@faircopy/rules-nlp 1.2.1 → 1.4.0

package/README.md

@@ -0,0 +1,32 @@
+# @faircopy/rules-nlp
+
+Optional NLP-powered ruleset for faircopy using `compromise`.
+
+```sh
+npm i -D @faircopy/rules-nlp
+```
+
+Load the ruleset once, then configure rules with bare rule IDs:
+
+```ts
+rulesets: ['@faircopy/rules-nlp'],
+rules: {
+  'no-filter-words': 'warn',
+  'no-passive-voice': 'warn',
+  'no-weak-modals': 'warn',
+  'no-stacked-adjectives': 'warn',
+  'no-nominalized-phrases': 'warn',
+}
+```
+
+Package-qualified IDs like `@faircopy/rules-nlp/no-passive-voice` still work and are required if another loaded ruleset exposes the same bare rule name.
+
+## Rules
+
+| Rule | Description |
+|---|---|
+| `no-filter-words` | Ban filter phrases like `I think` and `it seems` |
+| `no-passive-voice` | Flag likely passive-voice constructions |
+| `no-weak-modals` | Flag hedged modal claims like `can help` and `might improve` |
+| `no-stacked-adjectives` | Flag noun phrases with multiple adjectives before the noun |
+| `no-nominalized-phrases` | Flag nominalized `X of Y` phrases like `optimization of onboarding` |

package/dist/index.d.ts

@@ -5,12 +5,29 @@ interface NoFilterWordsOptions {
 }
 declare const noFilterWords: Rule<NoFilterWordsOptions>;
 
+interface NoNominalizedPhrasesOptions {
+    suffixes?: string[];
+    allowedWords?: string[];
+}
+declare const noNominalizedPhrases: Rule<NoNominalizedPhrasesOptions>;
+
 interface NoPassiveVoiceOptions {
     allowedAuxiliaries?: string[];
 }
 declare const noPassiveVoice: Rule<NoPassiveVoiceOptions>;
 
+interface NoStackedAdjectivesOptions {
+    allowedPhrases?: string[];
+}
+declare const noStackedAdjectives: Rule<NoStackedAdjectivesOptions>;
+
+interface NoWeakModalsOptions {
+    modals?: string[];
+    verbs?: string[];
+}
+declare const noWeakModals: Rule<NoWeakModalsOptions>;
+
 /** All NLP rules keyed by their rule ID. */
 declare const ruleRegistry: Map<string, Rule>;
 
-export { type NoFilterWordsOptions, type NoPassiveVoiceOptions, noFilterWords, noPassiveVoice, ruleRegistry };
+export { type NoFilterWordsOptions, type NoNominalizedPhrasesOptions, type NoPassiveVoiceOptions, type NoStackedAdjectivesOptions, type NoWeakModalsOptions, noFilterWords, noNominalizedPhrases, noPassiveVoice, noStackedAdjectives, noWeakModals, ruleRegistry };

package/dist/index.js

@@ -18,6 +18,12 @@ function getMatchOccurrences(text, matches) {
     }];
   });
 }
+function getOccurrenceRange(sourceMap, occurrence) {
+  const start = sourceMap[occurrence.start];
+  const end = sourceMap[occurrence.end - 1];
+  if (start === void 0 || end === void 0) return null;
+  return { start, end: end + 1 };
+}
 function sumTermLengths(terms) {
   if (!terms?.length) return void 0;
   let total = 0;
@@ -64,6 +70,53 @@ var noFilterWords = {
   }
 };
 
+// src/no-nominalized-phrases.ts
+var DEFAULT_SUFFIXES = ["tion", "sion", "ment", "ance", "ence", "ity"];
+var DEFAULT_ALLOWED_WORDS = [
+  "accessibility",
+  "availability",
+  "capacity",
+  "community",
+  "identity",
+  "opportunity",
+  "privacy",
+  "quality",
+  "reliability",
+  "security"
+];
+var noNominalizedPhrases = {
+  id: "no-nominalized-phrases",
+  description: 'Flag nominalized "X of Y" phrases that hide the action in a noun',
+  defaults: { suffixes: DEFAULT_SUFFIXES, allowedWords: DEFAULT_ALLOWED_WORDS },
+  help: "Nominalized phrases bury the action. Rewrite the phrase with a verb so the sentence says who does what.",
+  check({ text, sourceMap, options }) {
+    const diagnostics = [];
+    const suffixes = options.suffixes?.length ? options.suffixes : DEFAULT_SUFFIXES;
+    const allowedWords = new Set((options.allowedWords?.length ? options.allowedWords : DEFAULT_ALLOWED_WORDS).map((value) => value.toLowerCase()));
+    const suffixPattern = suffixes.map(escapeRegex).join("|");
+    if (!suffixPattern) return diagnostics;
+    const doc = createDoc(text);
+    const matches = doc.match(`/[a-z]+(${suffixPattern})/ of .`);
+    for (const occurrence of getMatchOccurrences(text, matches)) {
+      const nominalization = occurrence.text.trim().split(/\s+/)[0]?.toLowerCase();
+      if (!nominalization || allowedWords.has(nominalization)) continue;
+      const range = getOccurrenceRange(sourceMap, occurrence);
+      if (!range) continue;
+      diagnostics.push({
+        ruleId: "no-nominalized-phrases",
