html-to-markdown-node 2.6.4 → 2.6.5

package/index.d.ts

@@ -0,0 +1,231 @@
+/* auto-generated by NAPI-RS */
+/* eslint-disable */
+/**
+ * Convert HTML to Markdown
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options
+ *
+ * # Example
+ *
+ * ```javascript
+ * const { convert } = require('html-to-markdown');
+ *
+ * const html = '<h1>Hello World</h1>';
+ * const markdown = convert(html);
+ * console.log(markdown); // # Hello World
+ * ```
+ */
+export declare function convert(html: string, options?: JsConversionOptions | undefined | null): string
+
+/**
+ * Convert HTML to Markdown while collecting inline images
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options
+ * * `image_config` - Configuration for inline image extraction
+ *
+ * # Example
+ *
+ * ```javascript
+ * const { convertWithInlineImages } = require('html-to-markdown');
+ *
+ * const html = '<img src="data:image/png;base64,..." alt="test">';
+ * const result = convertWithInlineImages(html, null, {
+ *   maxDecodedSizeBytes: 1024 * 1024,
+ *   inferDimensions: true
+ * });
+ * console.log(result.markdown);
+ * console.log(result.inlineImages.length);
+ * ```
+ */
+export declare function convertWithInlineImages(html: string, options?: JsConversionOptions | undefined | null, imageConfig?: JsInlineImageConfig | undefined | null): JsHtmlExtraction
+
+/** Code block style */
+export declare const enum JsCodeBlockStyle {
+  /** Indented code blocks (4 spaces) - CommonMark default */
+  Indented = 'Indented',
+  /** Fenced code blocks with backticks (```) */
+  Backticks = 'Backticks',
+  /** Fenced code blocks with tildes (~~~) */
+  Tildes = 'Tildes'
+}
+
+/** Main conversion options */
+export interface JsConversionOptions {
+  /** Heading style */
+  headingStyle?: JsHeadingStyle
+  /** List indentation type */
+  listIndentType?: JsListIndentType
+  /** List indentation width (spaces) */
+  listIndentWidth?: number
+  /** Bullet characters for unordered lists */
+  bullets?: string
+  /** Symbol for strong/emphasis (* or _) */
+  strongEmSymbol?: string
+  /** Escape asterisks in text */
+  escapeAsterisks?: boolean
+  /** Escape underscores in text */
+  escapeUnderscores?: boolean
+  /** Escape misc markdown characters */
+  escapeMisc?: boolean
+  /** Escape all ASCII punctuation */
+  escapeAscii?: boolean
+  /** Default code language */
+  codeLanguage?: string
+  /** Use autolinks for bare URLs */
+  autolinks?: boolean
+  /** Add default title if none exists */
+  defaultTitle?: boolean
+  /** Use <br> in tables instead of spaces */
+  brInTables?: boolean
+  /** Enable spatial table reconstruction in hOCR documents */
+  hocrSpatialTables?: boolean
+  /** Highlight style for <mark> elements */
+  highlightStyle?: JsHighlightStyle
+  /** Extract metadata from HTML */
+  extractMetadata?: boolean
+  /** Whitespace handling mode */
+  whitespaceMode?: JsWhitespaceMode
+  /** Strip newlines from HTML before processing */
+  stripNewlines?: boolean
+  /** Enable text wrapping */
+  wrap?: boolean
+  /** Text wrap width */
+  wrapWidth?: number
+  /** Treat block elements as inline */
+  convertAsInline?: boolean
+  /** Subscript symbol */
+  subSymbol?: string
+  /** Superscript symbol */
+  supSymbol?: string
+  /** Newline style */
+  newlineStyle?: JsNewlineStyle
+  /** Code block style */
+  codeBlockStyle?: JsCodeBlockStyle
