The HotDocs instruction “FILTER” is one that I use in almost every system that I’ve designed.  Its purpose is exactly what it sounds like – to filter (a repeat), based upon a certain criteria, so that the data output from the repeat is reduced – only the repeats that match the filter come out.  Like most instructions, it is best explained by example.

Lets say we have a repeating dialog named “Party RPT” that collects information regarding, well, parties in our fictitious matter. A party can be a corporation or individual (PARTY Entity Type MC), but never both. Throughout our system, we are constantly required to insert a sequence of all party names, as well as a sequence of all corporation names.  While we COULD use REPEAT instructions in our templates, this will not work in repeating templates, nor repeating sections.  So lets use the filter command to meet our requirements AND simplify code (its much easier to pre-calculate such lists instead of coding them in every template!).

Party RPT (dialog)
PARTY Name TE (name of party)
PARTY Entity Type MC (type of entity - individual or corporate).

GEN Parties All Names TE (list of all names of all parties)
GEN Parties Corp Names TE (list of all corporations only)

So lets do our “list of all parties” first – this would go in a computation…

FORMAT "a, b and c"
SET GEN Parties All Names TE TO RESULT

That’s it – all done.  Very simple, and now GEN Parties All Names TE contains something like “Seth Rowland, Basha Systems LLC, Ian Burrows, Routine Automation Pty Ltd and John Doe”

Now, lets use our FILTER instruction to slim that list down to only the companies.  Here is our filter computation, named “fil Party Companies CO”

PARTY Entity Type MC = “Corporation”

That’s it – a straight forward test: is the party entity type a corporation? yes/no.  So lets use this, again in a computation…

FILTER fil Party Companies CO
FORMAT "a, b and c"
SET GEN Parties Corp Names TE TO RESULT

GEN Parties Corp Names TE now has a value of "Basha Systems LLC and Routine Automation Pty Ltd".

Filters are absolutely fantastic, as they allow you to eliminate a lot of repetitive code and variables.  Take a litigation system for example, with Plaintiffs, Defendants, Third Parties, Plaintiffs by Counterclaim, Defendants by Counterclaim.  You COULD have a separate repeat for each of these parties.  Alternatively, you could have a repeat for Plaintiffs, with a checkbox to indicate that they are also optionally a defendant by counterclaim – then use filters on this variable.  Taking it one step further, you could simply have a dialog Parties that contains ALL parties in the matter, then use multiple-choice, multiple select variables to specify all the roles that a given party may have.  Another example is real estate documents – a given party may be a borrower, but optionally a mortgagee.  If it is an investment property, they may also be a lessor/landlord.  Occassionally, they may ALSO be a caveatee.

Filters allow you to collect information once centrally and then filter it in various ways to produce exact results.

ERASE Var and ERASE Dialog

The ERASE instruction is one of the handier instructions in HotDocs if you are populating data dynamically more than once, or are using ‘temporary dialogs’ during your interviews.  Use of this instruction will completely erase the contents of a variable or dialog, but be warned: it will erase all iterations of that variable or dialog!

Let’s say we’re doing an Estate Planning matter – you have the client, their spouse, a list of children, trustees, executors and some random beneficiaries.  There’s a lot of entities in there.  If you’re working with a system by Basha Systems, there would likely be a central party list that functions as a contact database.  If you’re not, you have several repeats containing contact information.  Either way, the approach is the same.  The example I will use is a generic letter.  You want to program a generic free-style letter that can be sent to any entity on the matter.  But because you quite often wish to send letters to the same party consecutively (a few days apart), you want the generic letter dialog to retain the information of who a letter was last sent to.  Here’s how you do it.

Create your dialog as you ordinarily would – name, address, salutation etc – lets call our dialog Generic Letter DLG.  I won’t get into the import mechanics – all we’re showing here is the ability to clear an entire dialog easily.  On that dialog, we’ll create a button that points to the computation Generic Letter Clear CO.  The script of that is:

ERASE Generic Letter DLG

That’s it.  If the variables on the dialog are saved in the answer file, then the next time you hit that dialog, the details of your last generic letter appear.  If the user wants to use those details – great!  If they don’t, they simply click the button and everything on that dialog is erased and the user can pick a new entity to direct a letter to.  Be warned that you don’t use this on repeating dialogs, as it will erase the entire dialog and all of its iterations/repetitions.

DEFAULT var TO value (and) SET var TO value

Today’s tip is about the DEFAULT and SET instruction models.  Both of these models will allow you to specify that a variable has a certain value, but in different circumstances and with different results.  This article provides the basics on when and how.