+        severity: "warn",
+        message: `rewrite "${occurrence.text}" with a verb`,
+        range,
+        help: noNominalizedPhrases.help
+      });
+    }
+    return diagnostics;
+  }
+};
+function escapeRegex(value) {
+  return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+}
+
 // src/no-passive-voice.ts
 var DEFAULT_ALLOWED_AUXILIARIES = ["is", "are", "was", "were", "be", "been", "being"];
 var noPassiveVoice = {
@@ -104,14 +157,92 @@ function dedupeDiagnostics(diagnostics) {
   });
 }
 
+// src/no-stacked-adjectives.ts
+var noStackedAdjectives = {
+  id: "no-stacked-adjectives",
+  description: "Flag noun phrases with multiple adjectives before the noun",
+  defaults: { allowedPhrases: [] },
+  help: "Stacked adjectives make copy feel generic. Keep the one descriptor that earns its place or replace the phrase with concrete evidence.",
+  check({ text, sourceMap, options }) {
+    const diagnostics = [];
+    const allowedPhrases = new Set((options.allowedPhrases ?? []).map((value) => value.toLowerCase()));
+    const doc = createDoc(text);
+    const matches = doc.match("#Adjective #Adjective+ #Noun");
+    for (const occurrence of getMatchOccurrences(text, matches)) {
+      if (allowedPhrases.has(occurrence.text.toLowerCase())) continue;
+      const range = getOccurrenceRange(sourceMap, occurrence);
+      if (!range) continue;
+      diagnostics.push({
+        ruleId: "no-stacked-adjectives",
+        severity: "warn",
+        message: `cut stacked adjectives in "${occurrence.text}"`,
+        range,
+        help: noStackedAdjectives.help
+      });
+    }
+    return diagnostics;
+  }
+};
+
+// src/no-weak-modals.ts
+var DEFAULT_MODALS = ["can", "could", "may", "might"];
+var DEFAULT_VERBS = [
+  "boost",
+  "drive",
+  "enable",
+  "help",
+  "improve",
+  "increase",
+  "make",
+  "reduce",
+  "support",
+  "transform",
+  "unlock"
+];
+var noWeakModals = {
+  id: "no-weak-modals",
+  description: 'Flag hedged modal claims like "can help" and "might improve"',
+  defaults: { modals: DEFAULT_MODALS, verbs: DEFAULT_VERBS },
+  help: "Hedged modal claims sound tentative. Replace them with the outcome, capability, or proof you can stand behind.",
+  check({ text, sourceMap, options }) {
+    const diagnostics = [];
+    const modals = new Set((options.modals?.length ? options.modals : DEFAULT_MODALS).map((value) => value.toLowerCase()));
+    const verbs = new Set((options.verbs?.length ? options.verbs : DEFAULT_VERBS).map((value) => value.toLowerCase()));
+    const doc = createDoc(text);
+    const matches = doc.match("#Modal #Adverb? #Verb");
+    for (const occurrence of getMatchOccurrences(text, matches)) {
+      const words = occurrence.text.trim().split(/\s+/);
+      const modal = words[0]?.toLowerCase();
+      const verb = words[words.length - 1]?.toLowerCase();
+      if (!modal || !verb || !modals.has(modal) || !verbs.has(verb)) continue;
+      const range = getOccurrenceRange(sourceMap, occurrence);
+      if (!range) continue;
+      diagnostics.push({
+        ruleId: "no-weak-modals",
+        severity: "warn",
+        message: `replace "${occurrence.text.toLowerCase()}" with a direct claim`,
+        range,
+        help: noWeakModals.help
+      });
+    }
+    return diagnostics;
+  }
+};
+
 // src/index.ts
 var ruleRegistry = /* @__PURE__ */ new Map([
   ["no-filter-words", noFilterWords],
-  ["no-passive-voice", noPassiveVoice]
+  ["no-nominalized-phrases", noNominalizedPhrases],
+  ["no-passive-voice", noPassiveVoice],
+  ["no-stacked-adjectives", noStackedAdjectives],
+  ["no-weak-modals", noWeakModals]
 ]);
 export {
   noFilterWords,
+  noNominalizedPhrases,
   noPassiveVoice,
+  noStackedAdjectives,
+  noWeakModals,
   ruleRegistry
 };
