python-insider-blog/src/plugins/remark-python-refs.ts at main · python/python-insider-blog · GitHub

History

338 lines (297 loc) · 10.1 KB

Raw

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

/**

* remark-python-refs

* A remark plugin that detects Python ecosystem URLs in Markdown links

* and transforms them into styled inline badges.

* Detected patterns:

* - PEP links (by URL pattern OR link text matching "PEP NNN")

* - CPython docs (docs.python.org/...)

* - PyPI links (pypi.org/project/NAME/)

* - GitHub issues/PRs (github.com/OWNER/REPO/issues|pull/NNN)

* - GitHub repos (github.com/OWNER/REPO — exactly 2 path segments)

* - GitHub users/orgs (github.com/NAME — exactly 1 segment, not reserved)

* - CVE references (nvd.nist.gov/vuln/detail/CVE-YYYY-NNNNN)

* - Python releases (python.org/downloads/release/python-XXXX/)

import type { Root, Link, Paragraph, PhrasingContent } from "mdast";

import { visit } from "unist-util-visit";

import {

pythonIcon,

docsIcon,

githubIcon,

packageIcon,

issueIcon,

shieldIcon,

downloadIcon,

externalIcon,

} from "../components/references/_icons.js";

// ── URL matchers ────────────────────────────────────────

const PEP_OLD = /^https?:\/\/(?:www\.)?python\.org\/dev\/peps\/pep-(\d+)\/?/i;

const PEP_NEW = /^https?:\/\/peps\.python\.org\/pep-(\d+)\/?/i;

const DOCS = /^https?:\/\/docs\.python\.org\//i;

const PYPI = /^https?:\/\/pypi\.org\/project\/([^/]+)\/?/i;

const GH_ISSUE = /^https?:\/\/github\.com\/([\w.-]+)\/([\w.-]+)\/(issues|pull)\/(\d+)\/?/i;

const CVE = /^https?:\/\/nvd\.nist\.gov\/vuln\/detail\/(CVE-[\d-]+)\/?/i;

const PY_RELEASE = /^https?:\/\/(?:www\.)?python\.org\/downloads\/release\/(python-[\w.]+)\/?/i;

const GITHUB =

/^https?:\/\/github\.com\/([\w.-]+)(?:\/([\w.-]+))?\/?$/i;

/** GitHub top-level paths that are NOT user/org profiles */

const GH_RESERVED = new Set([

"about",

"collections",

"contact",

"customer-stories",

"enterprise",

"events",

"explore",

"features",

"issues",

"login",

"marketplace",

"new",

"nonprofit",

"notifications",

"open-source",

"organizations",

"orgs",

"pricing",

"pulls",

"readme",

"search",

"security",

"settings",

"site",

"sponsors",

"team",

"topics",

"trending",

]);

// ── Helpers ─────────────────────────────────────────────

interface Match {

type: RefType;

icon: string;

label: string; // author's original link text — preserved

url: string;

}

function extractText(children: PhrasingContent[]): string {

return children

.map((c) => {

if (c.type === "text" || c.type === "inlineCode") return c.value;

if ("children" in c) return extractText(c.children as PhrasingContent[]);

return "";

})

.join("");

}

/** Match by link text — catches "PEP 649", "PEP 649," etc. */

const PEP_TEXT = /^PEP\s+\d+$/i;

function classify(url: string, linkText?: string): Omit<Match, "label" | "url"> | null {

let m: RegExpMatchArray | null;

// URL-based PEP detection

if ((m = url.match(PEP_OLD)) || (m = url.match(PEP_NEW))) {

return { type: "pep", icon: pythonIcon };

}

// Text-based PEP detection — catches links like [PEP 799](https://docs.python.org/...)

if (linkText && PEP_TEXT.test(linkText.trim())) {

return { type: "pep", icon: pythonIcon };

}

// CPython docs — checked after PEP text so [PEP NNN](docs.python.org/...) stays a PEP badge

if (DOCS.test(url)) {

return { type: "docs", icon: docsIcon };

}

if ((m = url.match(PYPI))) {

return { type: "pypi", icon: packageIcon };

}

if ((m = url.match(GH_ISSUE))) {

return { type: "gh-issue", icon: issueIcon };

}

if ((m = url.match(CVE))) {

return { type: "cve", icon: shieldIcon };

}

if ((m = url.match(PY_RELEASE))) {

return { type: "py-release", icon: downloadIcon };

}

if ((m = url.match(GITHUB))) {

const [, owner, repo] = m;

if (repo) {

return { type: "gh-repo", icon: githubIcon };

}

if (!GH_RESERVED.has(owner.toLowerCase())) {

return { type: "gh-user", icon: githubIcon };

}

return null;

}

function buildBadgeHtml(match: Match): string {

const cls = `ref-badge ref-${match.type}`;

// Escape HTML in label

const safeLabel = match.label

.replace(/&/g, "&")

.replace(/</g, "<")

.replace(/>/g, ">")

.replace(/"/g, """);

return (

`<a href="${match.url}" class="${cls}" target="_blank" rel="noopener noreferrer">` +

`${match.icon}` +

`${safeLabel}` +

`${externalIcon}` +

`</a>`

);

}

// ── Markdoc tag handling ────────────────────────────────

// Keystatic serializes inline components as Markdoc tags:

