Line breaks in templated HTML scriptlets - google-apps-script

If I have an HTML file test.html:
<p><?= str ?></p>
And a script function:
var t = HtmlService.createTemplateFromFile("test.html");
t.str = "test\nstring";
var content = t.evaluate().setSandboxMode(...).getContent();
Logger.log(content);
Is there any way to safely replace the newline with an HTML line break? I can use String.prototype.replace() to replace \n with <br/>, but then I'd have to use <?!= to disable the HTML templating engine's contextual escaping. I'm dealing with untrusted input and so I need both escaping and smart handling of line breaks. Having it contextually would be nice. As things stand, I wrote my own escaper, but it is only good for one context.

I see two options for your scenario, the simple one is to forget the substitution entirely and use a <pre> tag, which will render your line breaks (and other formatting chars)
<pre> <?= str ?> </pre>
The second is to perform the substitution and sanitize your input with a custom function, so that you can safely use the force print scriptlet.
In your html:
<?!= sanitize(str); ?>
and in your .gs:
function sanitize(val){
var vals = val.split('\n'); //split string into an array on newlines
for(var i in vals){
vals[i] = Encoder.htmlEncode(vals[i]); //sanitize each element in the array
}
return vals.join('<br />'); //join the elements as a string, with <br /> as glue.
}
Note, in my example I'm using the library located here to sanitize the strings: http://www.strictly-software.com/scripts/downloads/encoder.js

If anyone's curious, this is the code I wound up using to sanitize the untrusted input for display. It's not safe for use inside tags, in script sections, etc. That's why Google's context aware sanitization is so handy.
function dumbEscapeAndBreak(str) {
return str.replace(/&/g, '&').replace(/</g, '<').replace(/>/g, '>').replace(/\n/g, '<br/>');
}
function _testEscape() {
GSUnit.assertEquals('my<test<string>is>not&good&okay<br/>fine<br/>okay', dumbEscapeAndBreak("my<test<string>is>not&good&okay\nfine\nokay"));
}

Or you can do this:
In your .gs script function:
var t = HtmlService.createTemplateFromFile("test.html");
var paragraphs = "test\nstring";
t.str = paragraphs.split("\n");
In your test.html:
<?
for (i = 0; i < str.length; i++) {
?>
<span><?= str[i]?></span><br />
<!-- <p><?= str[i]?></p> -->
<?
}
?>
You can use the <span> or <p> tag, that is up to you. You can also write some simple logic to decide when to put the <br />, e.g. before <span> when i > 0. That will completely fulfil your "test\nstring".

Related

Passing variable to HTML output and then into a scriptlet

Code.gs
function doPost(e) {
...
template.data += getCustomerData + "<br>";
}
return template.evaluate().setSandboxMode(HtmlService.SandboxMode.IFRAME);
index.html
...
<?= data ?>
The code shown does display the correct values. However, it doesn't translate <br> into HTML. I'm not sure why it isn't working since template.evaluate() is supposed to return an HtmlOutput object.
By default the strings are sanitized, converting special characters to their HTML encoded equivalents (such as < becoming <).
When outputting HTML you must use <?!= to avoid sanitizing the data.
<?!= data ?>
See the details on standard & force-printing scriptlets here:
https://developers.google.com/apps-script/guides/html/templates#standard_scriptlets

