mupdf webassembly 试用

mupdf 是基于c++编写的pdf 读写以及预览工具，同时也提供了sdk，官方基于emscripten 开发了一个
webassembly的扩展

参考使用

 

const fs = require("fs")

const mupdf = require("mupdf")

let data = fs.readFileSync("demo.pdf")

let doc = mupdf.Document.openDocument(data, "application/pdf")

console.log(doc.countPages())

let page = doc.loadPage(0);

var sText = page.toStructuredText();

const myinfo =  sText.asJSON()

console.log(myinfo)

wasm 生成简单说明

• 项目结构

mupdf.c 为需要暴露的c 服务，mupdf.js 是基于wasm 实现的方便pdf 操作的工具方法（支持web 以及node 运行）

• 文档打开
  调用了的暴露的webassembly 方法

 

static openDocument(from, magic) {

    checkType(magic, "string")

 

    let pointer = 0

 

    if (from instanceof ArrayBuffer || from instanceof Uint8Array)

        from = new Buffer(from)

    if (from instanceof Buffer)

       //

        pointer = libmupdf._wasm_open_document_with_buffer(STRING(magic), from)

    else if (from instanceof Stream)

        pointer = libmupdf._wasm_open_document_with_stream(STRING(magic), from)

    else

        throw new Error("not a Buffer or Stream")

 

    let pdf_ptr = libmupdf._wasm_pdf_document_from_fz_document(pointer)

    if (pdf_ptr)

        return new PDFDocument(pointer)

    return new Document(pointer)

}

c 实现

// --- Document ---

 

EXPORT

fz_document * wasm_open_document_with_buffer(char *magic, fz_buffer *buffer)

{

    POINTER(fz_open_document_with_buffer, magic, buffer)

}

• 构建
  基于了emscripten 同时包含了不少参数

 

#!/bin/bash

 

MUPDF_DIR=../..

EMSDK_DIR=/opt/emsdk

 

MUPDF_OPTS="-Os -DTOFU -DTOFU_CJK -DFZ_ENABLE_XPS=0 -DFZ_ENABLE_SVG=0 -DFZ_ENABLE_CBZ=0 -DFZ_ENABLE_IMG=0 -DFZ_ENABLE_HTML=0 -DFZ_ENABLE_EPUB=0 -DFZ_ENABLE_JS=0 -DFZ_ENABLE_OCR_OUTPUT=0 -DFZ_ENABLE_DOCX_OUTPUT=0 -DFZ_ENABLE_ODT_OUTPUT=0"

 

export EMSDK_QUIET=1

source $EMSDK_DIR/emsdk_env.sh

echo

 

echo BUILDING MUPDF CORE

make -j4 -C $MUPDF_DIR build=release OS=wasm XCFLAGS="$MUPDF_OPTS" libs

echo

 

echo BUILDING MUPDF WASM

emcc -o lib/mupdf-wasm.js -I $MUPDF_DIR/include lib/mupdf.c \

    --no-entry \

    -sABORTING_MALLOC=0 \

    -sALLOW_MEMORY_GROWTH=1 \

    -sMODULARIZE=1 \

    -sNODEJS_CATCH_EXIT=0 \

    -sWASM_ASYNC_COMPILATION=0 \

    -sEXPORT_NAME='"libmupdf"' \

    -sEXPORTED_RUNTIME_METHODS='["ccall","UTF8ToString","lengthBytesUTF8","stringToUTF8"]' \

     $MUPDF_DIR/build/wasm/release/libmupdf.a \

     $MUPDF_DIR/build/wasm/release/libmupdf-third.a

echo

说明

mupdf 对于webassembly 的支持实现上是很值得学习的，属于一个比较完整的emscripten集成项目

参考资料

https://github.com/ArtifexSoftware/mupdf
https://mupdf.readthedocs.io/en/latest/quick-start-guide.html
https://mupdf.readthedocs.io/en/latest/mupdf-wasm.html
https://github.com/ArtifexSoftware/mupdf/blob/master/platform/wasm/lib/mupdf.js
https://github.com/ArtifexSoftware/mupdf/blob/master/platform/wasm/lib/mupdf.c
https://mupdf.com/wasm/demo/view.html?file=../../docs/mupdf_explored.pdf#page2
https://emscripten.org/docs/porting/connecting_cpp_and_javascript/Interacting-with-code.html#interacting-with-code-ccall-cwrap