etherlink-intern · etherlink-intern · May 31, 2026 · May 31, 2026
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -0,0 +1,25 @@
+name: CI
+
+on:
+  pull_request:
+  push:
+    branches:
+      - main
+  workflow_dispatch:
+
+jobs:
+  test:
+    name: Swift tests and smoke tests
+    runs-on: macos-15
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Show Swift version
+        run: swift --version
+
+      - name: Run unit tests
+        run: swift test
+
+      - name: Run CLI smoke tests
+        run: Scripts/smoke-test.sh
diff --git a/README.md b/README.md
@@ -43,6 +43,17 @@ print(document.markdown)
 swift run swift-markitdown path/to/file.html
 ```
 
+## Testing
+
+Run the unit test suite and CLI fixture smoke tests before opening a PR:
+
+```bash
+swift test
+Scripts/smoke-test.sh
+```
+
+GitHub Actions runs the same checks on pushes to `main`, pull requests, and manual workflow dispatches.
+
 ## Roadmap
 
 1. Expand the text/HTML/CSV/JSON converters with richer Markdown normalization and metadata extraction.

diff --git a/Scripts/smoke-test.sh b/Scripts/smoke-test.sh
@@ -0,0 +1,31 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+BINARY="${ROOT_DIR}/.build/debug/swift-markitdown"
+
+swift build --package-path "${ROOT_DIR}" --product swift-markitdown
+
+run_case() {
+  local name="$1"
+  local input="${ROOT_DIR}/Tests/Fixtures/${name}"
+  local expected="${ROOT_DIR}/Tests/Expected/${name%.*}.md"
+  local actual
+  actual="$(mktemp)"
+
+  "${BINARY}" "${input}" > "${actual}"
+
+  if ! diff -u "${expected}" "${actual}"; then
+    echo "Smoke test failed for ${name}" >&2
+    rm -f "${actual}"
+    return 1
+  fi
+
+  rm -f "${actual}"
+  echo "✓ ${name}"
+}
+
+run_case note.txt
+run_case page.html
+run_case table.csv
+run_case data.json
diff --git a/Sources/SwiftMarkItDown/Converters/HTMLConverter.swift b/Sources/SwiftMarkItDown/Converters/HTMLConverter.swift
@@ -10,6 +10,7 @@ public struct HTMLConverter: DocumentConverter {
         html = html.replacingOccurrences(of: "\r\n", with: "\n")
 
         let rules: [(String, String)] = [
+            ("(?is)<head\\b[^>]*>.*?</head>", ""),
             ("(?is)<script\\b[^>]*>.*?</script>", ""),
             ("(?is)<style\\b[^>]*>.*?</style>", ""),
             ("(?is)<h1\\b[^>]*>(.*?)</h1>", "\n# $1\n"),
@@ -36,6 +37,9 @@ public struct HTMLConverter: DocumentConverter {
             .smid_decodingHTMLEntities()
             .replacingOccurrences(of: "[ \t]+\n", with: "\n", options: .regularExpression)
             .replacingOccurrences(of: "\n{3,}", with: "\n\n", options: .regularExpression)
+            .split(separator: "\n", omittingEmptySubsequences: false)
+            .map { $0.trimmingCharacters(in: .whitespaces) }
+            .joined(separator: "\n")
             .smid_trimmedBlankLines
 
         return MarkdownDocument(markdown: markdown, sourceFormat: .html)

diff --git a/Tests/Expected/data.md b/Tests/Expected/data.md
@@ -0,0 +1,4 @@
+- **formats**:
+  - txt
+  - html
+- **title**: Roadmap
diff --git a/Tests/Expected/note.md b/Tests/Expected/note.md
@@ -0,0 +1,2 @@
+Hello SwiftMarkItDown
+This is plain text.
diff --git a/Tests/Expected/page.md b/Tests/Expected/page.md
@@ -0,0 +1,5 @@
+# Smoke Test
+
+Hello **native Swift** & Markdown.
+
+[Example](https://example.com)
diff --git a/Tests/Expected/table.md b/Tests/Expected/table.md
@@ -0,0 +1,4 @@
+| Name | Note |
+| --- | --- |
+| Swift | Native |
+| Mark, It Down | CSV \| escaped |
diff --git a/Tests/Fixtures/data.json b/Tests/Fixtures/data.json
@@ -0,0 +1 @@
+{"title":"Roadmap","formats":["txt","html"]}
diff --git a/Tests/Fixtures/note.txt b/Tests/Fixtures/note.txt
@@ -0,0 +1,2 @@
+Hello SwiftMarkItDown
+This is plain text.
diff --git a/Tests/Fixtures/page.html b/Tests/Fixtures/page.html
@@ -0,0 +1,13 @@
+<!doctype html>
+<html>
+  <head>
+    <title>Ignored title</title>
+    <style>body { color: red; }</style>
+    <script>console.log("ignore");</script>
+  </head>
+  <body>
+    <h1>Smoke Test</h1>
+    <p>Hello <strong>native Swift</strong> &amp; Markdown.</p>
+    <p><a href="https://example.com">Example</a></p>
+  </body>
+</html>
diff --git a/Tests/Fixtures/table.csv b/Tests/Fixtures/table.csv
@@ -0,0 +1,3 @@
+Name,Note
+Swift,Native
+"Mark, It Down","CSV | escaped"
diff --git a/Tests/SwiftMarkItDownTests/SwiftMarkItDownTests.swift b/Tests/SwiftMarkItDownTests/SwiftMarkItDownTests.swift
@@ -20,7 +20,7 @@ struct SwiftMarkItDownTests {
 
     @Test("converts simple HTML to Markdown")
     func convertsHTML() throws {
-        let html = "<h1>Title</h1><p>Hello <strong>Swift</strong> &amp; iOS.</p><a href=\"https://example.com\">Link</a>"
+        let html = "<html><head><title>Ignored</title></head><body><h1>Title</h1><p>Hello <strong>Swift</strong> &amp; iOS.</p><a href=\"https://example.com\">Link</a></body></html>"
         let request = ConversionRequest(data: Data(html.utf8), fileName: "index.html")
         let document = try MarkItDown().convert(request)
         #expect(document.markdown == "# Title\nHello **Swift** & iOS.\n\n[Link](https://example.com)")