 //# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/utils.ts","../src/no-filter-words.ts","../src/no-passive-voice.ts","../src/index.ts"],"sourcesContent":["import nlp from 'compromise'\nimport type { DocView, JsonOffsetEntry, JsonOffsetTerm, MatchView } from './types.js'\n\nexport interface MatchOccurrence {\n  text: string\n  start: number\n  end: number\n}\n\nexport function createDoc(text: string): DocView {\n  return nlp(text) as unknown as DocView\n}\n\nexport function getMatchOccurrences(text: string, matches: MatchView): MatchOccurrence[] {\n  const json = matches.json({ offset: true, text: true, terms: { offset: true } }) as JsonOffsetEntry[]\n\n  return json.flatMap((entry) => {\n    const start = entry.offset?.start ?? entry.terms?.[0]?.offset?.start\n    const length = entry.offset?.length ?? sumTermLengths(entry.terms)\n\n    if (typeof start !== 'number' || typeof length !== 'number' || length <= 0) {\n      return []\n    }\n\n    return [{\n      text: entry.text ?? text.slice(start, start + length),\n      start,\n      end: start + length,\n    }]\n  })\n}\n\nfunction sumTermLengths(terms: JsonOffsetTerm[] | undefined): number | undefined {\n  if (!terms?.length) return undefined\n\n  let total = 0\n  for (const term of terms) {\n    const length = term.offset?.length\n    if (typeof length !== 'number') return undefined\n    total += length\n  }\n\n  return total\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences } from './utils.js'\n\nexport interface NoFilterWordsOptions {\n  phrases?: string[]\n}\n\nconst DEFAULT_PHRASES = [\n  'I think',\n  'it seems',\n  'basically',\n  'in order to',\n]\n\nexport const noFilterWords: Rule<NoFilterWordsOptions> = {\n  id: 'no-filter-words',\n  description: 'Ban filter phrases that distance the claim from the reader',\n  defaults: { phrases: DEFAULT_PHRASES },\n  help: 'Filter phrases announce a perspective or pad the sentence instead of making the point. Delete the phrase or rewrite the sentence so the claim stands on its own.',\n\n  check({ text, sourceMap, options }: RuleInput<NoFilterWordsOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const phrases = options.phrases?.length ? options.phrases : DEFAULT_PHRASES\n    const doc = createDoc(text)\n\n    for (const phrase of phrases) {\n      const matches = doc.match(phrase)\n      for (const occurrence of getMatchOccurrences(text, matches)) {\n        const start = sourceMap[occurrence.start]\n        const end = sourceMap[occurrence.end - 1]\n        if (start === undefined || end === undefined) continue\n\n        diagnostics.push({\n          ruleId: 'no-filter-words',\n          severity: 'error',\n          message: `remove \"${occurrence.text.toLowerCase()}\" — state the claim directly`,\n          range: { start, end: end + 1 },\n          help: noFilterWords.help,\n        })\n      }\n    }\n\n    return diagnostics\n  },\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences } from './utils.js'\n\nexport interface NoPassiveVoiceOptions {\n  allowedAuxiliaries?: string[]\n}\n\nconst DEFAULT_ALLOWED_AUXILIARIES = ['is', 'are', 'was', 'were', 'be', 'been', 'being']\n\nexport const noPassiveVoice: Rule<NoPassiveVoiceOptions> = {\n  id: 'no-passive-voice',\n  description: 'Flag likely passive-voice constructions using POS tagging patterns',\n  defaults: { allowedAuxiliaries: DEFAULT_ALLOWED_AUXILIARIES },\n  help: 'Passive voice often hides the actor and adds drag. Prefer naming who did the action unless the actor genuinely does not matter.',\n\n  check({ text, sourceMap, options }: RuleInput<NoPassiveVoiceOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const auxiliaries = new Set((options.allowedAuxiliaries?.length ? options.allowedAuxiliaries : DEFAULT_ALLOWED_AUXILIARIES).map(value => value.toLowerCase()))\n    const doc = createDoc(text)\n    const matches = doc.match('(#Copula|#Auxiliary) #PastTense')\n\n    for (const occurrence of getMatchOccurrences(text, matches)) {\n      const words = occurrence.text.trim().split(/\\s+/)\n      if (words.length < 2) continue\n      if (!auxiliaries.has(words[0]!.toLowerCase())) continue\n\n      const start = sourceMap[occurrence.start]\n      const end = sourceMap[occurrence.end - 1]\n      if (start === undefined || end === undefined) continue\n\n      diagnostics.push({\n        ruleId: 'no-passive-voice',\n        severity: 'warn',\n        message: `rewrite passive construction \"${occurrence.text}\" with a named actor`,\n        range: { start, end: end + 1 },\n        help: noPassiveVoice.help,\n      })\n    }\n\n    return dedupeDiagnostics(diagnostics)\n  },\n}\n\nfunction dedupeDiagnostics(diagnostics: Diagnostic[]): Diagnostic[] {\n  const seen = new Set<string>()\n  return diagnostics.filter((diagnostic) => {\n    const key = `${diagnostic.range.start}:${diagnostic.range.end}`\n    if (seen.has(key)) return false\n    seen.add(key)\n    return true\n  })\n}\n","import type { Rule } from '@faircopy/core'\nimport { noFilterWords } from './no-filter-words.js'\nimport { noPassiveVoice } from './no-passive-voice.js'\n\nexport { noFilterWords } from './no-filter-words.js'\nexport { noPassiveVoice } from './no-passive-voice.js'\nexport type { NoFilterWordsOptions } from './no-filter-words.js'\nexport type { NoPassiveVoiceOptions } from './no-passive-voice.js'\n\n/** All NLP rules keyed by their rule ID. */\nexport const ruleRegistry: Map<string, Rule> = new Map([\n  ['no-filter-words', noFilterWords as Rule],\n  ['no-passive-voice', noPassiveVoice as Rule],\n])\n"],"mappings":";AAAA,OAAO,SAAS;AAST,SAAS,UAAU,MAAuB;AAC/C,SAAO,IAAI,IAAI;AACjB;AAEO,SAAS,oBAAoB,MAAc,SAAuC;AACvF,QAAM,OAAO,QAAQ,KAAK,EAAE,QAAQ,MAAM,MAAM,MAAM,OAAO,EAAE,QAAQ,KAAK,EAAE,CAAC;AAE/E,SAAO,KAAK,QAAQ,CAAC,UAAU;AAC7B,UAAM,QAAQ,MAAM,QAAQ,SAAS,MAAM,QAAQ,CAAC,GAAG,QAAQ;AAC/D,UAAM,SAAS,MAAM,QAAQ,UAAU,eAAe,MAAM,KAAK;AAEjE,QAAI,OAAO,UAAU,YAAY,OAAO,WAAW,YAAY,UAAU,GAAG;AAC1E,aAAO,CAAC;AAAA,IACV;AAEA,WAAO,CAAC;AAAA,MACN,MAAM,MAAM,QAAQ,KAAK,MAAM,OAAO,QAAQ,MAAM;AAAA,MACpD;AAAA,MACA,KAAK,QAAQ;AAAA,IACf,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,eAAe,OAAyD;AAC/E,MAAI,CAAC,OAAO,OAAQ,QAAO;AAE3B,MAAI,QAAQ;AACZ,aAAW,QAAQ,OAAO;AACxB,UAAM,SAAS,KAAK,QAAQ;AAC5B,QAAI,OAAO,WAAW,SAAU,QAAO;AACvC,aAAS;AAAA,EACX;AAEA,SAAO;AACT;;;ACpCA,IAAM,kBAAkB;AAAA,EACtB;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAEO,IAAM,gBAA4C;AAAA,EACvD,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,SAAS,gBAAgB;AAAA,EACrC,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAkD;AACjF,UAAM,cAA4B,CAAC;AACnC,UAAM,UAAU,QAAQ,SAAS,SAAS,QAAQ,UAAU;AAC5D,UAAM,MAAM,UAAU,IAAI;AAE1B,eAAW,UAAU,SAAS;AAC5B,YAAM,UAAU,IAAI,MAAM,MAAM;AAChC,iBAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,cAAM,QAAQ,UAAU,WAAW,KAAK;AACxC,cAAM,MAAM,UAAU,WAAW,MAAM,CAAC;AACxC,YAAI,UAAU,UAAa,QAAQ,OAAW;AAE9C,oBAAY,KAAK;AAAA,UACf,QAAQ;AAAA,UACR,UAAU;AAAA,UACV,SAAS,WAAW,WAAW,KAAK,YAAY,CAAC;AAAA,UACjD,OAAO,EAAE,OAAO,KAAK,MAAM,EAAE;AAAA,UAC7B,MAAM,cAAc;AAAA,QACtB,CAAC;AAAA,MACH;AAAA,IACF;AAEA,WAAO;AAAA,EACT;AACF;;;ACrCA,IAAM,8BAA8B,CAAC,MAAM,OAAO,OAAO,QAAQ,MAAM,QAAQ,OAAO;AAE/E,IAAM,iBAA8C;AAAA,EACzD,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,oBAAoB,4BAA4B;AAAA,EAC5D,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAmD;AAClF,UAAM,cAA4B,CAAC;AACnC,UAAM,cAAc,IAAI,KAAK,QAAQ,oBAAoB,SAAS,QAAQ,qBAAqB,6BAA6B,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AAC7J,UAAM,MAAM,UAAU,IAAI;AAC1B,UAAM,UAAU,IAAI,MAAM,iCAAiC;AAE3D,eAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,YAAM,QAAQ,WAAW,KAAK,KAAK,EAAE,MAAM,KAAK;AAChD,UAAI,MAAM,SAAS,EAAG;AACtB,UAAI,CAAC,YAAY,IAAI,MAAM,CAAC,EAAG,YAAY,CAAC,EAAG;AAE/C,YAAM,QAAQ,UAAU,WAAW,KAAK;AACxC,YAAM,MAAM,UAAU,WAAW,MAAM,CAAC;AACxC,UAAI,UAAU,UAAa,QAAQ,OAAW;AAE9C,kBAAY,KAAK;AAAA,QACf,QAAQ;AAAA,QACR,UAAU;AAAA,QACV,SAAS,iCAAiC,WAAW,IAAI;AAAA,QACzD,OAAO,EAAE,OAAO,KAAK,MAAM,EAAE;AAAA,QAC7B,MAAM,eAAe;AAAA,MACvB,CAAC;AAAA,IACH;AAEA,WAAO,kBAAkB,WAAW;AAAA,EACtC;AACF;AAEA,SAAS,kBAAkB,aAAyC;AAClE,QAAM,OAAO,oBAAI,IAAY;AAC7B,SAAO,YAAY,OAAO,CAAC,eAAe;AACxC,UAAM,MAAM,GAAG,WAAW,MAAM,KAAK,IAAI,WAAW,MAAM,GAAG;AAC7D,QAAI,KAAK,IAAI,GAAG,EAAG,QAAO;AAC1B,SAAK,IAAI,GAAG;AACZ,WAAO;AAAA,EACT,CAAC;AACH;;;ACzCO,IAAM,eAAkC,oBAAI,IAAI;AAAA,EACrD,CAAC,mBAAmB,aAAqB;AAAA,EACzC,CAAC,oBAAoB,cAAsB;AAC7C,CAAC;","names":[]}
+{"version":3,"sources":["../src/utils.ts","../src/no-filter-words.ts","../src/no-nominalized-phrases.ts","../src/no-passive-voice.ts","../src/no-stacked-adjectives.ts","../src/no-weak-modals.ts","../src/index.ts"],"sourcesContent":["import nlp from 'compromise'\nimport type { Diagnostic } from '@faircopy/core'\nimport type { DocView, JsonOffsetEntry, JsonOffsetTerm, MatchView } from './types.js'\n\nexport interface MatchOccurrence {\n  text: string\n  start: number\n  end: number\n}\n\nexport function createDoc(text: string): DocView {\n  return nlp(text) as unknown as DocView\n}\n\nexport function getMatchOccurrences(text: string, matches: MatchView): MatchOccurrence[] {\n  const json = matches.json({ offset: true, text: true, terms: { offset: true } }) as JsonOffsetEntry[]\n\n  return json.flatMap((entry) => {\n    const start = entry.offset?.start ?? entry.terms?.[0]?.offset?.start\n    const length = entry.offset?.length ?? sumTermLengths(entry.terms)\n\n    if (typeof start !== 'number' || typeof length !