8Error) -> Self {
+        SyncLibError::Base64DecodingError {
+            reason: e.to_string(),
+        }
+    }
+}
+
+impl From<SyncLibError> for JsValue {
+    fn from(val: SyncLibError) -> Self { JsValue::from_str(&val.to_string()) }
+}

backend/sync_lib/src/lib.rs

@@ -0,0 +1,136 @@
+//! This crate provides utilities for easily communicating between backend &
+//! frontend and ensuring the same logic for encoding and decoding binary data,
+//! and 3-way-merging documents in Rust and JavaScript.
+//!
+//! The crate is designed to be used as a Rust library and as a
+//! TypeScript/JavaScript package through WebAssembly (WASM).
+//!
+//! # Modules
+//!
+//! - `errors`: Contains error types used in this crate.
+use core::str;
+
+use base64::{Engine as _, engine::general_purpose::STANDARD};
+use errors::SyncLibError;
+use wasm_bindgen::prelude::*;
+
+pub mod errors;
+
+/// Encode binary data for easy transport over HTTP. Inverse of
+/// `base64_to_bytes`.
+///
+/// # Arguments
+///
+/// - `input`: The binary data to encode.
+///
+/// # Returns
+///
+/// The base64-encoded string.
+///
+/// # Panics
+///
+/// If the input is not valid UTF-8.
+#[wasm_bindgen(js_name = bytesToBase64)]
+#[must_use]
+pub fn bytes_to_base64(input: &[u8]) -> String {
+    set_panic_hook();
+
+    STANDARD.encode(input)
+}
+
+/// Inverse of `bytes_to_base64`.
+/// Decode base64-encoded data into binary data.
+///
+/// # Arguments
+///
+/// - `input`: The base64-encoded string.
+///
+/// # Returns
+///
+/// The decoded binary data.
+///
+/// # Errors
+///
+/// If the input is not valid base64.
+#[wasm_bindgen(js_name = base64ToBytes)]
+pub fn base64_to_bytes(input: &str) -> Result<Vec<u8>, SyncLibError> {
+    set_panic_hook();
+
+    STANDARD.decode(input).map_err(SyncLibError::from)
+}
+
+/// Merge two documents with a common parent. Relies on `reconcile::reconcile`
+/// for texts and returns the right document as-is if either of the updated
+/// documents is binary.
+///
+/// # Arguments
+///
+/// - `parent`: The common parent document.
+/// - `left`: The left document updated by one user.
+/// - `right`: The right document updated by another user.
+///
+/// # Returns
+///
+/// The merged document.
+///
+/// # Panics
+///
+/// If any of the input documents are not valid UTF-8 strings.
+#[wasm_bindgen]
+#[must_use]
+pub fn merge(parent: &[u8], left: &[u8], right: &[u8]) -> Vec<u8> {
+    set_panic_hook();
+
+    if is_binary(parent) || is_binary(left) || is_binary(right) {
+        right.to_vec()
+    } else {
+        reconcile::reconcile(
+            str::from_utf8(parent).expect("parent must be valid UTF-8 because it's not binary"),
+            str::from_utf8(left).expect("left must be valid UTF-8 because it's not binary"),
+            str::from_utf8(right).expect("right must be valid UTF-8 because it's not binary"),
+        )
+        .into_bytes()
+    }
+}
+
+/// WASM wrapper around `reconcile::reconcile` for text merging.
+#[wasm_bindgen(js_name = mergeText)]
+#[must_use]
+pub fn merge_text(parent: &str, left: &str, right: &str) -> String {
+    set_panic_hook();
+
+    reconcile::reconcile(parent, left, right)
+}
+
+/// Heuristically determine if the given data is a binary or a text file's
+/// content.
+#[wasm_bindgen(js_name = isBinary)]
+#[must_use]
+pub fn is_binary(data: &[u8]) -> bool {
+    set_panic_hook();
+
+    if data.contains(&0) {
+        // Even though the NUL character is valid in UTF-8, it's highly suspicious in
+        // human-readable text.
+        return true;
+    }
+
+    std::str::from_utf8(data).is_err()
+}
+
+/// We don't want to support merging structured data like JSON, YAML, etc.
+#[wasm_bindgen(js_name = isFileTypeMergable)]
+#[must_use]
+pub fn is_file_type_mergable(path_or_file_name: &str) -> bool {
+    set_panic_hook();
+
+    let file_extension = path_or_file_name.split('.').next_back().unwrap_or_default();
+
+    matches!(file_extension.to_lowercase().as_str(), "md" | "txt")
+}
+
+fn set_panic_hook() {
+    // https://github.com/rustwasm/console_error_panic_hook#readme
+    #[cfg(feature = "console_error_panic_hook")]
+    console_error_panic_hook::set_once();
+}

backend/sync_lib/tests/snapshots/web__base64_to_bytes_error.snap

@@ -0,0 +1,10 @@
+---
+source: sync_lib/tests/web.rs
+expression: base64_to_bytes(input)
+snapshot_kind: text
+---
+Err(
+    Base64DecodingError {
+        reason: "Invalid symbol 61, offset 0.",
+    },
+)

backend/sync_lib/tests/web.rs

@@ -0,0 +1,63 @@
+use insta::assert_debug_snapshot;
+use sync_lib::*;
+use wasm_bindgen_test::*;
+
+#[wasm_bindgen_test(unsupported = test)]
+fn test_bytes_to_base64() {
+    let input = b"hello";
+    let expected = "aGVsbG8=";
+    assert_eq!(bytes_to_base64(input), expected);
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn test_base64_to_bytes() {
+    let input = "aGVsbG8=";
+    let expected = b"hello".to_vec();
+    assert_eq!(base64_to_bytes(input).unwrap(), expected);
+}
+
+#[test] // insta doesn't support wasm-bindgen-test
+fn test_base64_to_bytes_error() {
+    let input = "===";
+    assert_debug_snapshot!(base64_to_bytes(input));
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn merge_text() {
+    let left = b"hello ";
+    let right = b"world";
+    let result = merge(b"", left, right);
+    assert_eq!(result, b"hello world");
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn merge_binary() {
+    let left = [0, 1, 2];
+    let right = [3, 4, 5];
+    assert_eq!(merge(b"", &left, &right), right);
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn test_is_binary() {
+    assert!(is_binary(&[0, 159, 146, 150]));
+    assert!(is_binary(&[0, 12]));
+    assert!(!is_binary(b"hello"));
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn test_is_binary_empty() {
+    assert!(!is_binary(b""));
+}
+
+#[wasm_bindgen_test(unsupported = test)]
+fn test_is_file_type_mergable() {
+    assert!(is_file_type_mergable(".md"));
+    assert!(is_file_type_mergable("hi.md"));
+    assert!(is_file_type_mergable("my/path/to/my/document.md"));
+    assert!(is_file_type_mergable("hi.MD"));
+    assert!(is_file_type_mergable("my/path/to/my/DOCUMENT.MD"));
+
+    assert!(!is_file_type_mergable(".json"));
+    assert!(!is_file_type_mergable("HELLO.JSON"));
+    assert!(!is_file_type_mergable("my/config.yml"));
+}

backend/sync_server/Cargo.toml

@@ -0,0 +1,42 @@
+[package]
+name = "sync_server"
+version.workspace = true
+edition.workspace = true
+authors.workspace = true
+license.workspace = true
+repository.workspace = true
+
+[dependencies]
+reconcile = { path = "../reconcile" }
+sync_lib = { path = "../sync_lib" }
+
+serde = { workspace = true }
+thiserror = { workspace = true }
+
+tokio = { version = "1.42.0", features = ["full"]}
+uuid = { version = "1.16.0", features = ["v4", "serde"] }
+log = { version = "0.4.22" }
+anyhow =  { version = "1.0.94", features = ["backtrace"] }
+axum = { version = "0.7.4", features = ["ws", "macros", "tracing", "multipart"]}
+axum-extra = { version = "0.9.6", features = ["typed-header"] }
+aide-axum-typed-multipart = "0.13.0"
+axum_typed_multipart = "0.11.0"
+tower-http = { version = "0.6.1", features = ["cors", "trace", "limit"] }
+tracing-subscriber = { version = "0.3.19", features = ["fmt", "env-filter"]}
+serde_yaml = "0.9.34"
+sqlx = { version = "0.8.2", features = ["sqlite", "runtime-tokio", "uuid", "chrono"] }
+chrono = { version = "0.4.40", features = ["serde"] }
+aide = { version = "0.13.4", features = ["axum", "axum-ws", "scalar", "axum-headers"] }
+schemars = { version = "0.8.21", features = ["chrono", "uuid1", "bytes"] }
+tracing = "0.1.41"
+rand = "0.8.5"
+sanitize-filename = "0.6.0"
+axum-jsonschema = { version = "0.8.0", features = ["aide"] }
+regex = "1.11.1"
+clap = { version = "4.5.32", features = ["derive"] }
+futures = "0.3.31"
+serde_json = "1.0.140"
+clap-verbosity-flag = "3.0.2"
+
+[lints]
+workspace = true

backend/sync_server/README.md

@@ -0,0 +1,4 @@
+
+cargo install sqlx-cli
+rm db.sqlite3; sqlx database create --database-url sqlite://db.sqlite3
+sqlx migrate run --source sync_server/src/app_state/database/migrations --database-url sqlite://db.sqlite3

backend/sync_server/src/app_state.rs

@@ -0,0 +1,34 @@
+pub mod broadcasts;
+pub mod database;
+
+use std::ffi::OsString;
+
+use anyhow::Result;
+use broadcasts::Broadcasts;
+use database::Database;
+
+use crate::{config::Config, consts::DEFAULT_CONFIG_PATH};
+
+#[derive(Clone, Debug)]
+pub struct AppState {
+    pub config: Config,
+    pub database: Database,
+    pub broadcasts: Broadcasts,
+}
+
+impl AppState {
+    pub async fn try_new(config_path: Option<OsString>) -> Result<Self> {
+        let config_path = config_path.unwrap_or_else(|| OsString::from(DEFAULT_CONFIG_PATH));
+        let path = std::path::PathBuf::from(config_path);
+
+        let config = Config::read_or_create(&path).await?;
+        let database = Database::try_new(&config.database).await?;
+        let broadcasts = Broadcasts::new(&config.server);
+
+        Ok(Self {
+            config,
+            database,
+            broadcasts,
+        })
+    }
+}

backend/sync_server/src/app_state/broadcasts.rs

@@ -0,0 +1,57 @@
+use std::{collections::HashMap, sync::Arc};
+
+use anyhow::Context;
+use tokio::sync::{Mutex, broadcast};
+
+use super::database::models::{DocumentVersionWithoutContent, VaultId};
+use crate::{config::server_config::ServerConfig, errors::server_error};
+
+#[derive(Debug, Clone)]
+pub struct Broadcasts {
+    max_clients_per_vault: usize,
+    tx: Arc<Mutex<HashMap<VaultId, broadcast::Sender<DocumentVersionWithoutContent>>>>,
+}
+
+impl Broadcasts {
+    pub fn new(server_config: &ServerConfig) -> Self {
+        Self {
+            max_clients_per_vault: server_config.max_clients_per_vault,
+            tx: Arc::new(Mutex::new(HashMap::new())),
+        }
+    }
+
+    pub async fn get_receiver(
+        &self,
+        vault: VaultId,
+    ) -> broadcast::Receiver<DocumentVersionWithoutContent> {
+        let tx = self.get_or_create(vault).await;
+
+        tx.subscribe()
+    }
+
+    /// Sent a document update to all clients subscribed to the vault.
+    /// We ignore & log failures.
+    pub async fn send(&self, vault: VaultId, document: DocumentVersionWithoutContent) {
+        let tx = self.get_or_create(vault).await;
+
+        let result = tx
+            .send(document)
+            .context("Cannot broadcast update message to websocket listeners")
+            .map_err(server_error);
+
+        if result.is_err() {
+            log::debug!("Failed to send message: {result:?}");
+        }
+    }
+
+    async fn get_or_create(
+        &self,
+        vault: VaultId,
+    ) -> broadcast::Sender<DocumentVersionWithoutContent> {
+        let mut tx = self.tx.lock().await;
+
+        tx.entry(vault)
+            .or_insert_with(|| broadcast::channel(self.max_clients_per_vault).0.clone())
+            .clone()
+    }
+}

backend/sync_server/src/app_state/database.rs

@@ -2,62 +2,38 @@ use core::time::Duration;
 use std::{collections::HashMap, sync::Arc};
 
 use anyhow::{Context as _, Result};
-use log::info;
 use models::{
     DocumentId, DocumentVersionWithoutContent, StoredDocumentVersion, VaultId, VaultUpdateId,
 };
-use sqlx::{ConnectOptions, sqlite::SqliteConnectOptions, types::chrono::Utc};
-
+use sqlx::{sqlite::SqliteConnectOptions, types::chrono::Utc};
 pub mod models;
 use sqlx::{Pool, Sqlite, sqlite::SqlitePoolOptions};
 use tokio::sync::Mutex;
-use tokio::time::Instant;
 use uuid::fmt::Hyphenated;
 
-use super::websocket::{
-    broadcasts::Broadcasts,
-    models::{WebSocketServerMessage, WebSocketServerMessageWithOrigin, WebSocketVaultUpdate},
-};
 use crate::config::database_config::DatabaseConfig;
 
-#[derive(Clone)]
-struct PoolWithTimestamp {
-    pool: Pool<Sqlite>,
-    last_accessed: Instant,
-}
-
-impl std::fmt::Debug for PoolWithTimestamp {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        f.debug_struct("PoolWithTimestamp")
-            .field("pool", &"Pool<Sqlite>")
-            .field("last_accessed", &self.last_accessed)
-            .finish()
-    }
-}
-
 #[derive(Clone, Debug)]
 pub struct Database {
     config: DatabaseConfig,
-    broadcasts: Broadcasts,
-    connection_pools: Arc<Mutex<HashMap<VaultId, PoolWithTimestamp>>>,
+    connection_pools: Arc<Mutex<HashMap<VaultId, Pool<Sqlite>>>>,
 }
 
 pub type Transaction<'a> = sqlx::Transaction<'a, Sqlite>;
 
 impl Database {
-    pub async fn try_new(config: &DatabaseConfig, broadcasts: &Broadcasts) -> Result<Self> {
+    pub async fn try_new(config: &DatabaseConfig) -> Result<Self> {
         tokio::fs::create_dir_all(&config.databases_directory_path)
             .await
             .with_context(|| {
                 format!(
-                    "Failed to create databases directory at `{}`",
+                    "Failed to create databases directory: {}",
                     config.databases_directory_path.to_string_lossy()
                 )
             })?;
 
         let mut connection_pools = std::collections::HashMap::new();
 
