My HTML table displays dynamic images properly on the webpage, however, an attempt to generate a PDF using jsPDF following this example created on codepen is failing me. I'm getting the below pdf without the image filling the entire width and the height appears to be reduced instead of getting the original image size- I'm using base64:
I have also looked at several questions and answers here on SO and elsewhere extensively with no avail like this one here How to get table row image in PDF using jsPDF? and followed this discussion here https://github.com/simonbengtsson/jsPDF-AutoTable/issues/173, but I don't seem to wrap my head around this.
function createPdf(){
var doc = new jsPDF();
doc.autoTable({
html: '#mytable',
bodyStyles: {minCellHeight : 30},
didDrawCell: function(data){
if(data.cell.section ==='body' && data.column.index ===1) {
var td = data.cell.raw;
var img = td.getElementsByTagName('img')[0];
var dim = data.cell.height - data.cell.padding('vertical');
var textPos = data.cell.textPos;
doc.addImage(img.src, textPos.x, textPos.y, dim, dim);
}
}
})
doc.save("table.pdf");
}
The issue appears to do with the width and height dimensions which is dynamic and also the images are further on the right. However, the priority issue is to set auto value for height to provide desirable outcome.
Anyone with expertise in this, kindly Check what I have used here https://jsfiddle.net/kibika2020/0rtoph6s/5/
function createPdf(){
var doc = new jsPDF();
doc.autoTable({
html: '#mytable',
bodyStyles: {minCellHeight : 50},
didDrawCell: function(data){
if(data.cell.section ==='body' && data.column.index ===1) {
var td = data.cell.raw;
var img = td.getElementsByTagName('img')[0];
var dim = data.cell.height - data.cell.padding('vertical');
var textPos = data.cell.textPos;
doc.addImage(img.src, textPos.x, textPos.y, dim, dim);
}
}
})
doc.save("autoble.pdf");
}
I'm a beginner and finding this jsPDF stuff really tough than I expected. Take a look and advise me accordingly.
Thank you.
Related
I have used chart.js to generate a report page that has multiple charts. I need to export this report to PDF. There are many solutions available via search, but I cannot find one which has multiple canvas elements.
The only available solution seems to be to loop through all the images, and recreate the report using the images, and then download that as a pdf.
Is there any simpler/more efficient way to accomplish this?
<body>
<h1> Chart 1 </h1>
<div style="width:800px; height:400px;">
<canvas id="chart_1" width="50" height="50"></canvas>
</div>
<h1> Chart 2 </h1>
<div style="width:800px; height:400px;">
<canvas id="chart_2" width="50" height="50"></canvas>
</div>
<h1> Chart 3 </h1>
<div style="width:800px; height:400px;">
<canvas id="chart_3" width="50" height="50"></canvas>
</div>
</body>
Honestly, it seems like the easiest approach would be to just provide a "download to PDF" link that pops up the browser's print page and instruct to user to select "print as pdf".
If that approach doesn't work for you (or your users), then here is a rough way to do it.
Basically, we create a new canvas element that is the size of your report page and incrementally paint the pixels from your existing chart.js canvas charts into the new canvas. Once that is done, then you can use jsPDF to add the new canvas to a pdf document as an image and download the file.
Here is an example implementation that does just that.
