@gobing-ai/knowledge-kit 0.1.8 → 0.1.10

package/plugins/publishings/surfdash-pub/dist/index.js

@@ -12345,6 +12345,10 @@ init_config();
 
 // ../../packages/kk-core/src/render-md.ts
 if (false) {}
+// ../../packages/kk-core/src/sentences.ts
+var CJK_TERMINATORS = new Set(["\u3002", "\uFF01", "\uFF1F"]);
+var LATIN_TERMINATORS = new Set([".", "!", "?"]);
+var ATTACHED_CLOSERS = new Set(["\u201D", "\u2019", "\u300D", "\u300F", "\uFF09", ")", '"', "'"]);
 // ../../plugins/publishings/surfdash-pub/src/index.ts
 init_dist();
 
@@ -12360,11 +12364,17 @@ function localeIndexPath(postPath, locale) {
     return;
   return `${postPath.slice(0, idx)}${sep}${locale}${sep}${postPath.slice(idx + marker.length)}`;
 }
+var SERIES_NAMES = {
+  zh: "\u6BCF\u65E5AI\u4E71\u7096",
+  en: "Daily AI Stew",
+  ja: "\u30C7\u30A4\u30EA\u30FCAI\u30B7\u30C1\u30E5\u30FC"
+};
 function crossLinkFor(locale, dateIso, showNotesUrl) {
+  const series = SERIES_NAMES[locale];
   const links = {
-    zh: `\uD83C\uDFA7 \u672C\u671F\u5DF2\u540C\u6B65\u5236\u4F5C\u4E3A\u64AD\u5BA2\u8282\u76EE\uFF0C[\u70B9\u51FB\u6536\u542C\u300A\u6BCF\u65E5AI\u4E71\u7096 \xB7 ${dateIso}\u300B](${showNotesUrl})\u3002`,
-    en: `\uD83C\uDFA7 This episode is also available as a podcast: [listen to \u6BCF\u65E5AI\u4E71\u7096 \xB7 ${dateIso}](${showNotesUrl}).`,
-    ja: `\uD83C\uDFA7 \u672C\u30A8\u30D4\u30BD\u30FC\u30C9\u306F\u30DD\u30C3\u30C9\u30AD\u30E3\u30B9\u30C8\u3067\u3082\u914D\u4FE1\u4E2D\uFF1A[\u6BCF\u65E5AI\u4E71\u7096 \xB7 ${dateIso} \u3092\u8074\u304F](${showNotesUrl})\u3002`
+    zh: `\uD83C\uDFA7 \u672C\u671F\u5DF2\u540C\u6B65\u5236\u4F5C\u4E3A\u64AD\u5BA2\u8282\u76EE\uFF0C[\u70B9\u51FB\u6536\u542C\u300A${series} \xB7 ${dateIso}\u300B](${showNotesUrl})\u3002`,
+    en: `\uD83C\uDFA7 This episode is also available as a podcast: [listen to ${series} \xB7 ${dateIso}](${showNotesUrl}).`,
+    ja: `\uD83C\uDFA7 \u672C\u30A8\u30D4\u30BD\u30FC\u30C9\u306F\u30DD\u30C3\u30C9\u30AD\u30E3\u30B9\u30C8\u3067\u3082\u914D\u4FE1\u4E2D\uFF1A[${series} \xB7 ${dateIso} \u3092\u8074\u304F](${showNotesUrl})\u3002`
   };
   return links[locale];
 }
@@ -12402,10 +12412,16 @@ async function patchPostFiles(request) {
     const end = frontmatterEnd(md);
     const fm = end < 0 ? md : md.slice(0, end);
     if (request.coverUrl !== "") {
+      if (!/^https:\/\//.test(request.coverUrl)) {
+        throw new Error(`coverUrl must be an absolute https:// URL, got: ${request.coverUrl}`);
+      }
       md = upsertFrontmatterKey(md, "image", request.coverUrl);
       md = upsertFrontmatterKey(md, "og_image", request.coverUrl);
       images[locale] = request.coverUrl;
-    } else if (!/^image:/m.test(fm) && await fs.exists(request.localCover)) {
+    } else if (!/^image:/m.test(fm)) {
+      if (!await fs.exists(request.localCover)) {
+        throw new Error(`${locale} post has no cover: coverUrl is empty, the frontmatter has no image key, and localCover does not exist (${request.localCover})`);
+      }
       const asset = join(dir, "assets", "cover.png");
       await fs.ensureDir(dirname2(asset));
       await fs.copy(request.localCover, asset);
@@ -12414,11 +12430,20 @@ async function patchPostFiles(request) {
       md = upsertFrontmatterKey(md, "og_image", "./assets/cover.png");
       images[locale] = "./assets/cover.png";
     }
-    if (request.showNotesUrl !== "" && !md.includes(request.showNotesUrl)) {
-      md = `${md.replace(/\s*$/, "")}
+    if (request.showNotesUrl !== "") {
+      const link = crossLinkFor(locale, request.dateIso, request.showNotesUrl);
+      const lines = md.split(`
+`);
+      const carriesUrl = lines.some((line) => line.includes(request.showNotesUrl));