-        info!("Applying pending database migrations");
         let mut entries = tokio::fs::read_dir(&config.databases_directory_path).await?;
         while let Some(entry) = entries.next_entry().await? {
             if !entry.file_name().to_string_lossy().ends_with(".sqlite") {
@@ -70,26 +46,16 @@ impl Database {
                 .trim_end_matches(".sqlite")
                 .to_owned();
 
-            let pool = Self::create_vault_database(config, &vault).await?;
             connection_pools.insert(
                 vault.clone(),
-                PoolWithTimestamp {
-                    pool,
-                    last_accessed: Instant::now(),
-                },
+                Self::create_vault_database(config, &vault).await?,
             );
         }
 
-        let database = Self {
+        Ok(Self {
             config: config.clone(),
             connection_pools: Arc::new(Mutex::new(connection_pools)),
-            broadcasts: broadcasts.clone(),
-        };
-
-        // Start background task to cleanup idle connection pools
-        database.start_idle_pool_cleanup();
-
-        Ok(database)
+        })
     }
 
     async fn create_vault_database(
@@ -103,18 +69,15 @@ impl Database {
         let connection_options = SqliteConnectOptions::new()
             .filename(file_name.clone())
             .create_if_missing(true)
-            .auto_vacuum(sqlx::sqlite::SqliteAutoVacuum::Full)
             .busy_timeout(Duration::from_secs(3600))
-            .journal_mode(sqlx::sqlite::SqliteJournalMode::Wal)
-            .log_slow_statements(log::LevelFilter::Warn, Duration::from_secs(30));
+            .journal_mode(sqlx::sqlite::SqliteJournalMode::Wal);
 
         let pool = SqlitePoolOptions::new()
-            .max_connections(config.max_connections_per_vault)
-            .acquire_slow_threshold(Duration::from_secs(30))
+            .max_connections(config.max_connections)
             .test_before_acquire(true)
             .connect_with(connection_options)
             .await
-            .with_context(|| format!("Cannot open database at `{}`", file_name.display()))?;
+            .with_context(|| format!("Cannot open database at {}", file_name.display()))?;
 
         Self::run_migrations(&pool).await?;
 
@@ -130,26 +93,16 @@ impl Database {
 
     async fn get_connection_pool(&self, vault: &VaultId) -> Result<Pool<Sqlite>> {
         let mut pools = self.connection_pools.lock().await;
-
         if !pools.contains_key(vault) {
             let pool = Self::create_vault_database(&self.config, vault).await?;
-            pools.insert(
-                vault.clone(),
-                PoolWithTimestamp {
-                    pool,
-                    last_accessed: Instant::now(),
-                },
-            );
+            pools.insert(vault.clone(), pool);
         }
 
-        let pool_with_timestamp = pools
-            .get_mut(vault)
+        let pool = pools
+            .get(vault)
             .expect("Pool was just inserted or already exists");
 
-        // Update last accessed time
-        pool_with_timestamp.last_accessed = Instant::now();
-
-        Ok(pool_with_timestamp.pool.clone())
+        Ok(pool.clone())
     }
 
     /// Attempting to write from this transaction might result in a
@@ -182,19 +135,17 @@ impl Database {
         vault: &VaultId,
         transaction: Option<&mut Transaction<'_>>,
     ) -> Result<Vec<DocumentVersionWithoutContent>> {
-        let query = sqlx::query!(
+        let query = sqlx::query_as!(
+            DocumentVersionWithoutContent,
             r#"
-            select
+            select 
                 vault_update_id,
-                document_id as "document_id: Hyphenated",
+                document_id as "document_id: Hyphenated", 
                 relative_path,
                 updated_date as "updated_date: chrono::DateTime<Utc>",
-                is_deleted,
-                user_id,
-                device_id,
-                length(content) as "content_size: u64"
+                is_deleted
             from latest_document_versions
-            order by vault_update_id
+            order by vault_update_id desc
             "#,
         );
 
@@ -206,22 +157,6 @@ impl Database {
                 .await
         }
         .context("Cannot fetch latest documents")
-        .map(|rows| {
-            rows.into_iter()
-                .map(|row| DocumentVersionWithoutContent {
-                    vault_update_id: row.vault_update_id,
-                    document_id: row.document_id.into(),
-                    relative_path: row.relative_path,
-                    updated_date: row.updated_date,
-                    is_deleted: row.is_deleted,
-                    user_id: row.user_id,
-                    device_id: row.device_id,
-                    content_size: row
-                        .content_size
-                        .expect("Content size can't be null but sqlx can't infer it"),
-                })
-                .collect()
-        })
     }
 
     /// Return the latest state of all documents (including deleted) in the
@@ -232,20 +167,18 @@ impl Database {
         vault_update_id: VaultUpdateId,
         transaction: Option<&mut Transaction<'_>>,
     ) -> Result<Vec<DocumentVersionWithoutContent>> {
-        let query = sqlx::query!(
+        let query = sqlx::query_as!(
+            DocumentVersionWithoutContent,
             r#"
             select
                 vault_update_id,
                 document_id as "document_id: Hyphenated",
                 relative_path,
                 updated_date as "updated_date: chrono::DateTime<Utc>",
-                is_deleted,
-                user_id,
-                device_id,
-                length(content) as "content_size: u64"
+                is_deleted
             from latest_document_versions
             where vault_update_id > ?
-            order by vault_update_id
+            order by vault_update_id desc
             "#,
             vault_update_id
         );
@@ -258,23 +191,7 @@ impl Database {
                 .await
         }
         .with_context(|| {
-            format!("Cannot fetch latest documents since vault_update_id `{vault_update_id}`")
-        })
-        .map(|rows| {
-            rows.into_iter()
-                .map(|row| DocumentVersionWithoutContent {
-                    vault_update_id: row.vault_update_id,
-                    document_id: row.document_id.into(),
-                    relative_path: row.relative_path,
-                    updated_date: row.updated_date,
-                    is_deleted: row.is_deleted,
-                    user_id: row.user_id,
-                    device_id: row.device_id,
-                    content_size: row
-                        .content_size
-                        .expect("Content size can't be null but sqlx can't infer it"),
-                })
-                .collect()
+            format!("Cannot fetch latest documents since vault_update_id {vault_update_id}")
         })
     }
 
@@ -310,21 +227,18 @@ impl Database {
         let query = sqlx::query_as!(
             StoredDocumentVersion,
             r#"
-            select
+            select 
                 vault_update_id,
-                document_id as "document_id: Hyphenated",
+                document_id as "document_id: Hyphenated", 
                 relative_path,
                 updated_date as "updated_date: chrono::DateTime<Utc>",
                 content,
-                is_deleted,
-                user_id,
-                device_id,
-                has_been_merged
+                is_deleted
             from latest_document_versions
-            where relative_path = ? and is_deleted = false
+            where relative_path = ?
             order by vault_update_id desc  -- `latest_document_versions` only contains a single latest version of each document, however,
-                                            -- multiple documents can have the same `relative_path`, if they have been deleted. That's
-                                            -- why we only care about the latest version of the document with the given relative path.
+                                           -- multiple documents can have the same `relative_path`, if they have been deleted. That's
+                                           -- why we only care about the latest version of the document with the given relative path.
             limit 1
             "#,
             relative_path
@@ -350,16 +264,13 @@ impl Database {
         let query = sqlx::query_as!(
             StoredDocumentVersion,
             r#"
-            select
+            select 
                 vault_update_id,
-                document_id as "document_id: Hyphenated",
+                document_id as "document_id: Hyphenated", 
                 relative_path,
                 updated_date as "updated_date: chrono::DateTime<Utc>",
                 content,
-                is_deleted,
-                user_id,
-                device_id,
-                has_been_merged
+                is_deleted
             from latest_document_versions
             where document_id = ?
             "#,
@@ -385,16 +296,13 @@ impl Database {
         let query = sqlx::query_as!(
             StoredDocumentVersion,
             r#"
-            select
+            select 
                 vault_update_id,
-                document_id as "document_id: Hyphenated",
+                document_id as "document_id: Hyphenated", 
                 relative_path,
                 updated_date as "updated_date: chrono::DateTime<Utc>",
                 content,
-                is_deleted,
-                user_id,
-                device_id,
-                has_been_merged
+                is_deleted
             from documents
             where vault_update_id = ?"#,
             vault_update_id
@@ -410,106 +318,40 @@ impl Database {
         .context("Cannot fetch document version")
     }
 
-    // inserting the document must be the last step of the transaction if there's one
     pub async fn insert_document_version(
         &self,
-        vault_id: &VaultId,
+        vault: &VaultId,
         version: &StoredDocumentVersion,
-        transaction: Option<Transaction<'_>>,
+        transaction: Option<&mut Transaction<'_>>,
     ) -> Result<()> {
         let document_id = version.document_id.as_hyphenated();
         let query = sqlx::query!(
             r#"
             insert into documents (
                 vault_update_id,
-                document_id,
+                document_id, 
                 relative_path,
                 updated_date,
                 content,
-                is_deleted,
-                user_id,
-                device_id
+                is_deleted
             )
-            values (?, ?, ?, ?, ?, ?, ?, ?)
+            values (?, ?, ?, ?, ?, ?)
             "#,
             version.vault_update_id,
             document_id,
             version.relative_path,
             version.updated_date,
             version.content,
-            version.is_deleted,
-            version.user_id,
-            version.device_id
+            version.is_deleted
         );
 
-        if let Some(mut transaction) = transaction {
-            query
-                .execute(&mut *transaction)
-                .await
-                .context("Cannot insert document version")?;
-
-            transaction
-                .commit()
-                .await
-                .context("Failed to commit transaction")?;
+        if let Some(transaction) = transaction {
+            query.execute(&mut **transaction).await
         } else {
-            query
-                .execute(&self.get_connection_pool(vault_id).await?)
-                .await
-                .context("Cannot insert document version")?;
+            query.execute(&self.get_connection_pool(vault).await?).await
         }
-
-        self.broadcasts
-            .send_document_update(
-                vault_id.clone(),
-                WebSocketServerMessageWithOrigin::with_origin(
-                    version.device_id.clone(),
-                    WebSocketServerMessage::VaultUpdate(WebSocketVaultUpdate {
-                        documents: vec![version.clone().into()],
-                        is_initial_sync: false,
-                    }),
-                ),
-            )
-            .await;
+        .context("Cannot insert document version")?;
 
         Ok(())
     }
-
-    /// Cleanup idle connection pools that haven't been accessed in more than 5 minutes
-    async fn cleanup_idle_pools(&self) {
-        let mut pools = self.connection_pools.lock().await;
-        let now = Instant::now();
-        let idle_timeout = Duration::from_secs(5 * 60); // 5 minutes
-
-        // Collect vaults to remove
-        let vaults_to_remove: Vec<VaultId> = pools
-            .iter()
-            .filter(|(_, pool_with_timestamp)| {
-                now.duration_since(pool_with_timestamp.last_accessed) > idle_timeout
-            })
-            .map(|(vault_id, _)| vault_id.clone())
-            .collect();
-
-        // Close and remove idle pools
-        for vault_id in &vaults_to_remove {
-            if let Some(pool_with_timestamp) = pools.remove(vault_id) {
-                info!("Closing idle database connection pool for vault `{vault_id}`");
-                pool_with_timestamp.pool.close().await;
-            }
-        }
-    }
-
-    /// Start a background task that periodically cleans up idle connection pools
-    fn start_idle_pool_cleanup(&self) {
-        let database = self.clone();
-        tokio::spawn(async move {
-            let mut interval = tokio::time::interval(Duration::from_secs(60)); // Check every minute
-            interval.set_missed_tick_behavior(tokio::time::MissedTickBehavior::Skip);
-
-            loop {
-                interval.tick().await;
-                database.cleanup_idle_pools().await;
-            }
-        });
-    }
 }

backend/sync_server/src/app_state/database/models.rs

@@ -1,14 +1,11 @@
-use base64::{Engine as _, engine::general_purpose::STANDARD};
 use chrono::{DateTime, Utc};
+use schemars::JsonSchema;
 use serde::Serialize;
-use ts_rs::TS;
+use sync_lib::bytes_to_base64;
 
 pub type VaultId = String;
 pub type VaultUpdateId = i64;
-
 pub type DocumentId = uuid::Uuid;
-pub type UserId = String;
-pub type DeviceId = String;
 
 #[derive(Debug, Clone)]
 pub struct StoredDocumentVersion {
@@ -18,33 +15,20 @@ pub struct StoredDocumentVersion {
     pub updated_date: DateTime<Utc>,
     pub content: Vec<u8>,
     pub is_deleted: bool,
-    pub user_id: UserId,
-    pub device_id: DeviceId,
-    #[allow(dead_code)] // This is for manual analysis
-    pub has_been_merged: bool,
 }
 
 impl PartialEq<Self> for StoredDocumentVersion {
-    fn eq(&self, other: &Self) -> bool {
-        self.vault_update_id == other.vault_update_id
-    }
+    fn eq(&self, other: &Self) -> bool { self.vault_update_id == other.vault_update_id }
 }
 
-#[derive(TS, Debug, Clone, Serialize)]
+#[derive(Debug, Clone, Serialize, JsonSchema)]
 #[serde(rename_all = "camelCase")]
 pub struct DocumentVersionWithoutContent {
-    #[ts(as = "i32")]
     pub vault_update_id: VaultUpdateId,
-
     pub document_id: DocumentId,
     pub relative_path: String,
     pub updated_date: DateTime<Utc>,
     pub is_deleted: bool,
-    pub user_id: UserId,
-    pub device_id: DeviceId,
-
-    #[ts(as = "i32")]
-    pub content_size: u64,
 }
 
 impl From<StoredDocumentVersion> for DocumentVersionWithoutContent {
@@ -55,26 +39,19 @@ impl From<StoredDocumentVersion> for DocumentVersionWithoutContent {
             relative_path: value.relative_path,
             updated_date: value.updated_date,
             is_deleted: value.is_deleted,
-            user_id: value.user_id,
-            device_id: value.device_id,
-            content_size: value.content.len() as u64,
         }
     }
 }
 
-#[derive(TS, Debug, Clone, Serialize)]
+#[derive(Debug, Clone, Serialize, JsonSchema)]
 #[serde(rename_all = "camelCase")]
 pub struct DocumentVersion {
-    #[ts(as = "i32")]
     pub vault_update_id: VaultUpdateId,
-
     pub document_id: DocumentId,
     pub relative_path: String,
     pub updated_date: DateTime<Utc>,
     pub content_base64: String,
     pub is_deleted: bool,
-    pub user_id: UserId,
-    pub device_id: DeviceId,
 }
 
 impl From<StoredDocumentVersion> for DocumentVersion {
@@ -84,10 +61,8 @@ impl From<StoredDocumentVersion> for DocumentVersion {
             document_id: value.document_id,
             relative_path: value.relative_path,
             updated_date: value.updated_date,
-            content_base64: STANDARD.encode(&value.content),
+            content_base64: bytes_to_base64(&value.content),
             is_deleted: value.is_deleted,
-            user_id: value.user_id,
-            device_id: value.device_id,
         }
     }
 }

backend/sync_server/src/cli/args.rs

@@ -1,6 +1,7 @@
 use std::ffi::OsString;
 
 use clap::Parser;
+use clap_verbosity_flag::{InfoLevel, Verbosity};
 
 use crate::cli::color_when::ColorWhen;
 