$('#downloadPdf').click(function(event) {
// get size of report page
var reportPageHeight = $('#reportPage').innerHeight();
var reportPageWidth = $('#reportPage').innerWidth();
// create a new canvas object that we will populate with all other canvas objects
var pdfCanvas = $('<canvas />').attr({
id: "canvaspdf",
width: reportPageWidth,
height: reportPageHeight
});
// keep track canvas position
var pdfctx = $(pdfCanvas)[0].getContext('2d');
var pdfctxX = 0;
var pdfctxY = 0;
var buffer = 100;
// for each chart.js chart
$("canvas").each(function(index) {
// get the chart height/width
var canvasHeight = $(this).innerHeight();
var canvasWidth = $(this).innerWidth();
// draw the chart into the new canvas
pdfctx.drawImage($(this)[0], pdfctxX, pdfctxY, canvasWidth, canvasHeight);
pdfctxX += canvasWidth + buffer;
// our report page is in a grid pattern so replicate that in the new canvas
if (index % 2 === 1) {
pdfctxX = 0;
pdfctxY += canvasHeight + buffer;
}
});
// create new pdf and add our new canvas as an image
var pdf = new jsPDF('l', 'pt', [reportPageWidth, reportPageHeight]);
pdf.addImage($(pdfCanvas)[0], 'PNG', 0, 0);
// download the pdf
pdf.save('filename.pdf');
});
You can see it in action at this codepen.
Now let's talk about some gotchas with this approach. First, you have to control the position of each chart.js canvas in the new canvas object. The only way to do that is to have an understanding of how your report page is structured and implement that same structure. In my example, my charts are in a 2x2 grid and the logic handles this accordingly. If you had a 3x2 grid or something different then you would have to change the positioning logic.
Lastly, the final pdf output file dimensions are much larger than the original chart page (from the web). I think the reason is because my chart "container" div stretches across the full page. Therefore, you probably want to use a different approach for setting the size of your new canvas.
So long story short, the above example is meant to demonstrate an approach and not be your final solution.
Good luck!
I have a working solution in vanilla javascript(although I used ts typing) and using the lib jsPdf, where you need a plot per pdf page:
let index = 1;
// create new pdf object
// if don't choose compress as true you will end up with a large pdf file
let pdf = new jsPDF({
orientation: 'landscape',
unit: 'px',
format: 'a4',
compress: true,
})
// search for the html element(s) you need
const canvas = document.querySelectorAll("canvas");
// here my size are in pixels since I configured that in the obj instance
let pageWidth = 400;
let pageHeight = 400;
let index = 1;
// traverse the array of canvas
canvas.forEach( (canva:HTMLCanvasElement) => {
// I added some options among others I added the type of the compression
// method: FAST
pdf.addImage(canva, 'PNG', 10, 10, pageWidth, pageHeight, `img${index}`, "FAST");
// so as to not end up with an extra pdf page at the end of the iteration
if (index < canvas.length) {
pdf.addPage();
}
index++;
});
// download the pdf
pdf.save('Reporte.pdf');
I am using jspdf and html2canvas combination to save html page as pdf. A pdf copy of current page is saved the moment you click a button. The problem is, if you zoom in the page, and then click the button, the saved pdf contains incomplete portion of the current page. Most of the part not visible on page due to zooming, gets cut off in the saved pdf page. What is the solution?
Below is the js code being invoked upon click of save button-
var pdf = new jsPDF('l', 'pt', 'a4');
var source = $('#someId')[0];
var options = {
background : '#eee'
};
pdf.addHTML(source, options, function(){
pdf.save('abcd.pdf');
});
EDIT
Taking idea from Saurabh's approach, I tried quite a similar thing, but without writing code for any extra div element. Before saving to pdf I made the screen size of a fixed width, and after printing I brought back the width back to default normal. It is working fine for, if it fails, we can always fix the height of the screen too, so that it appears fine in generated pdf despite zooming. Below is the code used by me:-
var pdf = new jsPDF('l', 'pt', 'a4');
var source = $('#someId')[0];
var options = {
background : '#eee'
};
var width = source.clientWidth;
source.style.width = '1700px';
pdf.addHTML(source, options,
function(){
pdf.save('abcd.pdf');
source.style.width = width+'px';
});
Here is how I managed to get the full page pdf while the page is zoomed in using jsPDF's new .html() method. First, I force the page zoom level back to 100% before converting it to pdf. It's important to reset the scale in html2canvas option after that, otherwise it'll returns a blank page.