+      if (!carriesUrl) {
+        md = `${md.replace(/\s*$/, "")}
 
-${crossLinkFor(locale, request.dateIso, request.showNotesUrl)}
+${link}
 `;
+      } else {
+        md = lines.map((line) => line.includes(request.showNotesUrl) && line.trimStart().startsWith("\uD83C\uDFA7") ? link : line).join(`
+`);
+      }
     }
     if (md !== before) {
       await fs.writeFile(indexMd, md);
@@ -12433,6 +12458,172 @@ ${crossLinkFor(locale, request.dateIso, request.showNotesUrl)}
   return { written, changed, absent, images, slug: basename(dirname2(request.postPath)) };
 }
 
+// ../../plugins/publishings/surfdash-pub/src/digest-check.ts
+var DIGEST_TAG = "sub-category:ai-daily";
+var PODCAST_INTROS = ["\u8FD9\u91CC\u662F\u7531Robin", "\u54B8\u8BDD\u54B8\u8BF4", "This is Robin", "Robin \u304C\u304A\u5C4A\u3051"];
+var SEO_TITLE_LIMIT = 60;
+var KEY_LINE = /^([A-Za-z_][A-Za-z0-9_-]*):(.*)$/;
+var escapeRe = (token) => token.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+function unquote(value) {
+  if (value.length >= 2 && value.startsWith('"') && value.endsWith('"')) {
+    try {
+      return JSON.parse(value);
+    } catch {
+      return value.slice(1, -1);
+    }
+  }
+  if (value.length >= 2 && value.startsWith("'") && value.endsWith("'")) {
+    return value.slice(1, -1).replaceAll("''", "'");
+  }
+  return value;
+}
+function parseFrontmatterScalars(markdown) {
+  const out = new Map;
+  const lines = markdown.split(/\r?\n/);
+  if (lines[0]?.trim() !== "---")
+    return out;
+  const end = lines.findIndex((line, i) => i > 0 && line.trim() === "---");
+  if (end < 0)
+    return out;
+  for (let i = 1;i < end; i++) {
+    const line = lines[i];
+    if (!line)
+      continue;
+    const match = line.match(KEY_LINE);
+    if (!match?.[1] || match[2] === undefined)
+      continue;
+    const key = match[1];
+    const value = match[2].trim();
+    if (value === ">-" || value === ">" || value === ">+" || value === "|-" || value === "|") {
+      const parts = [];
+      let j = i + 1;
+      while (j < end) {
+        const continuation = lines[j];
+        if (!continuation || !/^[ \t]+\S/.test(continuation))
+          break;
+        parts.push(continuation.trim());
+        j++;
+      }
+      out.set(key, parts.join(" "));
+      i = j - 1;
+      continue;
+    }
+    if (value === "") {
+      const items = [];
+      let j = i + 1;
+      while (j < end) {
+        const itemLine = lines[j];
+        if (!itemLine || !/^\s*-\s+/.test(itemLine))
+          break;
+        items.push(unquote(itemLine.replace(/^\s*-\s+/, "").trim()));
+        j++;
+      }
+      if (items.length > 0) {
+        out.set(key, items);
+        i = j - 1;
+      } else {
+        out.set(key, "");
+      }
+      continue;
+    }
+    if (value.startsWith("[") && value.endsWith("]")) {
+      out.set(key, value.slice(1, -1).split(",").map((item) => unquote(item.trim())).filter((item) => item !== ""));
+      continue;
+    }
+    out.set(key, unquote(value));
+  }
+  return out;
+}
+function scalar(frontmatter, key) {
+  const value = frontmatter.get(key);
+  if (value === undefined)
+    return;
+  return Array.isArray(value) ? value.join(" ") : value;
+}
+function endsAsSentence(text, locale) {
+  return locale === "en" ? /[.!?]["')\u201D\u2019]*$/u.test(text.trimEnd()) : /[\u3002\uFF01\uFF1F][\u300D\u300F\u201D\uFF09]*$/u.test(text.trimEnd());
+}
+function checkDigestFrontmatter(markdown, locale) {
+  const frontmatter = parseFrontmatterScalars(markdown);
+  const tags = frontmatter.get("tags");
+  const tagList = Array.isArray(tags) ? tags : typeof tags === "string" && tags !== "" ? [tags] : [];
