PHP Classes


Recommend this page to a friend!
  Classes of Timothy Edwards   PHP Parse DOCX to HTML with Images   Download  
Role: Documentation
Content type: text/markdown
Description: Documentation
Class: PHP Parse DOCX to HTML with Images
Convert Microsoft Word DOCX document to HTML
Author: By
Last change: Update

Now recognise various underline styles
Date: 4 months ago
Size: 10,474 bytes


Class file image Download

This is an improved class to read a Microsoft Word DOCX file and output it in HTML format for the web.


This improved Microsoft Word DOCX to HTML class will recognise nearly all the formatting, themes, images etc. in the original Word DOCX document. It will now also display any Mathematical Equations in the document either standalone or inline with normal text. The only significant exception is tabs as these are very difficult to replicate in HTLM due to a web page page width being very flexible and difference sizes of screens. The resultant HTML should look very much like the original.

For update notes detailing changes up to the latest version of 2.1.13 see below.

NOTE - Needs at least php 5 and will run on up to (at least) php 8.1.

NOTE:- It will not read the older 'DOC' Word format. In this case it will give an error message saying that it is a 'non zip file'.


1. It will use the correct font and font size (assuming that a common font is being used).

2. Text formating - Bold, Underlining (various styles/colours), Italic, Single & Double Strikethrough, Superscript and Subscript are all replicated, along with text alignment : Left, Centre, Right, Justified. Also indented and hanging text.

3. It will display multi-level lists with the correct alpha-numeric numbering as per the original word document.

4. It will now recognise paragraphs numbered with the Word paragraph numbering function as well as the list numbering function.

5. It will display tables with merged cells correctly along with border and cell colours etc.

6. In the default mode, tables are replicated as near as possible to the original DOCX word document formatting and size. However an option is provided for the tables to take up 100% of screen width (to allow for better display in mobile devices etc.).

7. Both footnote and endnote references are located in the correct place in the text. All the actual footnotes and endnotes are located at the end of the text. Links are provided to jump from a reference to actual note and then back again.

8. The bookmarks in a 'Table of Contents' or similar provide a link to the correct section of the document as per the original Word document. A return link is also provided.

9. As many Word documents are quite long a link is provided to jump back to the top of the document. Located at the middle of the right side of the screen.

10. In the default mode, images are formatted and sized very similarly to the original DOCX word document, which is fine for desktop computers. However an option is provided to allow for external CSS formatting to be used instead (e.g. to allow for better display in mobile devices etc.). In this mode each image is given a unique CSS class name - 'Wimg1' for the first image. 'Wimg2' for the second image, etc. to enable formatting of each image as desired. There is also an option to omit images from the resultant HTML if this is desired.

11. Will now recognise when images are cropped and display the correct cropped image.

12. By default images are saved into the 'images' directory, which is automatically created if it does not exist. An option is provided to enable the name of this directory to be changed if desired.

13. It will now recognise symbols from most of the symbol character sets used in Word (Wingdings, Wingdings 2, Wingdings 3, Webdings, Symbol, Zapf Dingbats). These in the main are not commonly available on the web. However most of the characters or equivalents are available in the Unicode character set so these are used instead. Available when using php 7.2 and above. Please note that not all browsers can display the full Unicode character set.

14. Will now recognise nearly all Word Mathematical Equations, both standalone and inline with text. It does this using the online version of Mathjax (so internet acces is required for this). Note that Mathjax does not support the Surface Integral and Volume Integral symbols, so multiple Line Integral Symbols are used instead. Also The Double Square Bracket is not supported, so any occurences of this are replaced by the Double Pipe.

15. The resultant html code is designed to be used either as is, or (after saving) included in another html file. However an option is provided to add a html header, so that after saving it can be used as a standalone file (along with any images that it contains).

If anyone finds any problems or has sugestions for enhancements, please contact me on


$rt = new WordPHP(false);
$text = $rt->readDocument('FILENAME','N');
echo $text;


Increase php memory limit -

Optional - Only needed when there are high resolution images in the Word document and PHP fatal errors due to using up all the available memory are produced

ini_set('memory_limit', '512M'); OR ini_set('memory_limit', '1G');

Include the class in your php script


debug mode

Will display the various zipped XML files in the DOCX file which are used by the class for the document being converted and will also display the resultant HTML.

