How to convert addresses into a Mail Merge Data Source

Article contributed by Beth Melton and Dave Rado

Suppose you are given a document of address labels in table cells, each containing one name and address:


Figure 1

Although, unfortunately, you're more likely to have to work with something like the following (if you aren't familiar with non-printing characters such as , you might want to read: What do all those funny marks, like the dots between the words in my document, and the square bullets in the left margin, mean?):


Figure 2

Or what if you are given a document containing addresses separated by blank paragraphs, as in:

JohnW.Doh
15AnywhereLane
Somewhere,SomeState,99999

Mr.HomerSimpson
NuclearSafetyInspector
742EvergreenTerrace
Springfield,
IL62707

Figure 3

Or a mixture of line and paragraph breaks:

JohnW.Doh
15AnywhereLane
Somewhere,SomeState99999

Mr.HomerSimpson
NuclearSafetyInspector
 742EvergreenTerrace
Springfield,
IL62707

Figure 4

Quite apart from the inconsistencies, and the unwanted spaces and so on, in the above examples, none of these formats are very helpful ways of storing data – what you really need is a Mail Merge Data Source. With that, you can easily add, delete, sort and filter records and you can use the data to create any type of mail merge document (letters, labels, faxes, emails, and so on).

If you aren't familiar with finding and replacing non-printing characters, you might want to look at Finding and replacing non-printing characters (such as paragraph marks), other special characters, and text formatting before continuing.

First find out whether a data file already exists ...

The chances are, when someone sends you a file like that, that whoever created the file originally performed a mail merge in order to create it. Ring them up and ask them. If they used a spreadsheet, or a Word datasource, you're in business – ask them to send you that. If they used a company database, ask them if they could get a CSV (comma separated variable) file exported from the database with those names and addresses. If they can get that arranged, you're in business. 

But if you're stuck with having to do things the hard way, here are the steps to follow (note that whilst the Find and Replace operations described here are easy to automate – just perform them manually with the macro recorder on, then play back the macro whenever you want – other parts of the procedure are impossible to automate).

1.

First, if there are any tab characters, get rid of them: press Ctrl+H to bring up the Find and Replace dialog:

   in the Find what box, type ^t
   leave the Replace with box blank
   click Replace All.

If any tabs are replaced, click Replace All again until nothing is found. Once you've done that ...

If you're starting with labels (as in Figures 1 and 2)

2.

We want the address items to be separated by tabs, so ...

If the address lines are separated by paragraph breaks as in Figure 1:

   in the Find what box, type ^p
   in the Replace with box, type ^t
   click Replace All.

... to replace the paragraph marks with tabs.

Or if they are separated by line breaks:

   in the Find what box, type ^l
   in the Replace with box, type ^t
   click Replace All”.

(If a mixture of both, perform both replace operations one after the other).

3.

Now select Table/Convert Table to Text, and where it says Separate text with, select Paragraph marks.

4.

Now strip out any redundant paragraph marks:

   in the Find what box, type ^p^p
   in the Replace with box, type ^p
   click Replace All”.

Without closing the dialog, repeat until nothing is found.

You will now have something like this:

JohnW.Doh ➝ 15AnywhereLane ➝ Somewhere,SomeState,99999
Mr.Homer Simpson ➝ NuclearSafetyInspector ➝ 742EvergreenTerrace ➝ Springfield, ➝ IL62707

If you're starting with paragraphs (as in Figures 3 or 4)

If your paragraphs are laid out as in Figure 3:

2.

Replace the double paragraph marks with $ symbols:

   in the Find what box, type ^p^p
   in the Replace with box, type $
   click Replace All”.

3.

Replace the remaining paragraph marks with tab characters:

   in the Find what box, type ^p
   in the Replace with box, type ^t
   click Replace All”.

4.

Now replace the dollar signs with paragraph marks: 

   in the Find what box, type $
   in the Replace with box, type ^p
   click Replace All”.

If your paragraphs are laid out as in Figure 4:

2.

Replace the line breaks with tab characters:

   in the Find what box, type ^l
   in the Replace with box, type ^t
   click Replace All”.

You will now have something like this:

JohnW.Doh ➝ 15 Anywhere Lane ➝ Somewhere,SomeState, 99999
Mr. Homer Simpson ➝ NuclearSafetyInspector ➝ 742Evergreen Terrace ➝ Springfield, ➝ IL62707

Tidying up before converting to a table

Whichever of the above sets of steps you followed, you will now have a set of paragraphs, each containing fields delimited by tabs, as shown above.