== 'number' || length <= 0) {\n      return []\n    }\n\n    return [{\n      text: entry.text ?? text.slice(start, start + length),\n      start,\n      end: start + length,\n    }]\n  })\n}\n\nexport function getOccurrenceRange(sourceMap: number[], occurrence: MatchOccurrence): Diagnostic['range'] | null {\n  const start = sourceMap[occurrence.start]\n  const end = sourceMap[occurrence.end - 1]\n  if (start === undefined || end === undefined) return null\n\n  return { start, end: end + 1 }\n}\n\nfunction sumTermLengths(terms: JsonOffsetTerm[] | undefined): number | undefined {\n  if (!terms?.length) return undefined\n\n  let total = 0\n  for (const term of terms) {\n    const length = term.offset?.length\n    if (typeof length !== 'number') return undefined\n    total += length\n  }\n\n  return total\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences } from './utils.js'\n\nexport interface NoFilterWordsOptions {\n  phrases?: string[]\n}\n\nconst DEFAULT_PHRASES = [\n  'I think',\n  'it seems',\n  'basically',\n  'in order to',\n]\n\nexport const noFilterWords: Rule<NoFilterWordsOptions> = {\n  id: 'no-filter-words',\n  description: 'Ban filter phrases that distance the claim from the reader',\n  defaults: { phrases: DEFAULT_PHRASES },\n  help: 'Filter phrases announce a perspective or pad the sentence instead of making the point. Delete the phrase or rewrite the sentence so the claim stands on its own.',\n\n  check({ text, sourceMap, options }: RuleInput<NoFilterWordsOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const phrases = options.phrases?.length ? options.phrases : DEFAULT_PHRASES\n    const doc = createDoc(text)\n\n    for (const phrase of phrases) {\n      const matches = doc.match(phrase)\n      for (const occurrence of getMatchOccurrences(text, matches)) {\n        const start = sourceMap[occurrence.start]\n        const end = sourceMap[occurrence.end - 1]\n        if (start === undefined || end === undefined) continue\n\n        diagnostics.push({\n          ruleId: 'no-filter-words',\n          severity: 'error',\n          message: `remove \"${occurrence.text.toLowerCase()}\" — state the claim directly`,\n          range: { start, end: end + 1 },\n          help: noFilterWords.help,\n        })\n      }\n    }\n\n    return diagnostics\n  },\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences, getOccurrenceRange } from './utils.js'\n\nexport interface NoNominalizedPhrasesOptions {\n  suffixes?: string[]\n  allowedWords?: string[]\n}\n\nconst DEFAULT_SUFFIXES = ['tion', 'sion', 'ment', 'ance', 'ence', 'ity']\nconst DEFAULT_ALLOWED_WORDS = [\n  'accessibility',\n  'availability',\n  'capacity',\n  'community',\n  'identity',\n  'opportunity',\n  'privacy',\n  'quality',\n  'reliability',\n  'security',\n]\n\nexport const noNominalizedPhrases: Rule<NoNominalizedPhrasesOptions> = {\n  id: 'no-nominalized-phrases',\n  description: 'Flag nominalized \"X of Y\" phrases that hide the action in a noun',\n  defaults: { suffixes: DEFAULT_SUFFIXES, allowedWords: DEFAULT_ALLOWED_WORDS },\n  help: 'Nominalized phrases bury the action. Rewrite the phrase with a verb so the sentence says who does what.',\n\n  check({ text, sourceMap, options }: RuleInput<NoNominalizedPhrasesOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const suffixes = options.suffixes?.length ? options.suffixes : DEFAULT_SUFFIXES\n    const allowedWords = new Set((options.allowedWords?.length ? options.allowedWords : DEFAULT_ALLOWED_WORDS).map(value => value.toLowerCase()))\n    const suffixPattern = suffixes.map(escapeRegex).join('|')\n    if (!suffixPattern) return diagnostics\n\n    const doc = createDoc(text)\n    const matches = doc.match(`/[a-z]+(${suffixPattern})/ of .`)\n\n    for (const occurrence of getMatchOccurrences(text, matches)) {\n      const nominalization = occurrence.text.trim().split(/\\s+/)[0]?.toLowerCase()\n      if (!nominalization || allowedWords.has(nominalization)) continue\n\n      const range = getOccurrenceRange(sourceMap, occurrence)\n      if (!range) continue\n\n      diagnostics.push({\n        ruleId: 'no-nominalized-phrases',\n        severity: 'warn',\n        message: `rewrite \"${occurrence.text}\" with a verb`,\n        range,\n        help: noNominalizedPhrases.help,\n      })\n    }\n\n    return diagnostics\n  },\n}\n\nfunction escapeRegex(value: string): string {\n  return value.replace(/[.*+?