<script src="https://cdnjs.cloudflare.com/ajax/libs/jspdf/1.5.3/jspdf.debug.js"
integrity="sha384-NaWTHo/8YCBYJ59830LTz/P4aQZK1sS0SneOgAvhsIl3zBu8r9RevNg5lHCHAuQ/"
crossorigin="anonymous"></script>
<script src="https://html2canvas.hertzen.com/dist/html2canvas.min.js"></script>
<!-- html2canvas 1.0.0-alpha.11 or higher version is needed -->
<script>
function download() {
// Bring the page zoom level back to 100%
const scale = window.innerWidth / window.outerWidth;
if (scale != 1) {
document.body.style.zoom = scale;
}
let pdf = new jsPDF('p', 'pt', 'a4');
pdf.html(document.getElementById('idName'), {
html2canvas: {
scale: 1 // default is window.devicePixelRatio
},
callback: function () {
// pdf.save('test.pdf');
window.open(pdf.output('bloburl')); // to debug
}
});
}
</script>
Update: A better way is to adjust the html2canvas.scale according to the scale factor.
function download() {
let pWidth = pdf.internal.pageSize.width; // 595.28 is the width of a4
let srcWidth = document.getElementById('idName').scrollWidth;
let margin = 18; // narrow margin - 1.27 cm (36);
let scale = (pWidth - margin * 2) / srcWidth;
let pdf = new jsPDF('p', 'pt', 'a4');
pdf.html(document.getElementById('idName'), {
x: margin,
y: margin,
html2canvas: {
scale: scale,
},
callback: function () {
window.open(pdf.output('bloburl'));
}
});
}
I was going through the same problem,
To do this what I did is I made a copy of printing div and while clicking print button I attached div copy to my dom with margin-top:500px
After I got its image then I hide this copy of the div, and set margin-top:0px
I hope this will work for you.
Hoping this doesn't get flagged as a duplicate because none of the other q/as on SO have helped me fix this, I think I need a more specific line of help.
I have a profile page on my site that allows the user to change their profile picture without page reloads (via AJAX / jQuery).
This all works fine. The user opens the "Change Profile Picture" modal, selects a file to upload and presses "Crop this image". When this button is pressed, it uploads a file to the website, using the typical way of sending the file and formData (which I append the file data to).
It gets sent backend with the following jQuery:
// Upload the image for cropping (Crop this Image!)
$("#image-upload").click(function(){
// File data
var fileData = $("#image-select").prop("files")[0];
// Set up a form
var formData = new FormData();
// Append the file to the new form for submission
formData.append("file", fileData);
// Send the file to be uploaded
$.ajax({
// Set the params
cache: false,
contentType: false,
processData: false,
// Page & file information
url: "index.php?action=uploadimage",
dataType: "text",
type: "POST",
// The data to send
data: formData,
// On success...
success: function(data){
// If no image was returned
// "not-image" is returned from the PHP script if we return it in case of an error
if(data == "not-image"){
alert("That's not an image, please upload an image file.");
return false;
}
// Else, load the image on to the page so we don't need to reload
$(profileImage).attr("src", data);
// If the API is already set, then we should apply a new image
if(jCropAPI){
jCropAPI.setImage(data + "?" + new Date().getTime());
}
// Initialise jCrop
setJCrop();
//$("#image-profile").show();
$("#send-coords").show();
}
})
});
setJcrop does the following
function setJCrop(){
// Get width / height of the image
var width = profileImage.width();
var height = profileImage.height();
// Var containing the source image
var imgSource = profileImage.attr("src");
// New image object to work on
var image = new Image();
image.src = imgSource;
// The SOURCE (ORIGINAL) width / height
var origWidth = image.width;
var origHeight = image.height;
// Set up the option to jCrop it
$(profileImage).Jcrop({
onSelect: setCoords,
onChange: setCoords,
setSelect: [0, 0, 51, 51],
aspectRatio: 1, // This locks it to a square image, so it fits the site better
boxWidth: width,
boxHeight: height, // Fixes the size permanently so that we can load new images
}, function(){jCropAPI = this});