@@ -11,6 +12,9 @@ pub struct Args {
     #[arg(index = 1)]
     pub config_path: Option<OsString>,
 
+    #[command(flatten)]
+    pub verbose: Verbosity<InfoLevel>,
+
     #[arg(
             long,
             value_name = "WHEN",

backend/sync_server/src/config.rs

@@ -2,15 +2,13 @@ use std::path::Path;
 
 use anyhow::{Context as _, Result};
 use database_config::DatabaseConfig;
-use log::info;
-use logging_config::LoggingConfig;
+use log::{info, warn};
 use serde::{Deserialize, Serialize};
 use server_config::ServerConfig;
 use tokio::fs;
 use user_config::UserConfig;
 
 pub mod database_config;
-pub mod logging_config;
 pub mod server_config;
 pub mod user_config;
 
@@ -22,40 +20,38 @@ pub struct Config {
     pub server: ServerConfig,
     #[serde(default)]
     pub users: UserConfig,
-    #[serde(default)]
-    pub logging: LoggingConfig,
 }
 
 impl Config {
     pub async fn read_or_create(path: &Path) -> Result<Self> {
-        let config = if path.exists() {
+        if path.exists() {
             info!(
-                "Loading configuration from `{}`",
-                path.canonicalize().unwrap().display()
+                "Loading configuration from {:?}",
+                path.canonicalize().unwrap()
             );
-            Self::load_from_file(path).await?
+            Self::load_from_file(path).await
         } else {
-            Self::default()
-        };
-
-        config.write(path).await?;
-        info!(
-            "Updated configuration at `{}`",
-            path.canonicalize().unwrap().display()
-        );
-
-        Ok(config)
+            let config = Self::default();
+            config.write(path).await?;
+            warn!(
+                "Configuration file not found, wrote default configuration to {:?}",
+                path.canonicalize().unwrap()
+            );
+            Ok(config)
+        }
     }
 
     pub async fn load_from_file(path: &Path) -> Result<Self> {
         let contents = fs::read_to_string(path).await.with_context(|| {
             format!(
-                "Cannot load configuration from disk from `{}`",
+                "Cannot load configuration from disk from {}",
                 path.display()
             )
         })?;
 
-        serde_yaml::from_str(&contents).context("Failed to parse configuration")
+        let config = serde_yaml::from_str(&contents).context("Failed to parse configuration")?;
+
+        Ok(config)
     }
 
     pub async fn write(&self, path: &Path) -> Result<()> {

backend/sync_server/src/config/database_config.rs

@@ -0,0 +1,34 @@
+use std::path::PathBuf;
+
+use log::debug;
+use serde::{Deserialize, Serialize};
+
+use crate::consts::{DEFAULT_DATABASES_DIRECTORY_PATH, DEFAULT_MAX_CONNECTIONS};
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct DatabaseConfig {
+    #[serde(default = "default_databases_directory_path")]
+    pub databases_directory_path: PathBuf,
+
+    #[serde(default = "default_max_connections")]
+    pub max_connections: u32,
+}
+
+fn default_databases_directory_path() -> PathBuf {
+    debug!("Using default databases directory path: {DEFAULT_DATABASES_DIRECTORY_PATH:?}");
+    PathBuf::from(DEFAULT_DATABASES_DIRECTORY_PATH)
+}
+
+fn default_max_connections() -> u32 {
+    debug!("Using default max connections: {DEFAULT_MAX_CONNECTIONS}");
+    DEFAULT_MAX_CONNECTIONS
+}
+
+impl Default for DatabaseConfig {
+    fn default() -> Self {
+        Self {
+            databases_directory_path: default_databases_directory_path(),
+            max_connections: default_max_connections(),
+        }
+    }
+}

backend/sync_server/src/config/server_config.rs

@@ -1,13 +1,11 @@
 use log::debug;
 use serde::{Deserialize, Serialize};
-use std::time::Duration;
 
 use crate::consts::{
-    DEFAULT_HOST, DEFAULT_MAX_BODY_SIZE_MB, DEFAULT_MAX_CLIENTS_PER_VAULT,
-    DEFAULT_MERGEABLE_FILE_EXTENSIONS, DEFAULT_PORT, DEFAULT_RESPONSE_TIMEOUT_SECONDS,
+    DEFAULT_HOST, DEFAULT_MAX_BODY_SIZE_MB, DEFAULT_MAX_CLIENTS_PER_VAULT, DEFAULT_PORT,
 };
 
-#[derive(Debug, Deserialize, Serialize, Clone, Default)]
+#[derive(Debug, Deserialize, Serialize, Clone)]
 pub struct ServerConfig {
     #[serde(default = "default_host")]
     pub host: String,
@@ -20,12 +18,6 @@ pub struct ServerConfig {
 
     #[serde(default = "default_max_clients_per_vault")]
     pub max_clients_per_vault: usize,
-
-    #[serde(default = "default_response_timeout", with = "humantime_serde")]
-    pub response_timeout: Duration,
-
-    #[serde(default = "default_mergeable_file_extensions")]
-    pub mergeable_file_extensions: Vec<String>,
 }
 
 fn default_host() -> String {
@@ -39,7 +31,7 @@ fn default_port() -> u16 {
 }
 
 fn default_max_body_size_mb() -> usize {
-    debug!("Using default max body size {DEFAULT_MAX_BODY_SIZE_MB} MB");
+    debug!("Using default max body size (MB): {DEFAULT_MAX_BODY_SIZE_MB}");
     DEFAULT_MAX_BODY_SIZE_MB
 }
 
@@ -48,15 +40,13 @@ fn default_max_clients_per_vault() -> usize {
     DEFAULT_MAX_CLIENTS_PER_VAULT
 }
 
-fn default_response_timeout() -> Duration {
-    debug!("Using default response timeout: {DEFAULT_RESPONSE_TIMEOUT_SECONDS:?}");
-    DEFAULT_RESPONSE_TIMEOUT_SECONDS
-}
-
-fn default_mergeable_file_extensions() -> Vec<String> {
-    debug!("Using default mergeable file extensions: {DEFAULT_MERGEABLE_FILE_EXTENSIONS:?}");
-    DEFAULT_MERGEABLE_FILE_EXTENSIONS
-        .iter()
-        .map(|s| (*s).to_owned())
-        .collect()
+impl Default for ServerConfig {
+    fn default() -> Self {
+        Self {
+            host: default_host(),
+            port: default_port(),
+            max_body_size_mb: default_max_body_size_mb(),
+            max_clients_per_vault: default_max_clients_per_vault(),
+        }
+    }
 }

backend/sync_server/src/config/user_config.rs

@@ -0,0 +1,43 @@
+use rand::{Rng as _, distributions::Alphanumeric, thread_rng};
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct UserConfig {
+    #[serde(default = "default_users")]
+    pub user_tokens: Vec<User>,
+}
+
+impl UserConfig {
+    pub fn get_user(&self, token: &str) -> Option<&User> {
+        self.user_tokens.iter().find(|u| u.token == token)
+    }
+}
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct User {
+    pub name: String,
+    pub token: String,
+}
+
+impl Default for UserConfig {
+    fn default() -> Self {
+        Self {
+            user_tokens: default_users(),
+        }
+    }
+}
+
+fn default_users() -> Vec<User> {
+    vec![User {
+        name: "admin".to_owned(),
+        token: get_random_token(),
+    }]
+}
+
+pub fn get_random_token() -> String {
+    thread_rng()
+        .sample_iter(&Alphanumeric)
+        .take(64)
+        .map(char::from)
+        .collect()
+}

backend/sync_server/src/consts.rs

@@ -0,0 +1,7 @@
+pub const DEFAULT_CONFIG_PATH: &str = "config.yml";
+pub const DEFAULT_DATABASES_DIRECTORY_PATH: &str = "databases";
+pub const DEFAULT_HOST: &str = "127.0.0.1";
+pub const DEFAULT_PORT: u16 = 3000;
+pub const DEFAULT_MAX_CONNECTIONS: u32 = 12;
+pub const DEFAULT_MAX_BODY_SIZE_MB: usize = 4096;
+pub const DEFAULT_MAX_CLIENTS_PER_VAULT: usize = 256;

backend/sync_server/src/errors.rs

@@ -1,14 +1,13 @@
-use std::fmt::Display;
-
+use aide::OperationOutput;
 use axum::{
     Json,
     http::StatusCode,
     response::{IntoResponse, Response},
 };
-use log::{debug, error};
+use log::{error, info};
+use schemars::JsonSchema;
 use serde::Serialize;
 use thiserror::Error;
-use ts_rs::TS;
 
 #[derive(Error, Debug)]
 pub enum SyncServerError {
@@ -25,7 +24,7 @@ pub enum SyncServerError {
     NotFound(#[source] anyhow::Error),
 
     #[error("Unauthorized: {0}")]
-    Unauthenticated(#[source] anyhow::Error),
+    Unauthorized(#[source] anyhow::Error),
 
     #[error("Permission denied error: {0}")]
     PermissionDeniedError(#[source] anyhow::Error),
@@ -38,35 +37,12 @@ impl SyncServerError {
             | Self::ClientError(error)
             | Self::ServerError(error)
             | Self::NotFound(error)
-            | Self::Unauthenticated(error)
+            | Self::Unauthorized(error)
             | Self::PermissionDeniedError(error) => error.into(),
         }
     }
 }
 
-#[derive(TS, Debug, Clone, Serialize)]
-#[serde(rename_all = "camelCase")]
-#[ts(export)]
-pub struct SerializedError {
-    pub error_type: &'static str,
-    pub message: String,
-    pub causes: Vec<String>,
-}
-
-impl Display for SerializedError {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(f, "{}: {}", self.error_type, self.message)?;
-        if !self.causes.is_empty() {
-            write!(f, "\nCauses:\n")?;
-            for cause in &self.causes {
-                write!(f, "{}", &format!("- {cause}\n"))?;
-            }
-        }
-
-        Ok(())
-    }
-}
-
 impl IntoResponse for SyncServerError {
     fn into_response(self) -> Response {
         let body = Json(self.serialize());
@@ -77,12 +53,18 @@ impl IntoResponse for SyncServerError {
             }
             Self::ClientError(_) => (StatusCode::BAD_REQUEST, body).into_response(),
             Self::NotFound(_) => (StatusCode::NOT_FOUND, body).into_response(),
-            Self::Unauthenticated(_) => (StatusCode::UNAUTHORIZED, body).into_response(),
+            Self::Unauthorized(_) => (StatusCode::UNAUTHORIZED, body).into_response(),
             Self::PermissionDeniedError(_) => (StatusCode::FORBIDDEN, body).into_response(),
         }
     }
 }
 
+#[derive(Debug, Clone, Serialize, JsonSchema)]
+pub struct SerializedError {
+    pub message: String,
+    pub causes: Vec<String>,
+}
+
 impl From<&anyhow::Error> for SerializedError {
     fn from(error: &anyhow::Error) -> SerializedError {
         let mut causes = vec![];
@@ -93,49 +75,42 @@ impl From<&anyhow::Error> for SerializedError {
         }
 
         SerializedError {
-            error_type: error.downcast_ref::<SyncServerError>().map_or(
-                "UnknownError",
-                |e| match e {
-                    SyncServerError::InitError(_) => "InitError",
-                    SyncServerError::ClientError(_) => "ClientError",
-                    SyncServerError::ServerError(_) => "ServerError",
-                    SyncServerError::NotFound(_) => "NotFound",
-                    SyncServerError::Unauthenticated(_) => "Unauthenticated",
-                    SyncServerError::PermissionDeniedError(_) => "PermissionDeniedError",
-                },
-            ),
             message: error.to_string(),
             causes,
         }
     }
 }
 
-pub fn init_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Initialization error: {error:?}");
+impl OperationOutput for SyncServerError {
+    type Inner = Self;
+}
+
+pub const fn init_error(error: anyhow::Error) -> SyncServerError {
     SyncServerError::InitError(error)
 }
 
 pub fn server_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Server error: {error:?}");
+    error!("Server error: {:?}", error);
     SyncServerError::ServerError(error)
 }
 
 pub fn client_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Client error: {error:?}");
+    info!("Client error: {:?}", error);
     SyncServerError::ClientError(error)
 }
 
 pub fn not_found_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Not found: {error:?}");
+    info!("Not found error: {:?}", error);
     SyncServerError::NotFound(error)
 }
 
-pub fn unauthenticated_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Unauthenticated user: {error:?}");
-    SyncServerError::Unauthenticated(error)
+pub fn unauthorized_error(error: anyhow::Error) -> SyncServerError {
+    info!("Unauthorized error: {:?}", error);
+    SyncServerError::Unauthorized(error)
 }
 