// {% Pep number=649 /%} or {% Pep number=649 %}PEP 649{% /Pep %}

const MARKDOC_TAG =

/\{%\s*(Pep|Docs|PyPi|GhRepo|GhUser)\s+([\s\S]*?)\s*(?:\/%\}|%\}[\s\S]*?\{%\s*\/\1\s*%\})/g;

const ATTR_RE = /(\w+)=(?:"([^"]*?)"|(\S+))/g;

function parseAttrs(str: string): Record<string, string> {

const attrs: Record<string, string> = {};

let m: RegExpExecArray | null;

while ((m = ATTR_RE.exec(str)) !== null) {

attrs[m[1]] = m[2] ?? m[3];

}

ATTR_RE.lastIndex = 0;

return attrs;

}

function markdocTagToBadge(tag: string, attrs: Record<string, string>): string | null {

switch (tag) {

case "Pep": {

const num = attrs.number;

if (!num) return null;

const padded = String(num).padStart(4, "0");

return buildBadgeHtml({

type: "pep",

icon: pythonIcon,

label: `PEP ${num}`,

url: `https://peps.python.org/pep-${padded}/`,

});

}

case "Docs": {

const path = attrs.path;

if (!path) return null;

const label = attrs.label ?? path.replace(/^\//, "").replace(/\.html$/, "").split("/").pop()!;

return buildBadgeHtml({

type: "docs",

icon: docsIcon,

label,

url: `https://docs.python.org/${path.replace(/^\//, "")}`,

});

}

case "PyPi": {

const name = attrs.name;

if (!name) return null;

return buildBadgeHtml({

type: "pypi",

icon: packageIcon,

label: name,

url: `https://pypi.org/project/${name}/`,

});

}

case "GhRepo": {

const repo = attrs.repo;

if (!repo) return null;

return buildBadgeHtml({

type: "gh-repo",

icon: githubIcon,

label: repo.split("/").pop()!,

url: `https://github.com/${repo}`,

});

}

case "GhUser": {

const name = attrs.name;

if (!name) return null;

return buildBadgeHtml({

type: "gh-user",

icon: githubIcon,

label: `@${name}`,

url: `https://github.com/${name}`,

});

}

default:

return null;

}

/** Split a text value on Markdoc tags, returning text + html nodes. */

function splitMarkdocTags(value: string): Array<{ type: "text"; value: string } | { type: "html"; value: string }> {

const parts: Array<{ type: "text"; value: string } | { type: "html"; value: string }> = [];

let lastIndex = 0;

MARKDOC_TAG.lastIndex = 0;

let m: RegExpExecArray | null;

while ((m = MARKDOC_TAG.exec(value)) !== null) {

if (m.index > lastIndex) {

parts.push({ type: "text", value: value.slice(lastIndex, m.index) });

}

const html = markdocTagToBadge(m[1], parseAttrs(m[2]));

if (html) {

parts.push({ type: "html", value: html });

} else {

parts.push({ type: "text", value: m[0] });

}

lastIndex = m.index + m[0].length;

}

if (lastIndex < value.length) {

parts.push({ type: "text", value: value.slice(lastIndex) });

}

return parts;

}

// ── Plugin ──────────────────────────────────────────────

export default function remarkPythonRefs() {

return (tree: Root, file: any) => {

// Collect unique references for the post footer

const seen = new Set<string>();

const refs: Array<{ type: RefType; label: string; url: string }> = [];

function collectRef(type: RefType, label: string, url: string) {

const key = `${type}:${url}`;

if (!seen.has(key)) {

seen.add(key);

refs.push({ type, label, url });

}

// Pass 1: Transform matching links → badges

visit(tree, "link", (node: Link, index, parent) => {

if (index == null || !parent) return;

const rawLabel = extractText(node.children);

const info = classify(node.url, rawLabel);

if (!info) return;

// For autolinked bare URLs, drop the protocol, leading "www.",

// and trailing slash so the badge reads "python.org/downloads/..."

// rather than the full "https://www.python.org/.../" form.

const label = rawLabel.startsWith("http")

? rawLabel.replace(/^https?:\/\/(?:www\.)?/i, "").replace(/\/$/, "")

: rawLabel;

collectRef(info.type, label, node.url);

const html = buildBadgeHtml({ ...info, label, url: node.url });

// Replace the link node with a raw HTML node

(parent.children as any)[index] = {

type: "html",

value: html,

};

});

// Pass 2: Transform Markdoc inline tags (from Keystatic) → badges

visit(tree, "paragraph", (node: Paragraph) => {

let changed = false;

const newChildren: PhrasingContent[] = [];

for (const child of node.children) {

if (child.type !== "text" || !MARKDOC_TAG.test(child.value)) {

MARKDOC_TAG.lastIndex = 0;

newChildren.push(child);

continue;

}

MARKDOC_TAG.lastIndex = 0;

changed = true;

const parts = splitMarkdocTags(child.value);

for (const part of parts) {

newChildren.push(part as any);

}

if (changed) {

node.children = newChildren;

}

});

// Expose collected references via remarkPluginFrontmatter

if (!file.data.astro) file.data.astro = {};

if (!file.data.astro.frontmatter) file.data.astro.frontmatter = {};

file.data.astro.frontmatter.references = refs;

};

}

python-insider-blog/src/plugins/remark-python-refs.ts at main · python/python-insider-blog · GitHub

Uh oh!

Uh oh!

python-insider-blog/src/plugins/remark-python-refs.ts at main · python/python-insider-blog · GitHub

Uh oh!

Uh oh!

Expand file tree

Breadcrumbs

remark-python-refs.ts

Latest commit

History

Breadcrumbs

remark-python-refs.ts

File metadata and controls