Word 2003 produces some pretty horrible code too and less web functionality – I’m informed that Word 2007 has improved this somewhat, but still has some way to go. (If you have experience of Word 2007 and Quark, please leave a comment on this page.)

Here’s an illustration. Open up Notepad (or similar) and type the following:

<p>Hello</p>

Select ‘Save As’ then type ‘file.htm’ into the filename box, which will produce an HTML file with the icon of your default browser. Double-click the file to view the results. As a comparison, open up MS Word, type in ‘Hello’ and save as a webpage. Then open up the file in Notepad to see the code. You’ll agree that it’s awful and there’s a lot of it. (QuarkXPress is no better.)

The straightforward way to publish online is to make a web-optimised PDF, but there are also good reasons to do it in HTML. There are proprietary tools that will produce nice code for you, but buying such a product might not be an option.

What I’m going to describe is tricky and a bit of a work-in-progress. It’s not an elegant solution, but it works okay and will hopefully save you some time. It’s a better alternative than making a text-only file and then formatting paragraph by paragraph.

Why is ‘nice’ code important?

The main reason is Cascading Style Sheets (CSS). Like the style functions in Word and Quark, they remove the need for manual formatting and keep the content and formatting separate. Without formatting to worry about, all you mostly need to do is produce bog-standard HTML (with perhaps some additional styles created by whoever wrote the CSS). This is the whole principle on which WordPress and other blogging platforms are founded and the reason why we can change our page templates so easily.

Additional reasons to use standard code are accessibility and the desirability of keeping file sizes down.

The aim of this tutorial

This tutorial will essentially describe how to convert a long text-string from Quark into HTML. It is likely that you’ll still need to go into the code to fix some glitches and you’ll have to add any images, tables and suchlike manually. It’s a bit of a hack solution, but it has worked for me. As always, if you have any suggestions, please email me or leave a comment.

Structural considerations

The structure of HTML is essentially similar to any large document. In a book, you’ll have a visual hierarchy of styles where, for example, a chapter heading is superior to a main heading, which is in turn superior to a subheading. Likewise in HTML there are headings <h1> to <h6>. Likewise you’ll be using ordered (numbered) lists and unordered lists (bullet points), corresponding to <li> tags, which in turn are wrapped in the <ol> and <ul> tags respectively. Your body copy will correspond to the HTML paragraph style <p>. In a book or report, the chances are those tags and styles will correspond to the vast majority of your document. The trick is to produce plain text that is wrapped in these tags.

Step-by-step HTML from Quark

1. Get rid of all formatting, numbers and bullet points

To guard against any residual fonts creeping into the HTML, convert all the text to 12pt Times New Roman. You can do either do this in the style sheets or by deleting fonts in the Usage box and replacing them with Times New Roman. Use find and replace to get rid of all the bullet points in the document. You’ll also need to get rid of all the numbers in numbered lists. (Of course, this assumes your bulleted lists and numbered lists are built using styles.)

Bear in mind the aim is to produce HTML — the CSS and the browser will do the formatting for you.

2. Save the text as MS Word

Save to Word using File>Save Text (if you’re unsure how to do this, it’s on the same principle as the method discussed here). Your Quark styles will follow through into the Word version

3. Apply Word styles to the text

Open up your new word file. In Word, Format>Styles and Formatting will display the styles. Now, this is where it gets a bit tricky as Microsoft has generally overcomplicated this function and it’s a mess. You may have to go to the bottom of the Styles and Formatting pane and select Show: Custom, then display all styles.

The heading styles you’re looking for are Heading 1 to Heading 6, which correspond to <h1> to <h6> in HTML. By convention, Heading 1/<h1> is the title. Click on your main heading style, select ‘Select All’ on the drop-down menu, then apply Heading 2. Apply Heading 3 to the next style and so on.

For body text, you need to apply the style Normal (Web).

4. Bullets and numbers

Bullets and numbers are tricky – try as I might, I couldn’t find a solution that produced the right tags. The solution is a hack – select the styles that apply to bullet and numbers and apply one of the near-redundant html styles on the list (I use HTML typewriter). You may want to apply others if you have well-used styles, although keep a note of what they are.