At this point, you may wish to split the first name and last names into two fields, using the technique described here: Split First and Last Name.

Next we need to remove as many of the remaining anomalies as possible using Find & Replace (again, you can record these steps as a macro, and play it back whenever you want; and in the unlikely event that your data has no inconistencies in it to start with, you may be able to miss out some of these steps ).

1.

Replace commas followed by a <space> with a tab character: 

   in the Find what box, type a comma and press the spacebar (, )
   in the Replace with box, type ^t
   click Replace All”.

Then remove any commas that are followed by a tab:

   in the Find what box, type ,^t
   in the Replace with box, type ^t
   click Replace All”.

Then remove any commas that are followed by a paragraph mark

   in the Find what box, type ,^p
   in the Replace with box, type ^p
   click Replace All”.

Finally, replace any remaining commas (which will now be of the Somewhere,SomeState variety) with tabs:

   in the Find what box, type ,
   in the Replace with box, type ^t
   click Replace All”.

2.

Replace any double spaces with single ones:

   in the Find what box, press the spacebar twice
   in the Replace with box, type press the spacebar once
   click Replace All”.

Without closing the dialog, repeat until nothing is found.

3.

Replace redundant spaces at the beginning or end of a field: 

   in the Find what box, type a <space> followed by ^t
   in the Replace with box, type ^t
   click Replace All”.

   in the Find what box, type ^t followed by a <space>
   in the Replace with box, type ^t
   click Replace All”.

   in the Find what box, type a <space> followed by ^p
   in the Replace with box, type ^p
   click Replace All”.

   in the Find what box, type ^p followed by a <space>
   in the Replace with box, type ^p
   click Replace All”.

4.

You may also want to replace Mr. with Mr and Mrs. with Mrs.

You will now have something like this:

JohnW.Doh ➝ 15 Anywhere Lane ➝ Somewhere ➝ SomeState ➝ 99999
Mr. Homer Simpson ➝ NuclearSafetyInspector ➝ 742Evergreen Terrace ➝ Springfield ➝ IL 62707
  

Converting to a table

Now begins the real nightmare ... 

Select Table/Convert Text to Table, and in the dialog, choose Separate text at Tabs. In our example, you'll get a table like this:

Figure 5

Although if you're incredibly lucky, you might have something more like this:

Figure 6

Splitting the columns where necessary

1.

You can split the First name and Surname into separate columns as follows:

a)

Select the column (Alt+Click) and Cut & Paste as a new table. (Add a blank paragraph after the first table before pasting the second one, or paste into a new blank document).

b)

Select the table (Alt+Double-click) and go to Table/Convert Table to Text. In Separate Text at, specify Paragraph Marks and click OK.

c)

Go to Table/Convert Text to Table In Separate Text at in the Other text box type a <space> and click OK. 

d)

If your data resembles Figure 6, you're in business; but if it is closer to Figure 5, with some rows containing middle names or titles and others not, you'll now have a table that looks something like this:

... in which case you need to:

i)

Highlight column 1, and click on the (context-sensitive) Tables button on the Standard toolbar, to insert a new column

ii)

Highlight Mr and Homer simultaneously and drag them one column to the left; and continue to drag and drop your data, row by row, into the appropriate columns, until all the fist names and surnames line up:

Now select all of the columns in the temporary table (by selecting the leftmost column and then dragging to the right to extend the selection – don't select the entire table, or the results won't paste correctly). Cut, click at the beginning of the first cell in the main table, and Paste.

2.

The same principle can be used to split any other columns that need splitting up; for instance, if you have City, State, Zip in one column.

3.

Next, insert any additional columns that may still be needed; for example, in Figure 5, you would need to add a column to cater for the JobTitle field.

4.

Insert a new row 1, and add the appropriate field names, giving you something like this:

5.

You may need to make the page landscape in order to see it all in Page/Print Layout View. You may also want to select the top row and select Table/Headings (Word 97), or Table/Heading Row Repeat (Word 2000 and above), so that the top row will appear at the top of every page.

6.

Finally, (and if you have a large amount of data, this is often the most time-consuming part of the entire procedure), drag and drop the data into the correct columns; so that you end up with something like this:

(In the process of doing so, you may find that you still don't have enough columns – but you can add an extra column at any stage).

You should now have a table that can be used as a mail merge data source. If it contains a large number of rows, you can paste it straight into Excel, which will give you much better performance.


Click to view Terms of Use page

Click to view Disclaimer page