DOMDocument issues: Escaping attributes and removing tags from javascript

I am not fan of DOMDocument because I believe it is not very good for real world usages. Yet in current project I need to replace all texts in a page (which I don't have access to source code) with other strings (some sort of translation); so I need to use it.
I tried doing this with DOMDocument and I didn't received the expected result. Here is the code I use:
function Translate_DoHTML($body, $replaceArray){
if ($replaceArray && is_array($replaceArray) && count($replaceArray) > 0){
$body2 = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8");
$doc = new DOMDocument();
$doc->resolveExternals = false;
$doc->substituteEntities = false;
$doc->strictErrorChecking = false;
if (#$doc->loadHTML($body2)){
Translate_DoHTML_Process($doc, $replaceArray);
$body = $doc->saveHTML();
}
}
return $body;
}
function Translate_DoHTML_Process($node, $replaceRules){
if($node->hasChildNodes()) {
$nodes = array();
foreach ($node->childNodes as $childNode)
$nodes[] = $childNode;
foreach ($nodes as $childNode)
if ($childNode instanceof DOMText) {
if (trim($childNode->wholeText)){
$text = str_ireplace(array_keys($replaceRules), array_values($replaceRules), $childNode->wholeText);
$node->replaceChild(new DOMText($text),$childNode);
}
}else
Translate_DoHTML_Process($childNode, $replaceRules);
}
}
And here are the problems:
Escaping attributes: There are data-X attributes in file that become escaped. This is not a major problem but it would be great if I could disable this behavior.
Before DOM:
data-link-content=" <a class="submenuitem" href=&quot
After DOM:
data-link-content=' <a class="submenuitem" href="
Removing of closing tags in javascript:
This is actually the main problem for me here. I don't know for what reason in the world DOMDocument may see any need to remove these tags. But it do. As you can clearly see in below example it remove closing tags in java-script string. It also removed last part of script. It seems like DOMDocument parse the java-script inside. Maybe because there is no CDATA tag? But any way it is HTML and we don't need CDDATA in HTML. I thought CDATA is for xHTML. Also I have no way to add CDDATA here. So can I ask it to not parse script tags?
Before DOM:
<script type="text/javascript"> document.write('<video src="http://x.webm"><p>You will need to Install the latest Flash plugin to view this page properly.</p></video>'); </script>
After DOM:
<script type="text/javascript"> document.write('<video src="http://x.webm"><p>You will need to <a href="http://www.adobe.com/go/getflashplayer" target="_blank">Install the latest Flash plugin to view this page properly.</script>
If there is no way for me to prevent these things, is there any way that I can port this code to SimpleHTMLDOM?
Thanks you very much.
Try this , and replace line content ;
$body2 = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8");
to ;
$body2 = convertor($body);
and insert in your code ;
function convertor($ToConvert)
{
$FromConvert = html_entity_decode($ToConvert,ENT_QUOTES,'ISO-8859-1');
$Convert = mb_convert_encoding($FromConvert, "ISO-8859-1", "UTF-8");
return ltrim($Convert);
}
But use the right encoding in the context.
Have a nice day.
Based on my search, reason of the second problem is actually what "Alex" told us in this question: DOM parser that allows HTML5-style </ in <script> tag
But based on their research there is no good parser out there capable of understanding today's HTML. Also, html5lib's last update was 2 years ago and it failed to work in real world situations based on my tests.
So I had only one way to solve the second problem. RegEx. Here is the code I use:
function Translate_DoHTML_GetScripts($body){
$res = array();
if (preg_match_all('/<script\b[^>]*>([\s\S]*?)<\/script>/m', $body, $matches) && is_array($matches) && isset($matches[0])){
foreach ($matches[0] as $key => $match)
$res["<!-- __SCRIPTBUGFIXER_PLACEHOLDER".$key."__ -->"] = $match;
$body = str_ireplace(array_values($res), array_keys($res), $body);
}
return array('Body' => $body, 'Scripts' => $res);
}
function Translate_DoHTML_SetScripts($body, $scripts){
return str_ireplace(array_keys($scripts), array_values($scripts), $body);
}
Using above two functions I will remove any script from HTML so I can use DomDocument to do my works. Then again at the end, I will add them back exactly where they were.
Yet I am not sure if regex is fast enough for this.
And don't tell me to not use RegEx for HTML. I know that HTML is not a regular language and so on; but if you read the problem your self, you will suggest the same approach.

Put a bit of HTML inside a <pre> tag?

How do I put a bit of HTML inside a tag without escaping it? Or am I using an incorrect tag?
P.S. I cannot escape the HTML, it is produced by the server.
If you have no control over the emitted HTML, you can still encode it on the client side.
Here is how you would escape all markup inside <pre> tags using the jQuery library:
$(function() {
var pre = $('pre');
pre.html(htmlEncode(pre.html()));
});
function htmlEncode(value){
return $('<div/>').text(value).html();
}
Edit: As requested, same code without using jQuery:
function encodePreElements() {
var pre = document.getElementsByTagName('pre');
for(var i = 0; i < pre.length; i++) {
var encoded = htmlEncode(pre[i].innerHTML);
pre[i].innerHTML = encoded;
}
};
function htmlEncode(value) {
var div = document.createElement('div');
var text = document.createTextNode(value);
div.appendChild(text);
return div.innerHTML;
}
And run the encodePreElements after the DOM has been loaded:
<body onLoad='encodePreElements()'>
<pre>Foo <b>bar</b></pre>
</body>
This:
<pre>
<‍div>Hello<‍/div>
</pre>
Prints this:
<div>Hello</div>
Zero Width Joiner = ‍
You need to escape the HTML, like so:
You can use HTML entities that are escaped and printed without compiling.
< // will print <
> // will print >
Using these two, you can print any HTML elements.
<div> // will print <div>
To solve your specific problem, you will need to parse the HTML string and convert the chevrons to these entities.

How to display code in plain text?