+  if (!tagList.some((tag) => tag.includes(DIGEST_TAG)))
+    return [];
+  const problems = [];
+  const series = SERIES_NAMES[locale];
+  const titleSep = locale === "en" ? ": " : "\uFF1A";
+  const listSep = locale === "en" ? ", " : "\u3001";
+  const title = scalar(frontmatter, "title");
+  if (title === undefined || title === "") {
+    problems.push("title: missing");
+  } else {
+    const match = title.match(new RegExp(`^${escapeRe(series)} \xB7 (\\d{4}-\\d{2}-\\d{2})${escapeRe(titleSep)}(.+)$`));
+    if (!match) {
+      problems.push(`title: expected "<series> \xB7 YYYY-MM-DD${titleSep}<h1>${listSep}<h2>" for ${series}, got "${title}"`);
+    } else {
+      const rest = match[2] ?? "";
+      const heads = locale === "en" ? rest.split(", ") : rest.split("\u3001");
+      const seps = locale === "en" ? (rest.match(/, /g) ?? []).length : (rest.match(/\u3001/g) ?? []).length;
+      if (seps !== 1 || heads.some((head) => head.trim() === "")) {
+        problems.push(`title: expected exactly two non-empty headlines joined "${listSep}", got "${rest}"`);
+      }
+    }
+    const seoTitle = `${title} | Surfing`;
+    const seoLength = Array.from(seoTitle).length;
+    if (seoLength > SEO_TITLE_LIMIT) {
+      problems.push(`title: "${seoTitle}" is ${seoLength} code points (max ${SEO_TITLE_LIMIT})`);
+    }
+    if (title.includes("\u2026") || title.includes("..."))
+      problems.push('title: contains "\u2026" or "..."');
+  }
+  const description = scalar(frontmatter, "description");
+  const [min, max] = locale === "en" ? [120, 155] : [70, 100];
+  if (description === undefined || description === "") {
+    problems.push("description: missing");
+  } else {
+    const length = Array.from(description).length;
+    if (length < min || length > max) {
+      problems.push(`description: ${length} code points (expected ${min}\u2013${max} for ${locale})`);
+    }
+    if (!endsAsSentence(description, locale)) {
+      problems.push("description: does not end with sentence punctuation");
+    }
+    const lower = description.toLowerCase();
+    const intro = PODCAST_INTROS.find((candidate) => lower.startsWith(candidate.toLowerCase()));
+    if (intro)
+      problems.push(`description: starts with the podcast intro ("${intro}")`);
+    if (description.includes("\u2026") || description.includes("...")) {
+      problems.push('description: contains "\u2026" or "..."');
+    }
+  }
+  const points = frontmatter.get("key_points");
+  if (!Array.isArray(points) || points.length === 0) {
+    problems.push("key_points: missing");
+  } else {
+    const labelMax = locale === "en" ? 28 : 14;
+    points.forEach((point, index) => {
+      const sep = point.indexOf(" \u2014 ");
+      if (sep < 0) {
+        problems.push(`key_points[${index}]: missing " \u2014 " separator`);
+        return;
+      }
+      const label = point.slice(0, sep);
+      const text = point.slice(sep + 3);
+      const labelLength = Array.from(label).length;
+      if (labelLength === 0 || labelLength > labelMax) {
+        problems.push(`key_points[${index}]: label "${label}" is ${labelLength} code points (max ${labelMax})`);
+      }
+      if (!endsAsSentence(text, locale)) {
+        problems.push(`key_points[${index}]: text does not end with sentence punctuation`);
+      }
+      if (point.includes("\u2026") || point.includes("...")) {
+        problems.push(`key_points[${index}]: contains "\u2026" or "..."`);
+      }
+    });
+  }
+  return problems;
+}
+
 // ../../plugins/publishings/surfdash-pub/src/index.ts
 var TARGET = "surfdash";
 var OPERATIONS = ["publish", "scaffold-locale", "sync", "patch"];