5. Save as html and fix the code

Saving as ‘Web page, filtered’ seems to produce better results than save as ‘Web page’. You can then open up your file in Notepad to do some more fixing.

• All the code before the first line of copy and after the final line can be deleted. If you use a content management system like WordPress, you don’t need them. If you’re passing on the html to a webdesigner, then he or she can sort it out. If you’re doing it yourself, you’ll know what code to use
• The bullets and numbers, to which you’ve applied an redundant style, will have code like this:

<p class=MsoNormal><tt>Bullet point 1</tt></p>

You can fix this using find and replace. In this example, replace <p class=MsoNormal><tt> with <li> and change </tt></p> to </li>. These will all display as bullets in a browser. To make strict HTML, you’ll have to manually add <ul> and </ul> before and after these list tags to make proper bullets and <ol> and </ol> to make numbers.

• You’ll have to do a similar hack for any other styles you’ve defined in this way with a different format. You may want to use <blockquote> for example.

6. Open up in a web browser and check your work

What you’ll see won’t look that special – the text will take on the default characteristics of your browser – but hopefully the structure will be okay. If anything looks awry, you’ll have to go into the code and fix it. The most usual problem is font tags creeping into the HTML, which you don’t want. Other tags in your original file, such as ones for superscript, italic and so on may also have gone missing.

If you place the cursor inside a text box and go to File>Save Text, QuarkXPress will let you save the entire story — the contents of the text box and all those linked to it. If you select some text additional options will appear; either to save the selected text or the entire story.

You are presented with a number of options in the ‘Save as Type’ drop-down menu, although you will presumably want to save to the most up-to-date version of Microsoft Word.

There are sometimes good reasons to do this — if authors decide to do a substantial rewrite, or wish to update a previous version of a publication, it makes sense to let them edit the text directly. Of course, this may only be a good thing if your publication is essentially a single, long text thread. You may wish to sit down with authors and explain how the formatting works, to avoid having to tidy up their text too much at a later stage.

Depressingly, the other reason you get asked is because some people assume there’s some magic feature that transforms a Quark document into a faithful Word version (or that you’ve managed to do it in Word in the first place). And there obviously isn’t. Any independent elements such as manual footnotes, tables, standalone text boxes and images will not export.

What to look out for in your Word export

• The good news is that exporting to Word from Quark brings all your Quark styles into Word. This is a huge time-saver, as your styles should be intact when you import it back again using File>Import Text
• It’s likely that you’ll be using fonts the author doesn’t have and these will carry through into the Word version. There are two options here. The first is to go to Tools>Options>Save in MS Word and embed the fonts in the document. If that’s unsatisfactory, then edit the Word style sheets to change the fonts to standard Windows (or Mac) ones. When you import the text back into Quark, it should change them back to the fonts you defined in your Quark style sheets
• Certain Quark-specific functions, for example thin spaces and non-breaking hyphens, aren’t recognised by Word. They will probably export as square-shaped characters in the text. However, they should be restored when you reimport into Quark, so tell your authors to ignore them
• The Word export will convert your defined colours to RGB. Although coloured text should be converted back to the colour defined in the style sheet, the RGB value will appear in your colours palette and may be lurking somewhere in your document. To be on the safe side, delete this colour from the palette and replace it with the correct one
• Any text not defined by a style in Quark tends to look horribly unformatted in the Word version
• The process is not foolproof and you should still check the Quark version carefully after you reimport the text. Don’t come running to me if it goes wrong!

Actually, Find and Replace is one of the better features of MS Word. I consider myself pretty decent on it, but there’s far more functions in the Help screens than I ever use. The important thing to bear in mind is that you can find and replace things far beyond letters, numbers and words. In addition, I would recommend doing all your finding and replacing in Word as, in terms of working on copy, the functions are more powerful than the equivalent ones in Quark.

Anyway, here’s the box. The full functions are revealed when you press the ‘more’ button.

