Skip to content

Open-source AI browser agent for Chrome and Firefox (monorepo) 🧠

fix(agent): carry negation across coordinated publish verbs and resolve CodeQL alert

Emre Sokullubb2de0b250b6
diff --git a/src/chrome/src/agent/agent.js b/src/chrome/src/agent/agent.js
index c2175d5cc..5a2a79719 100644
--- a/src/chrome/src/agent/agent.js
+++ b/src/chrome/src/agent/agent.js
@@ -245,7 +245,7 @@ const SOCIAL_NOUN_LIKE_PUBLISH = new RegExp(
'iu',
);
const SOCIAL_NEGATION = new RegExp(
- `(?<![${SOCIAL_WORD_EDGE}])(?:not|never|don't|dont|do\\s+not|cannot|can't|cant|shouldn't|shouldnt|should\\s+not|mustn't|mustnt|must\\s+not|won't|wont|will\\s+not|avoid|refrain|stop|prevent|prohibit|no(?=\\s+(?!attachments?|photos?|images?|pictures?|videos?|files?|media|delays?|doubt|worries|hashtags?|tags?|links?|urls?\\b))|ne|pas|ne\\s+pas|nunca|jamás|jamais|nicht|nie|kein|keine|non|mai|não|nao|hayır|asla|sakın|yapmayın|yapma|не|никогда|нет)(?![${SOCIAL_WORD_EDGE}])`
+ `(?<![${SOCIAL_WORD_EDGE}])(?:not|never|neither|nor|don't|dont|do\\s+not|cannot|can't|cant|shouldn't|shouldnt|should\\s+not|mustn't|mustnt|must\\s+not|won't|wont|will\\s+not|avoid|refrain|stop|prevent|prohibit|no(?=\\s+(?!attachments?|photos?|images?|pictures?|videos?|files?|media|delays?|doubt|worries|hashtags?|tags?|links?|urls?\\b))|ne|pas|ne\\s+pas|nunca|jamás|jamais|nicht|nie|kein|keine|non|mai|não|nao|hayır|asla|sakın|yapmayın|yapma|не|никогда|нет)(?![${SOCIAL_WORD_EDGE}])`
+ '|不要|别|不能|不可|不得|不用|请勿|勿|严禁|禁止|决不|绝不|決して'
+ '|하지\\s*마|하지\\s*마세요|금지',
'iu',
@@ -254,7 +254,38 @@ const SOCIAL_POST_NEGATION = new RegExp(
`^(?:\\s*(?:nothing|nowhere|none)|しないで|してはいけない|してはならない|はいけない|はならない|はだめ|はいけません|はなりません|すんな|するな|禁止|ないで|\\s*(?:하지\\s*마|하지\\s*마세요|하지\\s*않|금지))`,
'iu',
);
-const SOCIAL_CLAUSE_BREAK = new RegExp('[.!?;:,\\n]|(?<![\\p{L}\\p{N}_])(?:and|then|but|or|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![\\p{L}\\p{N}_])|然后|然後|接着|そして|それから|、|。', 'iu');
+const SOCIAL_CLAUSE_DELIMITER = new RegExp(
+ `([.!?;:\\n]|(?<![${SOCIAL_WORD_EDGE}])(?:and|then|but|or|nor|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![${SOCIAL_WORD_EDGE}])|然后|然後|接着|そして|それから|または|、|。|,)`,
+ 'iu',
+);
+const SOCIAL_COORDINATING_DELIMITER = new RegExp(
+ `^(?:or|nor|and|neither|o|ou|oder|ni|nem|né|veya|ya\\s+da|или|ни|或|或者|または|もしくは|や|또는|이나|거나|,|、)$`,
+ 'iu',
+);
+const SOCIAL_CLAUSE_BREAK = new RegExp(
+ `[.!?;:,\\n]|(?<![${SOCIAL_WORD_EDGE}])(?:and|then|but|or|nor|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![${SOCIAL_WORD_EDGE}])|然后|然後|接着|そして|それから|または|、|。`,
+ 'iu',
+);
+
+const GENERIC_ATTACHMENT_WORDS = new Set([
+ 'true', 'yes', 'attached', 'attachment', 'attachments', 'media',
+ 'image', 'images', 'photo', 'photos', 'picture', 'pictures', 'pic', 'pics',
+ 'video', 'videos', 'clip', 'clips', 'recording', 'recordings',
+ 'file', 'files', 'graphic', 'graphics',
+ 'a', 'an', 'the', 'of', 'in', 'with', 'some', 'any',
+ 'one', 'two', 'three', 'four', 'five', 'six', 'seven', 'eight', 'nine', 'ten',
+ 'single', 'multiple', 'both',
+ 'de', 'du', 'des', 'el', 'la', 'los', 'las', 'der', 'die', 'das', 'di', 'il', 'lo', 'gli', 'le',
+ 'o', 'os', 'as', 'do', 'da', 'dos', 'das', 'un', 'une', 'deux', 'trois', 'quatre', 'cinq',
+ 'uno', 'una', 'unos', 'unas', 'dos', 'tres', 'cuatro', 'cinco', 'due', 'tre', 'quattro', 'cinque',
+ 'ein', 'eine', 'einen', 'einer', 'zwei', 'drei', 'vier', 'fünf', 'um', 'uma', 'dois', 'duas', 'três',