To quote the HotDocs help file, “This instruction (DEFAULT) suggests a value for a variable if the variable is unanswered”.  Unsurprisingly, the instruction is exactly what it appears to be – a DEFAULT value.  It will not, in any circumstance, overwrite any existing variable value, even if that variable value is a single space character, or 0 in a number variable.

Personally, I find DEFAULT useful for exactly that – providing default or “starting” answers to variables that may be overlooked by users.  Typically, its something I use with “fresh” answer sets, at the start of a matter.  The simple rule is to generally not bother with DEFAULT, unless you’re using it in a scenario where the dialog isn’t (or shouldn’t) have (m)any values in it.

SET however, is a completely different kettle of fish.  I use SET wherever I can.  The way SET works is very simple – ask no questions, take no prisoners and absolutely give this variable the supplied value.  It will override anything that previously existed in that variable.  There is one major drawback to SET – if you use it in a dialog script, it will prohibit the user from changing the SET variable.  Even if you wrap it in a conditional statement, there is no guarantee that HotDocs will re-process the condition properly and allow you to edit the variable. A minor bugbear.  The workaround (if absolutely necessary) is to use SET instructions in a button (dialog element, script link) on your dialog.  Because the SET instruction is not run by the script dynamically, but rather, is being run ONCE only on user click, HotDocs will not gray out the variable and prohibit changes. For the price of one click, you can set/clear/change/populate and manipulate your data as much as you want, without worrying about grayed out variables, or processing times (heavy script leads to slow response times.  heavy script on buttons only results in a one-off hit when a user actually clicks the button.  Its far more elegant).

And yes, I ignore all the HotDocs “errors” that tell me I am asking a variable in an interview and SETting it to have a value.  Why do I do this? Because I presume that good code gets you a correct answer, but relying on a user to get it right can get you inconsistent answers.  Lets say that everytime you specify you’re wearing a red shirt/blouse, you want black pants.  If you’re male, you also want a red and white tie.  If you’re female, you want ruby earings.  Without SET, you are relying on a user to know the ensemble and, just as importantly, to key it in correctly, every time.  Mistakes happen. Interruptions occur. And sometimes, people just plain get it wrong; its a part of the human condition.

I rely on human correctness (or, more accurately, avoid human error) as often as possible.  So lets show how SET can be used to automatically calculate answers based upon user data entry.  Here’s what my dialog script may look like:

//hide what we're going to force anyway.
HIDE Earings MC

//force the user to answer enough questions to calculate the answers

IF Blouse MC"Red"
SET Pants MC TO "Black"
IF Gender MC"Male"
SET Tie MC TO "Black and White"
ELSE IF Gender MC"Female"
SET Earings MC TO "Ruby"
ELSE IF Blouse MC"Black"
SET Pants MC TO "Black"
IF Gender MC"Male"
SET Tie MC TO "White"
ELSE IF Gender MC"Female"
SET Earings MC TO "Diamond"
//its not a red shirt/blouse and its not black either, let the user pick these ones....
//of course, you could specify as many ensembles as you wished.
IF Gender MC"Male"
ELSE IF Gender MC"Female"
SHOW Earings MC

In the above instance, you may be a little colour challenged, so you’d add this at the top of your code:

DEFAULT Blouse MC TO "Black"

This would thereby ensure that the first time you were presented with the dialog (for each different answer file), you could have your black ensemble, without touching a thing – just let the answers ride.  At the point you specified anything in the Blouse MC variable, the DEFAULT instruction would be ignored forever more.  But the first time round – it would do the job.

Looking at this, its a lot of code to avoid having the user answer a few extra questions.  However, lets put this in perspective – let’s say this WAS a program to get dressed in the morning, and it would get you ready for work in half the time you currently spend.  But a wrong answer would mean that you would have to go back to the start, re-specify your answers and let the program do its thing again. Suddenly, this isn’t “too much code”, but rather “time saving code”, because it calculates 2 answers based upon the answer to the first one.

The real benefit to such an approach (where possible) is that suddenly, your proof reading times are reduced.  Sure, saving your support staff time (and improving data integrity) is great. But saving an attorney an extra few minutes on proof reading (because the attorney only has to check one answer, not three…), now that’s profitable.

Calculate data from existing answers where you can. The LESS questions you ask, the better; so long as your answers allow you to calculate everything you need to know.  It is not uncommon for me to ask 10 questions and produce 13 answers.  SET and DEFAULT will allow you to do this.  DEFAULT will set up your most common answer set, so they don’t have to worry about doing anything at least SOME of the time.  SET will allow you to implement “sets” of instructions, so that one answer produces multiple correct answers.  That saves everyone time.