The non-keyboard elements you can search for are all located in the ‘special’ button and enter search terms in the relevant box, usually in the form of the ‘^’ symbol followed by a single letter. For example, ‘^p’ is the term for a paragraph mark (i.e a carriage return).

Some standard fixes

There are three find and replace functions I do on any new document. I actually have macros set up to do these on individual buttons on my toolbar. How to do macros is a bit beyond the scope of this post, but this guide on about.com is pretty good.

1. Replace all double spaces with single ones
   There shouldn’t really be any double spaces at all in a document. Many people believe there should be a double space after a full stop, although this is typographically incorrect. I believe it’s something people were taught to do in the days of typewriters. (A double space looks particularly horrible in a fully justified column.)
2. Replace hyphens with en dashes
   Another common typographical mistake. To avoid getting rid of genuine hyphens, enter spaces on either side of the hyphen and the dash in the find and replace box. (If your house style is to use em dashes, then the space either side of the em dash isn’t necessary.)
3. Get rid of instances where a space follows a paragraph mark
   This won’t appear to be that important, but it will make more sense when you read the example below. In MS Word, you’re replace ‘^p[space]’ with ‘^p’ in the find and replace box.

A typical problem

Sometimes — usually the result of putting text through too many clipboard functions or converting formats too many times — you get a Word document where every line ends in a carriage return. Namely, when you show the hidden formatting, it looks like this:

This would obviously play havoc with formatting and I expect many people have been reduced to fixing this line-by-line. The good news is that there’s a straightforward fix for this. The approach is an example of the lateral thinking you sometimes need when faced with similar problems.

Okay here’s the fix. From looking at the copy, each paragraph proper is separated by two paragraph marks (the invisible ¶ symbols). Firstly, I’d recommend running the three find and replaces explained above.

As we’ve seen, these non-printing characters are searchable. Selecting the paragraph mark option twice from the ‘special’ menu will give you the codes ‘^p^p’ in the find box. You then need to replace this with a character (any character) that you’re confident doesn’t appear in the document. I usually use the dollar symbol, although US and Australian readers may wish to use something else. When you press replace, everything may turn a bit scary, but don’t worry.

You can then replace every single paragraph mark with a space. Once you’ve done that, change the new character you’ve introduced back into a paragraph mark (I recommend using a single paragraph mark, as paragraphs should really be separated by setting a distance in the paragraph formatting box). And there you go. So, to summarise:

• Replace all instances of two paragraph marks with an unusual character, such as $, # or {
• Replace all single paragraph marks with a space
• Replace the new character you’ve introduced with a paragraph mark.

Try this out yourself to experiment — you’ll soon get the idea.

The product is called qXport and the link to the English version of the site is here (although not all the links on this area of the site seem to work). There is a free trial version, which I might have a play with. If anyone’s any experience of this, please post a comment.

If it’s at all possible, it can be a real timesaver in the long run to get together with authors and point them in the right direction before they start writing. And obviously, the value of preparation is in direct proportion to how long and complicated the publication is going to be. Anyway, here are some suggestions for making the process a bit less painful.

Styles and formatting

This is probably the most important thing of all — getting your authors to use the Styles and Formatting functions in MS Word saves an awful lot of time. You can use this basic template I posted on an earlier entry.

Spend ten or fifteen minutes explaining the concepts and make sure the authors understand how to use them. Even if you don’t want the numbered headings that are on the template, I would still advocate them, as they make authors think about the structure of their documents — if left alone, they tend to throw in headings willy-nilly, without thinking how they fit into the bigger picture. (And it only takes a minute to turn off all the numbering when the writing’s finished.) Updating the table of contents allows them to see how their publications are building up — hopefully saving you the bother of trying to fix the structure later.

Style guides

Your organisation will no doubt have a style guide, to encourage good writing practice and consistency. However, less-experienced authors have a tendency to be daunted by them, or to ignore them completely. The best compromise may well be to go through the most common mistakes and inconsistencies. This is especially important if a publication has multiple authors.