+ 'imagen', 'imagenes', 'imágenes', 'foto', 'fotos', 'vidéo', 'vidéos', 'bild', 'bilder',
+ 'fichier', 'fichiers', 'archivo', 'archivos', 'datei', 'dateien', 'allegato', 'allegati', 'anexo', 'anexos',
+ 'один', 'одна', 'одно', 'два', 'две', 'три', 'четыре', 'пять',
+ 'вложения', 'вложение', 'фотографии', 'фотография', 'изображения', 'изображение', 'видео',
+]);
+
+const CJK_GENERIC_ATTACHMENT_REGEX = /^[0-9一二两三四五六七八九十添付画像写真動画メディア已上传附件图片照片视频媒体枚つの本张條条个個장での\s\-_,.:;!?/\\()]+$/u;
// "On <url>, publish this" names a destination just as plainly as
// "publish this on <url>", but only when the URL is presented as a place.
@@ -2298,14 +2329,10 @@ export class Agent extends LoopDetector {
}
_cleanSpecificAttachmentTarget(value) {
- let target = String(value || '').trim().toLowerCase();
- target = target
- .replace(/^https?:\/\/[^\s]+/i, '')
- .replace(/\b(?:two|three|four|five|six|seven|eight|nine|ten|\d+)\s+(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with name)\s+/i, '')
- .replace(/\b(?:an?|the)\s+(?:image|photo|picture|video|attachment|file)\s+(?:of|named|called|with name)\s+/i, '')
- .replace(/\b(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with name)\s+/i, '')
- .trim();
- return target;
+ const text = String(value || '').trim();
+ const match = text.match(/(?:(?:an?|the|\d+|one|two|three|four|five|six|seven|eight|nine|ten)\s+)?(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with\s+name)\s+(.+)$/i);
+ if (match) return match[1].trim().toLowerCase();
+ return text.toLowerCase();
}
_parseWorkflowAttachmentRequirement(value) {
@@ -2339,18 +2366,14 @@ export class Agent extends LoopDetector {
}
}
- const stripped = text
- .replace(/^(?:true|yes|attached)\b/i, '')
- .replace(/\b(?:\d+|one|two|three|four|five|six|seven|eight|nine|ten|single|multiple|both|un|une|deux|trois|quatre|cinq|uno|una|unos|unas|dos|tres|cuatro|cinco|due|tre|quattro|cinque|ein|eine|einen|einer|zwei|drei|vier|fünf|um|uma|dois|duas|três|quatro|один|одна|одно|два|две|три|четыре|пять)\b/gi, '')
- .replace(/\b(?:a|an|the|of|in|with|some|any|de|du|des|el|la|los|las|der|die|das|di|il|lo|gli|le|o|os|as|do|da|dos|das)\b/gi, '')
- .replace(/\b(?:image|images|photo|photos|picture|pictures|pic|pics|video|videos|clip|clips|recording|recordings|media|attachment|attachments|file|files|graphic|graphics)\b/gi, '')
- .replace(/\b(?:imagen(?:es)?|imágenes|foto|fotos|vidéo|vidéos|bild(?:er)?|fichier|fichiers|archivo|archivos|datei(?:en)?|allegat[oi]|anexo|anexos)\b/gi, '')
- .replace(/\b(?:вложени[яе]|фотографи[ия]|изображени[яе]|видео)\b/gi, '')
- .replace(/[一二两三四五六七八九十]/g, '')
- .replace(/(?:添付|画像|写真|動画|メディア|已上传|附件|图片|照片|视频|媒体|枚|つの|本|张|條|条|个|個|장|개|の)/gu, '')
- .replace(/[\s\-_,.:;!?/\\()]+/g, '');
-
- const isGeneric = stripped.length === 0;
+ let isGeneric = false;
+ if (CJK_GENERIC_ATTACHMENT_REGEX.test(text)) {
+ isGeneric = true;
+ } else {
+ const nonPunctuation = text.replace(/[\s\-_,.:;!?/\\()]+/g, ' ').trim();
+ const words = nonPunctuation ? nonPunctuation.split(/\s+/) : [];
+ isGeneric = words.length > 0 && words.every(w => /^\d+$/.test(w) || GENERIC_ATTACHMENT_WORDS.has(w) || CJK_GENERIC_ATTACHMENT_REGEX.test(w));
+ }
return { isGeneric, expectedCount, wantsImage, wantsVideo, normalized: text };
}
@@ -14829,19 +14852,63 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
return url;
