mothwing 0.0.3 → 0.0.4

package/README.md

@@ -1,5 +1,5 @@
 # mothwing
-A small static site builder for pages as light as a moth's wing. Plain HTML in, plain HTML out, with hashed asset links and modulepreloads for code split JS. The rest is there when you ask for it: SEO tags, a Content-Security-Policy that allows only your own inline scripts, sitemaps dated from git, robots.txt and minified HTML.
+**Build, don't re-learn.** A small static site builder for pages as light as a moth's wing. Plain HTML in, plain HTML out, with hashed asset links and modulepreloads for code split JS. The rest is there when you ask for it: SEO tags, a Content-Security-Policy that allows only your own inline scripts, sitemaps dated from git, robots.txt and minified HTML.
 
 It was born as the build script of [ffxiv.watch](https://ffxiv.watch), and got tweaked over time until I was a little too proud of it to keep it there. So it branched out into its own thing, for any site that wants the same.
 
@@ -95,7 +95,7 @@ Every tag is ordinary HTML to an editor or formatter. Templates formatted by Pre
 ## Pages
 `src/pages/about.html` is `/about/`, `src/pages/blog/index.html` is `/blog/`.
 
-`[projects.id].html` makes a page per item of `projects` in the data folder, with its `id` as the URL: `src/pages/projects/[projects.id].html` is `/projects/hexclock/` for `{ "id": "hexclock" }`. The item is `{{ page }}`, and the list links to the pages by the same field, `/projects/{{ project.id }}/`. An item with a `sources` list is dated by those files instead of the whole data file, so editing one item doesn't date every page.
+`[projects.id].html` makes a page per item of `projects` in the data folder, with its `id` as the URL: `src/pages/projects/[projects.id].html` is `/projects/hexclock/` for `{ "id": "hexclock" }`. An id that's fine in a URL is kept as it is, `create_ssh_key` included, anything else is made into one: `Hex Clock` is `hex-clock`. The item is `{{ page }}`, and the list links to the pages by the same field, `/projects/{{ project.id }}/`. An item with a `sources` list is dated by those files instead of the whole data file, so editing one item doesn't date every page.
 
 ### .data.js
 For values that take more than the data folder, a page can have a `.data.js` next to it, whose default export returns them. Its values come before the data folder's:
@@ -151,7 +151,7 @@ Off until you turn them on. `true` turns one on as it comes, an object changes h
 | --- | --- | --- |
 | `minify` | Minifies the HTML of production builds | html-minifier-terser options |
 | `csp` | A `<meta>` Content-Security-Policy: only your own files, and your inline scripts by their hash | Directives over the defaults, `{ "style-src": "'self' 'unsafe-inline'" }` |
-| `sitemap` | `sitemap.xml`, each page dated by the last commit to its files | `sections` that get a sitemap each, `dates: false` to leave git out |
+| `sitemap` | `sitemap.xml` with every page, each dated by the last commit to its files | `sections` that get a sitemap each, behind a sitemap index, for big sites. `dates: false` to leave git out |
 | `robots` | `robots.txt` that allows everything, pointing at the sitemap | `block` crawlers by name, `disallow` paths |
 | `cname` | `CNAME` with the domain, for GitHub Pages | |
 | `activeLinks` | `aria-current="page"` on links to the page they're on, `"true"` on links to its section | A function `(href, route)` returning `"page"`, `"section"` or `false`, `undefined` keeps the default |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mothwing",
-  "version": "0.0.3",
+  "version": "0.0.4",
   "description": "A small static site builder for pages as light as a moth's wing",
   "keywords": [
     "static-site-generator",

package/src/config.js

@@ -55,7 +55,8 @@ export const CONFIG_FILE = "mothwing.config.js"
 
 /**
  * @typedef {object} SitemapConfig
- * @property {string[]} [sections] Sections that get a sitemap of their own, the rest share sitemap-pages.xml
+ * @property {string[]} [sections] Sections that get a sitemap of their own behind a sitemap index, the rest share
+ *   sitemap-pages.xml. Without any, sitemap.xml has every page itself
  * @property {boolean} [dates] Date each page by the last git commit to its files. true
  */
 

package/src/routes.js

@@ -116,7 +116,9 @@ function pagesFromData(site, siteData, source, list, key) {
   const seen = new Set()
   return items.map((item, i) => {
     if (item?.[key] === undefined) throw new Error(`${from}: item ${i} of ${list} has no ${key}`)
-    const pageSlug = slug(String(item[key]))
+    // A value that's fine in a URL already is kept as it is, so existing links like /docs/create_ssh_key/ keep working
+    const value = String(item[key])
+    const pageSlug = /^[\w.-]+$/.test(value) ? value : slug(value)
     if (seen.has(pageSlug)) throw new Error(`${from}: two of ${list} are both "${pageSlug}"`)
     seen.add(pageSlug)
     const output = source.replace(/\[[\w.]+\]\.html$/, `${pageSlug}/index.html`)

