How to extract untagged text from html string with PHP

Question

Consider the following HTML code :

<strong>title</strong>
Hello World
<strong>Sub-Title</strong>
<div>This is just stuff</div>

How can I clean up the string to just return the string with no tags, i.e. 'Hello World'. I presume this is with DOM, and would prefer a non-regex answer if anyone has a way, and without using javascript or jquery.

[EDIT] Code it fails on.

<span style="color: #677b8d"><strong>Short Description</strong><br/>Microsoft Office Home and Business 2013-Word, Excel, PowerPoint, OneNote and Outlook(Does not include Publisher or Access), DSP , No Warranty on Software <br/><br/><strong>Long description<br/></strong><div>Microsoft Office Home and Business 2013 32-bit/x64 DSP No Warranty on Software </div> <font face="Arial"> <div><br/><strong>Product Overview </strong> <div><font face="Arial">The New Microsoft Office Home &amp; Business 2013 is designed to help you create and communicate faster with new, time-saving features and a clean, modern look. Plus, you can save your documents in the cloud on SkyDrive and access them virtually anywhere. The latest versions of Word, Excel, PowerPoint, OneNote plus Outlook on 1 PC.</font></div> </div> <div><strong><br/> Features<br/></strong><font face="Arial">•One time purchase for the life of your PC; non-transferrable.<br/> •Office on one PC for business and household use.<br/> •The latest versions of Word, Excel, PowerPoint, OneNote, and Outlook.<br/> •7 GB of online storage in SkyDrive.<br/> •Free Office Web Apps* for accessing, editing, and sharing documents.<br/> •An improved user interface optimized for touch, pen, and keyboard.</font> <div> </div> <div><font face="Arial"><strong>Specifications<br/></strong>Operating System Windows <br/> Office/Productivity Software Office Suites &amp; Tools <br/> Purchase Method Boxed <br/> Users/Devices per License 1-User <br/></font></div> </div> <div><font face="Arial"><strong>System Requirements:<br/></strong>Computer and Processor 1 GHz or faster x86 or 64-bit processor with SSE2 instruction set</font></div> <div> <p><font face="Arial"><strong>Memory<br/></strong>1 GB RAM (32-bit); 2 GB RAM (64-bit) recommended for graphics features, Outlook Instant Search, and certain advanced functionality**</font></p> <p><font face="Arial"><strong>Hard Disk<br/></strong>3.0 GB available disk space</font></p> <p><font face="Arial"><strong>Display<br/></strong>1366 x 768 resolution</font></p> <p><font face="Arial"><strong>Operating System<br/></strong>Windows® 7, Windows 8, Windows Server 2008 R2 with .NET 3.5 or later</font></p> <p><font face="Arial"><strong>Graphics<br/></strong>Graphics hardware acceleration requires a DirectX10 graphics card</font></p> <p><font face="Arial"><strong>Additional Requirements<br/></strong>Internet connection. Fees may apply.</font></p> <p><font face="Arial">Microsoft Internet Explorer 8, 9, or 10; Mozilla Firefox 10.x or a later version; Apple Safari 5; or Google Chrome 17.x.</font></p> <p><font face="Arial">A touch-enabled device is required to use any multi-touch functionality. However, all features and functionality are always available by using a keyboard, mouse, or other standard or accessible input device. New touch features are optimized for use with Windows 8.</font></p> <p><font face="Arial">Information Rights Management features require access to a Windows 2003 Server with SP1 or later running Windows Rights Management Services.</font></p> <p><font face="Arial">Microsoft and Skype accounts.</font></p> <p><font face="Arial"><strong>Other<br/></strong>Product functionality and graphics may vary based on your system configuration. Some features may require additional or advanced hardware or server connectivity.</font></p> <p><font face="Arial">*An appropriate device, Internet connection and Internet Explorer, Firefox or Safari browser are required.<br/> **512 MB RAM recommended for accessing Outlook data files larger than 1GB<br/></font></p> </div> </font></span>

@mudasobwa, nope, cos striptags will leave 'title', 'Sub-Title', etc — Gavin Simpson
– Gavin Simpson, Commented Nov 24, 2014 at 12:43
Yes, that's it. The bottom line is if I can remove all tags plus their nodeValue I should be left with a result. — Gavin Simpson
– Gavin Simpson, Commented Nov 24, 2014 at 12:46

Aleksei Matiushkin · Accepted Answer · 2014-11-24 12:59:53Z

1

I would suggest you to surround the code with kinda exotic tag, which is definitely not occured in the code itself, like:

 $a="<body><strong>title</strong>\nHello World\n<strong>Sub-Title</strong>\n<div>This is just stuff</div></body>";

Then use DOM:

$doc = new DOMDocument();
$doc->loadHTML($a);
$xpath = new DOMXPath($doc);
$textnodes = $xpath->evaluate('//body/text()[not(normalize-space() = '')]');

Now you may get whatever you want:

foreach( $textnodes as $el ) {
  print_r($el);
}

/*
DOMText Object
(
    [wholeText] => 
Hello World

    [data] => 
Hello World

    [length] => 13
    [nodeName] => #text
    [nodeValue] => 
Hello World

    [nodeType] => 3
    [parentNode] => (object value omitted)
    [childNodes] => 
    [firstChild] => 
    [lastChild] => 
    [previousSibling] => (object value omitted)
    [nextSibling] => (object value omitted)
    [attributes] => 
    [ownerDocument] => (object value omitted)
    [namespaceURI] => 
    [prefix] => 
    [localName] => 
    [baseURI] => 
    [textContent] => 
Hello World
*/

answered Nov 24, 2014 at 12:59

Aleksei Matiushkin

121k12 gold badges109 silver badges174 bronze badges

Sign up to request clarification or add additional context in comments.

8 Comments

Gavin Simpson Over a year ago

Odd one - when using your sample data it works fine, but when I use my html loaded via file_get_contents it returns a blank. Same with Ghost's answer below. I have added the <body> tags. I am still looking into it but just wanted to update and thank you both for your help.

Gavin Simpson Over a year ago

Close, but no (I think). I think you both made a typos as I do get all the text from all tags if the query contains two slashes, ie "$textnodes = $xpath->evaluate('//body//text()[not(normalize-space() = '')]');", but I want all text with tags removed.

Aleksei Matiushkin Over a year ago

Can’t say for another answer’s author, but I definitely did not make any typo. //body/text() states for “any body element immediately followed by text node.” Would you mind to print out the data from file_get_contents?

Gavin Simpson Over a year ago

Ok, thanks. Other guy deleted his answer. For whatever reason. I cannot get it to return and data with the one slash, but if I add two I get the same results as your example. It is still not helping me remove all the text that do contain tags. It would help if I could be assured the format of all the html input was the same, but I cannot, so cannot just use array indices to decide on what to use. I must remove the other text with tags.

Aleksei Matiushkin Over a year ago

Of course it fails on this code. This code has no top-level text nodes. The whole code is surrounded with span. So, if you want to find text nodes with this portion of code, use '//span/text()[not(normalize-space() = '')]'—it will return each text node, directly nested by top-level span.

|

Collectives™ on Stack Overflow

How to extract untagged text from html string with PHP

1 Answer 1

8 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

8 Comments

Your Answer

Sign up or log in

Post as a guest

Related