}
+ _socialPublicationClauses(text) {
+ const parts = String(text || '').split(SOCIAL_CLAUSE_DELIMITER);
+ const clauses = [];
+ for (let i = 0; i < parts.length; i += 2) {
+ const clauseText = parts[i] || '';
+ const delim = i > 0 ? (parts[i - 1] || '').trim() : '';
+ clauses.push({ text: clauseText, delim });
+ }
+
+ let carriedNegation = false;
+ for (let i = 0; i < clauses.length; i++) {
+ const c = clauses[i];
+ const isCoordinated = i > 0 && SOCIAL_COORDINATING_DELIMITER.test(c.delim);
+ if (!isCoordinated) carriedNegation = false;
+
+ const hasExplicitNeg = SOCIAL_NEGATION.test(c.text);
+ if (hasExplicitNeg) {
+ c.isNegated = true;
+ carriedNegation = true;
+ } else if (carriedNegation) {
+ c.isNegated = true;
+ } else {
+ c.isNegated = false;
+ }
+ }
+
+ let carriedPostNegation = false;
+ for (let i = clauses.length - 1; i >= 0; i--) {
+ const c = clauses[i];
+ const hasExplicitPostNeg = SOCIAL_POST_NEGATION.test(c.text.trim());
+ if (hasExplicitPostNeg) {
+ c.isNegated = true;
+ carriedPostNegation = true;
+ } else if (carriedPostNegation && i < clauses.length - 1 && SOCIAL_COORDINATING_DELIMITER.test(clauses[i + 1].delim)) {
+ c.isNegated = true;
+ } else {
+ carriedPostNegation = false;
+ }
+ }
+
+ return clauses;
+ }
+
// Publication language only counts as a command when nothing in its own
// clause is asking to read or forbidding publication. "Read posts on <feed>"
// and "阅读推文 <feed>" name content; "do not post this on <feed>" forbids
// publication; "read the summary and publish it on <feed>" asks for a post in
// a clause of its own.
_socialPublicationCommandIn(text) {
- return String(text || '').split(SOCIAL_CLAUSE_BREAK).some((clause) => {
- const publish = clause ? clause.match(SOCIAL_PUBLISH_VERBS) : null;
+ return this._socialPublicationClauses(text).some((clause) => {
+ if (clause.isNegated) return false;
+ const publish = clause.text ? clause.text.match(SOCIAL_PUBLISH_VERBS) : null;
if (!publish) return false;
- const before = clause.slice(0, publish.index);
+ const before = clause.text.slice(0, publish.index);
if (SOCIAL_READ_VERBS.test(before)) return false;
if (SOCIAL_NEGATION.test(before)) return false;
- const after = clause.slice(publish.index + publish[0].length);
+ const after = clause.text.slice(publish.index + publish[0].length);
if (SOCIAL_POST_NEGATION.test(after)) return false;
if (SOCIAL_NOUN_LIKE_PUBLISH.test(publish[0]) && SOCIAL_READ_VERBS.test(after)) return false;
return true;
@@ -14895,8 +14962,9 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
match.index + match[0].length,
match.index + match[0].length + 60,
);
- const clauseBefore = before.split(SOCIAL_CLAUSE_BREAK).pop() || '';
- if (SOCIAL_NEGATION.test(clauseBefore)) continue;
+ const beforeClauses = this._socialPublicationClauses(before);
+ const lastBeforeClause = beforeClauses[beforeClauses.length - 1];
+ if (lastBeforeClause?.isNegated) continue;
const governed = /\/(?:compose|intent|i\/flow)\//.test(url)
|| this._socialPublicationCommandIn(before)
|| (SOCIAL_PUBLISH_DESTINATION_LEAD.test(before) && this._socialPublicationCommandIn(after));
@@ -14912,6 +14980,7 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
// publication target. Verbs and prepositions are the same multilingual
// sets the URL scan uses, so "Publícalo en X" counts while English-only
// matching would leave it unbound.