setOthers(width, height, origWidth, origHeight);
}
And once backend, it does the following:
public function uploadImage($file){
// See if there is already an error
if(0 < $file["file"]["error"]){
return $file["file"]["error"] . " (error)";
}else{
// Set up the image
$image = $file["file"];
$imageSizes = getimagesize($image["tmp_name"]);
// If there are no image sizes, return the not-image error
if(!$imageSizes){
return "not-image";
}
// SIZE LIMIT HERE SOON (TBI)
// Set a name for the image
$username = $_SESSION["user"]->getUsername();
$fileName = "images/profile/$username-profile-original.jpg";
// Move the image which is guaranteed a unique name (unless it is due to overwrite), to the profile pictures folder
move_uploaded_file($image["tmp_name"], $fileName);
// Return the new filename
return $fileName;
}
}
Then, the user selects their area on the image with the selector and pressed "Change Profile Picture" which does the following
// Send the Coords and upload the new image
$("#send-coords").click(function(){
$.ajax({
type: "POST",
url: "index.php?action=uploadprofilepicture",
data: {
coordString: $("#coords").text() + $("#coords2").text(),
imgSrc: $("#image-profile").attr("src")
},
success: function(data){
if(data == "no-word"){
alert("Can not work with this image type, please try with another image");
}else{
// Append a date to make sure it reloads the image without using a cached version
var dateNow = new Date();
var newImageLink = data + "?" + dateNow.getTime();
$("#profile-picture").attr("src", newImageLink);
// Hide the modal
$("#profile-picture-modal").modal("hide");
}
}
});
})
The backend is:
public function uploadProfilePicture($coordString, $imgSrc){
// Target dimensions
$tarWidth = $tarHeight = 150;
// Split the coords in to an array (sent by a string that was created by JS)
$coordsArray = explode(",", $coordString);
//Set them all from the array
$x = $coordsArray[0];
$y = $coordsArray[1];
$width = $coordsArray[2];
$height = $coordsArray[3];
$newWidth = $coordsArray[4];
$newHeight = $coordsArray[5];
$origWidth = $coordsArray[6];
$origHeight = $coordsArray[7];
// Validate the image and decide which image type to create the original resource from
$imgDetails = getimagesize($imgSrc);
$imgMime = $imgDetails["mime"];
switch($imgMime){
case "image/jpeg":
$originalImage = imagecreatefromjpeg($imgSrc);
break;
case "image/png":
$originalImage = imagecreatefrompng($imgSrc);
break;
default:
return "no-work";
}
// Target image resource
$imgTarget = imagecreatetruecolor($tarWidth, $tarHeight);
$img = imagecreatetruecolor($newWidth, $newHeight);
// Resize the original image to work with our coords
imagecopyresampled($img, $originalImage, 0, 0, 0, 0,
$newWidth, $newHeight, $origWidth, $origHeight);
// Now copy the CROPPED image in to the TARGET resource
imagecopyresampled(
$imgTarget, // Target resource
$img, // Target image
0, 0, // X / Y Coords of the target image; this will always be 0, 0 as we do not want any black nothingness
$x, $y, // X / Y Coords (top left) of the target area
$tarWidth,
$tarHeight, // width / height of the target
$width,
$height // Width / height of the source image crop
);
$username = $_SESSION["user"]->getUsername();
$newPath = "images/profile/$username-profile-cropped.jpg";
// Create that shit!
imagejpeg($imgTarget, $newPath);
// Return the path
return $newPath;
}
So basically this the returns the path of the new file, which gets changed to the user's profile picture (same name every time) and uploaded live with a time appended after ? to refresh the image properly (no cache).
This all works fine, however if the user selects another image to upload, after already uploading one, the coords get all messed up (e.g. they go from 50 to 250) and they end up cropping a totally different part of the image, leaving most of it black nothing-ness.
Really sorry for the ridiculous amount of code that is in this question but I'd appreciate any help from people who might have worked around this before.
Some of the code may seem out of place but that's just me trying to debug it.
