@inshapardaz/likhari-converters 0.1.45-dev.334 → 0.1.45-dev.335

package/dist/merge/index.d.ts

@@ -11,6 +11,8 @@ export interface MergePagesOptions {
     inputFormat?: FormatId;
     /** The format of the combined document. */
     outputFormat: FormatId;
+    /** Leave out pages with no content: nothing, whitespace, or only empty paragraphs. Off by default. */
+    skipBlankPages?: boolean;
     /** Passed to every parse and to the final serialize (image URL resolver and the like). */
     context?: ConverterContext;
     /** Defaults to the registry with all four built-in formats. */
@@ -23,6 +25,8 @@ export declare function mergePagesToState(pages: PageInput[], options?: Omit<Mer
  * with a page break between each pair of pages, in the requested output format. Needs no
  * editor or DOM, so it can run in a server, a worker or a build script.
  *
+ * With `skipBlankPages`, blank pages are left out and no break is added for them.
+ *
  * Footnote ids are only unique within a page, so a page whose footnote ids clash with an
  * earlier page's gets them renamed (`1` becomes `1-2`), in references and notes alike; the
  * notes of all pages end up in one list at the end of the document.

package/dist/merge/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/merge/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,qBAAqB,EAAE,MAAM,SAAS,CAAC;AACrD,OAAO,EAAyB,KAAK,cAAc,EAAE,MAAM,aAAa,CAAC;AACzE,OAAO,KAAK,EAAE,gBAAgB,EAAE,QAAQ,EAAE,MAAM,UAAU,CAAC;AAG3D,kGAAkG;AAClG,MAAM,MAAM,SAAS,GAAG,MAAM,GAAG;IAAE,OAAO,EAAE,MAAM,CAAC;IAAC,MAAM,CAAC,EAAE,QAAQ,CAAA;CAAE,CAAC;AAExE,MAAM,WAAW,iBAAiB;IAChC,+EAA+E;IAC/E,WAAW,CAAC,EAAE,QAAQ,CAAC;IACvB,2CAA2C;IAC3C,YAAY,EAAE,QAAQ,CAAC;IACvB,0FAA0F;IAC1F,OAAO,CAAC,EAAE,gBAAgB,CAAC;IAC3B,+DAA+D;IAC/D,QAAQ,CAAC,EAAE,cAAc,CAAC;CAC3B;AAED,mGAAmG;AACnG,wBAAgB,iBAAiB,CAAC,KAAK,EAAE,SAAS,EAAE,EAAE,OAAO,GAAE,IAAI,CAAC,iBAAiB,EAAE,cAAc,CAAM,GAAG,qBAAqB,CA4BlI;AAED;;;;;;;;GAQG;AACH,wBAAgB,UAAU,CAAC,KAAK,EAAE,SAAS,EAAE,EAAE,OAAO,EAAE,iBAAiB,GAAG,MAAM,CAGjF"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/merge/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,qBAAqB,EAAE,MAAM,SAAS,CAAC;AACrD,OAAO,EAAyB,KAAK,cAAc,EAAE,MAAM,aAAa,CAAC;AACzE,OAAO,KAAK,EAAE,gBAAgB,EAAE,QAAQ,EAAE,MAAM,UAAU,CAAC;AAG3D,kGAAkG;AAClG,MAAM,MAAM,SAAS,GAAG,MAAM,GAAG;IAAE,OAAO,EAAE,MAAM,CAAC;IAAC,MAAM,CAAC,EAAE,QAAQ,CAAA;CAAE,CAAC;AAExE,MAAM,WAAW,iBAAiB;IAChC,+EAA+E;IAC/E,WAAW,CAAC,EAAE,QAAQ,CAAC;IACvB,2CAA2C;IAC3C,YAAY,EAAE,QAAQ,CAAC;IACvB,sGAAsG;IACtG,cAAc,CAAC,EAAE,OAAO,CAAC;IACzB,0FAA0F;IAC1F,OAAO,CAAC,EAAE,gBAAgB,CAAC;IAC3B,+DAA+D;IAC/D,QAAQ,CAAC,EAAE,cAAc,CAAC;CAC3B;AAED,mGAAmG;AACnG,wBAAgB,iBAAiB,CAAC,KAAK,EAAE,SAAS,EAAE,EAAE,OAAO,GAAE,IAAI,CAAC,iBAAiB,EAAE,cAAc,CAAM,GAAG,qBAAqB,CAgClI;AAED;;;;;;;;;;GAUG;AACH,wBAAgB,UAAU,CAAC,KAAK,EAAE,SAAS,EAAE,EAAE,OAAO,EAAE,iBAAiB,GAAG,MAAM,CAGjF"}