@@ -12449,39 +12640,31 @@ function normalizeTag(raw) {
   }
   return slugify2(raw);
 }
-function mapContentToSurfingMarkdown(content, options = {}) {
-  let title = content.title;
-  if (!title) {
-    const match = content.body.match(/^#\s+(.+)$/m);
-    title = match?.[1] ? match[1].trim() : "Untitled Post";
-  }
-  let description;
-  if (typeof content.metadata?.description === "string") {
-    description = content.metadata.description;
-  } else {
-    const paragraphs = content.body.split(/\n\s*\n/).map((p) => p.replace(/^#+\s+.*$/, "").trim()).filter((p) => p.length > 0);
-    if (paragraphs.length > 0) {
-      description = paragraphs[0];
-    }
-  }
+function contentTags(content) {
   const rawTags = Array.isArray(content.metadata?.tags) ? [...content.metadata.tags] : Array.isArray(content.metadata?.keywords) ? [...content.metadata.keywords] : [];
   if (typeof content.metadata?.subCategory === "string" && content.metadata.subCategory.trim().length > 0) {
     const subRaw = content.metadata.subCategory.trim();
     rawTags.push(/^sub-category\s*:/i.test(subRaw) ? subRaw : `sub-category:${subRaw}`);
   }
-  const tags = [
+  return [
     ...new Set(rawTags.filter((t) => typeof t === "string" && t.trim().length > 0).map((t) => normalizeTag(t)))
   ].filter((tag) => tag.length > 0);
+}
+function mapContentToSurfingMarkdown(content, options = {}) {
+  let title = content.title;
+  if (!title) {
+    const match = content.body.match(/^#\s+(.+)$/m);
+    title = match?.[1] ? match[1].trim() : "Untitled Post";
+  }
+  const description = typeof content.metadata?.description === "string" && content.metadata.description.length > 0 ? content.metadata.description : "";
+  const tags = contentTags(content);
   const isoDate = options.publishDate ?? new Date().toISOString().slice(0, 10);
-  const firstRefUrl = content.references?.[0]?.url;
-  const canonical = typeof firstRefUrl === "string" ? firstRefUrl : typeof content.metadata?.canonicalUrl === "string" ? content.metadata.canonicalUrl : undefined;
   const draft = typeof content.metadata?.published === "boolean" ? !content.metadata.published : false;
   const frontmatterObj = {
     title,
-    description: description ?? "",
+    description,
     tags,
     publishDate: isoDate,
-    ...canonical ? { canonical } : {},
     draft,
     ...options.image ? { image: options.image, og_image: options.image } : {}
   };
@@ -12492,6 +12675,9 @@ function mapContentToSurfingMarkdown(content, options = {}) {
 ${content.body.trim()}
 `;
 }
+function isDigestContent(content) {
+  return contentTags(content).some((tag) => tag.includes(DIGEST_TAG));
+}
 function fail(error) {
   return ResultSchema.parse({ ok: false, target: TARGET, error });
 }
@@ -12552,7 +12738,8 @@ async function spawnCli(cmd, cwd) {
   const proc = Bun.spawn(cmd, {
     cwd,
     stdout: "pipe",
-    stderr: "pipe"
+    stderr: "pipe",
+    env: process.env
   });
   const [exitCode, stdout, stderr] = await Promise.all([
     proc.exited,
@@ -12568,6 +12755,46 @@ function cliErrorText(spawned) {
 function withEmdashTokenHint(errText) {
   return /INVALID_TOKEN|API error 401/.test(errText) ? `${errText} \u2014 hint: regenerate the EmDash token (surfing.salty.vip admin console) and update EMDASH_API_TOKEN in ~/.config/kk/config.yaml` : errText;
 }
+function readRefinedSnapshot(filePath) {
+  if (!existsSync2(filePath))
+    return;
+  const raw = readFileSync2(filePath, "utf-8");
+  const block = raw.match(/^---\r?\n([\s\S]*?)\r?\n---/);
+  const fm = block?.[1] ?? "";
+  if (!fm || !/^refined:\s*true$/m.test(fm))
+    return;
+  const read = (key) => {
+    const m = fm.match(new RegExp(`^${key}:\\s*(?:>-?
