使用pdf.js和ImageData将.pdf渲染为单个Canvas

时间:2013-03-11 14:25:08

标签: javascript html5-canvas getimagedata pdf.js putimagedata

我正在尝试使用PDF.js阅读整个.pdf文档,然后在一个画布上渲染所有页面。

我的想法:将每个页面渲染到画布上并获取ImageData(context.getImageData()),清除画布做下一页。我将所有ImageDatas存储在一个数组中,一旦所有页面都在那里,我想把数组中的所有ImageDatas放到一个画布上。

var pdf = null;
PDFJS.disableWorker = true;
var pages = new Array();
    //Prepare some things
    var canvas = document.getElementById('cv');
    var context = canvas.getContext('2d');
    var scale = 1.5;
    PDFJS.getDocument(url).then(function getPdfHelloWorld(_pdf) {
        pdf = _pdf;
        //Render all the pages on a single canvas
        for(var i = 1; i <= pdf.numPages; i ++){
            pdf.getPage(i).then(function getPage(page){
                var viewport = page.getViewport(scale);
                canvas.width = viewport.width;
                canvas.height = viewport.height;
                page.render({canvasContext: context, viewport: viewport});
                pages[i-1] = context.getImageData(0, 0, canvas.width, canvas.height);
                context.clearRect(0, 0, canvas.width, canvas.height);
                p.Out("pre-rendered page " + i);
            });
        }

    //Now we have all 'dem Pages in "pages" and need to render 'em out
    canvas.height = 0;
    var start = 0;
    for(var i = 0; i < pages.length; i++){
        if(canvas.width < pages[i].width) canvas.width = pages[i].width;
        canvas.height = canvas.height + pages[i].height;
        context.putImageData(pages[i], 0, start);
        start += pages[i].height;
    }
    });

所以从我的方式来看,这应该有用,对吧? 当我运行这个时,我最终得到的画布很大,包含pdf的所有页面,但没有显示pdf ......

感谢您的帮助。

3 个答案:

答案 0 :(得分:20)

PDF操作在所有阶段都是异步的。这意味着您还需要在最后一次渲染时捕获承诺。如果你没有抓住它,你将只得到一个空白的画布,因为在循环继续到下一页之前渲染没有完成。

提示:我还建议您使用除getImageData以外的其他内容,因为这将存储未压缩的位图,例如data-uri,而不是压缩数据。

这是一种略微不同的方法,消除了for循环,并为此目的更好地使用了承诺:

LIVE FIDDLE

var canvas = document.createElement('canvas'), // single off-screen canvas
    ctx = canvas.getContext('2d'),             // to render to
    pages = [],
    currentPage = 1,
    url = 'path/to/document.pdf';              // specify a valid url

PDFJS.getDocument(url).then(iterate);   // load PDF document

/* To avoid too many levels, which easily happen when using chained promises,
   the function is separated and just referenced in the first promise callback
*/

function iterate(pdf) {

    // init parsing of first page
    if (currentPage <= pdf.numPages) getPage();

    // main entry point/function for loop
    function getPage() {

        // when promise is returned do as usual
        pdf.getPage(currentPage).then(function(page) {

            var scale = 1.5;
            var viewport = page.getViewport(scale);

            canvas.height = viewport.height;
            canvas.width = viewport.width;

            var renderContext = {
                canvasContext: ctx,
                viewport: viewport
            };

            // now, tap into the returned promise from render:
            page.render(renderContext).then(function() {

                // store compressed image data in array
                pages.push(canvas.toDataURL());

                if (currentPage < pdf.numPages) {
                    currentPage++;
                    getPage();        // get next page
                }
                else {
                    done();           // call done() when all pages are parsed
                }
            });
        });
    }

}

当您需要检索页面时,您只需创建一个图像元素并将data-uri设置为源:

function drawPage(index, callback) {
    var img = new Image;
    img.onload = function() {
        /* this will draw the image loaded onto canvas at position 0,0
           at the optional width and height of the canvas.
           'this' is current image loaded 
        */
        ctx.drawImage(this, 0, 0, ctx.canvas.width, ctx.canvas.height);
        callback();          // invoke callback when we're done
    }
    img.src = pages[index];  // start loading the data-uri as source
}

由于图像加载,它本质上也是异步的,这就是我们需要回调的原因。如果您不想要异步性质,那么您也可以在渲染承诺上执行此步骤(创建和设置图像元素),然后存储图像元素而不是data-uris。

希望这有帮助!

答案 1 :(得分:8)

我不能说你的代码中将pdf渲染成画布的部分,但我确实看到了一些问题。

  • 每次重置canvas.width或canvas.height会自动清除画布内容。因此,在顶部,不需要clearRect,因为画布在每个pages.render之前被canvas.width清除。
  • 更重要的是,在底部,所有以前的pdf图纸都会被每个画布调整大小清除(哎呀!)。
  • getImageData()获取数组,其中每个像素由该数组的4个连续元素表示(红色然后是绿色,然后是蓝色,然后是alpha)。因为getImageData()是一个数组,所以它没有pages [i] .width或pages [i] .height-它只有一个pages [i] .length。该数组长度不能用于确定宽度或高度。