+ const clauses = this._socialPublicationClauses(trustedContext);
const publishesTo = (platform) => {
const platformPattern = new RegExp(
`(?<![${SOCIAL_WORD_EDGE}])(?:on|onto|to|via|in|at|en|sur|sobre|\u00e0|au|auf|su|em|na|para|\u0432|\u043d\u0430)(?![${SOCIAL_WORD_EDGE}])\\s+(?:the\\s+)?${platform}(?![${SOCIAL_WORD_EDGE}])`
@@ -14921,26 +14990,25 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
'iu',
);
const verbPattern = new RegExp(SOCIAL_PUBLISH_VERBS.source, 'giu');
- const clauses = trustedContext.split(SOCIAL_CLAUSE_BREAK);
return clauses.some((clause, clauseIdx) => {
- if (!clause) return false;
- for (const verb of clause.matchAll(verbPattern)) {
+ if (!clause.text || clause.isNegated) return false;
+ for (const verb of clause.text.matchAll(verbPattern)) {
const verbIndex = verb.index ?? 0;
- const beforeVerbClause = clause.slice(0, verbIndex);
+ const beforeVerbClause = clause.text.slice(0, verbIndex);
if (SOCIAL_READ_VERBS.test(beforeVerbClause)) continue;
if (SOCIAL_NEGATION.test(beforeVerbClause)) continue;
const verbWord = verb[0] || '';
- const afterVerbText = clause.slice(verbIndex + verbWord.length);
+ const afterVerbText = clause.text.slice(verbIndex + verbWord.length);
if (SOCIAL_POST_NEGATION.test(afterVerbText)) continue;
if (SOCIAL_NOUN_LIKE_PUBLISH.test(verbWord) && SOCIAL_READ_VERBS.test(afterVerbText)) continue;
- const afterVerb = clause.slice(verbIndex, verbIndex + verbWord.length + 60);
- const beforeVerb = clause.slice(Math.max(0, verbIndex - 60), verbIndex);
+ const afterVerb = clause.text.slice(verbIndex, verbIndex + verbWord.length + 60);
+ const beforeVerb = clause.text.slice(Math.max(0, verbIndex - 60), verbIndex);
platformPattern.lastIndex = 0;
if (platformPattern.test(afterVerb) || platformPattern.test(beforeVerb)) return true;
if (clauseIdx > 0) {
const prevClause = clauses[clauseIdx - 1];
platformPattern.lastIndex = 0;
- if (platformPattern.test(prevClause) && !SOCIAL_READ_VERBS.test(prevClause) && !SOCIAL_NEGATION.test(prevClause)) {
+ if (platformPattern.test(prevClause.text) && !SOCIAL_READ_VERBS.test(prevClause.text) && !prevClause.isNegated) {
return true;
}
}
@@ -14948,12 +15016,13 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
return false;
});
};
- const tweetsThis = String(trustedContext || '').split(SOCIAL_CLAUSE_BREAK).some((clause) => {
- const match = clause.match(/\btweet\s+(?:this|that|it|the\s+following)\b/i);
+ const tweetsThis = clauses.some((clause) => {
+ if (clause.isNegated) return false;
+ const match = clause.text.match(/\btweet\s+(?:this|that|it|the\s+following)\b/i);
if (!match) return false;
- const before = clause.slice(0, match.index);
+ const before = clause.text.slice(0, match.index);
if (SOCIAL_READ_VERBS.test(before) || SOCIAL_NEGATION.test(before)) return false;
- const after = clause.slice(match.index + match[0].length);
+ const after = clause.text.slice(match.index + match[0].length);
if (SOCIAL_POST_NEGATION.test(after)) return false;
return true;
});
diff --git a/src/firefox/src/agent/agent.js b/src/firefox/src/agent/agent.js
index ee94b21de..ab00eadc6 100644
--- a/src/firefox/src/agent/agent.js
+++ b/src/firefox/src/agent/agent.js
@@ -178,7 +178,7 @@ const SOCIAL_NOUN_LIKE_PUBLISH = new RegExp(
'iu',
);
const SOCIAL_NEGATION = new RegExp(