package/dist/merge/index.js

@@ -2,10 +2,11 @@ import { defaultFormatRegistry } from '../registry.js';
 import { elementBase, makeState, rootChildren } from '../shared/serialized.js';
 /** Parses each page and returns the combined document as Lexical state, without serializing it. */
 export function mergePagesToState(pages, options = {}) {
-    const { inputFormat = 'markdown', context, registry = defaultFormatRegistry } = options;
+    const { inputFormat = 'markdown', skipBlankPages = false, context, registry = defaultFormatRegistry } = options;
     const body = [];
     const footnotes = [];
     const takenFootnoteIds = new Set();
+    let kept = 0;
     pages.forEach((page, index) => {
         const { content, format = inputFormat } = typeof page === 'string' ? { content: page } : page;
         let blocks;
@@ -16,14 +17,17 @@ export function mergePagesToState(pages, options = {}) {
             const reason = error instanceof Error ? error.message : String(error);
             throw new Error(`mergePages: page ${index + 1} is not valid ${format}: ${reason}`);
         }
+        if (skipBlankPages && blocks.every(isBlank))
+            return;
         // Each page brings its own footnote list; they are combined into one at the end.
         const pageLists = blocks.filter((block) => block.type === 'footnote-list');
         const pageBody = blocks.filter((block) => block.type !== 'footnote-list');
         uniquifyFootnoteIds([...pageBody, ...pageLists], takenFootnoteIds);
         for (const list of pageLists)
             footnotes.push(...(list.children ?? []));
-        if (index > 0)
+        if (kept > 0)
             body.push({ type: 'page-break', version: 1 });
+        kept += 1;
         body.push(...pageBody);
     });
     return makeState(footnotes.length > 0 ? [...body, elementBase('footnote-list', footnotes)] : body);
@@ -33,6 +37,8 @@ export function mergePagesToState(pages, options = {}) {
  * with a page break between each pair of pages, in the requested output format. Needs no
  * editor or DOM, so it can run in a server, a worker or a build script.
  *
+ * With `skipBlankPages`, blank pages are left out and no break is added for them.
+ *
  * Footnote ids are only unique within a page, so a page whose footnote ids clash with an
  * earlier page's gets them renamed (`1` becomes `1-2`), in references and notes alike; the
  * notes of all pages end up in one list at the end of the document.
@@ -41,6 +47,12 @@ export function mergePages(pages, options) {
     const registry = options.registry ?? defaultFormatRegistry;
     return registry.serialize(options.outputFormat, mergePagesToState(pages, options), options.context);
 }
+/** A block with nothing in it: a paragraph holding only whitespace and line breaks. Anything else (an image, a rule, a table, a break) is content. */
+function isBlank(block) {
+    if (block.type !== 'paragraph')
+        return false;
+    return (block.children ?? []).every((child) => (child.type === 'text' ? String(child.text ?? '').trim() === '' : child.type === 'linebreak'));
+}
 /** Renames this page's footnote ids that an earlier page already uses, then claims all its ids. */
 function uniquifyFootnoteIds(blocks, taken) {
     const nodes = [];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@inshapardaz/likhari-converters",
-  "version": "0.1.45-dev.334",
+  "version": "0.1.45-dev.335",
   "description": "Format transformers (Markdown/HTML/plain-text/Lexical JSON) against the canonical Lexical EditorState, isolated from UI.",
   "license": "MIT",
   "repository": {