为了让您入门,我首先将您的代码更改为(非常非常未经测试!):

var pdf = null;
PDFJS.disableWorker = true;
var pages = new Array();
//Prepare some things
var canvas = document.getElementById('cv');
var context = canvas.getContext('2d');
var scale = 1.5;
var canvasWidth=0;
var canvasHeight=0;
var pageStarts=new Array();
pageStarts[0]=0;

PDFJS.getDocument(url).then(function getPdfHelloWorld(_pdf) {
    pdf = _pdf;
    //Render all the pages on a single canvas
    for(var i = 1; i <= pdf.numPages; i ++){
        pdf.getPage(i).then(function getPage(page){
            var viewport = page.getViewport(scale);
            // changing canvas.width and/or canvas.height auto-clears the canvas
            canvas.width = viewport.width;
            canvas.height = viewport.height;
            page.render({canvasContext: context, viewport: viewport});
            pages[i-1] = context.getImageData(0, 0, canvas.width, canvas.height);
            // calculate the width of the final display canvas
            if(canvas.width>maxCanvasWidth){
              maxCanvasWidth=canvas.width;
            }
            // calculate the accumulated with of the final display canvas
            canvasHeight+=canvas.height;
            // save the "Y" starting position of this pages[i]
            pageStarts[i]=pageStarts[i-1]+canvas.height;
            p.Out("pre-rendered page " + i);
        });
    }


    canvas.width=canvasWidth; 
    canvas.height = canvasHeight;  // this auto-clears all canvas contents
    for(var i = 0; i < pages.length; i++){
        context.putImageData(pages[i], 0, pageStarts[i]);
    }

});

或者,这是完成任务的更传统方式:

使用单个“显示”画布,并允许用户“翻阅”每个所需的页面。

由于您已经开始将每个页面绘制到画布中,为什么不为每个页面保留一个单独的隐藏画布。然后,当用户想要查看第6页时,您只需将隐藏的画布#6复制到显示画布上。

Mozilla开发人员在这里的pdfJS演示中使用了这种方法:http://mozilla.github.com/pdf.js/web/viewer.html

您可以在此处查看查看器的代码:http://mozilla.github.com/pdf.js/web/viewer.js

答案 2 :(得分:0)

您可以将数字页面传递给promises,获取该页面画布数据并在画布上以正确的顺序呈现

    var renderPageFactory = function (pdfDoc, num) {
        return function () {

            var localCanvas = document.createElement('canvas');

            ///return pdfDoc.getPage(num).then(renderPage);
            return  pdfDoc.getPage(num).then((page) => {
                renderPage(page, localCanvas, num);
            });


        };
    };

    var renderPages = function (pdfDoc) {
        var renderedPage = $q.resolve();
        for (var num = 1; num <= pdfDoc.numPages; num++) {
            // Wait for the last page t render, then render the next
            renderedPage = renderedPage.then(renderPageFactory(pdfDoc, num));
        }
    };

    renderPages(pdf);

完整示例

   function renderPDF(url, canvas) {



    var pdf = null;
    PDFJS.disableWorker = true;
    var pages = new Array();

    var context = canvas.getContext('2d');
    var scale = 1;

    var canvasWidth = 256;
    var canvasHeight = 0;
    var pageStarts = new Array();
    pageStarts[0] = 0;





    var k = 0;

    function finishPage(localCanvas, num) {
        var ctx = localCanvas.getContext('2d');

        pages[num] = ctx.getImageData(0, 0, localCanvas.width, localCanvas.height);

        // calculate the accumulated with of the final display canvas
        canvasHeight += localCanvas.height;
        // save the "Y" starting position of this pages[i]
        pageStarts[num] = pageStarts[num -1] + localCanvas.height;

        if (k + 1 >= pdf.numPages)
        {


            canvas.width = canvasWidth;
            canvas.height = canvasHeight;  // this auto-clears all canvas contents
            for (var i = 0; i < pages.length; i++) {
                context.putImageData(pages[i+1], 0, pageStarts[i]);
            }

            var img = canvas.toDataURL("image/png");
            $scope.printPOS(img);
        }

        k++;


    }

    function renderPage(page, localCanvas, num) {

        var ctx = localCanvas.getContext('2d');

        var viewport = page.getViewport(scale);


        // var viewport = page.getViewport(canvas.width / page.getViewport(1.0).width);
        // changing canvas.width and/or canvas.height auto-clears the canvas
        localCanvas.width = viewport.width;

        /// viewport.width = canvas.width;
        localCanvas.height = viewport.height;

        var renderTask = page.render({canvasContext: ctx, viewport: viewport});


        renderTask.then(() => {
            finishPage(localCanvas, num);
        });


    }





    PDFJS.getDocument(url).then(function getPdfHelloWorld(_pdf) {

        pdf = _pdf;



        var renderPageFactory = function (pdfDoc, num) {
            return function () {

                var localCanvas = document.createElement('canvas');

                ///return pdfDoc.getPage(num).then(renderPage);
                return  pdfDoc.getPage(num).then((page) => {
                    renderPage(page, localCanvas, num);
                });


            };
        };

        var renderPages = function (pdfDoc) {
            var renderedPage = $q.resolve();
            for (var num = 1; num <= pdfDoc.numPages; num++) {
                // Wait for the last page t render, then render the next
                renderedPage = renderedPage.then(renderPageFactory(pdfDoc, num));
            }
        };

        renderPages(pdf);






    });





}