+#[allow(dead_code)]
 pub fn permission_denied_error(error: anyhow::Error) -> SyncServerError {
-    debug!("Permission denied: {error:?}");
+    info!("Permission denied error: {:?}", error);
     SyncServerError::PermissionDeniedError(error)
 }

backend/sync_server/src/main.rs

@@ -0,0 +1,86 @@
+mod app_state;
+mod cli;
+mod config;
+mod consts;
+mod errors;
+mod server;
+mod utils;
+
+use std::process::ExitCode;
+
+use anyhow::{Context as _, Result};
+use clap::Parser;
+use cli::args::Args;
+use errors::{SyncServerError, init_error};
+use log::info;
+use server::create_server;
+use tracing_subscriber::{EnvFilter, fmt::format, util::SubscriberInitExt};
+
+#[tokio::main]
+async fn main() -> ExitCode {
+    let args = Args::parse();
+
+    let mut result = set_up_logging(&args);
+
+    if result.is_ok() {
+        result = start_server(args).await;
+    }
+
+    match result {
+        Ok(()) => ExitCode::SUCCESS,
+        Err(e) => {
+            eprintln!("Failed to set up logging: {e}");
+            ExitCode::FAILURE
+        }
+    }
+}
+
+fn set_up_logging(args: &Args) -> Result<(), SyncServerError> {
+    let level_filter = match args.verbose.log_level_filter() {
+        // We don't want to allow disabling all logging
+        log::LevelFilter::Off | log::LevelFilter::Error => tracing::Level::ERROR,
+        log::LevelFilter::Warn => tracing::Level::WARN,
+        log::LevelFilter::Info => tracing::Level::INFO,
+        log::LevelFilter::Debug => tracing::Level::DEBUG,
+        log::LevelFilter::Trace => tracing::Level::TRACE,
+    };
+
+    let env_filter = EnvFilter::builder()
+        .with_default_directive(level_filter.into())
+        .from_env()
+        .context("Failed to create logging env filter")
+        .map_err(init_error)?;
+
+    let use_colors = args.color.use_colors();
+
+    let is_debug_mode = args.verbose.log_level_filter() >= log::LevelFilter::Debug;
+
+    tracing_subscriber::fmt()
+        .with_ansi(use_colors)
+        .with_env_filter(env_filter)
+        .event_format(
+            format()
+                .without_time()
+                .with_target(is_debug_mode)
+                .with_line_number(is_debug_mode)
+                .compact(),
+        )
+        .finish()
+        .try_init()
+        .context("Failed to initialise tracing")
+        .map_err(init_error)?;
+
+    Ok(())
+}
+
+async fn start_server(args: Args) -> Result<(), SyncServerError> {
+    info!(
+        "Starting VaultLink server version {}",
+        env!("CARGO_PKG_VERSION")
+    );
+
+    create_server(args.config_path)
+        .await
+        .context("Failed to start server")
+        .map_err(init_error)
+}

backend/sync_server/src/server.rs

@@ -1,36 +1,41 @@
-pub mod auth;
+mod auth;
 mod create_document;
 mod delete_document;
-mod device_id_header;
 mod fetch_document_version;
 mod fetch_document_version_content;
 mod fetch_latest_document_version;
 mod fetch_latest_documents;
-mod index;
 mod ping;
 mod requests;
 mod responses;
 mod update_document;
 mod websocket;
 
+use std::{ffi::OsString, sync::Arc};
+
+use aide::{
+    axum::{
+        ApiRouter,
+        routing::{delete, get, post, put},
+    },
+    openapi::{Info, OpenApi},
+    scalar::Scalar,
+    transform::TransformOpenApi,
+};
 use anyhow::{Context as _, Result, anyhow};
-use auth::auth_middleware;
 use axum::{
-    Router,
+    Extension, Json,
     extract::{DefaultBodyLimit, Request},
     http::{self, HeaderValue, Method},
-    middleware,
     response::IntoResponse,
-    routing::{IntoMakeService, delete, get, post, put},
+    routing::IntoMakeService,
 };
-use device_id_header::DEVICE_ID_HEADER_NAME;
-use log::info;
+use log::{error, info};
 use tokio::signal;
 use tower_http::{
     LatencyUnit,
     cors::CorsLayer,
     limit::RequestBodyLimitLayer,
-    timeout::TimeoutLayer,
     trace::{
         DefaultOnBodyChunk, DefaultOnEos, DefaultOnFailure, DefaultOnRequest, DefaultOnResponse,
         TraceLayer,
@@ -40,42 +45,67 @@ use tracing::{Level, info_span};
 
 use crate::{
     app_state::AppState,
-    config::{Config, server_config::ServerConfig},
-    errors::{client_error, not_found_error},
+    config::server_config::ServerConfig,
+    errors::{SerializedError, not_found_error},
 };
 
-pub async fn create_server(config: Config) -> Result<()> {
-    let app_state = AppState::try_new(config)
+pub async fn create_server(config_path: Option<OsString>) -> Result<()> {
+    aide::r#gen::on_error(|err| error!("{err}"));
+    aide::r#gen::extract_schemas(true);
+
+    let app_state = AppState::try_new(config_path)
         .await
         .context("Failed to initialise app state")?;
 
     let server_config = app_state.config.server.clone();
 
-    let app = Router::new()
-        .nest("/", get_authed_routes(app_state.clone()))
-        .route("/", get(index::index))
-        .route("/vaults/:vault_id/ping", get(ping::ping))
-        .route("/vaults/:vault_id/ws", get(websocket::websocket_handler))
-        .layer(DefaultBodyLimit::disable())
-        .layer(RequestBodyLimitLayer::new(
-            app_state.config.server.max_body_size_mb * 1024 * 1024,
-        ))
-        .layer(TimeoutLayer::new(server_config.response_timeout))
-        .layer(
-            CorsLayer::new()
-                .allow_origin("*".parse::<HeaderValue>().expect("Failed to parse origin"))
-                .allow_headers([
-                    http::header::CONTENT_TYPE,
-                    http::header::AUTHORIZATION,
-                    DEVICE_ID_HEADER_NAME.clone(),
-                ])
-                .allow_methods([Method::GET, Method::POST, Method::PUT, Method::DELETE]),
+    let mut api = create_open_api();
+    let app = ApiRouter::new()
+        .api_route("/ping", get(ping::ping))
+        .api_route(
+            "/vaults/:vault_id/documents",
+            get(fetch_latest_documents::fetch_latest_documents),
         )
+        .route("/vaults/:vault_id/ws", get(websocket::websocket_handler))
+        .api_route(
+            "/vaults/:vault_id/documents",
+            post(create_document::create_document_multipart),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/json",
+            post(create_document::create_document_json),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id",
+            get(fetch_latest_document_version::fetch_latest_document_version),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id",
+            put(update_document::update_document_multipart),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id/json",
+            put(update_document::update_document_json),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id/versions/:version_id",
+            put(fetch_document_version::fetch_document_version),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id/versions/:version_id/content",
+            put(fetch_document_version_content::fetch_document_version_content),
+        )
+        .api_route(
+            "/vaults/:vault_id/documents/:document_id",
+            delete(delete_document::delete_document),
+        )
+        .route("/", Scalar::new("/api.json").axum_route())
+        .route("/api.json", axum::routing::get(serve_api))
         .layer(
             TraceLayer::new_for_http()
                 .make_span_with(|request: &Request<_>| {
                     info_span!(
-                        "http",
+                        "http_request",
                         method = ?request.method(),
                         uri = ?request.uri(),
                     )
@@ -90,49 +120,49 @@ pub async fn create_server(config: Config) -> Result<()> {
                 .on_eos(DefaultOnEos::new())
                 .on_failure(DefaultOnFailure::new().level(Level::ERROR)),
         )
+        .layer(DefaultBodyLimit::disable())
+        .layer(RequestBodyLimitLayer::new(
+            app_state.config.server.max_body_size_mb * 1024 * 1024,
+        ))
+        .layer(
+            CorsLayer::new()
+                .allow_origin("*".parse::<HeaderValue>().expect("Failed to parse origin"))
+                .allow_headers([http::header::CONTENT_TYPE, http::header::AUTHORIZATION])
+                .allow_methods([Method::GET, Method::POST, Method::PUT, Method::DELETE]),
+        )
         .with_state(app_state)
-        .fallback(handle_404)
-        .fallback(handle_405)
+        .finish_api_with(&mut api, add_api_docs_error_example)
+        .layer(Extension(Arc::new(api))) // https://github.com/tamasfe/aide/blob/507f4a8822bc0c13cbda0f589da1e0f4cbcdb812/examples/example-axum/src/main.rs#L39
+        .fallback(handler_404)
         .into_make_service();
 
     start_server(app, &server_config).await
 }
 
-fn get_authed_routes(app_state: AppState) -> Router<AppState> {
-    Router::new()
-        .route(
-            "/vaults/:vault_id/documents",
-            get(fetch_latest_documents::fetch_latest_documents),
-        )
-        .route(
-            "/vaults/:vault_id/documents",
-            post(create_document::create_document),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id",
-            get(fetch_latest_document_version::fetch_latest_document_version),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id/binary",
-            put(update_document::update_binary),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id/text",
-            put(update_document::update_text),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id/versions/:vault_update_id",
-            get(fetch_document_version::fetch_document_version),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id/versions/:vault_update_id/content",
-            get(fetch_document_version_content::fetch_document_version_content),
-        )
-        .route(
-            "/vaults/:vault_id/documents/:document_id",
-            delete(delete_document::delete_document),
-        )
-        .layer(middleware::from_fn_with_state(app_state, auth_middleware))
+async fn serve_api(Extension(api): Extension<Arc<OpenApi>>) -> impl IntoResponse { Json(api) }
+
+fn create_open_api() -> OpenApi {
+    OpenApi {
+        info: Info {
+            title: "VaultLink sync server".to_owned(),
+            summary: Some(
+                "Simple API for syncing documents between concurrent clients.".to_owned(),
+            ),
+            description: Some(include_str!("../README.md").to_owned()),
+            version: env!("CARGO_PKG_VERSION").to_owned(),
+            ..Info::default()
+        },
+        ..OpenApi::default()
+    }
+}
+
+fn add_api_docs_error_example(api: TransformOpenApi<'_>) -> TransformOpenApi<'_> {
+    api.default_response_with::<Json<SerializedError>, _>(|res| {
+        res.example(SerializedError {
+            message: "An error has occurred".to_owned(),
+            causes: vec![],
+        })
+    })
 }
 
 async fn start_server(app: IntoMakeService<axum::Router>, config: &ServerConfig) -> Result<()> {
@@ -179,10 +209,4 @@ async fn shutdown_signal() {
     }
 }
 
-async fn handle_404() -> impl IntoResponse {
-    not_found_error(anyhow!("Page not found"))
-}
-
-async fn handle_405() -> impl IntoResponse {
-    client_error(anyhow!("Method not allowed"))
-}
+async fn handler_404() -> impl IntoResponse { not_found_error(anyhow!("Page not found")) }

backend/sync_server/src/server/auth.rs

@@ -0,0 +1,15 @@
+use crate::{
+    app_state::AppState,
+    config::user_config::User,
+    errors::{SyncServerError, unauthorized_error},
+};
+
+// TODO: turn this into a middleware
+pub fn auth(app_state: &AppState, token: &str) -> Result<User, SyncServerError> {
+    app_state
+        .config
+        .users
+        .get_user(token)
+        .cloned()
+        .ok_or_else(|| unauthorized_error(anyhow::anyhow!("Invalid token")))
+}

backend/sync_server/src/server/create_document.rs

@@ -0,0 +1,152 @@
+use aide_axum_typed_multipart::TypedMultipart;
+use anyhow::Context as _;
+use axum::extract::{Path, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
+};
+use axum_jsonschema::Json;
+use schemars::JsonSchema;
+use serde::Deserialize;
+use sync_lib::base64_to_bytes;
+
+use super::{
+    auth::auth,
+    requests::{CreateDocumentVersion, CreateDocumentVersionMultipart},
+};
+use crate::{
+    app_state::{
+        AppState,
+        database::models::{
+            DocumentId, DocumentVersionWithoutContent, StoredDocumentVersion, VaultId,
+        },
+    },
+    errors::{SyncServerError, client_error, server_error},
+    utils::sanitize_path,
+};
+
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
+pub struct CreateDocumentPathParams {
+    vault_id: VaultId,
+}
+
+/// Create a new document in case a document with the same doesn't exist
+/// already. If a document with the same path exists, a new version is created
+/// with their content merged.
+#[axum::debug_handler]
+pub async fn create_document_multipart(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
+    Path(CreateDocumentPathParams { vault_id }): Path<CreateDocumentPathParams>,
+    State(state): State<AppState>,
+    TypedMultipart(axum_typed_multipart::TypedMultipart(request)): TypedMultipart<
+        CreateDocumentVersionMultipart,
+    >,
+) -> Result<Json<DocumentVersionWithoutContent>, SyncServerError> {
+    internal_create_document(
+        auth_header,
+        state,
+        vault_id,
+        request.document_id,
+        request.relative_path,
+        request.content.contents.to_vec(),
+    )
+    .await
+}
+
+/// Create a new document in case a document with the same doesn't exist
+/// already. If a document with the same path exists, a new version is created
+/// with their content merged.
+#[axum::debug_handler]
+pub async fn create_document_json(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
+    Path(CreateDocumentPathParams { vault_id }): Path<CreateDocumentPathParams>,
+    State(state): State<AppState>,
+    Json(request): Json<CreateDocumentVersion>,
+) -> Result<Json<DocumentVersionWithoutContent>, SyncServerError> {
+    let content_bytes = base64_to_bytes(&request.content_base64)
+        .context("Failed to decode base64 content in request")
+        .map_err(client_error)?;
+
+    internal_create_document(
+        auth_header,
+        state,
+        vault_id,
+        request.document_id,
+        request.relative_path,
+        content_bytes,
+    )
+    .await
+}
+
+async fn internal_create_document(
+    auth_header: Authorization<Bearer>,
+    state: AppState,
+    vault_id: VaultId,
+    document_id: Option<DocumentId>,
+    relative_path: String,
+    content: Vec<u8>,
+) -> Result<Json<DocumentVersionWithoutContent>, SyncServerError> {
+    auth(&state, auth_header.token())?;
+
+    let mut transaction = state
+        .database
+        .create_write_transaction(&vault_id)
+        .await
+        .map_err(server_error)?;
+
+    let document_id = match document_id {
+        Some(document_id) => {
+            let existing_version = state
+                .database
+                .get_latest_document(&vault_id, &document_id, Some(&mut transaction))
+                .await
+                .map_err(server_error)?;
+
+            if existing_version.is_some() {
+                return Err(client_error(anyhow::anyhow!(
+                    "Document with the same ID already exists"
+                )));
+            }
+
+            document_id
+        }
+        None => uuid::Uuid::new_v4(),
+    };
+
+    let last_update_id = state
+        .database
+        .get_max_update_id_in_vault(&vault_id, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    let sanitized_relative_path = sanitize_path(&relative_path);
+
+    let new_version = StoredDocumentVersion {
+        vault_update_id: last_update_id + 1,
+        document_id,
+        relative_path: sanitized_relative_path,
+        content,
+        updated_date: chrono::Utc::now(),
+        is_deleted: false,
+    };
+
+    state
+        .database
+        .insert_document_version(&vault_id, &new_version, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    transaction
+        .commit()
+        .await
+        .context("Failed to commit successful transaction")
+        .map_err(server_error)?;
+
+    state
+        .broadcasts
+        .send(vault_id, new_version.clone().into())
+        .await;
+
+    Ok(Json(new_version.into()))
+}

backend/sync_server/src/server/delete_document.rs

@@ -0,0 +1,81 @@
+use anyhow::Context as _;
+use axum::extract::{Path, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
+};
+use axum_jsonschema::Json;
+use schemars::JsonSchema;
+use serde::Deserialize;
+
+use super::{auth::auth, requests::DeleteDocumentVersion};
+use crate::{
+    app_state::{
+        AppState,
+        database::models::{
+            DocumentId, DocumentVersionWithoutContent, StoredDocumentVersion, VaultId,
+        },
+    },
+    errors::{SyncServerError, server_error},
+    utils::sanitize_path,
+};
+
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
+pub struct DeleteDocumentPathParams {
+    vault_id: VaultId,
+    document_id: DocumentId,
+}
+
+#[axum::debug_handler]
+pub async fn delete_document(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
+    Path(DeleteDocumentPathParams {
+        vault_id,
+        document_id,
+    }): Path<DeleteDocumentPathParams>,
+    State(state): State<AppState>,
+    Json(request): Json<DeleteDocumentVersion>,
+) -> Result<Json<DocumentVersionWithoutContent>, SyncServerError> {
+    auth(&state, auth_header.token())?;
+
+    let mut transaction = state
+        .database
+        .create_write_transaction(&vault_id)
+        .await
+        .map_err(server_error)?;
+
+    let last_update_id = state
+        .database
+        .get_max_update_id_in_vault(&vault_id, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    let new_version = StoredDocumentVersion {
+        vault_update_id: last_update_id + 1,
+        document_id,
+        relative_path: sanitize_path(&request.relative_path),
+        content: vec![],
+        updated_date: chrono::Utc::now(),
+        is_deleted: true,
+    };
+
+    state
+        .database
+        .insert_document_version(&vault_id, &new_version, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    transaction
+        .commit()
+        .await
+        .context("Failed to commit successful transaction")
+        .map_err(server_error)?;
+
+    state
+        .broadcasts
+        .send(vault_id, new_version.clone().into())
+        .await;
+
+    Ok(Json(new_version.into()))
+}

backend/sync_server/src/server/fetch_document_version.rs

@@ -1,31 +1,33 @@
 use anyhow::anyhow;
-use axum::{
-    Json,
-    extract::{Path, State},
+use axum::extract::{Path, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
 };
-use log::debug;
+use axum_jsonschema::Json;
+use schemars::JsonSchema;
 use serde::Deserialize;
 
