| Description | Following an excel import of a survey, I go to the survey, and then to the last group [I happened to be testing, there].
I execute a preview of the Group, and after some thinking, receive the error: "Invalid group number for this survey: 318"
If, instead, I first do a QA, or an edit, FIRST, the error does not appear.
If, following the error, I do a QA or an edit, the error disappears, thereafter, with a subsequent preview of that group.
I suspect that a variant of this error is known already to Tom W etal...though I thought this but report might narrow down it ultimate correction path. |
|---|
| Additional Information | I am using Tom Whites 192 w/ the excel import code - his master of 7-26, coded 0720
I notice that, at line +- 1944, [sorry, it's not exact, as I've made some changes, already, to the code] there is a commented out insertdata of the "qid",
[code]
// $insertdata['qid'] = $sqid; // this was causing key duplications - removing it seems to have fixed that
[/code]
which is where my instinct says the error comes from. I assume that executing QA or other code resets an index or something, curing the bug/error, which appears to be otherwise non-fatal.
I also noticed this quirk, not related, discovered on trying to track down the error source: that there are two full sweeps of the excel file, on import, first to search for S/SLs, and then to search for the balance of class codes -- A quick start to that import coding, I suspect..
I am working on an amendment that modifies that, and creates an import/export schema. The coding for it would allow for backwards compatibility, and forward looking to additional improvements: I plan to add a "XX" class as the first line that would have a second column 'blank', if original schema, or a number for the following on schema. This will allow for additional improvements without crashing earlier saves. I'll share it here, as another bug/improvement, when done, so I can ask someone to review and pull if it meets guidelines.
Unrelated: Can someone clue me in on which files have the code/codeChain is that pulls in the fieldmapping when a subquestion is rendered on survey execution? Thanks ! |
|---|
related to
child of
duplicate of
