Changing structure of table via events - when can it be done, and why only then?

jdewells25

<p>Greetings one and all.</p><p> </p><p><span style="font-size:10px;">(this is the background part that I somehow feel obligated to tell you, even though if I was reading this I'd skip it too)</span></p><p>I am developing something in BIRT that will allow us to specify a table from our database. Hence, we have parameters for each column such as the order number, the header name, et cetera. So! We pull down all that lovely info, and now we need to add columns to a table, one for each result, and in the order supplied from the database.</p><p> </p><p>So I go into the onFetch method of the dataset and tell it to add a column to a table for each record it gets. Code's looking good. Only nothing actually...<em>happens</em>.</p><p> </p><p><span style="font-size:10px;">(skip to here if wanting to cut to the meat of the matter, this is the part that actually needs an answer)</span></p><p>As it turns out, I cannot actually modify the table's structure at all, unless done via the events on the report document itself (e.g. initialize, [before/after]Factory, [before/after]Render, etc). Additionally, trying to use breakpoints on anything other than those events simply does not break the program, even though I have verified that the code is indeed being run. I need to run a table-structure-modifying script for each row pulled from the database, but am worrying that it cannot be done. What's going on?</p><p> </p><p>I've done what research I can and have tried my best to find some crucial bit of info that I've missed but so far I've found nothing. What's the rhyme and reason here? Is it something to do with the various report phases (e.g. presentation, generation)? A failing on my part to understand the key differences between the events and their relationship to the Run and Render tasks? What am I missing?</p><p> </p><p>Any and all input is very much appreciated.</p><p>- JDE</p>

micajblock

<blockquote class="ipsBlockquote" data-author="jdewells25" data-cid="127226" data-time="1397043604"><div><p>I am developing something in BIRT that will allow us to specify a table from our database. Hence, we have parameters for each column such as the order number, the header name, et cetera.</p></div></blockquote><p>Why? What is the business reason? What are you trying to accomplish (not the technical way you solved it)? Why not ue multiple data sets and multiple tables. Then you drop the tables you do not want. Much simpler. Example here:</p><p> </p><p><a data-ipb='nomediaparse' href='http://developer.actuate.com/community/devshare/_/designing-birt-reports/simple-example-of-dropping-a-tablechart-r1636'>http://developer.actuate.com/community/devshare/_/designing-birt-reports/simple-example-of-dropping-a-tablechart-r1636</a></p><p> </p><p></p><div style="font-size:12px;"><p>[color=rgb(40,40,40);font-family:'Source Sans Pro', sans-serif;]What's going on?[/color]</p><p> </p><p>[color=rgb(40,40,40);font-family:'Source Sans Pro', sans-serif;]See this video for order of events:[/color]</p><p> </p><p>https://www.youtube.com/watch?v=FQ9VjByPNWM&lt;/p&gt;&lt;/div&gt;