+((?:[ \\t]+.*
+?)+)|(.*))$`, "m"));
+    if (!m)
+      return;
+    const value = (m[1] ?? m[2] ?? "").replace(/\n[ \t]+/g, " ").trim();
+    return value.replace(/^["']|["']$/g, "") || undefined;
+  };
+  return {
+    title: read("title"),
+    seoTitle: read("seo_title"),
+    description: read("description"),
+    canonical: read("canonical")
+  };
+}
+function requireRefined(filePath) {
+  if (readRefinedSnapshot(filePath))
+    return;
+  return `surfdash refine step did not run for ${filePath} (no \`refined: true\`): the publishing pipeline is broken \u2014 the post would ship with generator-supplied SEO metadata`;
+}
+function digestCheckGate(filePath, locale) {
+  const problems = checkDigestFrontmatter(readFileSync2(filePath, "utf-8"), locale);
+  if (problems.length === 0)
+    return;
+  if (process.env.SURFDASH_DIGEST_CHECK === "warn") {
+    for (const problem of problems)
+      echoError(`digest check (warn): ${problem}`);
+    return;
+  }
+  return `digest frontmatter check failed: ${problems.join("; ")}`;
+}
 async function pushPostGit(surfdashRoot, relPostDir, message) {
   return commitPostPaths(surfdashRoot, [relPostDir], message);
 }
@@ -12584,12 +12811,40 @@ async function assertMainBranch(surfdashRoot) {
   }
   return { ok: true };
 }
+async function fastForwardToOrigin(surfdashRoot) {
+  const fetched = await spawnCli(["git", "fetch", "--quiet", "origin", "main"], surfdashRoot);
+  if (fetched.exitCode !== 0)
+    return { ok: true };
+  const counts = await spawnCli(["git", "rev-list", "--left-right", "--count", "origin/main...HEAD"], surfdashRoot);
+  if (counts.exitCode !== 0)
+    return { ok: true };
+  const [behindRaw = "0", aheadRaw = "0"] = counts.stdout.trim().split(/\s+/);
+  const behind = Number(behindRaw);
+  const ahead = Number(aheadRaw);
+  if (!Number.isFinite(behind) || !Number.isFinite(ahead) || behind === 0)
+    return { ok: true };
+  if (ahead > 0) {
+    return {
+      ok: false,
+      error: `surfdash checkout diverged from origin/main (${ahead} local, ${behind} remote commit(s)): reconcile it by hand before publishing`
+    };
+  }
+  const merged = await spawnCli(["git", "merge", "--ff-only", "--quiet", "origin/main"], surfdashRoot);
+  if (merged.exitCode !== 0) {
+    return { ok: false, error: `surfdash fast-forward to origin/main failed: ${cliErrorText(merged)}` };
+  }
+  echoError(`surfdash: fast-forwarded ${behind} commit(s) to origin/main before writing`);
+  return { ok: true };
+}
 async function commitPostPaths(surfdashRoot, relPaths, message, push = true) {
   if (relPaths.length === 0)
     return { ok: true, hash: "unchanged" };
   const branch = await assertMainBranch(surfdashRoot);
   if (!branch.ok)
     return branch;
+  const base = await fastForwardToOrigin(surfdashRoot);
+  if (!base.ok)
+    return base;
   const add = await spawnCli(["git", "add", "--", ...relPaths], surfdashRoot);
   if (add.exitCode !== 0)
     return { ok: false, error: `surfdash git add failed: ${cliErrorText(add)}` };
@@ -12716,6 +12971,12 @@ async function syncPost(options) {
     if (spawned.exitCode !== 0) {
       return fail(`publish-orchestrator sync failed (exit ${spawned.exitCode}): ${withEmdashTokenHint(cliErrorText(spawned))}`);
     }
+    const localeInPath = postPath.postPath.match(/[\\/]content[\\/]posts[\\/]articles[\\/](zh|en|ja)[\\/]/);
+    if (localeInPath) {
+      const digestProblem = digestCheckGate(postPath.postPath, localeInPath[1]);
+      if (digestProblem)
+        return fail(digestProblem);
