At 01:14 PM 5/24/2005, Brandon Goodin wrote:
You can easily create an account. I'd prefer that you enter it in so that we can ask you questions as we begin to review it.
On another project that I work on that uses Jira anyone reporting an issue is asked to create an account and create the issue. That person then becomes the "reporter" of the issue and it assumed that the reported would review the resolution of the issue as well. Not that would happen on *this* project but occasionally an bug might be filed on a project, someone responsible for the component fails to duplicate the problem and just closes the issue and marks it as resolved. Occasionally, a fix to a bug might be committed but doesn't actually fix the problem as originally reported. If the person that initially reports an issue is also responsible for confirming that has been resolved to their satisfaction it saves a lot of headaches.
Brandon On 5/24/05, Rob Butler <[EMAIL PROTECTED]> wrote: > Isn't that what I just did? :) > > I don't have a Jira account for iBatis, so could > someone add this for me? > > Thanks > Rob > --- Brandon Goodin <[EMAIL PROTECTED]> wrote: > > > Oh duh. haha. sorry. Read it quick. Should have > > taken a few more seconds. > > > > Follow the second temp file suggestion for now. It > > is not a terribly > > difficult way to solve this at all. Please put a > > feature request in > > for this. I'm not sure we will add this feature. > > But, at least get it > > in the que so that we can discuss it. > > > > Brandon > > > > On 5/24/05, Rob Butler <[EMAIL PROTECTED]> wrote: > > > Because, as I said below, SqlMapClientBuilder > > requires > > > the SqlMapConfig file. So if I use StringReader > > to > > > pass in the SqlMapConfig file, that works fine. > > But > > > the SqlMapConfig file must then reference the > > SqlMap > > > file, and this file must be loaded either as a > > > resource from the classpath, or as a URL. Since > > this > > > file doesn't actually exist on disk, or in the > > > classpath loading as a resource isn't going to > > work. > > > > > > So that leaves URL as the last hope. Java > > supports > > > http, ftp, and file protocols by default. Since > > the > > > file isn't on disk, file isn't going to help. I > > don't > > > want to embed an http or ftp server in my app just > > to > > > load a dynamicly generated segment of xml from > > memory, > > > so http and ftp are out too. > > > > > > URL could still work if I implement my own > > protocol > > > and add that to the one's the JVM can use, > > something > > > like mem://fake-file-name. Java supports this, > > but it > > > seems a bit extreme that I need to create an > > entirely > > > new protocol to load some dynamically generated > > XML > > > from memory into iBatis doesn't it? > > > > > > So while I could implement something to do what I > > need > > > all in memory, or could use temp files as someone > > > suggested, the best way to do it in my opinion is > > add > > > the capability to iBatis and not have to work > > around > > > this unnecessary limitation. > > > > > > Seeing as I need this capability, and someone else > > has > > > resorted to using temp files to accomplish the > > same > > > thing, it sounds like it would be a useful feature > > to > > > have. > > > > > > Thanks for the suggestions though. > > > Rob > > > > > > > > > --- Brandon Goodin <[EMAIL PROTECTED]> > > wrote: > > > > SqlMapClientBuilder takes a reader. If you have > > a > > > > String in memory > > > > then why couldn't you use a StringReader? > > > > > > > > Brandon > > > > > > > > On 5/23/05, Rob Butler <[EMAIL PROTECTED]> > > wrote: > > > > > Hello, > > > > > > > > > > I am building an application that will need to > > > > > dynamically build an SQL statement at run > > time, > > > > and > > > > > then populate the statement's fields from a > > map. > > > > > Sounds perfect for iBatis, right. That's what > > I > > > > > thought, but I don't think iBatis does quite > > what > > > > I > > > > > need. Normally iBatis can dynamically change > > it's > > > > SQL > > > > > but you need to know things like the number, > > name > > > > & > > > > > type of columns in advance when the SqlMap is > > > > created. > > > > > > > > > > Unfortunately in my application, the SQL > > statement > > > > > that is built at runtime is completely > > dynamic. > > > > The > > > > > name of the columns, their data types and the > > > > number > > > > > of columns will only be known at run time. So > > > > what I > > > > > would like to do is dynamically instantiate an > > > > > SqlMapClient from a String, which contains the > > > > SqlMap > > > > > that is dynamically generated at run time. > > Then > > > > let > > > > > iBatis do the work of converting the map's > > values > > > > into > > > > > the correct data types and populating the > > SQL's > > > > field > > > > > values. It would be nice if the table name > > could > > > > be > > > > > passed as a property instead of in the map of > > > > data. > > > > > > > > > > This looks like it would be possible with the > > > > > SqlMapClientBuilder, but I don't think that is > > > > going > > > > > to work. The problem is it wants to read an > > > > > SqlMapConfig, which then of course points to > > an > > > > SqlMap > > > > > file. However, the SqlMap file obviously > > won't > > > > exist > > > > > anywhere but as a string in ram, and it's not > > > > possible > > > > > (as far as I know) to directly load an SqlMap > > > > using > > > > > SqlMapClientBuilder, nor is it possible to > > load a > > > > > single "file" (string in ram) that is a > > combined > > > > > SqlMapConfig & SqlMap. > > > > > > > > > > I could probably cobble something together > > using > > > > the > > > > > internal non-user exposed classes within > > iBatis, > > > > but > > > > > then if those classes change from one release > > to > > > > the > > > > > next I would have to make the appropriate > > changes > > > > to > > > > > my code, and I would prefer to use something > > that > > > > is > > > > > part of iBatis. > > > > > > > > > > So, I would like to request a feature to make > > > > > dynamically building the SqlMap at runtime > > > > possible. > > > > > I suppose this could be done by developing a > > class > > > > > like SqlMapClientBuilder that accepted a > > String, > > > > and > > > > > didn't need an SqlMapConfig, just a SqlMap > > XML. > > > > > Alternately, a fixed API could be exposed to > > users > > > > to > > > > > manually construct a SqlClientMap. > > > > > > > > > > Thoughts? Is there a better approach that > > works > > > > now? > > > > > Did I miss something obvious? > > > > > > > > > > Thanks, and iBatis rocks! > > > > > Rob > > > > > > > > > > > > > > > > > > > > __________________________________ > > > > > Do you Yahoo!? > > > > > Yahoo! Small Business - Try our new Resources > > site > > > > > http://smallbusiness.yahoo.com/resources/ > > > > > > > > > > > > > > > __________________________________________________ > > > Do You Yahoo!? > > > Tired of spam? Yahoo! Mail has the best spam > > protection around > > > http://mail.yahoo.com > > > > > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com >