Thanks, and again, sorry for the size of this question.
-Edit-
My setCoords() and setOthers() functions look like so:
//Set the coords with this method, that is called every time the user makes / changes a selection on the crop panel
function setCoords(c){
$("#coords").text(c.x + "," + c.y + "," + c.w + "," + c.h + ",");
}
//This one adds the other parts to the second div; they will be concatenated in to the POST string
function setOthers(width, height, origWidth, origHeight){
$("#coords2").text(width + "," + height + "," + origWidth + "," + origHeight);
}
I have now resolved this issue.
The problem for me was that when using setJCrop(); - it was not re-loading the image. The reason for this is that the image uploaded and then loaded in to the JCrop window had the same name every time (username as a prefix, and then profile-cropped.jpg).
So to try and resolve this, I used the setImage method which loaded a full-sized image instead.
I got around this by setting the boxWidth / boxHeight params but they just left me with the issue of the coordinates being incorrect every time I loaded a new image in.
Turns out, it was loading the image from the cache every time, even when I was using new Image(); within jQuery.
To solve this, I have now used destroy(); on the jCropAPI and then re-initialised it every time, witout using setImage();
I set a max-width in the CSS on the image itself, which stopped it from being locked to a specific width.
The next problem was that every time I loaded an image a second time, it left the width / height of the old image there, which made the image look all skewed and wrong.
To solve this, I reset the width & height of the image that I use jCrop on, to "" with $(profileImage).css("width", ""); $(profileImage).css("height", ""); before re-setting the source of the image from the new uploaded image.
But I was still left with the issue of using the same name on the images, and then causing it to load from cache every time.
My solution to this was to add an "avatar" column in the database and save the image name in the Database each time. The image was named as $username-$time.jpg and $username-$time.jpg-cropped.jpg where $username is the username of the user (derp) and $time is simply time(); inside PHP.
This meant that every time I uploaded an image, it had a new name, so when any calls were made to this image, there was no cache of it.
Appending like imageName + ".jpg?" + new Date.getTime(); worked for some things but then when sending the image name backend, it didn't work properly, and deciding when to append it / not append it was a pain, and then one thing required it to be appended to force a re-load, but then when appended it didn't work properly, so I had to re-work it.
So the key: (TL;DR)
Don't use the same image name with jCrop, if you are loading a new image; upload an image with a different name and then refer to that one. Cache problems are a pain, and you can't really work around them properly without just using a new name every time, as this ensures that there will be absolutely no more problems (so long as the name is always unique).
Then, when you initialise jCrop, destroy the previous one beforehand if there is one. Use max-width instead of width on an image to stop it from locking the width, and re-set the width / height of the image if you're loading a new one in to the same <img> or <div>
Hope this helps somebody!
I used jcrop and I think this has happened to me. When there is a new image, you have to "reset" jcrop. Try something like this:
function resetJCrop()
{
if (jCropAPI) {
jCropAPI.disable();
jCropAPI.release();
jCropAPI.destroy();
}
}
$("#image-upload").click(function(){
success: function(data){
...
resetJCrop(); // RESETTING HERE
// If the API is already set, then we should apply a new image
if(jCropAPI){
jCropAPI.setImage(data + "?" + new Date().getTime());
}
// Initialise jCrop
setJCrop();
...
}
});
I can't remember the details about why I have to use disable() AND release() AND destroy() in my particular case. May be you can use only one of those. Just try it, and see if that works for you!
I am looking at a way to divide the image into pieces.
The no of pieces that the image can be split into is configurable using tile_height and tile_width.Looking for help.