In my experience, the biggest issue is usually capitalisation and the second, for more technical publications, is referencing. Unfortunately you can’t stop them writing drivel, but if you’re an editor, then that’s what you’re paid for.

Graphs, images and tables

This may make you unpopular, but there are several good reasons for keeping graphs, images and tables out of MS Word files.

• Upon import into Quark, they at best get mangled and at worst disappear. And often leave stray captions in the text
• A lot of authors have a terrible habit of writing things like ‘the table below’ or ‘the graph above”, which would be fine if you were printing on a roll of paper, but useless if you’re laying out a book, especially one with columns. Removing them will hopefully remind them to cite tables and pictures properly (as in ‘see Table 3’) and also to caption them.

Images

Bitmap images in Word don’t import into Quark and placing them in Word degrades them anyway.

Graphs

How authors generate graphs isn’t of vital importance as they’ll probably need redrawing anyway. What is important is ensuring that the person creating the graph for print can see the data that was used to generate it. Despite Microsoft’s Object Linking and Embedding, a lot of Excel graphs become pictures when they’re pasted into Word.

Tables

Upon import, tables get converted into tab-separated text. In the old days, this was the only way to make tables in Quark, but nowadays it’s better to use the automatic table functions present on later versions. Unless the table is very small, I’ve actually found the best way is to get authors to do their tables in Excel which, in my experience, imports into Quark quite well (more here).

Flowcharts

Unfortunately you’re almost certainly going to have to recreate these from scratch. Sorry!

Some basic typography

Try to get authors to avoid incorrect and unnecessary formatting. Some examples are:

• Not using double spaces after full stops
• Not pressing enter twice to start a new paragraph
• No unnecessary formatting such as title pages, manual page breaks, section breaks and the like
• Not using a hyphen where a dash should go (MS Word will convert two hyphens into a dash)

Footnotes

Not that there’s anything wrong with using footnotes, but unfortunately Quark doesn’t really support them. They will import, but Quark puts them all at the end of the imported text string. If there are only a few footnotes, it’s not a massive effort to place them manually, but for a document with a lot of references, this can be very time consuming. For this reason, you might want to suggest the author puts references at the end of a section or chapter.

It’s certainly possible to output a PDF (or PostScript file) from MS Word, but it presents several technical and aesthetic problems:

1. MS Word just doesn’t handle fonts very well, compared to a proper DTP package, and lacks the fine-tuning controls. Your finished product is always gong to look like an office document, albeit on nicer paper
2. The quality of images imported into MS Word gets degraded quite significantly. You’d probably be disappointed by what they’d look like on the finished product
3. Word struggles with image formats such as TIFF and EPS
4. The colour model of MS Word is RGB, whereas a printer needs the job in CMYK. There’s also no support for Pantone colours.

So, to summarise, unless you’re doing a simple black and white text-based project, don’t touch it with a bargepole. In fact, this blog post explains it much better than me.

I think the only possible exception would be an entirely text-based publication in black type (such as a novel), if you were self-publishing a book, for example. However, if you’ve set up your Word file properly, using Styles and Formatting, it wouldn’t be a big job for a designer to typeset it and it would look much nicer. And you’d need them to do your book cover anyway.

As you can see, I’ve kept the file as simple as possible (and the Word formatting is irrelevant to what the final Quark layout will look like anyway). It was created in MS Word 2003 — it will open on a PC, I’ve opened it in OpenOffice without any problems and it should open on Word for Macintosh. What you need to do is to get your authors to follow the template as closely as possible.

At the top of the page is an automatic table of contents. It’s generated from the headings in the document — to update it, periodically right click within the TOC and select Update Field. (You can choose to update the page numbers or the whole TOC.)

The document has been put together using Word’s Styles and Formatting features, which can be displayed using the View>Styles and Formatting menu. I’ve prefixed the names with “TOMP” for the simple reason that it keeps the styles together when they get listed in alphabetical order.

So, the principle is straightforward — write some text, click on the style in the right-hand pane and all the formatting is done for you.