^${}()|[\\]\\\\]/g, '\\\\$&')\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences } from './utils.js'\n\nexport interface NoPassiveVoiceOptions {\n  allowedAuxiliaries?: string[]\n}\n\nconst DEFAULT_ALLOWED_AUXILIARIES = ['is', 'are', 'was', 'were', 'be', 'been', 'being']\n\nexport const noPassiveVoice: Rule<NoPassiveVoiceOptions> = {\n  id: 'no-passive-voice',\n  description: 'Flag likely passive-voice constructions using POS tagging patterns',\n  defaults: { allowedAuxiliaries: DEFAULT_ALLOWED_AUXILIARIES },\n  help: 'Passive voice often hides the actor and adds drag. Prefer naming who did the action unless the actor genuinely does not matter.',\n\n  check({ text, sourceMap, options }: RuleInput<NoPassiveVoiceOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const auxiliaries = new Set((options.allowedAuxiliaries?.length ? options.allowedAuxiliaries : DEFAULT_ALLOWED_AUXILIARIES).map(value => value.toLowerCase()))\n    const doc = createDoc(text)\n    const matches = doc.match('(#Copula|#Auxiliary) #PastTense')\n\n    for (const occurrence of getMatchOccurrences(text, matches)) {\n      const words = occurrence.text.trim().split(/\\s+/)\n      if (words.length < 2) continue\n      if (!auxiliaries.has(words[0]!.toLowerCase())) continue\n\n      const start = sourceMap[occurrence.start]\n      const end = sourceMap[occurrence.end - 1]\n      if (start === undefined || end === undefined) continue\n\n      diagnostics.push({\n        ruleId: 'no-passive-voice',\n        severity: 'warn',\n        message: `rewrite passive construction \"${occurrence.text}\" with a named actor`,\n        range: { start, end: end + 1 },\n        help: noPassiveVoice.help,\n      })\n    }\n\n    return dedupeDiagnostics(diagnostics)\n  },\n}\n\nfunction dedupeDiagnostics(diagnostics: Diagnostic[]): Diagnostic[] {\n  const seen = new Set<string>()\n  return diagnostics.filter((diagnostic) => {\n    const key = `${diagnostic.range.start}:${diagnostic.range.end}`\n    if (seen.has(key)) return false\n    seen.add(key)\n    return true\n  })\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences, getOccurrenceRange } from './utils.js'\n\nexport interface NoStackedAdjectivesOptions {\n  allowedPhrases?: string[]\n}\n\nexport const noStackedAdjectives: Rule<NoStackedAdjectivesOptions> = {\n  id: 'no-stacked-adjectives',\n  description: 'Flag noun phrases with multiple adjectives before the noun',\n  defaults: { allowedPhrases: [] },\n  help: 'Stacked adjectives make copy feel generic. Keep the one descriptor that earns its place or replace the phrase with concrete evidence.',\n\n  check({ text, sourceMap, options }: RuleInput<NoStackedAdjectivesOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const allowedPhrases = new Set((options.allowedPhrases ?? []).map(value => value.toLowerCase()))\n    const doc = createDoc(text)\n    const matches = doc.match('#Adjective #Adjective+ #Noun')\n\n    for (const occurrence of getMatchOccurrences(text, matches)) {\n      if (allowedPhrases.has(occurrence.text.toLowerCase())) continue\n\n      const range = getOccurrenceRange(sourceMap, occurrence)\n      if (!range) continue\n\n      diagnostics.push({\n        ruleId: 'no-stacked-adjectives',\n        severity: 'warn',\n        message: `cut stacked adjectives in \"${occurrence.text}\"`,\n        range,\n        help: noStackedAdjectives.help,\n      })\n    }\n\n    return diagnostics\n  },\n}\n","import type { Diagnostic, Rule, RuleInput } from '@faircopy/core'\nimport { createDoc, getMatchOccurrences, getOccurrenceRange } from './utils.js'\n\nexport interface NoWeakModalsOptions {\n  modals?: string[]\n  verbs?: string[]\n}\n\nconst DEFAULT_MODALS = ['can', 'could', 'may', 'might']\nconst DEFAULT_VERBS = [\n  'boost',\n  'drive',\n  'enable',\n  'help',\n  'improve',\n  'increase',\n  'make',\n  'reduce',\n  'support',\n  'transform',\n  'unlock',\n]\n\nexport const noWeakModals: Rule<NoWeakModalsOptions> = {\n  id: 'no-weak-modals',\n  description: 'Flag hedged modal claims like \"can help\" and \"might improve\"',\n  defaults: { modals: DEFAULT_MODALS, verbs: DEFAULT_VERBS },\n  help: 'Hedged modal claims sound tentative. Replace them with the outcome, capability, or proof you can stand behind.',\n\n  check({ text, sourceMap, options }: RuleInput<NoWeakModalsOptions>): Diagnostic[] {\n    const diagnostics: Diagnostic[] = []\n    