+    }
     return ResultSchema.parse({
       ok: true,
       target: TARGET,
@@ -12820,12 +13081,18 @@ async function publishPost(payload) {
     return fail(`post path escapes ${articlesRoot}`);
   }
   const coverPath = payload.content.metadata?.coverImage;
-  if (typeof coverPath === "string" && existsSync2(coverPath) && !COVER_EXTS.has(extname(coverPath).toLowerCase())) {
+  if (typeof coverPath === "string" && !existsSync2(coverPath)) {
+    return fail(`metadata.coverImage not found: ${coverPath}`);
+  }
+  if (typeof coverPath === "string" && !COVER_EXTS.has(extname(coverPath).toLowerCase())) {
     return fail(`metadata.coverImage must be one of ${[...COVER_EXTS].join(", ")}`);
   }
+  if (isDigestContent(payload.content) && typeof coverPath !== "string") {
+    return fail("digest post has no cover image (metadata.coverImage)");
+  }
   await createNodeFileSystem().ensureDir(postDir);
   let imageFrontmatter;
-  if (typeof coverPath === "string" && existsSync2(coverPath)) {
+  if (typeof coverPath === "string") {
     const assetsDir = join2(postDir, "assets");
     await createNodeFileSystem().ensureDir(assetsDir);
     const coverName = `cover${extname(coverPath).toLowerCase()}`;
@@ -12843,6 +13110,13 @@ async function publishPost(payload) {
     if (spawned.exitCode !== 0) {
       return fail(`postsurfing CLI failed (exit ${spawned.exitCode}): ${withEmdashTokenHint(cliErrorText(spawned))}`);
     }
+    const notRefined = requireRefined(filePath);
+    if (notRefined)
+      return fail(notRefined);
+    const digestProblem = digestCheckGate(filePath, locale);
+    if (digestProblem)
+      return fail(digestProblem);
+    const refined = readRefinedSnapshot(filePath);
     const baseUrl = (process.env.SURFDASH_URL ?? "https://surfing.salty.vip").replace(/\/$/, "");
     const articleUrl = `${baseUrl}${locale === "en" ? "" : `/${locale}`}/posts/${slug}`;
     if (process.env.SURFDASH_AUTO_PUSH === "0") {
@@ -12851,7 +13125,7 @@ async function publishPost(payload) {
         target: TARGET,
         url: articleUrl,
         message: "Published via postsurfing CLI (auto-push disabled)",
-        metadata: { postPath: filePath }
+        metadata: { postPath: filePath, refinedTitle: refined?.title, refinedCanonical: refined?.canonical }
       });
     }
     const relPostDir = join2("content", "posts", "articles", locale, slug);
@@ -12864,7 +13138,7 @@ async function publishPost(payload) {
       target: TARGET,
       url: articleUrl,
       message: live ? "Published, committed, pushed, and live on the site" : "Published and pushed; URL not live within the verify window \u2014 check the surfdash deploy workflow",
-      metadata: { postPath: filePath, gitCommit: pushed.hash, live }
+      metadata: { postPath: filePath, gitCommit: pushed.hash, live, refinedTitle: refined?.title }
     });
   } catch (err) {
     return fail(`Postsurfing transport error: ${err instanceof Error ? err.message : String(err)}`);
@@ -12941,9 +13215,14 @@ if (import.meta.main) {
   process.exit(await main2());
 }
 export {
+  DIGEST_TAG,
   PostsurfingFileCliTransport,
   awaitArticleLive,
+  checkDigestFrontmatter,
   commitPostPaths,
+  contentTags,
+  digestCheckGate,
+  fastForwardToOrigin,
   getPublishTransport,
   main2 as main,
   mapContentToSurfingMarkdown,
@@ -12951,6 +13230,8 @@ export {
   normalizeTag,
   processPublishIO,
   pushPostGit,
+  readRefinedSnapshot,
+  requireRefined,
   slugify2 as slugify,
   withEmdashTokenHint
 };