+use super::auth::auth;
 use crate::{
     app_state::{
         AppState,
         database::models::{DocumentId, DocumentVersion, VaultId, VaultUpdateId},
     },
     errors::{SyncServerError, not_found_error, server_error},
-    utils::normalize::normalize,
 };
 
-#[derive(Deserialize)]
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
 pub struct FetchDocumentVersionPathParams {
-    #[serde(deserialize_with = "normalize")]
     vault_id: VaultId,
-
     document_id: DocumentId,
     vault_update_id: VaultUpdateId,
 }
 
 #[axum::debug_handler]
 pub async fn fetch_document_version(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
     Path(FetchDocumentVersionPathParams {
         vault_id,
         document_id,
@@ -33,9 +35,7 @@ pub async fn fetch_document_version(
     }): Path<FetchDocumentVersionPathParams>,
     State(state): State<AppState>,
 ) -> Result<Json<DocumentVersion>, SyncServerError> {
-    debug!(
-        "Fetching document version `{vault_update_id}` for document `{document_id}` in vault `{vault_id}`"
-    );
+    auth(&state, auth_header.token())?;
 
     let result = state
         .database
@@ -54,7 +54,7 @@ pub async fn fetch_document_version(
     if result.document_id != document_id {
         return Err(not_found_error(anyhow!(
             "Document with document id `{document_id}` does not have a version with id \
-            `{vault_update_id}`",
+             `{vault_update_id}`",
         )));
     }
 

backend/sync_server/src/server/fetch_document_version_content.rs

@@ -3,29 +3,33 @@ use axum::{
     body::Bytes,
     extract::{Path, State},
 };
-use log::debug;
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
+};
+use schemars::JsonSchema;
 use serde::Deserialize;
 
+use super::auth::auth;
 use crate::{
     app_state::{
         AppState,
         database::models::{DocumentId, VaultId, VaultUpdateId},
     },
     errors::{SyncServerError, not_found_error, server_error},
-    utils::normalize::normalize,
 };
 
-#[derive(Deserialize)]
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
 pub struct FetchDocumentVersionContentPathParams {
-    #[serde(deserialize_with = "normalize")]
     vault_id: VaultId,
-
     document_id: DocumentId,
     vault_update_id: VaultUpdateId,
 }
 
 #[axum::debug_handler]
 pub async fn fetch_document_version_content(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
     Path(FetchDocumentVersionContentPathParams {
         vault_id,
         document_id,
@@ -33,9 +37,7 @@ pub async fn fetch_document_version_content(
     }): Path<FetchDocumentVersionContentPathParams>,
     State(state): State<AppState>,
 ) -> Result<Bytes, SyncServerError> {
-    debug!(
-        "Fetching document version `{vault_update_id}` for document `{document_id}` in vault `{vault_id}`"
-    );
+    auth(&state, auth_header.token())?;
 
     let result = state
         .database
@@ -54,7 +56,7 @@ pub async fn fetch_document_version_content(
     if result.document_id != document_id {
         return Err(not_found_error(anyhow!(
             "Document with document id `{document_id}` does not have a version with id \
-            `{vault_update_id}`",
+             `{vault_update_id}`",
         )));
     }
 

backend/sync_server/src/server/fetch_latest_document_version.rs

@@ -1,37 +1,39 @@
 use anyhow::anyhow;
-use axum::{
-    Json,
-    extract::{Path, State},
+use axum::extract::{Path, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
 };
-use log::debug;
+use axum_jsonschema::Json;
+use schemars::JsonSchema;
 use serde::Deserialize;
 
+use super::auth::auth;
 use crate::{
     app_state::{
         AppState,
         database::models::{DocumentId, DocumentVersion, VaultId},
     },
     errors::{SyncServerError, not_found_error, server_error},
-    utils::normalize::normalize,
 };
 
-#[derive(Deserialize)]
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
 pub struct FetchLatestDocumentVersionPathParams {
-    #[serde(deserialize_with = "normalize")]
     vault_id: VaultId,
-
     document_id: DocumentId,
 }
 
 #[axum::debug_handler]
 pub async fn fetch_latest_document_version(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
     Path(FetchLatestDocumentVersionPathParams {
         vault_id,
         document_id,
     }): Path<FetchLatestDocumentVersionPathParams>,
     State(state): State<AppState>,
 ) -> Result<Json<DocumentVersion>, SyncServerError> {
-    debug!("Fetching latest document version for document `{document_id}` in vault `{vault_id}`");
+    auth(&state, auth_header.token())?;
 
     let latest_version = state
         .database

backend/sync_server/src/server/fetch_latest_documents.rs

@@ -1,38 +1,41 @@
-use axum::{
-    Json,
-    extract::{Path, Query, State},
+use axum::extract::{Path, Query, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
 };
-use log::debug;
+use axum_jsonschema::Json;
+use schemars::JsonSchema;
 use serde::Deserialize;
 
-use super::responses::FetchLatestDocumentsResponse;
+use super::{auth::auth, responses::FetchLatestDocumentsResponse};
 use crate::{
     app_state::{
         AppState,
         database::models::{VaultId, VaultUpdateId},
     },
     errors::{SyncServerError, server_error},
-    utils::normalize::normalize,
 };
 
-#[derive(Deserialize)]
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
 pub struct FetchLatestDocumentsPathParams {
-    #[serde(deserialize_with = "normalize")]
     vault_id: VaultId,
 }
 
-#[derive(Deserialize)]
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
 pub struct QueryParams {
     since_update_id: Option<VaultUpdateId>,
 }
 
 #[axum::debug_handler]
 pub async fn fetch_latest_documents(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
     Path(FetchLatestDocumentsPathParams { vault_id }): Path<FetchLatestDocumentsPathParams>,
     Query(QueryParams { since_update_id }): Query<QueryParams>,
     State(state): State<AppState>,
 ) -> Result<Json<FetchLatestDocumentsResponse>, SyncServerError> {
-    debug!("Fetching latest documents in vault `{vault_id}` since update ID `{since_update_id:?}`");
+    auth(&state, auth_header.token())?;
 
     let documents = if let Some(since_update_id) = since_update_id {
         state

backend/sync_server/src/server/ping.rs

@@ -0,0 +1,22 @@
+use axum::{Json, extract::State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
+};
+
+use super::{auth::auth, responses::PingResponse};
+use crate::{app_state::AppState, errors::SyncServerError};
+
+#[axum::debug_handler]
+pub async fn ping(
+    maybe_auth_header: Option<TypedHeader<Authorization<Bearer>>>,
+    State(state): State<AppState>,
+) -> Result<Json<PingResponse>, SyncServerError> {
+    let is_authenticated =
+        maybe_auth_header.is_some_and(|auth_header| auth(&state, auth_header.token()).is_ok());
+
+    Ok(Json(PingResponse {
+        server_version: env!("CARGO_PKG_VERSION").to_owned(),
+        is_authenticated,
+    }))
+}

backend/sync_server/src/server/requests.rs

@@ -1,13 +1,13 @@
+use aide_axum_typed_multipart::FieldData;
 use axum::body::Bytes;
-use axum_typed_multipart::{FieldData, TryFromMultipart};
-use reconcile_text::NumberOrText;
+use axum_typed_multipart::TryFromMultipart;
+use schemars::JsonSchema;
 use serde::{self, Deserialize};
-use ts_rs::TS;
 
 use crate::app_state::database::models::{DocumentId, VaultUpdateId};
 
-#[derive(TS, Debug, TryFromMultipart)]
-#[ts(export)]
+#[derive(Debug, Deserialize, JsonSchema)]
+#[serde(rename_all = "camelCase")]
 pub struct CreateDocumentVersion {
     /// The client can decide the document id (if it wishes to) in order
     /// to help with syncing. If the client does not provide a document id,
@@ -15,37 +15,36 @@ pub struct CreateDocumentVersion {
     /// it must not already exist in the database.
     pub document_id: Option<DocumentId>,
     pub relative_path: String,
+    pub content_base64: String,
+}
 
-    #[ts(as = "Vec<u8>")]
+#[derive(Debug, TryFromMultipart, JsonSchema)]
+pub struct CreateDocumentVersionMultipart {
+    pub document_id: Option<DocumentId>,
+    pub relative_path: String,
     #[form_data(limit = "unlimited")]
     pub content: FieldData<Bytes>,
 }
 
-#[derive(Debug, TryFromMultipart)]
-pub struct UpdateBinaryDocumentVersion {
+#[derive(Debug, Deserialize, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct UpdateDocumentVersion {
     pub parent_version_id: VaultUpdateId,
     pub relative_path: String,
+    pub content_base64: String,
+}
 
+#[derive(Debug, TryFromMultipart, JsonSchema)]
+#[serde(rename_all = "camelCase")]
+pub struct UpdateDocumentVersionMultipart {
+    pub parent_version_id: VaultUpdateId,
+    pub relative_path: String,
     #[form_data(limit = "unlimited")]
     pub content: FieldData<Bytes>,
 }
 
-#[derive(TS, Debug, Deserialize)]
+#[derive(Debug, Deserialize, JsonSchema)]
 #[serde(rename_all = "camelCase")]
-#[ts(export)]
-pub struct UpdateTextDocumentVersion {
-    #[ts(as = "i32")]
-    pub parent_version_id: VaultUpdateId,
-
-    pub relative_path: String,
-
-    #[ts(type = "Array<number | string>")]
-    pub content: Vec<NumberOrText>,
-}
-
-#[derive(TS, Debug, Deserialize)]
-#[serde(rename_all = "camelCase")]
-#[ts(export)]
 pub struct DeleteDocumentVersion {
     pub relative_path: String,
 }

backend/sync_server/src/server/responses.rs

@@ -1,14 +1,13 @@
+use schemars::JsonSchema;
 use serde::{self, Serialize};
-use ts_rs::TS;
 
 use crate::app_state::database::models::{
     DocumentVersion, DocumentVersionWithoutContent, VaultUpdateId,
 };
 
 /// Response to a ping request.
-#[derive(TS, Debug, Clone, Serialize)]
+#[derive(Debug, Clone, Serialize, JsonSchema)]
 #[serde(rename_all = "camelCase")]
-#[ts(export)]
 pub struct PingResponse {
     /// Semantic version of the server.
     pub server_version: String,
@@ -16,19 +15,11 @@ pub struct PingResponse {
     /// Whether the client is authenticated based on the sent Authorization
     /// header.
     pub is_authenticated: bool,
-
-    /// List of file extensions that are allowed to be merged.
-    pub mergeable_file_extensions: Vec<String>,
-
-    /// API version ensuring backwards & forwards compatibility between the client
-    /// and server.
-    pub supported_api_version: u32,
 }
 
 /// Response to a fetch latest documents request.
-#[derive(TS, Debug, Clone, Serialize)]
+#[derive(Debug, Clone, Serialize, JsonSchema)]
 #[serde(rename_all = "camelCase")]
-#[ts(export)]
 pub struct FetchLatestDocumentsResponse {
     pub latest_documents: Vec<DocumentVersionWithoutContent>,
 
