[phpesp-dev] [ phpesp-Bugs-2669428 ] Copying an existing survey does not copy conditions

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

Bugs item #2669428, was opened at 2009-03-06 20:45
Message generated for change (Comment added) made by nobody
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=108956&aid=2669428&group_id=8956

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: kswartz (kswartz)
Assigned to: Nobody/Anonymous (nobody)
Summary: Copying an existing survey does not copy conditions

Initial Comment:
I'm working with version 2.0.2 of phpESP, and noticed that if you copy an existing survey, it copies everything BUT the conditions. If this isn't fixed in the latest version, please consider it for the next update.

(If this IS fixed in 2.1.2, it will probably speed up my upgrade. Only reason I haven't so far is because I made a lot of changes and fixed some bugs in 2.0.2, and wasn't sure how long it would take to copy the changes over.)

----------------------------------------------------------------------

Comment By: Nobody/Anonymous (nobody)
Date: 2009-04-09 21:18

Message:
Sorry, the SQL in the last update was missing a line. It should read:

INSERT INTO conditions (id,survey_id,q1_id,q2_id,cond,cond_value)
SELECT NULL, $new_survey_id, qnew1.id, qnew2.id, cond, cond_value
FROM question qold1, question qold2, question qnew1, question qnew2,
conditions c
WHERE qold1.id=c.q1_id AND qold2.id=c.q2_id
AND qold1.survey_id=3 and qold2.survey_id=3
AND qnew1.survey_id=$new_survey_id AND qnew1.position=qold1.position
AND qnew1.deleted=qold1.deleted
AND qnew2.survey_id=$new_survey_id AND qnew2.position=qold2.position
AND qnew2.deleted=qold2.deleted;

Just a reminder, too, this enhancement REQUIRES the fix to 2672129 also
(which makes copied surveys start with the same position as the original).
This SQL cannot be manually applied to previously copied surveys unless the
positions between the original and copy are re-aligned by hand, which may
or may not be part of the fix to 2672129. (Probably not.)

----------------------------------------------------------------------

Comment By: kswartz (kswartz)
Date: 2009-03-08 05:23

Message:
(Sorry for the double post before.)

Okay, I've looked at survey_aggregate.inc, and I think what I described
previously is actually a bug. I've logged a separate bug for this.

With that fix in place, this is the SQL I used to copy the conditions from
the old survey to the new survey:

INSERT INTO conditions (id,survey_id,q1_id,q2_id,cond,cond_value)
  SELECT NULL, $new_survey_id, qnew1.id, qnew2.id, cond, cond_value
    FROM question qold1, question qold2, question qnew1, question qnew2,
conditions c
   WHERE qold1.id=c.q1_id AND qold2.id=c.q2_id
     AND qnew1.survey_id=$new_survey_id AND qnew1.position=qold1.position
AND qnew1.deleted=qold1.deleted
     AND qnew2.survey_id=$new_survey_id AND qnew2.position=qold2.position
AND qnew2.deleted=qold2.deleted;

I'll work on producing a patch that incorporates this into
survey_aggregate.inc

Incidentally, I'm not sure why you chose to insert the questions one at a
time while cycling through the results of a select. I think all the changes
you made to the old data could have been made in SQL. I elected to do an
INSERT INTO ... SELECT FROM ... instead, as you can see, as it's much
faster.

----------------------------------------------------------------------

Comment By: kswartz (kswartz)
Date: 2009-03-08 04:24

Message:
Okay, I've set up 2.1.2, and I may have a patch for this coming, but I've
discovered a weird quirk that's affecting the SQL.

The starting position for all my surveys (in the question table) is 0.
However, when I copy a survey, the starting position is 1. Is that intended
behavior? Can I always assume that the value of "position" for a question
in a copied survey is always one greater than its counterpart in the
original survey?

----------------------------------------------------------------------

Comment By: kswartz (kswartz)
Date: 2009-03-07 19:45

Message:
Okay, I've set up 2.1.2, and I may have a patch for this coming, but I've
discovered a weird quirk that's affecting the SQL.

The starting position for all my surveys (in the question table) is 0.
However, when I copy a survey, the starting position is 1. Is that intended
behavior? Can I always assume that the value of "position" for a question
in a copied survey is always one greater than its counterpart in the
original survey?

----------------------------------------------------------------------

Comment By: kswartz (kswartz)
Date: 2009-03-07 00:01

Message:
That's cool - it's not a feature I use very often, which is why it took so
long to say anything.

Next time I go to do this, I'll add a comment to the request with the SQL
necessary to do the copy (which is the hard part). If I have time to
actually fix the code, I'll gladly supply a patch.

(And yes, I'll upgrade to 2.1.2 first. :) )

----------------------------------------------------------------------

Comment By: Franky Van Liedekerke (liedekef)
Date: 2009-03-06 23:40

Message:
Hi,

sorry, but I don't think it is fixed in 2.1.2 (it was known to me though,
and I'm pretty sure I documented it somewhere...)
But feel free to make any contribution, be it patches, bug fixes, etc ...

Franky

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=108956&aid=2669428&group_id=8956