The parsing of MOBI and KF8 files is based on this MOBI parsing implementation.

Introduction

The MOBI file format is based on the PDB file format, while the KF8 (Kindle Format 8) format is built on top of MOBI. Therefore, the parsing of both MOBI and KF8 is handled together in the mobi-parser.

Install

pnpm install @lingo-reader/mobi-parser

MobiFile

Usage in browser

import { initMobiFile } from '@lingo-reader/mobi-parser'
import type { MobiSpine } from '@lingo-reader/mobi-parser'

function initMobi(file: File) {
  const mobi: Mobi = await initMobiFile(file)
  // spine
  const spine: MobiSpine = mobi.getSpine()
  // loadChapter
  const firstChapter = mobi.loadChapter(spine[0].id)
}

// see Mobi class
// .....

Usage in node

import { initMobiFile } from '@lingo-reader/mobi-parser'
import type { MobiSpine } from '@lingo-reader/mobi-parser'

const mobi: Mobi = await initMobiFile('./example/taoyong.mobi')
// spine
const spine: MobiSpine = mobi.getSpine()
// loadChapter
const firstChapter = mobi.loadChapter(spine[0].id)

// see Mobi class
// .....

initMobiFile(file: string | File |Uint8Array, resourceSaveDir?: string): Promise<Mobi>

API for initializing a MOBI file. You can pass in a file path, a File object, or a Uint8Array to receive a Mobi instance, which provides APIs for reading metadata, spine information, and more.

Parameters:

file: string | File | Uint8Array – The MOBI file input, which can be a file path (Node.js), a File object (browser), or a Uint8Array in either environment.
resourceSaveDir?: string – (Optional) Applicable mainly in Node.js. Specifies the directory to save images and other resources. Defaults to ./images.

Returns:

Promise – A Promise that resolves to the initialized Mobi object.

Note: For the file parameter:

In the browser, it must be a File or Uint8Array. Passing a string will throw an error.
In Node.js, it must be a string or Uint8Array. Passing a File will throw an error.

Mobi class

class Mobi {
  getFileInfo(): MobiFileInfo
  getSpine(): MobiSpine
  loadChapter(id: string): MobiProcessedChapter | undefined
  getToc(): MobiToc
  getCoverImage(): string
  getMetadata(): MobiMetadata
  resolveHref(href: string): MobiResolvedHref | undefined
  destroy(): void
}

getFileInfo(): MobiFileInfo

interface MobiFileInfo {
  // MOBI file name, including the file extension
  fileName: string
}

This returns file information, currently with only the fileName field.

getSpine(): MobiSpine

interface MobiChapter {
  // Chapter ID
  id: string
  // Raw HTML text with unmodified resource paths
  text: string
  // The byte position where the chapter starts in the file
  start: number
  // The byte position where the chapter ends in the file
  end: number
  // The byte size of the chapter
  size: number
}
type MobiSpine = MobiChapter[]

The spine lists all the chapters to be displayed in sequence.

In the original HTML text of the chapters, image resource paths point to internal file addresses. These need to be replaced with blob URLs in order to display them directly in the browser.

loadChapter(id: string): MobiProcessedChapter | undefined

interface MobiCssPart {
  id: string
  href: string
}
interface MobiProcessedChapter {
  html: string
  css: MobiCssPart[]
}

The loadChapter function takes a chapter ID as its argument and returns the processed chapter object. If the return value is undefined, it means the chapter does not exist.

The raw HTML is split into CSS and text (MobiProcessedChapter), where the CSS is provided as a blob URL (MobiCssPart), and the HTML is the modified text after resource replacement, mainly in the <body> tag content.

getToc(): MobiToc

interface MobiTocItem {
  // TOC item label
  label: string
  // Internal MOBI href
  href: string
  // Sub TOC item
  children?: MobiTocItem[]
}
type MobiToc = MobiTocItem[]

This function retrieves the table of contents (TOC) of the book.

getCoverImage(): string

This retrieves the cover image of the book as a blob URL. When an empty string is returned, it means that the CoverImage does not exist.

getMetadata(): MobiMetadata

interface MobiMetadata {
  // Unique identifier
  identifier: string
  // Book title
  title: string
  // Author(s)
  author: string[]
  // Publisher
  publisher: string
  // Language of the book
  language: string
  // Publication date
  published: string
  // Description of the book
  description: string
  // Categories or subjects
  subject: string[]
  // Copyright information
  rights: string
  // Contributors
  contributor: string[]
}

This retrieves the metadata for the book.

resolveHref(href: string): MobiResolvedHref | undefined

interface MobiResolvedHref {
  // Chapter ID
  id: string
  // HTML selector
  selector: string
}

This function parses links that point to other chapters and returns the result in the format shown above.

destroy(): void

This function clears blob URLs and resources created during the file parsing process to prevent memory leaks.