- `(?<![${SOCIAL_WORD_EDGE}])(?:not|never|don't|dont|do\\s+not|cannot|can't|cant|shouldn't|shouldnt|should\\s+not|mustn't|mustnt|must\\s+not|won't|wont|will\\s+not|avoid|refrain|stop|prevent|prohibit|no(?=\\s+(?!attachments?|photos?|images?|pictures?|videos?|files?|media|delays?|doubt|worries|hashtags?|tags?|links?|urls?\\b))|ne|pas|ne\\s+pas|nunca|jamás|jamais|nicht|nie|kein|keine|non|mai|não|nao|hayır|asla|sakın|yapmayın|yapma|не|никогда|нет)(?![${SOCIAL_WORD_EDGE}])`
+ `(?<![${SOCIAL_WORD_EDGE}])(?:not|never|neither|nor|don't|dont|do\\s+not|cannot|can't|cant|shouldn't|shouldnt|should\\s+not|mustn't|mustnt|must\\s+not|won't|wont|will\\s+not|avoid|refrain|stop|prevent|prohibit|no(?=\\s+(?!attachments?|photos?|images?|pictures?|videos?|files?|media|delays?|doubt|worries|hashtags?|tags?|links?|urls?\\b))|ne|pas|ne\\s+pas|nunca|jamás|jamais|nicht|nie|kein|keine|non|mai|não|nao|hayır|asla|sakın|yapmayın|yapma|не|никогда|нет)(?![${SOCIAL_WORD_EDGE}])`
+ '|不要|别|不能|不可|不得|不用|请勿|勿|严禁|禁止|决不|绝不|決して'
+ '|하지\\s*마|하지\\s*마세요|금지',
'iu',
@@ -187,7 +187,38 @@ const SOCIAL_POST_NEGATION = new RegExp(
`^(?:\\s*(?:nothing|nowhere|none)|しないで|してはいけない|してはならない|はいけない|はならない|はだめ|はいけません|はなりません|すんな|するな|禁止|ないで|\\s*(?:하지\\s*마|하지\\s*마세요|하지\\s*않|금지))`,
'iu',
);
-const SOCIAL_CLAUSE_BREAK = new RegExp('[.!?;:,\\n]|(?<![\\p{L}\\p{N}_])(?:and|then|but|or|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![\\p{L}\\p{N}_])|然后|然後|接着|そして|それから|、|。', 'iu');
+const SOCIAL_CLAUSE_DELIMITER = new RegExp(
+ `([.!?;:\\n]|(?<![${SOCIAL_WORD_EDGE}])(?:and|then|but|or|nor|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![${SOCIAL_WORD_EDGE}])|然后|然後|接着|そして|それから|または|、|。|,)`,
+ 'iu',
+);
+const SOCIAL_COORDINATING_DELIMITER = new RegExp(
+ `^(?:or|nor|and|neither|o|ou|oder|ni|nem|né|veya|ya\\s+da|или|ни|或|或者|または|もしくは|や|또는|이나|거나|,|、)$`,
+ 'iu',
+);
+const SOCIAL_CLAUSE_BREAK = new RegExp(
+ `[.!?;:,\\n]|(?<![${SOCIAL_WORD_EDGE}])(?:and|then|but|or|nor|after|before|y|luego|puis|et|ensuite|und|dann|poi|sonra|ve|затем|и)(?![${SOCIAL_WORD_EDGE}])|然后|然後|接着|そして|それから|または|、|。`,
+ 'iu',
+);
+
+const GENERIC_ATTACHMENT_WORDS = new Set([
+ 'true', 'yes', 'attached', 'attachment', 'attachments', 'media',
+ 'image', 'images', 'photo', 'photos', 'picture', 'pictures', 'pic', 'pics',
+ 'video', 'videos', 'clip', 'clips', 'recording', 'recordings',
+ 'file', 'files', 'graphic', 'graphics',
+ 'a', 'an', 'the', 'of', 'in', 'with', 'some', 'any',
+ 'one', 'two', 'three', 'four', 'five', 'six', 'seven', 'eight', 'nine', 'ten',
+ 'single', 'multiple', 'both',
+ 'de', 'du', 'des', 'el', 'la', 'los', 'las', 'der', 'die', 'das', 'di', 'il', 'lo', 'gli', 'le',
+ 'o', 'os', 'as', 'do', 'da', 'dos', 'das', 'un', 'une', 'deux', 'trois', 'quatre', 'cinq',
+ 'uno', 'una', 'unos', 'unas', 'dos', 'tres', 'cuatro', 'cinco', 'due', 'tre', 'quattro', 'cinque',
+ 'ein', 'eine', 'einen', 'einer', 'zwei', 'drei', 'vier', 'fünf', 'um', 'uma', 'dois', 'duas', 'três',
+ 'imagen', 'imagenes', 'imágenes', 'foto', 'fotos', 'vidéo', 'vidéos', 'bild', 'bilder',
+ 'fichier', 'fichiers', 'archivo', 'archivos', 'datei', 'dateien', 'allegato', 'allegati', 'anexo', 'anexos',
+ 'один', 'одна', 'одно', 'два', 'две', 'три', 'четыре', 'пять',
+ 'вложения', 'вложение', 'фотографии', 'фотография', 'изображения', 'изображение', 'видео',
+]);
+
+const CJK_GENERIC_ATTACHMENT_REGEX = /^[0-9一二两三四五六七八九十添付画像写真動画メディア已上传附件图片照片视频媒体枚つの本张條条个個장での\s\-_,.:;!?/\\()]+$/u;
// "On <url>, publish this" names a destination just as plainly as
// "publish this on <url>", but only when the URL is presented as a place.