+  /** Elements where images should remain as markdown */
+  keepInlineImagesIn?: Array<string>
+  /** Preprocessing options */
+  preprocessing?: JsPreprocessingOptions
+  /** Source encoding (informational) */
+  encoding?: string
+  /** Enable debug mode with diagnostic warnings */
+  debug?: boolean
+  /** List of HTML tags to strip */
+  stripTags?: Array<string>
+  /** List of HTML tags to preserve as-is in the output */
+  preserveTags?: Array<string>
+}
+
+/** Heading style options */
+export declare const enum JsHeadingStyle {
+  /** Underlined style (=== for h1, --- for h2) */
+  Underlined = 'Underlined',
+  /** ATX style (# for h1, ## for h2, etc.) */
+  Atx = 'Atx',
+  /** ATX closed style (# title #) */
+  AtxClosed = 'AtxClosed'
+}
+
+/** Highlight style for `<mark>` elements */
+export declare const enum JsHighlightStyle {
+  /** ==text== */
+  DoubleEqual = 'DoubleEqual',
+  /** <mark>text</mark> */
+  Html = 'Html',
+  /** **text** */
+  Bold = 'Bold',
+  /** Plain text (no formatting) */
+  None = 'None'
+}
+
+/** Result of HTML extraction with inline images */
+export interface JsHtmlExtraction {
+  /** Converted markdown */
+  markdown: string
+  /** Extracted inline images */
+  inlineImages: Array<JsInlineImage>
+  /** Warnings encountered during extraction */
+  warnings: Array<JsInlineImageWarning>
+}
+
+/** Inline image data */
+export interface JsInlineImage {
+  /** Raw image data */
+  data: Buffer
+  /** Image format (png, jpeg, gif, etc.) */
+  format: string
+  /** Generated or provided filename */
+  filename?: string
+  /** Alt text / description */
+  description?: string
+  /** Image dimensions (width, height) if available */
+  dimensions?: Array<number>
+  /** Source type (img_data_uri or svg_element) */
+  source: string
+  /** HTML attributes from the source element */
+  attributes: Record<string, string>
+}
+
+/** Inline image configuration */
+export interface JsInlineImageConfig {
+  /** Maximum decoded size in bytes (default: 5MB) */
+  maxDecodedSizeBytes?: bigint
+  /** Filename prefix for generated filenames */
+  filenamePrefix?: string
+  /** Capture inline SVG elements (default: true) */
+  captureSvg?: boolean
+  /** Infer image dimensions (default: false) */
+  inferDimensions?: boolean
+}
+
+/** Warning about inline image processing */
+export interface JsInlineImageWarning {
+  /** Index of the image that caused the warning */
+  index: number
+  /** Warning message */
+  message: string
+}
+
+/** List indentation type */
+export declare const enum JsListIndentType {
+  Spaces = 'Spaces',
+  Tabs = 'Tabs'
+}
+
+/** Newline style */
+export declare const enum JsNewlineStyle {
+  /** Two spaces at end of line */
+  Spaces = 'Spaces',
+  /** Backslash at end of line */
+  Backslash = 'Backslash'
+}
+
+/** HTML preprocessing options */
+export interface JsPreprocessingOptions {
+  /** Enable preprocessing */
+  enabled?: boolean
+  /** Preprocessing preset */
+  preset?: JsPreprocessingPreset
+  /** Remove navigation elements */
+  removeNavigation?: boolean
+  /** Remove form elements */
+  removeForms?: boolean
+}
+
+/** Preprocessing preset levels */
+export declare const enum JsPreprocessingPreset {
+  Minimal = 'Minimal',
+  Standard = 'Standard',
+  Aggressive = 'Aggressive'
+}
+
+/** Whitespace handling mode */
+export declare const enum JsWhitespaceMode {
+  Normalized = 'Normalized',
+  Strict = 'Strict'
+}