You’ll notice the headings are numbered and there are two reasons for this. Firstly, numbered headings are useful if you’re producing a long, more technical document that will be cross-referenced. Secondly, it forces authors to think about the structure of their documents, as in my experience, they usually tend to throw in headings and subheadings at random. If you don’t want them in your final publication, it’s very quick and straightforward to turn the numbering off.

It is possible that you might need to alter the look of the Word document. Let’s say you want to make the Level A headings red. The easiest way is to recolour one of the headings red, go to the Level A style in the right-hand pane, click the drop-down box and select ‘Update to match selection’. All the Level A headings will now be red.

I have to be honest, this feature in Word can be a real pain, although despite all the glitches with it, it’s still worth sticking with. MS Word often invents new, irrelevant styles, which quickly fill up the right-hand pane. The only way to get rid of them is to click on the ‘Show’ drop-down box right at the bottom of the Styles and Formatting pane, then select ‘Custom’. You can then just display the ones you need — in this template, just the ones preceded by ‘TOMP’.

However, even with this layout, you will need to impress on your authors (and maybe yourself) several things:

• This template essentially formats itself, in terms of spaces between headings and paragraphs. Don’t start a new paragraph by hitting return twice. If you really need a new line, do a manual line break (Shift-Return)
• Using bold and italics is fine (although from an editorial point of view, too much bold text is best avoided)
• Don’t use double spaces after full-stops
• Set up the Autocorrect options on Word to ensure hyphens are converted into dashes and three dots into ellipses. It’s a good idea to turn off the automatic hyperlinks as well
• There is a style for doing normal numbered lists. There is one thing to look out for — if, for example, you’ve numbered three paragraphs 1-3, then start another numbered list later in the document, the numbering will start at four. To get round this, right-click the relevant paragraph and select ‘Restart numbering’
• Stick to the preset styles absolutely — of course you may need a few more paragraph styles, but set these up in the Styles and Formatting as opposed to doing them manually
• Avoid things like page breaks and section breaks
• If possible, put tables, graphs, figures and images in separate files. Big tables can be done in Excel and imported into Quark. Graphs will probably have to be redrawn (and won’t import into Quark anyway).
  At best, things like this will disappear on import; at worst, they’ll just get mangled, interrupt the flow of text and have to be removed.

Once you or your author has finished the document, it can be imported into Quark using Get>Text. Ideally, you’ll have set up styles in Quark with identical names and they’ll get converted to their new formats.

Before you do, however, please read my posts here and here, which address some issues with importing text and show how to fix some glitches.

For more advice on MS Word, Shauna Kelly’s site is very comprehensive.

The bullets and numbering problem

Importing text has one big problem — for some reason, using Import Text causes you to lose the automatic bullet points and numbering in MS Word, which can be a real headache especially if your publication has numbered headings and lots of bullet points.

Thankfully, there is a way around this and it means setting up a macro in Word. There doesn’t seem to be a straightforward way of saving macros, so I’ll have to talk you through some code.

Open up Notepad and paste in the following text:

Attribute VB_Name = "NewMacro" 

Sub ConvertAutoNumsandBullets()
'
' Numsbullets Macro
' Macro recorded 07/04/2006 by Brookes
'
    ActiveDocument.ConvertNumbersToText
    Selection.MoveRight Unit:=wdCharacter, Count:=1, Extend:=wdExtend
    Selection.Copy
    Selection.MoveLeft Unit:=wdCharacter, Count:=1
    Selection.Find.ClearFormatting
    Selection.Find.Replacement.ClearFormatting
    With Selection.Find
        .Text = ChrW(61623)
        .Replacement.Text = "·"
        .Forward = True
        .Wrap = wdFindContinue
        .Format = True
        .MatchCase = False
        .MatchWholeWord = False
        .MatchWildcards = False
        .MatchSoundsLike = False
        .MatchAllWordForms = False
    End With
    Selection.Find.Execute Replace:=wdReplaceAll
End Sub

Save the file as NewMacro onto your desktop.

In MS Word, go to Tools>Macro>Visual Basic Editor. In the left hand pane, right-click on the ‘Normal’ icon (you may have to select View > Project Explorer to make this visible). Select Import File, then select the Notepad file you’ve just created (you’ll have to select Files of Type>All Files to make it visible). Close the editor, clicking Save if prompted to do so.