$rt = new WordPHP(true);

without debug (Normal Use)

$rt = new WordPHP(false); or $rt = new WordPHP();

Set output encoding (Default is UTF-8)

You can alter the encoding of the resultant HTML - eg. 'ISO-8859-1', 'windows-1252', etc. Although note that many special chacters and symbols may not then display correctly so the default should be used whenever practical.

$rt = new WordPHP(false, 'ISO-8859-1');

Change directory for images (Default is 'images')

Will change the directory used for any images in the document.

$rt = new WordPHP(false, null, 'dir_name');

Read docx file and return the html code - Default mode

$text = $rt->readDocument('FILENAME','N');

Read docx file and return the html code - Option 1a

Images can be formatted using external CSS

$text = $rt->readDocument('FILENAME','Y');

Read docx file and return the html code - Option 1b

Images will be omitted

$text = $rt->readDocument('FILENAME','O');

Read docx file and return the html code - Option 2

Tables will be formatted to 100% of screen width

$text = $rt->readDocument('FILENAME','NY');

Read docx file and return the html code - Option 3

Resultant HTML code will have a standard HTML header added so that when saved, the saved file can be run directly in a browser. N.B if you wish to use this option in conjunction with option 1a then the CSS file should be named - word-htm.css

$text = $rt->readDocument('FILENAME','NNY');

Display the html code on screen

echo $text;

Save the html code to a file (if required)

$myfile = fopen("newfile.php", "w") or die("Unable to open file!");

fwrite($myfile, $text)


Version 2.1.13 - It will now recognise many of the various underlining styles in Word. It will duplicate the following styles:- Single, Double, Dotted, Dashed and Wavy together with their heavy/thick versions and coloured underlining. It will also now recognise 'double strikethrough. In addition a few bugs have been cleared.

Version 2.1.12 - Elements with subscripts etc. are now allowed in matrices. Also standalone mathematical equations can now be left or right aligned to the page or element (eg table cell) as per the Word document instead of always in the centre.

Version 2.1.11 - Enhancements and bug fixes - It will now recognise and display Mathematical Equations (using Mathjax). Also clears a bug in list/paragraph numbering and a bug with merged cells in tables. An option is now provided to omit any images from the resultant html code if this is required. Also an option to include a html header to the resultant code is provided for when the html code is saved and used as a standalone file.

Version 2.1.10 - It will now recognise when images are cropped and display the correct cropped image. Previously it displayed the full original image.

Version 2.1.9 - Enhancement - It will now recognise symbols from most of the symbol character sets used in Word (Wingdings, Wingdings 2, Wingdings 3, Webdings, Symbol, Zapf Dingbats) when using php 7.2 and above. These character sets in the main are not commonly available on the web/browsers. However most of the characters or equivalents are available in Unicode so these are used instead. Bug fixes:- 1. If the first few characters in a paragraph were bold and/or underlined, then the whole paragraph was. 2. If a blank line contained a single space, then the blank line was ignored.

Version 2.1.8 - Various enhancements and bug fixes. 1. Will now recognise multiple spaces, - 2. Will recognise pdf images. - 3. Will now tell you if the .docx file cannot be found. - 4. Will now recognise checkboxes. - 5. Will now recognise most common list bullets (assuming that UTF-8 is used). - 6. Multiple bugfixes (Table widths and alignments not always correct - Top and bottom margins of text not always displayed correctly - Several other minor fixes).

Version 2.1.7 - Modified to clear a lot of php Warning Messages that appear in some environments

Version 2.1.6 - Will now work on up to (at least) php 8.1

Version 2.1.5 - Various enhancements. 1. Will now operate much faster. Previously, was often several seconds, now will give the output in a fraction of a second. - 2. Will now recognise Drop Capitals. - 3. Will now recognise a Cross-Reference link in the document. - 4. Text formatting now implemented in the Footnotes and endnotes. - 5. Rectified a bug where sometimes hyperlinks were not recognised properly in footnotes and endnotes.

Version 2.1.4 - Clearance of a minor bug.

Version 2.1.3 - Clearance of a few bugs: 1. The endnote references in the text were not in roman numerals as they should be. - 2. In certain circumstances the correct font and font size were not being used. - 3. In certain circumstances a bookmark link did not work.

Version 2.1.2 - Will now recognise higher resolution images (Word deals with these differently to lower resolution images).

Version 2.1.1 - Will now recognise paragraph numbering as well as list numbering.

Version 2.1.0 - Improved method of displaying lists. Can now display footnotes and endnotes and their text references. The bookmarks in a 'Table of Contents' or similar, now link to the appropriate section of the document.