@@ -2012,14 +2043,10 @@ export class Agent extends LoopDetector {
}
_cleanSpecificAttachmentTarget(value) {
- let target = String(value || '').trim().toLowerCase();
- target = target
- .replace(/^https?:\/\/[^\s]+/i, '')
- .replace(/\b(?:two|three|four|five|six|seven|eight|nine|ten|\d+)\s+(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with name)\s+/i, '')
- .replace(/\b(?:an?|the)\s+(?:image|photo|picture|video|attachment|file)\s+(?:of|named|called|with name)\s+/i, '')
- .replace(/\b(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with name)\s+/i, '')
- .trim();
- return target;
+ const text = String(value || '').trim();
+ const match = text.match(/(?:(?:an?|the|\d+|one|two|three|four|five|six|seven|eight|nine|ten)\s+)?(?:images?|photos?|pictures?|videos?|attachments?|files?)\s+(?:of|named|called|with\s+name)\s+(.+)$/i);
+ if (match) return match[1].trim().toLowerCase();
+ return text.toLowerCase();
}
_parseWorkflowAttachmentRequirement(value) {
@@ -2053,18 +2080,14 @@ export class Agent extends LoopDetector {
}
}
- const stripped = text
- .replace(/^(?:true|yes|attached)\b/i, '')
- .replace(/\b(?:\d+|one|two|three|four|five|six|seven|eight|nine|ten|single|multiple|both|un|une|deux|trois|quatre|cinq|uno|una|unos|unas|dos|tres|cuatro|cinco|due|tre|quattro|cinque|ein|eine|einen|einer|zwei|drei|vier|fünf|um|uma|dois|duas|três|quatro|один|одна|одно|два|две|три|четыре|пять)\b/gi, '')
- .replace(/\b(?:a|an|the|of|in|with|some|any|de|du|des|el|la|los|las|der|die|das|di|il|lo|gli|le|o|os|as|do|da|dos|das)\b/gi, '')
- .replace(/\b(?:image|images|photo|photos|picture|pictures|pic|pics|video|videos|clip|clips|recording|recordings|media|attachment|attachments|file|files|graphic|graphics)\b/gi, '')
- .replace(/\b(?:imagen(?:es)?|imágenes|foto|fotos|vidéo|vidéos|bild(?:er)?|fichier|fichiers|archivo|archivos|datei(?:en)?|allegat[oi]|anexo|anexos)\b/gi, '')
- .replace(/\b(?:вложени[яе]|фотографи[ия]|изображени[яе]|видео)\b/gi, '')
- .replace(/[一二两三四五六七八九十]/g, '')
- .replace(/(?:添付|画像|写真|動画|メディア|已上传|附件|图片|照片|视频|媒体|枚|つの|本|张|條|条|个|個|장|개|の)/gu, '')
- .replace(/[\s\-_,.:;!?/\\()]+/g, '');
-
- const isGeneric = stripped.length === 0;
+ let isGeneric = false;
+ if (CJK_GENERIC_ATTACHMENT_REGEX.test(text)) {
+ isGeneric = true;
+ } else {
+ const nonPunctuation = text.replace(/[\s\-_,.:;!?/\\()]+/g, ' ').trim();
+ const words = nonPunctuation ? nonPunctuation.split(/\s+/) : [];
+ isGeneric = words.length > 0 && words.every(w => /^\d+$/.test(w) || GENERIC_ATTACHMENT_WORDS.has(w) || CJK_GENERIC_ATTACHMENT_REGEX.test(w));
+ }
return { isGeneric, expectedCount, wantsImage, wantsVideo, normalized: text };
}
@@ -12629,19 +12652,63 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
return url;
}
+ _socialPublicationClauses(text) {
+ const parts = String(text || '').split(SOCIAL_CLAUSE_DELIMITER);
+ const clauses = [];
+ for (let i = 0; i < parts.length; i += 2) {
+ const clauseText = parts[i] || '';
+ const delim = i > 0 ? (parts[i - 1] || '').trim() : '';
+ clauses.push({ text: clauseText, delim });
+ }
