@deadair/plugin-sdk 0.54.0 → 0.56.0

package/dist/article.parse.d.ts

@@ -20,9 +20,12 @@ import type { HostFetchInit, PluginHost } from './plugin.host.js';
  * host can check a claim against the text it came from, and prose that has been
  * through a plugin's own paraphrase is prose nothing can check.
  *
- * The one thing removed that is not furniture is a paragraph announcing itself as an advertisement
- * — see {@link SPONSOR_OPENERS}. It is still a cut rather than a rewrite: what is left is the
- * publisher's own prose, and what went is prose the publisher was paid for.
+ * Three things removed are not furniture: a paragraph announcing itself as an advertisement (see
+ * {@link SPONSOR_OPENERS}), a paragraph in which the writer pitches their own book or newsletter
+ * (see {@link SELF_PROMOTION}), and a paragraph set entirely in italics (see
+ * {@link setEntirelyInEmphasis}), which is the publisher's note about the story rather than the
+ * story. All three are still cuts rather than rewrites: what is left is the publisher's own prose,
+ * and what went is prose the publisher was paid for or wrote about itself.
  *
  * ## Wrong in one direction on purpose
  *

package/dist/article.parse.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"article.parse.d.ts","sourceRoot":"","sources":["../src/article.parse.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAElE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAkCG;AAEH;;;;;;;;;GASG;AACH,eAAO,MAAM,iBAAiB,OAAQ,CAAC;AAwIvC;;;;;;;;GAQG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,GAAE,MAA0B,GAAG,MAAM,GAAG,SAAS,CAgCrG;AA2BD;;;;;;;;;;;;;;;;;;;;GAoBG;AACH,wBAAsB,YAAY,CAC9B,IAAI,EAAE,UAAU,EAChB,GAAG,EAAE,MAAM,EACX,IAAI,CAAC,EAAE,aAAa,EACpB,OAAO,CAAC,EAAE;IAAE,QAAQ,CAAC,EAAE,MAAM,CAAA;CAAE,GAChC,OAAO,CAAC,MAAM,GAAG,SAAS,CAAC,CA0B7B"}
+{"version":3,"file":"article.parse.d.ts","sourceRoot":"","sources":["../src/article.parse.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAElE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAqCG;AAEH;;;;;;;;;GASG;AACH,eAAO,MAAM,iBAAiB,OAAQ,CAAC;AA8LvC;;;;;;;;GAQG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,GAAE,MAA0B,GAAG,MAAM,GAAG,SAAS,CAgCrG;AA2BD;;;;;;;;;;;;;;;;;;;;GAoBG;AACH,wBAAsB,YAAY,CAC9B,IAAI,EAAE,UAAU,EAChB,GAAG,EAAE,MAAM,EACX,IAAI,CAAC,EAAE,aAAa,EACpB,OAAO,CAAC,EAAE;IAAE,QAAQ,CAAC,EAAE,MAAM,CAAA;CAAE,GAChC,OAAO,CAAC,MAAM,GAAG,SAAS,CAAC,CA0B7B"}

package/dist/index.js

@@ -578,13 +578,15 @@ var SPONSOR_OPENERS = [
   "paid partnership",
   "presented by",
   "promoted by",
+  "read more:",
   "sponsored",
   "sponsored by",
   "sponsored content",
   "sponsored survey",
   "support for this",
   "support comes from",
-  "this post is sponsored"
+  "this post is sponsored",
+  "your support helps"
 ];
 function opensWithASponsor(text2) {
   const said = text2.toLowerCase();
@@ -595,6 +597,13 @@ function opensWithASponsor(text2) {
   });
 }
 __name(opensWithASponsor, "opensWithASponsor");
+var SELF_PROMOTION = /\b(?:check out my (?:book|newsletter|podcast)|for more of my work|sign up for (?:my|our) newsletter|subscribe to (?:my|our))\b/i;
+function setEntirelyInEmphasis(markup) {
+  if (!/<(em|i)\b/i.test(markup)) return false;
+  const outside = markup.replace(/<(em|i)\b[^>]*>[\s\S]*?<\/\1>/gi, "").replace(/<[^>]*>/g, "").replace(/&nbsp;|&#160;/gi, "").replace(/[\s\p{P}]/gu, "");
+  return outside.length === 0;
+}
+__name(setEntirelyInEmphasis, "setEntirelyInEmphasis");
 var CONTAINERS = [
   /<article\b[^>]*>([\s\S]*?)<\/article>/i,
   /<main\b[^>]*>([\s\S]*?)<\/main>/i
@@ -607,7 +616,7 @@ function extractArticle(html, maxChars = ARTICLE_MAX_CHARS) {
   for (const match of body.matchAll(PARAGRAPH)) {
     const text2 = withoutApparatus(plainText(match[1] ?? ""));
     if (text2 === void 0 || text2.length < MIN_PARAGRAPH_CHARS) continue;
-    if (opensWithASponsor(text2)) continue;
+    if (opensWithASponsor(text2) || SELF_PROMOTION.test(text2) || setEntirelyInEmphasis(match[1] ?? "")) continue;
     if (!paragraphs.includes(text2)) paragraphs.push(text2);
   }
   if (paragraphs.length === 0) return void 0;