@@ -37,9 +28,8 @@ pub struct FetchLatestDocumentsResponse {
 }
 
 /// Response to an update document request.
-#[derive(TS, Debug, Clone, Serialize)]
+#[derive(Debug, Clone, Serialize, JsonSchema)]
 #[serde(tag = "type")]
-#[ts(export)]
 pub enum DocumentUpdateResponse {
     /// Returned when the created/updated document's content is the same as was
     /// sent in the create/update request and thus the response doesn't contain

backend/sync_server/src/server/update_document.rs

@@ -0,0 +1,231 @@
+use aide_axum_typed_multipart::TypedMultipart;
+use anyhow::{Context as _, anyhow};
+use axum::extract::{Path, State};
+use axum_extra::{
+    TypedHeader,
+    headers::{Authorization, authorization::Bearer},
+};
+use axum_jsonschema::Json;
+use log::info;
+use schemars::JsonSchema;
+use serde::Deserialize;
+use sync_lib::{base64_to_bytes, is_file_type_mergable, merge};
+
+use super::{
+    auth::auth,
+    requests::{UpdateDocumentVersion, UpdateDocumentVersionMultipart},
+    responses::DocumentUpdateResponse,
+};
+use crate::{
+    app_state::{
+        AppState,
+        database::models::{DocumentId, StoredDocumentVersion, VaultId, VaultUpdateId},
+    },
+    errors::{SyncServerError, client_error, not_found_error, server_error},
+    utils::{deduped_file_paths, sanitize_path},
+};
+
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
+pub struct UpdateDocumentPathParams {
+    vault_id: VaultId,
+    document_id: DocumentId,
+}
+
+#[axum::debug_handler]
+pub async fn update_document_multipart(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
+    Path(UpdateDocumentPathParams {
+        vault_id,
+        document_id,
+    }): Path<UpdateDocumentPathParams>,
+    State(state): State<AppState>,
+    TypedMultipart(axum_typed_multipart::TypedMultipart(request)): TypedMultipart<
+        UpdateDocumentVersionMultipart,
+    >,
+) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
+    internal_update_document(
+        auth_header,
+        state,
+        vault_id,
+        document_id,
+        request.parent_version_id,
+        request.relative_path,
+        request.content.contents.to_vec(),
+    )
+    .await
+}
+
+#[axum::debug_handler]
+pub async fn update_document_json(
+    TypedHeader(auth_header): TypedHeader<Authorization<Bearer>>,
+    Path(UpdateDocumentPathParams {
+        vault_id,
+        document_id,
+    }): Path<UpdateDocumentPathParams>,
+    State(state): State<AppState>,
+    Json(request): Json<UpdateDocumentVersion>,
+) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
+    let content_bytes = base64_to_bytes(&request.content_base64)
+        .context("Failed to decode base64 content in request")
+        .map_err(client_error)?;
+
+    internal_update_document(
+        auth_header,
+        state,
+        vault_id,
+        document_id,
+        request.parent_version_id,
+        request.relative_path,
+        content_bytes,
+    )
+    .await
+}
+
+#[allow(clippy::too_many_arguments, clippy::too_many_lines)]
+async fn internal_update_document(
+    auth_header: Authorization<Bearer>,
+    state: AppState,
+    vault_id: VaultId,
+    document_id: DocumentId,
+    parent_version_id: VaultUpdateId,
+    relative_path: String,
+    content: Vec<u8>,
+) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
+    auth(&state, auth_header.token())?;
+
+    // No need for a transaction as document versions are immutable
+    let parent_document = state
+        .database
+        .get_document_version(&vault_id, parent_version_id, None)
+        .await
+        .map_err(server_error)?
+        .map_or_else(
+            || {
+                Err(not_found_error(anyhow!(
+                    "Parent version with id `{}` not found",
+                    parent_version_id
+                )))
+            },
+            Ok,
+        )?;
+
+    let mut transaction = state
+        .database
+        .create_write_transaction(&vault_id)
+        .await
+        .map_err(server_error)?;
+
+    let last_update_id = state
+        .database
+        .get_max_update_id_in_vault(&vault_id, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    let latest_version = state
+        .database
+        .get_latest_document(&vault_id, &document_id, Some(&mut transaction))
+        .await
+        .map_err(server_error)?
+        .map_or_else(
+            || {
+                Err(not_found_error(anyhow!(
+                    "Document with id `{document_id}` not found",
+                )))
+            },
+            Ok,
+        )?;
+
+    if latest_version.is_deleted {
+        transaction
+            .rollback()
+            .await
+            .context("Failed to roll back transaction")
+            .map_err(server_error)?;
+
+        return Ok(Json(DocumentUpdateResponse::FastForwardUpdate(
+            latest_version.into(),
+        )));
+    }
+
+    let sanitized_relative_path = sanitize_path(&relative_path);
+
+    // Return the latest version if the content and path are the same as the latest
+    // version
+    if content == latest_version.content && sanitized_relative_path == latest_version.relative_path
+    {
+        info!("Document content is the same as the latest version, skipping update");
+        transaction
+            .rollback()
+            .await
+            .context("Failed to roll back transaction")
+            .map_err(server_error)?;
+
+        return Ok(Json(DocumentUpdateResponse::FastForwardUpdate(
+            latest_version.into(),
+        )));
+    }
+
+    let merged_content = if is_file_type_mergable(&sanitized_relative_path) {
+        merge(&parent_document.content, &latest_version.content, &content)
+    } else {
+        content.clone()
+    };
+
+    let is_different_from_request_content = merged_content != content;
+
+    // We can only update the relative path if we're the first one to do so
+    let new_relative_path = if parent_document.relative_path == latest_version.relative_path
+        && latest_version.relative_path != sanitized_relative_path
+    {
+        let mut new_relative_path = String::default();
+        for candidate in deduped_file_paths(&sanitized_relative_path) {
+            if state
+                .database
+                .get_latest_document_by_path(&vault_id, &candidate, Some(&mut transaction))
+                .await
+                .map_err(server_error)?
+                .is_none()
+            {
+                new_relative_path = candidate;
+                break;
+            }
+        }
+
+        new_relative_path
+    } else {
+        latest_version.relative_path.clone()
+    };
+
+    let new_version = StoredDocumentVersion {
+        document_id,
+        vault_update_id: last_update_id + 1,
+        relative_path: new_relative_path,
+        content: merged_content,
+        updated_date: chrono::Utc::now(),
+        is_deleted: false,
+    };
+
+    state
+        .database
+        .insert_document_version(&vault_id, &new_version, Some(&mut transaction))
+        .await
+        .map_err(server_error)?;
+
+    transaction
+        .commit()
+        .await
+        .context("Failed to commit successful transaction")
+        .map_err(server_error)?;
+
+    state
+        .broadcasts
+        .send(vault_id, new_version.clone().into())
+        .await;
+
+    Ok(Json(if is_different_from_request_content {
+        DocumentUpdateResponse::MergingUpdate(new_version.into())
+    } else {
+        DocumentUpdateResponse::FastForwardUpdate(new_version.into())
+    }))
+}

backend/sync_server/src/server/websocket.rs

@@ -0,0 +1,147 @@
+use anyhow::Context;
+use axum::{
+    extract::{
+        Path, Query, State,
+        ws::{Message, WebSocket, WebSocketUpgrade},
+    },
+    response::Response,
+};
+use futures::{
+    sink::SinkExt,
+    stream::{SplitSink, StreamExt},
+};
+use log::{error, info, warn};
+use schemars::JsonSchema;
+use serde::Deserialize;
+
+use super::auth::auth;
+use crate::{
+    app_state::{
+        AppState,
+        database::models::{DocumentVersionWithoutContent, VaultId, VaultUpdateId},
+    },
+    errors::{SyncServerError, server_error, unauthorized_error},
+};
+
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
+pub struct WebsocketPathParams {
+    vault_id: VaultId,
+}
+
+// This is required for aide to infer the path parameter types and names
+#[derive(Deserialize, JsonSchema)]
+pub struct QueryParams {
+    since_update_id: Option<VaultUpdateId>,
+}
+
+pub async fn websocket_handler(
+    ws: WebSocketUpgrade,
+    Path(WebsocketPathParams { vault_id }): Path<WebsocketPathParams>,
+    Query(QueryParams { since_update_id }): Query<QueryParams>,
+    State(state): State<AppState>,
+) -> Result<Response, SyncServerError> {
+    Ok(ws.on_upgrade(move |socket| websocket_wrapped(state, socket, vault_id, since_update_id)))
+}
+
+async fn websocket_wrapped(
+    state: AppState,
+    stream: WebSocket,
+    vault_id: VaultId,
+    since_update_id: Option<VaultUpdateId>,
+) {
+    info!("Websocket connection opened on vault '{}'", vault_id);
+
+    let result = websocket(state, stream, vault_id.clone(), since_update_id).await;
+
+    if let Err(err) = result {
+        error!(
+            "Websocket connection error on vault '{}': {}",
+            vault_id, err
+        );
+    }
+
+    warn!("Websocket connection closed on vault '{}'", vault_id);
+}
+
+async fn websocket(
+    state: AppState,
+    stream: WebSocket,
+    vault_id: VaultId,
+    since_update_id: Option<VaultUpdateId>,
+) -> Result<(), SyncServerError> {
+    let (mut sender, mut receiver) = stream.split();
+
+    if let Some(Ok(Message::Text(token))) = receiver.next().await {
+        auth(&state, &token)?;
+    } else {
+        return Err(unauthorized_error(anyhow::anyhow!(
+            "Failed to authenticate"
+        )));
+    }
+
+    let mut rx = state.broadcasts.get_receiver(vault_id.clone()).await;
+
+    let documents = if let Some(since_update_id) = since_update_id {
+        state
+            .database
+            .get_latest_documents_since(&vault_id, since_update_id, None)
+            .await
+            .map_err(server_error)
+    } else {
+        state
+            .database
+            .get_latest_documents(&vault_id, None)
+            .await
+            .map_err(server_error)
+    }?;
+
+    for document in documents {
+        send_document_over_websocket(document, &mut sender).await?;
+    }
+
+    let mut send_task = tokio::spawn(async move {
+        while let Ok(update) = rx.recv().await {
+            send_document_over_websocket(update, &mut sender).await?;
+        }
+
+        Ok::<(), SyncServerError>(())
+    });
+
+    let mut recv_task =
+        tokio::spawn(
+            async move { while let Some(Ok(Message::Text(_text))) = receiver.next().await {} },
+        );
+
+    tokio::select! {
+        _ = &mut send_task => recv_task.abort(),
+        _ = &mut recv_task => send_task.abort(),
+    };
+
+    send_task
+        .await
+        .context("Websocket send task failed")
+        .map_err(server_error)??;
+
+    recv_task
+        .await
+        .context("Websocket receive task failed")
+        .map_err(server_error)?;
+
+    Ok(())
+}
+
+async fn send_document_over_websocket(
+    document: DocumentVersionWithoutContent,
+    sender: &mut SplitSink<WebSocket, Message>,
+) -> Result<(), SyncServerError> {
+    let serialized_update = serde_json::to_string(&document)
+        .context("Failed to serialize update")
+        .map_err(server_error)?;
+
+    sender
+        .send(Message::Text(serialized_update))
+        .await
+        .context("Failed to send message over websocket")
+        .map_err(server_error)
+}

backend/sync_server/src/utils.rs

@@ -0,0 +1,118 @@
+use regex::Regex;
+
+/// Sanitize the document's path to allow all clients to create the same path in
+/// their filesystem. If we didn't do this server-side, client's would need to
+/// deal with mapping invalid names to valid ones and then back.
+pub fn sanitize_path(path: &str) -> String {
+    let options = sanitize_filename::Options {
+        truncate: true,
+        windows: true, // Windows is the lowest common denominator
+        replacement: "",
+    };
+
+    path.split('/')
+        .map(|part| {
+            let proposal = sanitize_filename::sanitize_with_options(part, options.clone());
+            if !part.is_empty() && proposal.is_empty() {
+                "_".to_owned()
+            } else {
+                proposal
+            }
+        })
+        .collect::<Vec<_>>()
+        .join("/")
+}
+
+pub fn deduped_file_paths(path: &str) -> impl Iterator<Item = String> {
+    let mut path_parts = path.split('/').collect::<Vec<_>>();
+    let file_name = path_parts.pop().unwrap().to_owned();
+
+    let mut directory = path_parts.join("/");
+    if !directory.is_empty() {
+        directory.push('/');
+    }
+
+    let name_parts = file_name.rsplitn(2, '.').collect::<Vec<_>>();
+    let mut reverse_parts = name_parts.into_iter().rev();
+    let (stem, extension) = match (reverse_parts.next(), reverse_parts.next()) {
+        (Some(stem), maybe_extension) => (
+            stem.to_owned(),
+            maybe_extension
+                .map(|ext| format!(".{ext}"))
+                .unwrap_or_default(),
+        ),
+        _ => unreachable!("Path must have at least one part"),
+    };
+
+    let regex = Regex::new(r" \((\d+)\)$").unwrap();
+    let start_number = regex
+        .captures(&stem)
+        .and_then(|caps| caps.get(1))
+        .and_then(|m| m.as_str().parse::<u32>().ok())
+        .unwrap_or(0);
+
+    let clean_stem = regex.replace(&stem, "").to_string();
+
+    (start_number..).map(move |dedup_number| {
+        if dedup_number == 0 {
+            format!("{directory}{clean_stem}{extension}")
+        } else {
+            format!("{directory}{clean_stem} ({dedup_number}){extension}")
+        }
+    })
+}
+
+#[cfg(test)]
+mod test {
+    use super::*;
+
+    #[test]
+    fn test_sanitize_path() {
+        assert_eq!(sanitize_path("/my/path/what?"), "/my/path/what");
+        assert_eq!(sanitize_path("file (1).md"), "file (1).md");
+        assert_eq!(sanitize_path("/my/path/\\\\:?"), "/my/path/_");
+    }
+
+    #[test]
+    fn test_deduped_file_paths() {
+        let mut deduped = deduped_file_paths("file.txt");
+        assert_eq!(deduped.next(), Some("file.txt".to_owned()));
+        assert_eq!(deduped.next(), Some("file (1).txt".to_owned()));
+        assert_eq!(deduped.next(), Some("file (2).txt".to_owned()));
+
+        let mut deduped = deduped_file_paths("file");
+        assert_eq!(deduped.next(), Some("file".to_owned()));
+        assert_eq!(deduped.next(), Some("file (1)".to_owned()));
+        assert_eq!(deduped.next(), Some("file (2)".to_owned()));
+
+        let mut deduped = deduped_file_paths("file (51).md");
+        assert_eq!(deduped.next(), Some("file (51).md".to_owned()));
+        assert_eq!(deduped.next(), Some("file (52).md".to_owned()));
+        assert_eq!(deduped.next(), Some("file (53).md".to_owned()));
+
+        let mut deduped = deduped_file_paths("file (5)");
+        assert_eq!(deduped.next(), Some("file (5)".to_owned()));
+        assert_eq!(deduped.next(), Some("file (6)".to_owned()));
+        assert_eq!(deduped.next(), Some("file (7)".to_owned()));
+
+        let mut deduped = deduped_file_paths("my/path.with.dots/file (5).md");
+        assert_eq!(
+            deduped.next(),
+            Some("my/path.with.dots/file (5).md".to_owned())
+        );
+        assert_eq!(
+            deduped.next(),
+            Some("my/path.with.dots/file (6).md".to_owned())
+        );
+
+        let mut deduped = deduped_file_paths("my/path.with.dots/file (5)");
+        assert_eq!(
+            deduped.next(),
+            Some("my/path.with.dots/file (5)".to_owned())
+        );
+        assert_eq!(
+            deduped.next(),
+            Some("my/path.with.dots/file (6)".to_owned())
+        );
+    }
+}