Kf8File

Usage in browser

import { initKf8File } from '@lingo-reader/mobi-parser'
import type { Kf8Spine } from '@lingo-reader/mobi-parser'

function initMobi(file: File) {
  const mobi: Kf8 = await initKf8File(file)
  // spine
  const spine: Kf8Spine = mobi.getSpine()
  // loadChapter
  const firstChapter: Kf8ProcessedChapter = mobi.loadChapter(spine[0].id)
}

// see Kf8 class
// .....

Usage in node

import { initKf8File } from '@lingo-reader/mobi-parser'
import type { Kf8Spine } from '@lingo-reader/mobi-parser'

const mobi: Kf8 = await initKf8File('./example/taoyong.azw3')
// spine
const spine: Kf8Spine = mobi.getSpine()
// loadChapter
const firstChapter: Kf8ProcessedChapter = mobi.loadChapter(spine[0].id)

// see Kf8 class
// .....

initKf8File(file: InputFile, resourceSaveDir?: string): Promise<Kf8>

API for initializing an AZW3 (KF8) file. You can pass in a file path, a File object, or a Uint8Array to receive a fully initialized Kf8 object, which provides APIs for reading metadata, spine information, and more.

Parameters:

file: string | File | Uint8Array – The input KF8 file. Can be a file path (Node.js), a File object (browser), or a Uint8Array in either environment.
resourceSaveDir?: string – (Optional) Mainly used in Node.js. Specifies the directory to save images and other resources. Defaults to ./images.

Returns:

Promise – A Promise that resolves to the initialized Kf8 object.

Note: For the file parameter:

In the browser, the type must be File or Uint8Array. Passing a string will throw an error.
In Node.js, the type must be string or Uint8Array. Passing a File will throw an error.

Kf8 class

class Kf8 {
  getFileInfo(): Kf8FileInfo
  getMetadata(): Kf8Metadata
  getCoverImage(): string
  getSpine(): Kf8Spine
  getToc(): Kf8Toc
  getGuide(): Kf8Guide | undefined
  loadChapter(id: string): Kf8ProcessedChapter | undefined
  resolveHref(href: string): Kf8ResolvedHref | undefined
  destroy(): void
}

getFileInfo(): Kf8FileInfo

type Kf8FileInfo = MobiFileInfo

This is the same as MobiFileInfo, see here

getMetadata(): Kf8Metadata

type Kf8Metadata = MobiMetadata

This is the same as MobiMetadata, see here

getCoverImage(): string

This retrieves the book's cover image as a blob URL. When an empty string is returned, it means that the CoverImage does not exist.

getSpine(): Kf8Spine

interface Kf8Chapter {
  id: string
  skel: SkelTableItem
  frags: FragTable
  fragEnd: number
  length: number
  totalLength: number
}
type Kf8Spine = Kf8Chapter[]

No explanation available yet.

getToc(): Kf8Toc

interface Kf8TocItem {
  label: string
  // Path to the chapter resource
  href: string
  children?: Kf8TocItem[]
}
type Kf8Toc = Kf8TocItem[]

This retrieves the table of contents (TOC).

getGuide(): Kf8Guide | undefined

interface Kf8GuideItem {
  label: string
  type: string[]
  href: string
}
type Kf8Guide = Kf8GuideItem[]

This retrieves parts of the book that are available for browsing.

loadChapter(id: string): Kf8ProcessedChapter | undefined

interface Kf8CssPart {
  id: string
  href: string
}
interface Kf8ProcessedChapter {
  html: string
  css: Kf8CssPart[]
}

This works similarly to MobiProcessedChapter, see here

resolveHref(href: string): Kf8ResolvedHref | undefined

type Kf8ResolvedHref = MobiResolvedHref

This works similarly to MobiResolvedHref, see here。

destroy(): void

Clears the blob URLs and resources created during the parsing process to prevent memory leaks.

Package detail

@lingo-reader/mobi-parser

readme

Introduction

Install

MobiFile

Usage in browser

Usage in node

initMobiFile(file: string | File |Uint8Array, resourceSaveDir?: string): Promise<Mobi>

Mobi class

getFileInfo(): MobiFileInfo

getSpine(): MobiSpine

loadChapter(id: string): MobiProcessedChapter | undefined

getToc(): MobiToc

getCoverImage(): string

getMetadata(): MobiMetadata

resolveHref(href: string): MobiResolvedHref | undefined

destroy(): void

Kf8File

Usage in browser

Usage in node

initKf8File(file: InputFile, resourceSaveDir?: string): Promise<Kf8>

Kf8 class

getFileInfo(): Kf8FileInfo

getMetadata(): Kf8Metadata

getCoverImage(): string

getSpine(): Kf8Spine

getToc(): Kf8Toc

getGuide(): Kf8Guide | undefined

loadChapter(id: string): Kf8ProcessedChapter | undefined

resolveHref(href: string): Kf8ResolvedHref | undefined

destroy(): void