package/src/seo.js

@@ -35,8 +35,8 @@ export function seoTags(site, route, attrs) {
   const pageDescription = attr("description")
   if (!pageTitle || !pageDescription) throw new Error(`${route.source}: <x-seo> needs a title and a description`)
   const description = fitDescription(pageDescription, maxDescription)
-  // The front page keeps it, for people searching for the site by its name
-  const named = `${pageTitle} - ${siteName}`
+  // The front page keeps it, for people searching for the site by its name, unless it's the site's name already
+  const named = unescapeHtml(pageTitle) === siteName ? pageTitle : `${pageTitle} - ${siteName}`
   const title = route.url === "/" || unescapeHtml(named).length <= maxTitle ? named : pageTitle
 
   const url = `${site.url}${route.url}`

package/src/site.js

@@ -34,7 +34,8 @@ function lastModified(site, cache, files) {
  * @param {boolean} [dated] Add each page's lastmod, from git, which takes a moment
  */
 export async function writeSiteFiles(site, routes, dated = true) {
-  const files = [site.cname && "CNAME", site.sitemap && "the sitemaps", site.robots && "robots.txt"].filter(Boolean)
+  const sitemapFiles = site.sitemap && (site.sitemap.sections.length ? "the sitemaps" : "the sitemap")
+  const files = [site.cname && "CNAME", sitemapFiles, site.robots && "robots.txt"].filter(Boolean)
   if (!files.length) return
   log("SITE", `Writing ${files.length > 1 ? `${files.slice(0, -1).join(", ")} and ${files.at(-1)}` : files[0]}`)
 
@@ -45,7 +46,7 @@ export async function writeSiteFiles(site, routes, dated = true) {
   const sitemaps = new Map()
   if (site.sitemap) {
     const cache = new Map()
-    // A sitemap per section behind a sitemap index, so Search Console shows how each one is indexed on its own
+    // With sections, a sitemap per section behind a sitemap index, so Search Console shows how each one is indexed on its own
     for (const route of routes.filter(route => route.source !== site.notFound)) {
       const section = route.url.split("/")[1]
       const name = site.sitemap.sections.includes(section) ? section : "pages"
@@ -72,15 +73,23 @@ export async function writeSiteFiles(site, routes, dated = true) {
      */
     const entry = (tag, loc, lastmod) => `  <${tag}><loc>${loc}</loc>${lastmod ? `<lastmod>${lastmod}</lastmod>` : ""}</${tag}>`
 
-    const index = []
-    for (const [name, pages] of sitemaps) {
-      const file = `sitemap-${name}.xml`
-      fs.writeFileSync(path.join(site.out, file), xml("urlset", pages.map(p => entry("url", `${site.url}${p.url}`, p.lastmod))))
-      // YYYY-MM-DD sorts as text, so the newest is the last
-      const newest = pages.map(p => p.lastmod).filter(Boolean).sort().at(-1)
-      index.push(entry("sitemap", `${site.url}/${file}`, newest))
+    /** @param {{ url: string, lastmod: string | null | false }[]} pages */
+    const urlset = (pages) => xml("urlset", pages.map(p => entry("url", `${site.url}${p.url}`, p.lastmod)))
+
+    // Without sections, it's one sitemap.xml with every page, all a small site needs
+    if (!site.sitemap.sections.length) {
+      fs.writeFileSync(path.join(site.out, "sitemap.xml"), urlset(sitemaps.get("pages") ?? []))
+    } else {
+      const index = []
+      for (const [name, pages] of sitemaps) {
+        const file = `sitemap-${name}.xml`
+        fs.writeFileSync(path.join(site.out, file), urlset(pages))
+        // YYYY-MM-DD sorts as text, so the newest is the last
+        const newest = pages.map(p => p.lastmod).filter(Boolean).sort().at(-1)
+        index.push(entry("sitemap", `${site.url}/${file}`, newest))
+      }
+      fs.writeFileSync(path.join(site.out, "sitemap.xml"), xml("sitemapindex", index))
     }
-    fs.writeFileSync(path.join(site.out, "sitemap.xml"), xml("sitemapindex", index))
   }
 
   if (site.robots) {

package/types/config.d.ts

@@ -152,7 +152,8 @@ export type SeoConfig = {
 };
 export type SitemapConfig = {
     /**
-     * Sections that get a sitemap of their own, the rest share sitemap-pages.xml
+     * Sections that get a sitemap of their own behind a sitemap index, the rest share
+     * sitemap-pages.xml. Without any, sitemap.xml has every page itself
      */
     sections?: string[] | undefined;
     /**