I want to display bare code on an HTML page, I tried this:
<script>
function getSize() {
var myFSO = new ActiveXObject("Scripting.FileSystemObject");
var filepath = document.upload.file.value;
var thefile = myFSO.getFile(filepath);
var size = thefile.size;
alert(size + " bytes");
}
</script>
The above JavaScript code is some code entered by the user. I can't figure out to show this bare code on the html page without being interpreted and screwed up by the browser.
How do I display bare code on an HTML page?
I'm not quite clear on the specifics of the issue, as pre tags should, in general, do the trick, but here's an alternative tag:
<xmp>[Code can be displayed here]</xmp>
If you're using a server-side language, though, I'd suggest converting to HTML entities before outputting, then using CSS to style it.
As well, be sure if you're accepting user input that any JavaScript is being filtered and never executed.
You can use the <pre> and <code> tags to display formatted code. But to prevent the code from executing and not displaying you'll need to convert the text to character entities. > becomes >, < becomes &lt, etc.
You could do this by using PHP, for example:
<?php echo htmlentities('function getSize() { var myFSO = new
ActiveXObject("Scripting.FileSystemObject");
var filepath =
document.upload.file.value; var
thefile = myFSO.getFile(filepath);
var size = thefile.size; alert(size
+ " bytes"); }'); ?>
As your system relies on user input, you might have to rely on AJAX to actually process the user input and convert it to HTML entities.
Use the <code></code> tag, and use javascript or your sever-side scripting language
Dump it into a textarea and render it like a div tag
This is a bit of a hack and parlor trick, but it works.
Get bare code rendered onto an HTML page is to place it in a text area and remove all the formatting around the textarea so it looks like a <div> tag:
Code:<br>
<textarea style="border: none;width:400;height:200;background-color:lightgrey;">
#include<iostream>
using namespace std;
class Box{
public:
int mymethod(){ cout << "is method"; }
};
int myfunction(){ cout << "is function"; }
int main(){
Box b;
b.mymethod();
myfunction();
}
</textarea>
<br>
Output:
<pre>is methodis function
</pre>
The above html code should render like this on the page:
What I've done is invalid HTML, it only works because customary error handling happens to handle it this way. You shouldn't put unescaped angle brackets in the content of a <textarea>. You get undefined behavior depending on how the browser chooses to interpret your textarea tag.
The most reliable method is to htmlencode the code to be displayed on the page.
For example
< into &lt
space into &nbsp
etc.

Sending values through links

Here is the situation: I have 2 pages.
What I want is to have a number of text links(<a href="">) on page 1 all directing to page 2, but I want each link to send a different value.
On page 2 I want to show that value like this:
Hello you clicked {value}
Another point to take into account is that I can't use any php in this situation, just html.
Can you use any scripting? Something like Javascript. If you can, then pass the values along in the query string (just add a "?ValueName=Value") to the end of your links. Then on the target page retrieve the query string value. The following site shows how to parse it out: Parsing the Query String.
Here's the Javascript code you would need:
var qs = new Querystring();
var v1 = qs.get("ValueName")
From there you should be able to work with the passed value.
Javascript can get it. Say, you're trying to get the querystring value from this url: http://foo.com/default.html?foo=bar
var tabvalue = getQueryVariable("foo");
function getQueryVariable(variable)
{
var query = window.location.search.substring(1);
var vars = query.split("&");
for (var i=0;i<vars.length;i++)
{
var pair = vars[i].split("=");
if (pair[0] == variable)
{
return pair[1];
}
}
}
** Not 100% certain if my JS code here is correct, as I didn't test it.
You might be able to accomplish this using HTML Anchors.
http://www.w3schools.com/HTML/html_links.asp
Append your data to the HREF tag of your links ad use javascript on second page to parse the URL and display wathever you want
http://java-programming.suite101.com/article.cfm/how_to_get_url_parts_in_javascript
It's not clean, but it should work.
Use document.location.search and split()
http://www.example.com/example.html?argument=value
var queryString = document.location.search();
var parts = queryString.split('=');
document.write(parts[0]); // The argument name
document.write(parts[1]); // The value
Hope it helps
Well this is pretty basic with javascript, but if you want more of this and more advanced stuff you should really look into php for instance. Using php it's easy to get variables from one page to another, here's an example:
the url:
localhost/index.php?myvar=Hello World
You can then access myvar in index.php using this bit of code:
$myvar =$_GET['myvar'];
Ok thanks for all your replies, i'll take a look if i can find a way to use the scripts.
It's really annoying since i have to work around a CMS, because in the CMS, all pages are created with a Wysiwyg editor which tend to filter out unrecognized tags/scripts.
Edit: Ok it seems that the damn wysiwyg editor only recognizes html tags... (as expected)
Using php
<?
$passthis = "See you on the other side";
echo '<form action="whereyouwantittogo.php" target="_blank" method="post">'.
'<input type="text" name="passthis1" value="'.
$passthis .' " /> '.
'<button type="Submit" value="Submit" >Submit</button>'.
'</form>';
?>
The script for the page you would like to pass the info to:
<?
$thispassed = $_POST['passthis1'];
echo '<textarea>'. $thispassed .'</textarea>';
echo $thispassed;
?>
Use this two codes on seperate pages with the latter at whereyouwantittogo.php and you should be in business.