const modals = new Set((options.modals?.length ? options.modals : DEFAULT_MODALS).map(value => value.toLowerCase()))\n    const verbs = new Set((options.verbs?.length ? options.verbs : DEFAULT_VERBS).map(value => value.toLowerCase()))\n    const doc = createDoc(text)\n    const matches = doc.match('#Modal #Adverb? #Verb')\n\n    for (const occurrence of getMatchOccurrences(text, matches)) {\n      const words = occurrence.text.trim().split(/\\s+/)\n      const modal = words[0]?.toLowerCase()\n      const verb = words[words.length - 1]?.toLowerCase()\n      if (!modal || !verb || !modals.has(modal) || !verbs.has(verb)) continue\n\n      const range = getOccurrenceRange(sourceMap, occurrence)\n      if (!range) continue\n\n      diagnostics.push({\n        ruleId: 'no-weak-modals',\n        severity: 'warn',\n        message: `replace \"${occurrence.text.toLowerCase()}\" with a direct claim`,\n        range,\n        help: noWeakModals.help,\n      })\n    }\n\n    return diagnostics\n  },\n}\n","import type { Rule } from '@faircopy/core'\nimport { noFilterWords } from './no-filter-words.js'\nimport { noNominalizedPhrases } from './no-nominalized-phrases.js'\nimport { noPassiveVoice } from './no-passive-voice.js'\nimport { noStackedAdjectives } from './no-stacked-adjectives.js'\nimport { noWeakModals } from './no-weak-modals.js'\n\nexport { noFilterWords } from './no-filter-words.js'\nexport { noNominalizedPhrases } from './no-nominalized-phrases.js'\nexport { noPassiveVoice } from './no-passive-voice.js'\nexport { noStackedAdjectives } from './no-stacked-adjectives.js'\nexport { noWeakModals } from './no-weak-modals.js'\nexport type { NoFilterWordsOptions } from './no-filter-words.js'\nexport type { NoNominalizedPhrasesOptions } from './no-nominalized-phrases.js'\nexport type { NoPassiveVoiceOptions } from './no-passive-voice.js'\nexport type { NoStackedAdjectivesOptions } from './no-stacked-adjectives.js'\nexport type { NoWeakModalsOptions } from './no-weak-modals.js'\n\n/** All NLP rules keyed by their rule ID. */\nexport const ruleRegistry: Map<string, Rule> = new Map([\n  ['no-filter-words', noFilterWords as Rule],\n  ['no-nominalized-phrases', noNominalizedPhrases as Rule],\n  ['no-passive-voice', noPassiveVoice as Rule],\n  ['no-stacked-adjectives', noStackedAdjectives as Rule],\n  ['no-weak-modals', noWeakModals as Rule],\n])\n"],"mappings":";AAAA,OAAO,SAAS;AAUT,SAAS,UAAU,MAAuB;AAC/C,SAAO,IAAI,IAAI;AACjB;AAEO,SAAS,oBAAoB,MAAc,SAAuC;AACvF,QAAM,OAAO,QAAQ,KAAK,EAAE,QAAQ,MAAM,MAAM,MAAM,OAAO,EAAE,QAAQ,KAAK,EAAE,CAAC;AAE/E,SAAO,KAAK,QAAQ,CAAC,UAAU;AAC7B,UAAM,QAAQ,MAAM,QAAQ,SAAS,MAAM,QAAQ,CAAC,GAAG,QAAQ;AAC/D,UAAM,SAAS,MAAM,QAAQ,UAAU,eAAe,MAAM,KAAK;AAEjE,QAAI,OAAO,UAAU,YAAY,OAAO,WAAW,YAAY,UAAU,GAAG;AAC1E,aAAO,CAAC;AAAA,IACV;AAEA,WAAO,CAAC;AAAA,MACN,MAAM,MAAM,QAAQ,KAAK,MAAM,OAAO,QAAQ,MAAM;AAAA,MACpD;AAAA,MACA,KAAK,QAAQ;AAAA,IACf,CAAC;AAAA,EACH,CAAC;AACH;AAEO,SAAS,mBAAmB,WAAqB,YAAyD;AAC/G,QAAM,QAAQ,UAAU,WAAW,KAAK;AACxC,QAAM,MAAM,UAAU,WAAW,MAAM,CAAC;AACxC,MAAI,UAAU,UAAa,QAAQ,OAAW,QAAO;AAErD,SAAO,EAAE,OAAO,KAAK,MAAM,EAAE;AAC/B;AAEA,SAAS,eAAe,OAAyD;AAC/E,MAAI,CAAC,OAAO,OAAQ,QAAO;AAE3B,MAAI,QAAQ;AACZ,aAAW,QAAQ,OAAO;AACxB,UAAM,SAAS,KAAK,QAAQ;AAC5B,QAAI,OAAO,WAAW,SAAU,QAAO;AACvC,aAAS;AAAA,EACX;AAEA,SAAO;AACT;;;AC7CA,IAAM,kBAAkB;AAAA,EACtB;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAEO,IAAM,gBAA4C;AAAA,EACvD,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,SAAS,gBAAgB;AAAA,EACrC,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAkD;AACjF,UAAM,cAA4B,CAAC;AACnC,UAAM,UAAU,QAAQ,SAAS,SAAS,QAAQ,UAAU;AAC5D,UAAM,MAAM,UAAU,IAAI;AAE1B,eAAW,UAAU,SAAS;AAC5B,YAAM,UAAU,IAAI,MAAM,MAAM;AAChC,iBAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,cAAM,QAAQ,UAAU,WAAW,KAAK;AACxC,cAAM,MAAM,UAAU,WAAW,MAAM,CAAC;AACxC,YAAI,UAAU,UAAa,QAAQ,OAAW;AAE9C,oBAAY,KAAK;AAAA,UACf,QAAQ;AAAA,UACR,UAAU;AAAA,UACV,SAAS,WAAW,WAAW,KAAK,YAAY,CAAC;AAAA,UACjD,OAAO,EAAE,OAAO,KAAK,MAAM,EAAE;AAAA,UAC7B,MAAM,cAAc;AAAA,QACtB,CAAC;AAAA,MACH;AAAA,IACF;AAEA,WAAO;AAAA,EACT;AACF;;;ACpCA,IAAM,mBAAmB,CAAC,QAAQ,QAAQ,QAAQ,QAAQ,QAAQ,KAAK;AACvE,IAAM,wBAAwB;AAAA,EAC5B;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAEO,IAAM,uBAA0D;AAAA,EACrE,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,UAAU,kBAAkB,cAAc,sBAAsB;AAAA,EAC5E,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAyD;AACxF,UAAM,cAA4B,CAAC;AACnC,UAAM,WAAW,QAAQ,UAAU,SAAS,QAAQ,WAAW;AAC/D,UAAM,eAAe,IAAI,KAAK,QAAQ,cAAc,SAAS,QAAQ,eAAe,uBAAuB,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AAC5I,UAAM,gBAAgB,SAAS,IAAI,WAAW,EAAE,KAAK,GAAG;AACxD,QAAI,CAAC,cAAe,QAAO;AAE3B,UAAM,MAAM,UAAU,IAAI;AAC1B,UAAM,UAAU,IAAI,MAAM,WAAW,aAAa,SAAS;AAE3D,eAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,YAAM,iBAAiB,WAAW,KAAK,KAAK,EAAE,MAAM,KAAK,EAAE,CAAC,GAAG,YAAY;AAC3E,UAAI,CAAC,kBAAkB,aAAa,IAAI,cAAc,EAAG;AAEzD,YAAM,QAAQ,mBAAmB,WAAW,UAAU;AACtD,UAAI,CAAC,MAAO;AAEZ,kBAAY,KAAK;AAAA,QACf,QAAQ;AAAA,QACR,UAAU;AAAA,QACV,SAAS,YAAY,WAAW,IAAI;AAAA,QACpC;AAAA,QACA,MAAM,qBAAqB;AAAA,MAC7B,CAAC;AAAA,IACH;AAEA,WAAO;AAAA,EACT;AACF;AAEA,SAAS,YAAY,OAAuB;AAC1C,SAAO,MAAM,QAAQ,uBAAuB,MAAM;AACpD;;;ACrDA,IAAM,8BAA8B,CAAC,MAAM,OAAO,OAAO,QAAQ,MAAM,QAAQ,OAAO;AAE/E,IAAM,iBAA8C;AAAA,EACzD,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,oBAAoB,4BAA4B;AAAA,EAC5D,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAmD;AAClF,UAAM,cAA4B,CAAC;AACnC,UAAM,cAAc,IAAI,KAAK,QAAQ,oBAAoB,SAAS,QAAQ,qBAAqB,6BAA6B,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AAC7J,UAAM,MAAM,UAAU,IAAI;AAC1B,UAAM,UAAU,IAAI,MAAM,iCAAiC;AAE3D,eAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,YAAM,QAAQ,WAAW,KAAK,KAAK,EAAE,MAAM,KAAK;AAChD,UAAI,MAAM,SAAS,EAAG;AACtB,UAAI,CAAC,YAAY,IAAI,MAAM,CAAC,EAAG,YAAY,CAAC,EAAG;AAE/C,YAAM,QAAQ,UAAU,WAAW,KAAK;AACxC,YAAM,MAAM,UAAU,WAAW,MAAM,CAAC;AACxC,UAAI,UAAU,UAAa,QAAQ,OAAW;AAE9C,kBAAY,KAAK;AAAA,QACf,QAAQ;AAAA,QACR,UAAU;AAAA,QACV,SAAS,iCAAiC,WAAW,IAAI;AAAA,QACzD,OAAO,EAAE,OAAO,KAAK,MAAM,EAAE;AAAA,QAC7B,MAAM,eAAe;AAAA,MACvB,CAAC;AAAA,IACH;AAEA,WAAO,kBAAkB,WAAW;AAAA,EACtC;AACF;AAEA,SAAS,kBAAkB,aAAyC;AAClE,QAAM,OAAO,oBAAI,IAAY;AAC7B,SAAO,YAAY,OAAO,CAAC,eAAe;AACxC,UAAM,MAAM,GAAG,WAAW,MAAM,KAAK,IAAI,WAAW,MAAM,GAAG;AAC7D,QAAI,KAAK,IAAI,GAAG,EAAG,QAAO;AAC1B,SAAK,IAAI,GAAG;AACZ,WAAO;AAAA,EACT,CAAC;AACH;;;AC5CO,IAAM,sBAAwD;AAAA,EACnE,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,gBAAgB,CAAC,EAAE;AAAA,EAC/B,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAwD;AACvF,UAAM,cAA4B,CAAC;AACnC,UAAM,iBAAiB,IAAI,KAAK,QAAQ,kBAAkB,CAAC,GAAG,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AAC/F,UAAM,MAAM,UAAU,IAAI;AAC1B,UAAM,UAAU,IAAI,MAAM,8BAA8B;AAExD,eAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,UAAI,eAAe,IAAI,WAAW,KAAK,YAAY,CAAC,EAAG;AAEvD,YAAM,QAAQ,mBAAmB,WAAW,UAAU;AACtD,UAAI,CAAC,MAAO;AAEZ,kBAAY,KAAK;AAAA,QACf,QAAQ;AAAA,QACR,UAAU;AAAA,QACV,SAAS,8BAA8B,WAAW,IAAI;AAAA,QACtD;AAAA,QACA,MAAM,oBAAoB;AAAA,MAC5B,CAAC;AAAA,IACH;AAEA,WAAO;AAAA,EACT;AACF;;;AC5BA,IAAM,iBAAiB,CAAC,OAAO,SAAS,OAAO,OAAO;AACtD,IAAM,gBAAgB;AAAA,EACpB;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAEO,IAAM,eAA0C;AAAA,EACrD,IAAI;AAAA,EACJ,aAAa;AAAA,EACb,UAAU,EAAE,QAAQ,gBAAgB,OAAO,cAAc;AAAA,EACzD,MAAM;AAAA,EAEN,MAAM,EAAE,MAAM,WAAW,QAAQ,GAAiD;AAChF,UAAM,cAA4B,CAAC;AACnC,UAAM,SAAS,IAAI,KAAK,QAAQ,QAAQ,SAAS,QAAQ,SAAS,gBAAgB,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AACnH,UAAM,QAAQ,IAAI,KAAK,QAAQ,OAAO,SAAS,QAAQ,QAAQ,eAAe,IAAI,WAAS,MAAM,YAAY,CAAC,CAAC;AAC/G,UAAM,MAAM,UAAU,IAAI;AAC1B,UAAM,UAAU,IAAI,MAAM,uBAAuB;AAEjD,eAAW,cAAc,oBAAoB,MAAM,OAAO,GAAG;AAC3D,YAAM,QAAQ,WAAW,KAAK,KAAK,EAAE,MAAM,KAAK;AAChD,YAAM,QAAQ,MAAM,CAAC,GAAG,YAAY;AACpC,YAAM,OAAO,MAAM,MAAM,SAAS,CAAC,GAAG,YAAY;AAClD,UAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,OAAO,IAAI,KAAK,KAAK,CAAC,MAAM,IAAI,IAAI,EAAG;AAE/D,YAAM,QAAQ,mBAAmB,WAAW,UAAU;AACtD,UAAI,CAAC,MAAO;AAEZ,kBAAY,KAAK;AAAA,QACf,QAAQ;AAAA,QACR,UAAU;AAAA,QACV,SAAS,YAAY,WAAW,KAAK,YAAY,CAAC;AAAA,QAClD;AAAA,QACA,MAAM,aAAa;AAAA,MACrB,CAAC;AAAA,IACH;AAEA,WAAO;AAAA,EACT;AACF;;;ACrCO,IAAM,eAAkC,oBAAI,IAAI;AAAA,EACrD,CAAC,mBAAmB,aAAqB;AAAA,EACzC,CAAC,0BAA0B,oBAA4B;AAAA,EACvD,CAAC,oBAAoB,cAAsB;AAAA,EAC3C,CAAC,yBAAyB,mBAA2B;AAAA,EACrD,CAAC,kBAAkB,YAAoB;AACzC,CAAC;","names":[]}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@faircopy/rules-nlp",
-  "version": "1.2.1",
+  "version": "1.4.0",
   "description": "Optional NLP-powered ruleset for faircopy using compromise",
   "type": "module",
   "exports": {
@@ -15,11 +15,11 @@
   "scripts": {
     "build": "tsup",
     "typecheck": "tsc --noEmit",
-    "test": "echo 'no tests yet'",
+    "test": "pnpm run build && node --test test/*.test.mjs",
     "prepublishOnly": "pnpm run build"
   },
   "dependencies": {
-    "@faircopy/core": "1.2.1",
+    "@faircopy/core": "1.4.0",
     "compromise": "^14.15.0"
   },
   "devDependencies": {