docs/.cspell.json

@@ -1,87 +0,0 @@
-{
-    "version": "0.2",
-    "language": "en-GB",
-    "dictionaries": ["en-gb"],
-    "ignorePaths": ["node_modules", ".vitepress/dist", ".vitepress/cache", "package-lock.json"],
-    "words": [
-        "VaultLink",
-        "Obsidian",
-        "WebSocket",
-        "SQLite",
-        "codebase",
-        "CRDT",
-        "CRDTs",
-        "YAML",
-        "nginx",
-        "Caddy",
-        "Traefik",
-        "systemd",
-        "localhost",
-        "vaultlink",
-        "Axum",
-        "Tokio",
-        "SQLx",
-        "reconcile",
-        "postgresql",
-        "VitePress",
-        "markdownlint",
-        "filesystem",
-        "backend",
-        "frontend",
-        "macOS",
-        "CLI",
-        "API",
-        "JSON",
-        "HTTP",
-        "HTTPS",
-        "SSL",
-        "TLS",
-        "WSS",
-        "TCP",
-        "VPS",
-        "Docker",
-        "Github",
-        "Dockerfile",
-        "dockerignore",
-        "Rustup",
-        "PostgreSQL",
-        "UUID",
-        "CORS",
-        "HSTS",
-        "CI",
-        "CD",
-        "OpenSSL",
-        "README",
-        "config",
-        "submodule",
-        "repo",
-        "autocomplete",
-        "autoformat",
-        "dedupe",
-        "diff",
-        "grep",
-        "stdout",
-        "stderr",
-        "chmod",
-        "mkdir",
-        "rclone",
-        "uuidgen",
-        "letsencrypt",
-        "fullchain",
-        "privkey",
-        "schmelczer",
-        "Schmelczer",
-        "ghcr",
-        "keepalive",
-        "healthcheck",
-        "writable",
-        "Cloudant",
-        "Syncthing",
-        "cadvisor",
-        "Caddyfile",
-        "nodelay",
-        "websecure",
-        "certresolver",
-        "rootfs"
-    ]
-}

docs/.gitignore

@@ -1,2 +0,0 @@
-.vitepress/dist/
-.vitepress/cache/

docs/.prettierignore

@@ -1,4 +0,0 @@
-node_modules/
-.vitepress/dist/
-.vitepress/cache/
-package-lock.json

docs/.prettierrc

@@ -1,19 +0,0 @@
-{
-    "printWidth": 120,
-    "tabWidth": 4,
-    "useTabs": false,
-    "semi": false,
-    "singleQuote": false,
-    "trailingComma": "none",
-    "endOfLine": "lf",
-    "proseWrap": "preserve",
-    "overrides": [
-        {
-            "files": "*.md",
-            "options": {
-                "proseWrap": "preserve",
-                "printWidth": 120
-            }
-        }
-    ]
-}

docs/.vitepress/config.mts

@@ -1,60 +0,0 @@
-import { defineConfig } from "vitepress"
-
-export default defineConfig({
-    title: "VaultLink",
-    description: "Self-hosted real-time synchronisation for Obsidian",
-    base: "/vault-link/",
-    themeConfig: {
-        logo: "/logo.svg",
-        nav: [
-            { text: "Home", link: "/" },
-            { text: "Guide", link: "/guide/getting-started" },
-            { text: "Architecture", link: "/architecture/" },
-            { text: "GitHub", link: "https://github.com/schmelczer/vault-link" }
-        ],
-        sidebar: [
-            {
-                text: "Introduction",
-                items: [
-                    { text: "What is VaultLink?", link: "/guide/what-is-vaultlink" },
-                    { text: "Getting Started", link: "/guide/getting-started" },
-                    { text: "Limitations", link: "/guide/limitations" },
-                    { text: "Comparison with Alternatives", link: "/guide/alternatives" }
-                ]
-            },
-            {
-                text: "Setup",
-                items: [
-                    { text: "Server Setup", link: "/guide/server-setup" },
-                    { text: "Obsidian Plugin", link: "/guide/obsidian-plugin" },
-                    { text: "CLI Client", link: "/guide/cli-client" }
-                ]
-            },
-            {
-                text: "Configuration",
-                items: [
-                    { text: "Server Configuration", link: "/config/server" },
-                    { text: "Authentication", link: "/config/authentication" },
-                    { text: "Advanced Options", link: "/config/advanced" }
-                ]
-            },
-            {
-                text: "Architecture",
-                items: [
-                    { text: "Overview", link: "/architecture/" },
-                    { text: "Sync Algorithm", link: "/architecture/sync-algorithm" },
-                    { text: "Data Flow", link: "/architecture/data-flow" }
-                ]
-            }
-        ],
-        socialLinks: [{ icon: "github", link: "https://github.com/schmelczer/vault-link" }],
-        footer: {
-            message: "Released under the MIT License.",
-            copyright: "Copyright © 2024-present Andras Schmelczer"
-        },
-        search: {
-            provider: "local"
-        }
-    },
-    head: [["link", { rel: "icon", type: "image/svg+xml", href: "/vault-link/logo.svg" }]]
-})

docs/README.md

@@ -1,159 +0,0 @@
-# VaultLink Documentation
-
-This directory contains the VaultLink documentation site built with [VitePress](https://vitepress.dev/).
-
-## Development
-
-### Prerequisites
-
-- Node.js 18+
-- npm
-
-### Setup
-
-```bash
-cd docs
-npm install
-```
-
-### Local Development
-
-Start the development server with hot reload:
-
-```bash
-npm run dev
-```
-
-The site will be available at `http://localhost:5173/vault-link/`
-
-### Build
-
-Build the static site:
-
-```bash
-npm run build
-```
-
-Output will be in `.vitepress/dist/`
-
-### Preview
-
-Preview the built site:
-
-```bash
-npm run preview
-```
-
-### Format
-
-Format all markdown and TypeScript files:
-
-```bash
-npm run format
-```
-
-Check formatting without making changes:
-
-```bash
-npm run format:check
-```
-
-### Spell Check
-
-Check spelling (British English):
-
-```bash
-npm run spell
-```
-
-The spell checker enforces British English spellings (e.g., "synchronisation", "optimise", "behaviour").
-
-## Deployment
-
-The documentation is automatically deployed to GitHub Pages when changes are pushed to the `main` branch.
-
-The deployment workflow is configured in `.github/workflows/deploy-docs.yml`.
-
-## Structure
-
-```
-docs/
-├── .vitepress/
-│   └── config.ts          # VitePress configuration
-├── public/                # Static assets
-│   └── logo.svg          # VaultLink logo
-├── guide/                # User guides
-│   ├── what-is-vaultlink.md
-│   ├── getting-started.md
-│   ├── server-setup.md
-│   ├── obsidian-plugin.md
-│   └── cli-client.md
-├── architecture/         # Architecture documentation
-│   ├── index.md
-│   ├── sync-algorithm.md
-│   └── data-flow.md
-├── config/              # Configuration reference
-│   ├── server.md
-│   ├── authentication.md
-│   └── advanced.md
-└── index.md            # Home page
-
-```
-
-## Writing Documentation
-
-### Language
-
-All documentation uses **British English**. The spell checker enforces this in CI.
-
-### Markdown Features
-
-VitePress supports:
-
-- GitHub Flavoured Markdown
-- Custom containers (tip, warning, danger)
-- Code syntax highlighting
-- Mermaid diagrams
-- Emoji :rocket:
-
-### Custom Containers
-
-```markdown
-::: tip
-This is a tip
-:::
-
-::: warning
-This is a warning
-:::
-
-::: danger
-This is a danger message
-:::
-```
-
-### Code Blocks
-
-````markdown
-```bash
-npm install
-```
-
-```yaml
-server:
-    port: 3000
-```
-````
-
-## Contributing
-
-When adding new pages:
-
-1. Create the markdown file in the appropriate directory
-2. Add it to the sidebar in `.vitepress/config.ts`
-3. Test locally with `npm run dev`
-4. Submit a pull request
-
-## License
-
-MIT - Same as VaultLink

docs/architecture/data-flow.md

@@ -1,553 +0,0 @@
-# Data Flow
-
-How data flows through VaultLink, from client to server and back.
-
-## Connection Lifecycle
-
-### 1. Initial Connection
-
-```mermaid
-sequenceDiagram
-    participant C as Client
-    participant S as Server
-    participant DB as Database
-
-    C->>S: WebSocket connect
-    S->>S: Accept connection
-    C->>S: Auth message (token + vault)
-    S->>S: Validate token
-    S->>S: Check vault access
-    S-->>C: Auth success
-    Note over C,S: Connection established
-```
-
-**Steps**:
-
-1. Client initiates WebSocket connection to server
-2. Server accepts connection
-3. Client sends authentication message with token and vault name
-4. Server validates token against `config.yml`
-5. Server checks if user has access to requested vault
-6. Server responds with success or error
-7. Connection is ready for syncing
-
-### 2. Initial Sync
-
-After authentication, the client performs initial synchronisation:
-
-```mermaid
-sequenceDiagram
-    participant C as Client
-    participant S as Server
-    participant DB as SQLite
-
-    C->>C: Scan local filesystem
-    C->>S: Request file list
-    S->>DB: Query all files
-    DB-->>S: File metadata
-    S-->>C: File list with versions
-
-    loop For each local file
-        C->>C: Check if file on server
-        alt File not on server
-            C->>S: Upload file
-            S->>DB: Store file + metadata
-        else File on server (different version)
-            C->>C: Compare versions
-            C->>S: Upload newer or merge
-        end
-    end
-
-    loop For each server file
-        C->>C: Check if file local
-        alt File not local
-            C->>S: Download file
-            S->>DB: Retrieve file
-            DB-->>S: File content
-            S-->>C: File content
-            C->>C: Write to disk
-        end
-    end
-
-    S-->>C: Sync complete message
-```
-
-**Process**:
-
-1. Client scans local filesystem
-2. Client requests file list from server
-3. Server queries database and returns metadata
-4. Client uploads missing or changed local files
-5. Client downloads missing files from server
-6. Server sends sync complete notification
-
-### 3. Real-Time Synchronization
-
-After initial sync, changes are pushed in real-time:
-
-```mermaid
-sequenceDiagram
-    participant FS as Filesystem
-    participant C1 as Client 1
-    participant S as Server
-    participant DB as Database
-    participant C2 as Client 2
-
-    FS->>C1: File changed (fs.watch)
-    C1->>C1: Read file content
-    C1->>S: Upload file
-    S->>DB: Store new version
-    S->>S: Apply OT if needed
-    S-->>C1: Upload ACK
-    S->>C2: File update notification
-    C2->>S: Download file
-    S->>DB: Retrieve file
-    DB-->>S: File content
-    S-->>C2: File content
-    C2->>FS: Write to disk
-```
-
-**Flow**:
-
-1. Filesystem watcher detects local change
-2. Client reads file content
-3. Client uploads file via WebSocket
-4. Server stores in database
-5. Server applies operational transformation if concurrent edits
-6. Server acknowledges upload to sender
-7. Server broadcasts update to other clients
-8. Other clients download and apply changes
-
-## File Operations
-
-### Upload
-
-```
-┌─────────┐
-│ Client  │
-└───┬─-───┘
-    │ 1. Detect file change
-    │
-    ├─► 2. Read file content
-    │
-    ├─► 3. Create upload message
-    │    {
-    │      type: "upload_file",
-    │      path: "notes/daily.md",
-    │      content: "...",
-    │      version: 42,
-    │      timestamp: "2024-01-01T12:00:00Z"
-    │    }
-    │
-    ▼
-┌─────────┐
-│ Server  │
-└───┬────-┘
-    │ 4. Validate message
-    │
-    ├─► 5. Check permissions
-    │
-    ├─► 6. Apply OT (if conflicts)
-    │
-    ├─► 7. Store in database
-    │
-    ├─► 8. Update version
-    │
-    ├─► 9. Broadcast to clients
-    │
-    └─► 10. Send ACK to uploader
-```
-
-### Download
-
-```
-┌─────────┐
-│ Server  │
-└───┬─-───┘
-    │ 1. File updated by another client
-    │
-    ├─► 2. Broadcast notification
-    │    {
-    │      type: "file_updated",
-    │      path: "notes/daily.md",
-    │      version: 43
-    │    }
-    │
-    ▼
-┌─────────┐
-│ Client  │
-└───┬─-───┘
-    │ 3. Receive notification
-    │
-    ├─► 4. Request file download
-    │    {
-    │      type: "download_file",
-    │      path: "notes/daily.md",
-    │      version: 43
-    │    }
-    │
-    ▼
-┌─────────┐
-│ Server  │
-└───┬─=───┘
-    │ 5. Retrieve from database
-    │
-    └─► 6. Send file content
-          {
-            type: "file_content",
-            path: "notes/daily.md",
-            content: "...",
-            version: 43
-          }
-          │
-          ▼
-    ┌─────────┐
-    │ Client  │
-    └───-─┬───┘
-          │ 7. Write to filesystem
-          │
-          └─► 8. Update local metadata
-```
-
-### Delete
-
-```
-┌─────────┐
-│ Client  │
-└────┬────┘
-    │ 1. File deleted locally
-    │
-    ├─► 2. Send delete message
-    │    {
-    │      type: "delete_file",
-    │      path: "notes/old.md"
-    │    }
-    │
-    ▼
-┌─────────┐
-│ Server  │
-└────┬────┘
-    │ 3. Mark as deleted in DB
-    │    (soft delete for history)
-    │
-    ├─► 4. Broadcast deletion
-    │
-    └─► 5. ACK to sender
-          │
-          ▼
-    ┌─────────┐
-    │ Other   │
-    │ Clients │
-    └────┬────┘
-          │ 6. Delete local file
-          │
-          └─► 7. Update metadata
-```
-
-## Conflict Resolution Flow
-
-### Concurrent Edits Scenario
-
-```
-Time →
-
-Client A                    Server                      Client B
-  │                          │                            │
-  │ Edit file v10            │                            │
-  │ "Add line A"             │                            │ Edit file v10
-  │                          │                            │ "Add line B"
-  │                          │                            │
-  ├─── Upload @ t1 ─────────►│                            │
-  │                          │◄────── Upload @ t2 ────────┤
-  │                          │                            │
-  │                          │ 1. Receive both edits      │
-  │                          │    (based on v10)          │
-  │                          │                            │
-  │                          │ 2. Apply first edit        │
-  │                          │    → v11 (line A added)    │
-  │                          │                            │
-  │                          │ 3. Transform second edit   │
-  │                          │    against first           │
-  │                          │                            │
-  │                          │ 4. Apply transformed edit  │
-  │                          │    → v12 (both lines)      │
-  │                          │                            │
-  │◄──── v12 content ────────┤                            │
-  │                          ├───── v12 content ─────────►│
-  │                          │                            │
-  │ Apply v12                │                            │ Apply v12
-  │ (has both lines)         │                            │ (has both lines)
-  │                          │                            │
-```
-
-### Conflict Resolution Steps
-
-1. **Detection**: Server receives two edits based on the same version
-2. **Ordering**: Determine which edit to apply first (by timestamp or client ID)
-3. **First edit**: Apply directly to database
-4. **Transformation**: Transform second edit against first using OT
-5. **Second edit**: Apply transformed edit to database
-6. **Broadcast**: Send merged result to all clients
-7. **Application**: Clients apply merged version locally
-
-## Database Schema
-
-### Core Tables
-
-```sql
--- Document metadata
-CREATE TABLE documents (
-    id INTEGER PRIMARY KEY,
-    path TEXT NOT NULL,
-    version INTEGER NOT NULL,
-    content_hash TEXT,
-    size INTEGER,
-    created_at TIMESTAMP,
-    updated_at TIMESTAMP,
-    deleted BOOLEAN DEFAULT FALSE
-);
-
--- Version history
-CREATE TABLE versions (
-    id INTEGER PRIMARY KEY,
-    document_id INTEGER,
-    version INTEGER,
-    content BLOB,
-    created_at TIMESTAMP,
-    FOREIGN KEY (document_id) REFERENCES documents(id)
-);
-
--- Client sync cursors
-CREATE TABLE cursors (
-    client_id TEXT PRIMARY KEY,
-    last_version INTEGER,
-    last_updated TIMESTAMP
-);
-```
-
-### Queries
-
-**Get files since version**:
-
-```sql
-SELECT * FROM documents
-WHERE version > ? AND deleted = FALSE
-ORDER BY version ASC;
-```
-
-**Store new version**:
-
-```sql
-INSERT INTO versions (document_id, version, content, created_at)
-VALUES (?, ?, ?, ?);
-
-UPDATE documents
-SET version = ?, updated_at = ?
-WHERE id = ?;
-```
-
-**Update cursor**:
-
-```sql
-INSERT OR REPLACE INTO cursors (client_id, last_version, last_updated)
-VALUES (?, ?, ?);
-```
-
-## Message Protocol
-
-### Client → Server Messages
-
-**Upload File**:
-
-```json
-{
-    "type": "upload_file",
-    "path": "notes/example.md",
-    "content": "File content here...",
-    "base_version": 10,
-    "timestamp": "2024-01-01T12:00:00Z"
-}
-```
-
-**Download File**:
-
-```json
-{
-    "type": "download_file",
-    "path": "notes/example.md"
-}
-```
-
-**Delete File**:
-
-```json
-{
-    "type": "delete_file",
-    "path": "notes/old.md"
-}
-```
-
-**List Files**:
-
-```json
-{
-    "type": "list_files",
-    "since_version": 0
-}
-```
-
-### Server → Client Messages
-
-**File Updated**:
-
-```json
-{
-    "type": "file_updated",
-    "path": "notes/example.md",
-    "version": 11,
-    "size": 1024,
-    "hash": "abc123..."
-}
-```
-
-**File Content**:
-
-```json
-{
-    "type": "file_content",
-    "path": "notes/example.md",
-    "content": "Updated content...",
-    "version": 11
-}
-```
-
-**File Deleted**:
-
-```json
-{
-    "type": "file_deleted",
-    "path": "notes/old.md",
-    "version": 12
-}
-```
-
-**Sync Complete**:
-
-```json
-{
-    "type": "sync_complete",
-    "total_files": 150,
-    "current_version": 200
-}
-```
-
-**Error**:
-
-```json
-{
-    "type": "error",
-    "message": "File too large",
-    "code": "FILE_TOO_LARGE"
-}
-```
-
-## Error Handling
-
-### Client-Side Errors
-
-**Network failure**:
-
-1. Detect WebSocket disconnect
-2. Queue pending operations
-3. Retry connection with exponential backoff
-4. Replay queued operations on reconnect
-
-**File read error**:
-
-1. Log error
-2. Skip file
-3. Continue with other files
-4. Report to user
-
-**Write conflict**:
-
-1. Receive updated version from server
-2. Apply OT merge locally
-3. Overwrite local file
-4. Continue syncing
-
-### Server-Side Errors
-
-**Database error**:
-
-1. Log error
-2. Return error to client
-3. Client retries operation
-
-**Invalid operation**:
-
-1. Validate message format
-2. Return specific error code
-3. Client handles error appropriately
-
-**Authentication failure**:
-
-1. Reject connection
-2. Send auth error
-3. Client prompts for new credentials
-
-## Performance Optimizations
-
-### Batching
-
-- Small, rapid changes are batched together
-- Reduces message overhead
-- Applied as single atomic update
-
-### Compression
-
-- Large files compressed before transmission
-- Reduces bandwidth usage
-- Transparent to application layer
-
-### Incremental Sync
-
-- Only changed portions of files sent
-- Uses content-based diffing
-- Significantly reduces data transfer
-
-### Caching
-
-- Server caches recent file versions
-- Reduces database queries
-- Improves response time
-
-## Monitoring Data Flow
-
-### Server Logs
-
-```
-2024-01-01 12:00:00 INFO WebSocket connection from 192.168.1.100
-2024-01-01 12:00:01 INFO User 'alice' authenticated for vault 'personal'
-2024-01-01 12:00:05 INFO Upload: notes/daily.md (v10 -> v11)
-2024-01-01 12:00:06 INFO Broadcast to 3 clients
-2024-01-01 12:00:10 INFO Conflict resolved: notes/shared.md (v12)
-```
-
-### Client Logs
-
-```
-2024-01-01 12:00:00 INFO Connecting to ws://sync.example.com
-2024-01-01 12:00:01 INFO Connected, authenticating...
-2024-01-01 12:00:01 INFO Authentication successful
-2024-01-01 12:00:02 INFO Starting initial sync
-2024-01-01 12:00:10 INFO Sync complete: 150 files, 200 MB
-2024-01-01 12:00:15 INFO Uploaded: notes/daily.md
-2024-01-01 12:00:20 INFO Downloaded: notes/shared.md (merged)
-```
-
-## Next Steps
-
-- [Understand the sync algorithm →](/architecture/sync-algorithm)
-- [Configure the server →](/config/server)
-- [Deploy VaultLink →](/guide/getting-started)

