Send data URI as base64

[mrfrase3]

this prevents invalid URI characters from messing with page rendering, as seen in #21

[cr7boulos]

@mrfrase3 This would be great functionality to include in pdf-puppetteer. Two comments on the PR. First, base64-encoding should be an opt-in feature so that users can explicitly determine how their html should be intepreted. Base64-encoded strings can get rather long and some users may not be willing to accept that tradeoff. Secondly, adding a unit test to this feature would be valuable to ensure that these two methods for creating a pdf return the same results. I wrote up a sample unit test below. Feel free to tweak if you have another way to test this functionality and let me know if you have any comments or questions regarding my feedback.

Thanks

1. Updated function signature:
   convertHTMLToPDF = async (html, callback, puppeteerOptions = null, puppeteerArgs=null, remoteContent=true, options = {})

then in the body, something like this could be done:

if (remoteContent === true) {
        if(options.base64Encode){
            console.log('base 64 encoded');
            await page.goto(`data:text/html;base64,${Buffer.from(html).toString('base64')}`, {
                waitUntil: 'networkidle0'
            });
        }
        else {
            console.log('regular');
            await page.goto(`data:text/html,${html}`);
        }
        
    }

2. Sample unit test:

test('Converts remote Base64 encoded HTML To PDF with Puppeteer Arguments', async done => {
        let base64 = null;
        let reg = null;
        let options = { 'base64Encode': true };
        let t1 = convertHTMLToPDF(remoteTemplate, pdf => {
            expect(Object.prototype.toString.call(pdf)).toBe('[object Uint8Array]');
            base64 = pdf;
        }, {format: 'A4'}, {
            args: [
                '--no-sandbox',
                '--disable-setuid-sandbox'
            ]
        }, true, options);

        let t2 =  convertHTMLToPDF(remoteTemplate, pdf => {
            expect(Object.prototype.toString.call(pdf)).toBe('[object Uint8Array]');
            reg = pdf;
        }, {format: 'A4'}, {
            args: [
                '--no-sandbox',
                '--disable-setuid-sandbox'
            ]
        }, true. options);

        await Promise.all([t1, t2]);

        expect(base64.length).toBe(reg.length);

        let result = true;

        for(let i = 0; i < base64.length; i++){
            if(base64[i] !== reg[i]){
                result = false;
            }
        }

        expect(result).toBe(true);

        done();
    });

[mrfrase3]

Hey @cr7boulos,

When you say base64 is long, do you mean time-wise or storage space-wise? Storage space is effectively irrelevant for sending to a local process, and the increase in size is ~1.37x the original.

The # error is an invisible problem, this took me hours of debugging to track down. At the end of the day, HTML can contain illegal characters for a URI, so your default should always operate on the assumption that it does contain those characters. One could implement something that scans and sends a warning, but that would be costly. If someone wants to take on the risk for a speed up, then they should knowingly enable that.

I'm pushed on a deadline at the moment, so I can't devote more time to this PR.

[cr7boulos]

@mrfrase3,

Sounds good. I am a new maintainer of the package so still learning how the code works. Please disregard my previous comments. I will approve and push out an updated version to npm. Thanks for submitting.

[cr7boulos]

Please see the comment on the PR.

[cr7boulos]

@mrfrase3,

Sounds good. I am a new maintainer of the package so still learning how the code works. Please disregard my previous comments. I will approve and push out an updated version to npm. Thanks for submitting.