package/index.js

@@ -0,0 +1,255 @@
+const { existsSync, readFileSync } = require('fs')
+const { join } = require('path')
+
+const { platform, arch } = process
+
+let nativeBinding = null
+let localFileExisted = false
+let loadError = null
+
+function isMusl() {
+  // For Node 10
+  if (!process.report || typeof process.report.getReport !== 'function') {
+    try {
+      const lddPath = require('child_process').execSync('which ldd').toString().trim()
+      return readFileSync(lddPath, 'utf8').includes('musl')
+    } catch (e) {
+      return true
+    }
+  } else {
+    const { glibcVersionRuntime } = process.report.getReport().header
+    return !glibcVersionRuntime
+  }
+}
+
+switch (platform) {
+  case 'android':
+    switch (arch) {
+      case 'arm64':
+        localFileExisted = existsSync(join(__dirname, 'html-to-markdown-node.android-arm64.node'))
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.android-arm64.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-android-arm64')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'arm':
+        localFileExisted = existsSync(join(__dirname, 'html-to-markdown-node.android-arm-eabi.node'))
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.android-arm-eabi.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-android-arm-eabi')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Android ${arch}`)
+    }
+    break
+  case 'win32':
+    switch (arch) {
+      case 'x64':
+        localFileExisted = existsSync(
+          join(__dirname, 'html-to-markdown-node.win32-x64-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.win32-x64-msvc.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-win32-x64-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'ia32':
+        localFileExisted = existsSync(
+          join(__dirname, 'html-to-markdown-node.win32-ia32-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.win32-ia32-msvc.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-win32-ia32-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'arm64':
+        localFileExisted = existsSync(
+          join(__dirname, 'html-to-markdown-node.win32-arm64-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.win32-arm64-msvc.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-win32-arm64-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Windows: ${arch}`)
+    }
+    break
+  case 'darwin':
+    localFileExisted = existsSync(join(__dirname, 'html-to-markdown-node.darwin-universal.node'))
+    try {
+      if (localFileExisted) {
+        nativeBinding = require('./html-to-markdown-node.darwin-universal.node')
+      } else {
+        try {
+          nativeBinding = require('html-to-markdown-node-darwin-universal')
+        } catch {
+          switch (arch) {
+            case 'x64':
+              localFileExisted = existsSync(join(__dirname, 'html-to-markdown-node.darwin-x64.node'))
+              try {
+                if (localFileExisted) {
+                  nativeBinding = require('./html-to-markdown-node.darwin-x64.node')
+                } else {
+                  nativeBinding = require('html-to-markdown-node-darwin-x64')
+                }
+              } catch (e) {
+                loadError = e
+              }
+              break
+            case 'arm64':
+              localFileExisted = existsSync(
+                join(__dirname, 'html-to-markdown-node.darwin-arm64.node')
+              )
+              try {
+                if (localFileExisted) {
+                  nativeBinding = require('./html-to-markdown-node.darwin-arm64.node')
+                } else {
+                  nativeBinding = require('html-to-markdown-node-darwin-arm64')
+                }
+              } catch (e) {
+                loadError = e
+              }
+              break
+            default:
+              throw new Error(`Unsupported architecture on macOS: ${arch}`)
+          }
+        }
+      }
+    } catch (e) {
+      loadError = e
+    }
+    break
+  case 'freebsd':
+    if (arch !== 'x64') {
+      throw new Error(`Unsupported architecture on FreeBSD: ${arch}`)
+    }
+    localFileExisted = existsSync(join(__dirname, 'html-to-markdown-node.freebsd-x64.node'))
+    try {
+      if (localFileExisted) {
+        nativeBinding = require('./html-to-markdown-node.freebsd-x64.node')
+      } else {
+        nativeBinding = require('html-to-markdown-node-freebsd-x64')
+      }
+    } catch (e) {
+      loadError = e
+    }
+    break
+  case 'linux':
+    switch (arch) {
+      case 'x64':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'html-to-markdown-node.linux-x64-musl.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./html-to-markdown-node.linux-x64-musl.node')
+            } else {
+              nativeBinding = require('html-to-markdown-node-linux-x64-musl')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'html-to-markdown-node.linux-x64-gnu.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./html-to-markdown-node.linux-x64-gnu.node')
+            } else {
+              nativeBinding = require('html-to-markdown-node-linux-x64-gnu')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 'arm64':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'html-to-markdown-node.linux-arm64-musl.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./html-to-markdown-node.linux-arm64-musl.node')
+            } else {
+              nativeBinding = require('html-to-markdown-node-linux-arm64-musl')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'html-to-markdown-node.linux-arm64-gnu.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./html-to-markdown-node.linux-arm64-gnu.node')
+            } else {
+              nativeBinding = require('html-to-markdown-node-linux-arm64-gnu')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 'arm':
+        localFileExisted = existsSync(
+          join(__dirname, 'html-to-markdown-node.linux-arm-gnueabihf.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./html-to-markdown-node.linux-arm-gnueabihf.node')
+          } else {
+            nativeBinding = require('html-to-markdown-node-linux-arm-gnueabihf')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Linux: ${arch}`)
+    }
+    break
+  default:
+    throw new Error(`Unsupported OS: ${platform}, architecture: ${arch}`)
+}
+
+if (!nativeBinding) {
+  if (loadError) {
+    throw loadError
+  }
+  throw new Error(`Failed to load native binding`)
+}
+
+const { convert } = nativeBinding
+
+module.exports.convert = convert

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "html-to-markdown-node",
-  "version": "2.6.4",
+  "version": "2.6.5",
   "description": "High-performance HTML to Markdown converter - Node.js native bindings",
   "main": "index.js",
   "types": "index.d.ts",
@@ -59,15 +59,15 @@
     "up": "^1.0.2"
   },
   "optionalDependencies": {
-    "html-to-markdown-node-darwin-x64": "2.6.4",
-    "html-to-markdown-node-darwin-arm64": "2.6.4",
-    "html-to-markdown-node-win32-x64-msvc": "2.6.4",
-    "html-to-markdown-node-win32-arm64-msvc": "2.6.4",
-    "html-to-markdown-node-linux-x64-gnu": "2.6.4",
-    "html-to-markdown-node-linux-x64-musl": "2.6.4",
-    "html-to-markdown-node-linux-arm64-gnu": "2.6.4",
-    "html-to-markdown-node-linux-arm64-musl": "2.6.4",
-    "html-to-markdown-node-linux-arm-gnueabihf": "2.6.4"
+    "html-to-markdown-node-darwin-x64": "2.6.5",
+    "html-to-markdown-node-darwin-arm64": "2.6.5",
+    "html-to-markdown-node-win32-x64-msvc": "2.6.5",
+    "html-to-markdown-node-win32-arm64-msvc": "2.6.5",
+    "html-to-markdown-node-linux-x64-gnu": "2.6.5",
+    "html-to-markdown-node-linux-x64-musl": "2.6.5",
+    "html-to-markdown-node-linux-arm64-gnu": "2.6.5",
+    "html-to-markdown-node-linux-arm64-musl": "2.6.5",
+    "html-to-markdown-node-linux-arm-gnueabihf": "2.6.5"
   },
   "scripts": {
     "artifacts": "napi artifacts",