The next step is to make a shortcut to this macro via a button in MS Word. Select Tools>Customize, then select the commands tab. Select Macros in the left hand pane then find the new file — it will be called Normal.NewMacro.ConvertAutoNumsandBullets.

Drag this onto one of your toolbars at the top of the screen. The new button will have a long and unwieldy name — with the Customize box still open, right-click on the new button and select Default Style. You may want to right-click again and select Change Button Image, to display a distinctive icon. (If you’re really motivated, there’s a function in the drop-down menu to make your own button.)

I found this macro sometime last year — someone had thoughtfully posted it on the web (I’d gladly credit them if I could find it again). Running the macro, by pressing the button you’ve just made, converts all automatic numbers and bullets into editable characters, which will import into Quark.

Of course, this command is not reversible at a later stage, so keep a backup of your original file just in case.

A strange problem with indents

Another odd glitch I’ve noticed comes when you’re importing numbered headings with indents, of the form 1[tab]Introduction, for example.

QuarkXPress has an odd habit of making the indent spacing smaller than you’ve defined it in your Quark styles.

Initially, I had to fix this heading-by-heading, but a better way is to go back to your Word file and amend all the paragraph and tab settings so they match your settings in QuarkXPress. It will only take a minute or two to amend the Word style sheets then reimport into Quark.

If you regularly use the same QuarkXPress templates, you can import MS Word files into Quark without too much fuss, as long as your styles have the same names in both programs.

When the Import Text dialog box appears, be sure to select ‘Include Style Sheets’. If you have set up corresponding styles, you’ll get a conflict box in Quark informing you the names are in use. Simply tick the ‘Repeat For All Conflicts’ box and select ‘Use Existing’.

If you’ve only set up styles in Word, you can still import them and edit them in Quark.

Which Word features import into QuarkXPress?

Actually, not a great deal, but you need to think of MS Word as essentially a text editor, not a page layout program.

Things that import well

• Anything associated with the font menu, such as typeface, type size and formatting such as bold, italic, underline, superscript, subscript and so on
• Paragraph breaks, line breaks, tabs, section breaks, page breaks
• Word processor styles, assuming you enable this on the File>Import Text dialog box

Things that don’t import

• Images
• Fields
• Headers
• MS Word’s automatic bullets and numbers (although there is a workaround for this problem, which is discussed here)

Things that import imperfectly

• Tables — the contents of tables do come through, but appear as characters separated by tabs. If the tables are fairly small, then you’re probably better off recreating them using Quark’s table function. If they’re big, it’s worth getting your authors to do them in Excel, as Quark can import Excel tables
• Footnotes — the footnote text is imported, but all the footnotes appear at the end of the text string and not on the relevant pages. There are Quark XTensions that enable automatic footnotes (such as this one), but realistically you’re probably stuck with inserting them manually. This is quite a serious deficiency in QuarkXPress
• Hyperlinks — MS Word’s automatic hyperlinks come through as blue, underlined text, which you’ll have to get rid of one way or another. It’s a good idea to get authors to turn this function off in Word.

Anything else?

• Many authors collaborate using Word’s Track Changes function. Before you import text into Quark, ensure this is disabled by using the Accept All Changes in Document and Delete All Comments in Document functions.

Now, it’s very easy to be down on Microsoft, but I’m going to try and be balanced. MS Word is very good for composition. Its Autocorrect and spellchecker functions are very good and even its grammar checker throws up useful suggestions sometimes. It has a very powerful find and replace function, which I’ll discuss in detail at a later date, and can generate things like tables of contents.

Where MS Word is really lousy is formatting. It’s really not a programme with which to attempt any sort of even moderately ambitious page layout. Word tries to be all things to all people — usually unsuccessfully — and the trick is to know which features to turn off and which features to ignore.

What you really need to get your head round are its Styles and Formatting features. They’re neither straightforward nor foolproof, but will really save a lot of time in the long run.