docs/architecture/index.md

@@ -1,334 +0,0 @@
-# Architecture Overview
-
-Central sync server with multiple clients. High-level architecture and design decisions.
-
-## System Components
-
-```
-┌─────────────────────────────────────────────────────────────┐
-│                        Clients                              │
-├─────────────────────┬───────────────────┬───────────────────┤
-│  Obsidian Plugin    │  Obsidian Plugin  │   CLI Client      │
-│  (User A - Device1) │  (User A - Device2│   (Server/Backup) │
-└──────────┬──────────┴─────────┬─────────┴──────────┬────────┘
-          │                    │                    │
-          │    WebSocket       │   WebSocket        │   WebSocket
-          │                    │                    │
-          └────────────────────┼────────────────────┘
-                                │
-                    ┌───────────▼───────────┐
-                    │   Sync Server         │
-                    │   (Rust + Axum)       │
-                    │                       │
-                    │  ┌─────────────────┐  │
-                    │  │  WebSocket Hub  │  │
-                    │  └────────┬────────┘  │
-                    │           │           │
-                    │  ┌────────▼────────┐  │
-                    │  │ Sync Engine     │  │
-                    │  │ (OT Algorithm)  │  │
-                    │  └────────┬────────┘  │
-                    │           │           │
-                    │  ┌────────▼────────┐  │
-                    │  │ SQLite Database │  │
-                    │  │ (Per Vault)     │  │
-                    │  └─────────────────┘  │
-                    └───────────────────────┘
-```
-
-## Core Components
-
-### Sync Server
-
-Central authority for synchronisation. Rust + Axum framework.
-
-**Responsibilities**:
-
-- Accept WebSocket connections from clients
-- Authenticate users via token-based auth
-- Store document versions in SQLite
-- Coordinate real-time updates between clients
-- Apply operational transformation for conflict resolution
-- Manage vault access control
-
-**Technology**:
-
-- **Language**: Rust 1.92+
-- **Framework**: Axum (async web framework)
-- **Database**: SQLite with SQLx
-- **Protocol**: WebSockets for real-time communication
-- **Sync Algorithm**: reconcile-text (operational transformation)
-
-### Sync Client Library
-
-TypeScript library with core sync logic. Used by Obsidian plugin and CLI client.
-
-**Responsibilities**:
-
-- Manage WebSocket connection to server
-- Watch local filesystem for changes
-- Upload and download files
-- Apply remote changes locally
-- Handle conflict resolution
-- Maintain sync metadata
-
-**Technology**:
-
-- **Language**: TypeScript
-- **Build**: Webpack
-- **Protocol**: WebSocket client
-- **File System**: Node.js `fs` API / Obsidian API
-
-### Obsidian Plugin
-
-Integration layer between sync client and Obsidian.
-
-**Responsibilities**:
-
-- Provide UI for configuration
-- Bridge sync client with Obsidian's file system API
-- Handle Obsidian lifecycle events
-- Display sync status to users
-
-**Technology**:
-
-- **Platform**: Obsidian Plugin API
-- **Core**: sync-client library
-- **UI**: Obsidian settings UI
-
-### CLI Client
-
-Standalone executable for syncing vaults without Obsidian.
-
-**Responsibilities**:
-
-- Command-line interface
-- File system access via Node.js
-- Daemon mode for continuous sync
-- Health check endpoint for monitoring
-
-**Technology**:
-
-- **Language**: TypeScript
-- **Runtime**: Node.js
-- **CLI**: Commander.js
-- **Core**: sync-client library
-
-## Data Flow
-
-### Initial Connection
-
-1. Client connects via WebSocket to server
-2. Server authenticates using provided token
-3. Server verifies user has access to requested vault
-4. Connection established, sync begins
-
-### File Upload Flow
-
-```
-Client                          Server
-  │                               │
-  │  1. File changed locally      │
-  │                               │
-  │  2. Read file content         │
-  │                               │
-  │  3. WebSocket: Upload file    │
-  ├──────────────────────────────►│
-  │                               │ 4. Store in SQLite
-  │                               │
-  │                               │ 5. Broadcast to other clients
-  │                               ├───────────────────────►
-  │  6. Ack upload                │
-  │◄──────────────────────────────┤
-```
-
-### File Download Flow
-
-```
-Client A                  Server                  Client B
-  │                         │                         │
-  │                         │  1. File uploaded       │
-  │                         │◄────────────────────────┤
-  │                         │                         │
-  │                         │ 2. Store in DB          │
-  │                         │                         │
-  │  3. Push notification   │                         │
-  │◄────────────────────────┤                         │
-  │                         │                         │
-  │  4. Download file       │                         │
-  ├────────────────────────►│                         │
-  │                         │                         │
-  │  5. Write locally       │                         │
-  │                         │                         │
-```
-
-### Conflict Resolution
-
-When two clients edit the same file simultaneously:
-
-```
-Client A                  Server                  Client B
-  │                         │                         │
-  │ 1. Edit file            │                         │ 1. Edit same file
-  │                         │                         │
-  │ 2. Upload changes       │                         │ 2. Upload changes
-  ├────────────────────────►│◄────────────────────────┤
-  │                         │                         │
-  │                         │ 3. Apply OT algorithm   │
-  │                         │    - Merge both edits   │
-  │                         │    - Preserve all changes│
-  │                         │                         │
-  │ 4. Receive merged ver.  │ 5. Receive merged ver.  │
-  │◄────────────────────────┤────────────────────────►│
-  │                         │                         │
-  │ 6. Apply locally        │                         │ 6. Apply locally
-```
-
-## Storage Architecture
-
-### Server Storage
-
-Each vault has its own SQLite database:
-
-```
-databases/
-├── vault-1.db
-├── vault-2.db
-└── shared-team.db
-```
-
-**Database Schema** (simplified):
-
-- **documents**: File metadata (path, size, modified time)
-- **versions**: Document content with version history
-- **cursors**: Client sync state
-
-### Client Storage
-
-Clients maintain sync metadata:
-
-```
-.vaultlink/
-├── metadata.json       # Sync state
-└── cache/              # Optional local cache
-```
-
-The `.vaultlink` directory tracks which files have been synced and their versions to enable efficient synchronisation.
-
-## Communication Protocol
-
-### WebSocket Messages
-
-Client-server communication uses JSON messages over WebSocket.
-
-**Message Types**:
-
-- `upload_file`: Client → Server (file upload)
-- `download_file`: Client → Server (request file)
-- `file_updated`: Server → Client (file changed notification)
-- `file_deleted`: Server → Client (file deleted notification)
-- `sync_complete`: Server → Client (initial sync finished)
-
-### Authentication
-
-Token-based authentication on connection:
-
-```typescript
-// Client sends token on connect
-{
-  type: "auth",
-  token: "user-auth-token",
-  vault: "vault-name"
-}
-
-// Server responds
-{
-  type: "auth_success"
-}
-// or
-{
-  type: "auth_error",
-  message: "Invalid token"
-}
-```
-
-## Scalability Considerations
-
-### Current Architecture
-
-- **SQLite per vault**: Simple, performant, limited to single server
-- **WebSocket connections**: Stateful, requires sticky sessions for load balancing
-- **Operational transformation**: Centralized on server
-
-### Scaling Approaches
-
-**Vertical Scaling**:
-
-- Increase server resources (CPU, RAM, storage)
-- Optimize database queries and indexing
-- Tune connection limits
-
-**Horizontal Scaling** (future):
-
-- Separate vault servers (vault sharding)
-- Load balancer with sticky sessions
-- Shared storage layer for SQLite databases
-- Consider alternative databases (PostgreSQL) for multi-server setups
-
-### Performance Characteristics
-
-- **Small vaults** (< 1000 files): Excellent performance
-- **Medium vaults** (1000-10000 files): Good performance with tuning
-- **Large vaults** (> 10000 files): May require optimisation
-- **Concurrent users**: Tested with dozens of simultaneous clients per vault
-
-## Security Model
-
-### Authentication
-
-- Token-based authentication
-- Tokens configured in server `config.yml`
-- No password hashing (tokens are secrets)
-
-### Authorization
-
-- Per-user vault access control
-- Allow-list or deny-list patterns
-- Global access or vault-specific access
-
-### Network Security
-
-- WebSocket over TLS (WSS) for encrypted transport
-- No built-in SSL (use reverse proxy)
-- CORS configured for web clients
-
-### Data Security
-
-- No encryption at rest (use encrypted filesystems if needed)
-- No end-to-end encryption (server sees all content)
-- Self-hosted model: you control the data
-
-## Technology Choices
-
-**Rust**: Low latency, memory safe, excellent async with Tokio, compile-time SQL verification
-
-**SQLite**: No separate database server, fast for reads, single file per vault, backups are file copies
-
-**WebSocket**: Bidirectional push, no polling overhead, built-in browser/Node.js support
-
-**Operational Transformation**: Automatic conflict resolution, preserves all edits, real-time collaboration
-
-## Design Principles
-
-1. **Self-hosted first**: Users control their data and infrastructure
-2. **Simplicity**: Easy to deploy and operate
-3. **Real-time**: Changes appear immediately
-4. **Reliability**: Handle network failures gracefully
-5. **Performance**: Fast sync for typical vault sizes
-6. **Privacy**: No third-party services or telemetry
-
-## Next Steps
-
-- [Learn about the sync algorithm →](/architecture/sync-algorithm)
-- [Understand data flow in detail →](/architecture/data-flow)
-- [Deploy the server →](/guide/server-setup)