I would not want any solution with any frameworks. Only using vanilla JavaScript
I tried the below
var _clipX =0;
var _clipY = _clipX;
var _clipHeight = IMAGE_HEIGHT/TILE_HEIGHT;
var _clipWidth = IMAGE_WIDTH/TILE_WIDTH;
var _nRows = Math.floor(IMAGE_HEIGHT/TILE_HEIGHT);
var _nColumns = Math.floor(IMAGE_WIDTH/TILE_WIDTH);
for(var i=0;i<_nRows;i++){
for(var j=0;j<_nColumns;j++){
el.getContext('2d').drawImage(image, _clipX, _clipY, _clipWidth, _clipHeight, _clipX, _clipY,_clipWidth , _clipHeight);
_clipX = _clipX + _clipWidth;
}
_clipX = 0
_clipY = _clipY + _clipHeight;
}
Below is the JSFiddle for the work
JSFiddle
Yes!!!
Found it. Instead of iterating over the image tag dimensions when you call draw image it does iterate over the natural dimensions of the image hence the issue I faced.
Now I have tried with the image size same as my img tag dimensions and it works perfectly fine.
Thanks everyone for the help offered
I am using highcharts to display several graphs on a webpage which display fine.
I have an export function that tries to combine the charts into a pdf. I am getting the svg of the chart and converting it to a jpeg image to be included in a pdf created by jsPDF.
Here is the code I am using to generate the images:
if ($('.chart').length > 0) {
var chartSVG = $('.chart').highcharts().getSVG(),
chartImg = new Image();
chartImg.src = "data:image/svg+xml;base64," + window.btoa(unescape(encodeURIComponent(chartSVG)));
var chartCanvas = document.createElement("canvas");
chartCanvas.width = 600;
chartCanvas.height = 400;
chartCanvas.getContext("2d").drawImage(chartImg, 0, 0, 600, 400);
var chartImgData = chartCanvas.toDataURL("image/jpeg");
}
This works perfectly in Chrome but in Firefox it just returns a black image.
Does anyone know what might be going wrong or has seen a similar issue?
Thanks for your help.
UPDATE
I've updated the code but now no image is appended to the pdf document, either in Chrome or Firefox.
if ($('.sales').length > 0) {
var chartSVG = $('.sales').highcharts().getSVG(),
chartImg = new Image();
chartImg.onload = function () {
var chartCanvas = document.createElement("canvas");
chartCanvas.width = 600;
chartCanvas.height = 400;
chartCanvas.getContext("2d").drawImage(chartImg, 0, 0, 600, 400);
var chartImgData = chartCanvas.toDataURL("image/jpeg");
}
chartImg.src = "data:image/svg+xml;base64," + window.btoa(unescape(encodeURIComponent(chartSVG)));
}
Not sure if I have the code in the correct place.
If I log 'chartImgData' to the console, both browsers generate a dataURI, but Firefox's version differs to Chromes.
UPDATE
Fixed the issue with black images. Now i'm struggling with how to return multiple images - how to nest multiple callbacks or is there another way?
Example: jsfiddle.net/wmuk489c/2
SOLVED
Thanks for your help #RobertLangson. fiddle updated with final working code should anyone need it: http://jsfiddle.net/wmuk489c/3/
FURTHER ISSUES:
My charts are dynamic and so may not always be present. I need to get an image from each graph that exists. If the graph does not exist, the 'getSVG' function fails, see example: http://jsfiddle.net/wmuk489c/4/
How should the img.onload work if the chart doesn't exist? The first chart in the callback may not be present either, so how would this work? Is there a better way to get the images?
setting chartImg.src causes an asynchronous load so you then need to do this...
chartImg.onload = function() {
var chartCanvas = document.createElement("canvas");
chartCanvas.width = 600;
chartCanvas.height = 400;
chartCanvas.getContext("2d").drawImage(chartImg, 0, 0, 600, 400);
var chartImgData = chartCanvas.toDataURL("image/jpeg");
doc.addImage(chartImgData, 'JPEG', 0, 0, 200, 100);
// You can only do this bit after you've added the image so it needs
// to be in the callback too
doc.save('test.pdf');
}
chartImg.src = ...
You've a race condition otherwise and I imagine you just happen to get away with it with the Chrome browser on your PC.
Here's your fiddle fixed.