+
+ let carriedNegation = false;
+ for (let i = 0; i < clauses.length; i++) {
+ const c = clauses[i];
+ const isCoordinated = i > 0 && SOCIAL_COORDINATING_DELIMITER.test(c.delim);
+ if (!isCoordinated) carriedNegation = false;
+
+ const hasExplicitNeg = SOCIAL_NEGATION.test(c.text);
+ if (hasExplicitNeg) {
+ c.isNegated = true;
+ carriedNegation = true;
+ } else if (carriedNegation) {
+ c.isNegated = true;
+ } else {
+ c.isNegated = false;
+ }
+ }
+
+ let carriedPostNegation = false;
+ for (let i = clauses.length - 1; i >= 0; i--) {
+ const c = clauses[i];
+ const hasExplicitPostNeg = SOCIAL_POST_NEGATION.test(c.text.trim());
+ if (hasExplicitPostNeg) {
+ c.isNegated = true;
+ carriedPostNegation = true;
+ } else if (carriedPostNegation && i < clauses.length - 1 && SOCIAL_COORDINATING_DELIMITER.test(clauses[i + 1].delim)) {
+ c.isNegated = true;
+ } else {
+ carriedPostNegation = false;
+ }
+ }
+
+ return clauses;
+ }
+
// Publication language only counts as a command when nothing in its own
// clause is asking to read or forbidding publication. "Read posts on <feed>"
// and "阅读推文 <feed>" name content; "do not post this on <feed>" forbids
// publication; "read the summary and publish it on <feed>" asks for a post in
// a clause of its own.
_socialPublicationCommandIn(text) {
- return String(text || '').split(SOCIAL_CLAUSE_BREAK).some((clause) => {
- const publish = clause ? clause.match(SOCIAL_PUBLISH_VERBS) : null;
+ return this._socialPublicationClauses(text).some((clause) => {
+ if (clause.isNegated) return false;
+ const publish = clause.text ? clause.text.match(SOCIAL_PUBLISH_VERBS) : null;
if (!publish) return false;
- const before = clause.slice(0, publish.index);
+ const before = clause.text.slice(0, publish.index);
if (SOCIAL_READ_VERBS.test(before)) return false;
if (SOCIAL_NEGATION.test(before)) return false;
- const after = clause.slice(publish.index + publish[0].length);
+ const after = clause.text.slice(publish.index + publish[0].length);
if (SOCIAL_POST_NEGATION.test(after)) return false;
if (SOCIAL_NOUN_LIKE_PUBLISH.test(publish[0]) && SOCIAL_READ_VERBS.test(after)) return false;
return true;
@@ -12695,8 +12762,9 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
match.index + match[0].length,
match.index + match[0].length + 60,
);
- const clauseBefore = before.split(SOCIAL_CLAUSE_BREAK).pop() || '';
- if (SOCIAL_NEGATION.test(clauseBefore)) continue;
+ const beforeClauses = this._socialPublicationClauses(before);
+ const lastBeforeClause = beforeClauses[beforeClauses.length - 1];
+ if (lastBeforeClause?.isNegated) continue;
const governed = /\/(?:compose|intent|i\/flow)\//.test(url)
|| this._socialPublicationCommandIn(before)
|| (SOCIAL_PUBLISH_DESTINATION_LEAD.test(before) && this._socialPublicationCommandIn(after));
@@ -12712,6 +12780,7 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
// publication target. Verbs and prepositions are the same multilingual
// sets the URL scan uses, so "Publícalo en X" counts while English-only
// matching would leave it unbound.
