dravoice 0.1.1 → 0.1.3

package/src/v2/analyzers/rhythm.js

@@ -1,47 +1,39 @@
-import { distribution, rate } from "../text-utils.js";
-
-export function analyzeRhythm(documents) {
-  const sentences = documents.flatMap((document) => document.sentences);
-  const paragraphs = documents.flatMap((document) => document.paragraphs);
-  const blocks = documents.flatMap((document) => document.blocks);
-
-  return {
-    family: "rhythm",
-    confidence: confidenceFor(sentences.length),
-    features: {
-      documentCount: documents.length,
-      sentenceCount: sentences.length,
-      paragraphCount: paragraphs.length,
-      sentenceWords: distribution(sentences.map((sentence) => sentence.tokens.length)),
-      paragraphWords: distribution(paragraphs.map((paragraph) => paragraph.text.split(/\s+/).filter(Boolean).length)),
-      paragraphSentences: distribution(paragraphs.map((paragraph) => countParagraphSentences(sentences, paragraph))),
-      headingDensity: rate(documents.reduce((sum, document) => sum + document.headings.length, 0), sentences.length, 2),
-      listDensity: rate(blocks.filter((block) => block.type === "list").length, paragraphs.length, 2),
-      quoteDensity: rate(blocks.filter((block) => block.type === "quote").length, paragraphs.length, 2),
-    },
-    examples: [
-      `sentenceWords.median: ${distribution(sentences.map((sentence) => sentence.tokens.length)).median}`,
-      `paragraphWords.median: ${distribution(paragraphs.map((paragraph) => paragraph.text.split(/\s+/).filter(Boolean).length)).median}`,
-    ],
-    warnings: sentences.length < 12 ? ["Rhythm confidence is limited because the corpus has fewer than 12 usable sentences."] : [],
-    revisionHandles: ["Compare sentence and paragraph pacing against the learned range."],
-  };
-}
-
-function countParagraphSentences(sentences, paragraph) {
-  return Math.max(1, sentences.filter((sentence) =>
-    sentence.line >= paragraph.line &&
-    sentence.line < paragraph.line + 8 &&
-    sentence.headingId === paragraph.headingId
-  ).length);
-}
-
-function confidenceFor(sentenceCount) {
-  if (sentenceCount >= 30) {
-    return "high";
-  }
-  if (sentenceCount >= 8) {
-    return "medium";
-  }
-  return "low";
-}
+import { distribution, rate, splitSentences } from "../text-utils.js";
+
+export function analyzeRhythm(documents) {
+  const sentences = documents.flatMap((document) => document.sentences);
+  const paragraphs = documents.flatMap((document) => document.paragraphs);
+  const blocks = documents.flatMap((document) => document.blocks);
+
+  return {
+    family: "rhythm",
+    confidence: confidenceFor(sentences.length),
+    features: {
+      documentCount: documents.length,
+      sentenceCount: sentences.length,
+      paragraphCount: paragraphs.length,
+      sentenceWords: distribution(sentences.map((sentence) => sentence.tokens.length)),
+      paragraphWords: distribution(paragraphs.map((paragraph) => paragraph.text.split(/\s+/).filter(Boolean).length)),
+      paragraphSentences: distribution(paragraphs.map((paragraph) => Math.max(1, splitSentences(paragraph.text).length))),
+      headingDensity: rate(documents.reduce((sum, document) => sum + document.headings.length, 0), sentences.length, 2),
+      listDensity: rate(blocks.filter((block) => block.type === "list").length, paragraphs.length, 2),
+      quoteDensity: rate(blocks.filter((block) => block.type === "quote").length, paragraphs.length, 2),
+    },
+    examples: [
+      `sentenceWords.median: ${distribution(sentences.map((sentence) => sentence.tokens.length)).median}`,
+      `paragraphWords.median: ${distribution(paragraphs.map((paragraph) => paragraph.text.split(/\s+/).filter(Boolean).length)).median}`,
+    ],
+    warnings: sentences.length < 12 ? ["Rhythm confidence is limited because the corpus has fewer than 12 usable sentences."] : [],
+    revisionHandles: ["Compare sentence and paragraph pacing against the learned range."],
+  };
+}
+
+function confidenceFor(sentenceCount) {
+  if (sentenceCount >= 30) {
+    return "high";
+  }
+  if (sentenceCount >= 8) {
+    return "medium";
+  }
+  return "low";
+}

package/src/v2/analyzers/structure.js

@@ -1,24 +1,24 @@
-import { distribution, rate } from "../text-utils.js";
-import { moveFor } from "./rhetorical-shape.js";
-
-export function analyzeStructure(documents) {
-  const sectionLengths = documents.flatMap((document) =>
-    document.sections.map((section) => section.blocks.reduce((sum, block) => sum + block.lines.join(" ").split(/\s+/).filter(Boolean).length, 0))
-  );
-  const openingMoves = documents.flatMap((document) => document.sentences.slice(0, 2).map((sentence) => moveFor(sentence.text)));
-
-  return {
-    family: "structure",
-    confidence: documents.length >= 3 ? "medium" : "low",
-    features: {
-      sectionWords: distribution(sectionLengths),
-      headingCount: distribution(documents.map((document) => document.headings.length)),
-      openingMoves,
-      listDocumentRate: rate(documents.filter((document) => document.blocks.some((block) => block.type === "list")).length, documents.length, 2),
-      quoteDocumentRate: rate(documents.filter((document) => document.blocks.some((block) => block.type === "quote")).length, documents.length, 2),
-    },
-    examples: openingMoves.slice(0, 5),
-    warnings: documents.length < 3 ? ["Structure confidence is limited because the corpus has fewer than 3 documents."] : [],
-    revisionHandles: ["Compare headings, list/quote use, section size, and opening structure."],
-  };
-}
+import { distribution, rate } from "../text-utils.js";
+import { moveFor } from "./rhetorical-shape.js";
+
+export function analyzeStructure(documents) {
+  const sectionLengths = documents.flatMap((document) =>
+    document.sections.map((section) => section.blocks.reduce((sum, block) => sum + block.lines.join(" ").split(/\s+/).filter(Boolean).length, 0))
+  );
+  const openingMoves = documents.flatMap((document) => document.sentences.slice(0, 2).map((sentence) => moveFor(sentence.text)));
+
+  return {
+    family: "structure",
+    confidence: documents.length >= 3 ? "medium" : "low",
+    features: {
+      sectionWords: distribution(sectionLengths),
+      headingCount: distribution(documents.map((document) => document.headings.length)),
+      openingMoves,
+      listDocumentRate: rate(documents.filter((document) => document.blocks.some((block) => block.type === "list")).length, documents.length, 2),
+      quoteDocumentRate: rate(documents.filter((document) => document.blocks.some((block) => block.type === "quote")).length, documents.length, 2),
+    },
+    examples: openingMoves.slice(0, 5),
+    warnings: documents.length < 3 ? ["Structure confidence is limited because the corpus has fewer than 3 documents."] : [],
+    revisionHandles: ["Compare headings, list/quote use, section size, and opening structure."],
+  };
+}