+ const clauses = this._socialPublicationClauses(trustedContext);
const publishesTo = (platform) => {
const platformPattern = new RegExp(
`(?<![${SOCIAL_WORD_EDGE}])(?:on|onto|to|via|in|at|en|sur|sobre|\u00e0|au|auf|su|em|na|para|\u0432|\u043d\u0430)(?![${SOCIAL_WORD_EDGE}])\\s+(?:the\\s+)?${platform}(?![${SOCIAL_WORD_EDGE}])`
@@ -12721,26 +12790,25 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
'iu',
);
const verbPattern = new RegExp(SOCIAL_PUBLISH_VERBS.source, 'giu');
- const clauses = trustedContext.split(SOCIAL_CLAUSE_BREAK);
return clauses.some((clause, clauseIdx) => {
- if (!clause) return false;
- for (const verb of clause.matchAll(verbPattern)) {
+ if (!clause.text || clause.isNegated) return false;
+ for (const verb of clause.text.matchAll(verbPattern)) {
const verbIndex = verb.index ?? 0;
- const beforeVerbClause = clause.slice(0, verbIndex);
+ const beforeVerbClause = clause.text.slice(0, verbIndex);
if (SOCIAL_READ_VERBS.test(beforeVerbClause)) continue;
if (SOCIAL_NEGATION.test(beforeVerbClause)) continue;
const verbWord = verb[0] || '';
- const afterVerbText = clause.slice(verbIndex + verbWord.length);
+ const afterVerbText = clause.text.slice(verbIndex + verbWord.length);
if (SOCIAL_POST_NEGATION.test(afterVerbText)) continue;
if (SOCIAL_NOUN_LIKE_PUBLISH.test(verbWord) && SOCIAL_READ_VERBS.test(afterVerbText)) continue;
- const afterVerb = clause.slice(verbIndex, verbIndex + verbWord.length + 60);
- const beforeVerb = clause.slice(Math.max(0, verbIndex - 60), verbIndex);
+ const afterVerb = clause.text.slice(verbIndex, verbIndex + verbWord.length + 60);
+ const beforeVerb = clause.text.slice(Math.max(0, verbIndex - 60), verbIndex);
platformPattern.lastIndex = 0;
if (platformPattern.test(afterVerb) || platformPattern.test(beforeVerb)) return true;
if (clauseIdx > 0) {
const prevClause = clauses[clauseIdx - 1];
platformPattern.lastIndex = 0;
- if (platformPattern.test(prevClause) && !SOCIAL_READ_VERBS.test(prevClause) && !SOCIAL_NEGATION.test(prevClause)) {
+ if (platformPattern.test(prevClause.text) && !SOCIAL_READ_VERBS.test(prevClause.text) && !prevClause.isNegated) {
return true;
}
}
@@ -12748,12 +12816,13 @@ Rules: no prose intro, no conclusion, no "this screenshot shows...", no layout d
return false;
});
};
- const tweetsThis = String(trustedContext || '').split(SOCIAL_CLAUSE_BREAK).some((clause) => {
- const match = clause.match(/\btweet\s+(?:this|that|it|the\s+following)\b/i);
+ const tweetsThis = clauses.some((clause) => {
+ if (clause.isNegated) return false;
+ const match = clause.text.match(/\btweet\s+(?:this|that|it|the\s+following)\b/i);
if (!match) return false;
- const before = clause.slice(0, match.index);
+ const before = clause.text.slice(0, match.index);
if (SOCIAL_READ_VERBS.test(before) || SOCIAL_NEGATION.test(before)) return false;
- const after = clause.slice(match.index + match[0].length);
+ const after = clause.text.slice(match.index + match[0].length);
if (SOCIAL_POST_NEGATION.test(after)) return false;
return true;
});
diff --git a/test/run.js b/test/run.js
index ef9718d61..2810a9d3c 100644
--- a/test/run.js
+++ b/test/run.js
@@ -86220,6 +86220,16 @@ test('social publication workflow follows the live X or Bluesky destination and
'Korean post-negated publication command was wrongly treated as a publish destination'],
['Post this on X with no attachments', ['twitter'],
'post with no-attachments phrase was wrongly rejected'],
+ ['Do not post or publish this on X; submit it in the survey', [],
+ 'coordinated publish verbs under negation were wrongly treated as a publish destination'],
+ ['Never publish or post to Bluesky', [],
+ 'coordinated publish verbs under never were wrongly treated as Bluesky publish destination'],
+ ['Do not post on X or publish on Bluesky', [],
+ 'multiple publish destinations joined by or under negation were wrongly treated as a publish destination'],
+ ['Do not post on X, but publish on Bluesky', ['bluesky'],
+ 'contrastive but clause after negated post was wrongly rejected for Bluesky'],
+ ['Post or publish this on X', ['twitter'],
+ 'affirmative coordinated publish verbs were wrongly rejected for X'],
]) {
assert.deepEqual(
[...agent._trustedSocialPublishTargetAdapters({ taskText })],