Age | Commit message (Collapse) | Author |
|
The precondition was too strong. Loosen it just enough to make the function
useful in all the ways it should be without allowing bad cases through.
|
|
`join` and `rename` both had the ability to introduce some ambiguity, whether
by renaming two fields to the same alias (which one should use that alias?) or
by joining two fields with the same alias (which one should use it?). Some
preconditions have been added to warn against some ambiguous use cases
(although I have just now realised the rename precondition is insufficient, so
a new commit will be coming soon).
|
|
|
|
Now a user can decide what the query's behaviour on deref should be. This means
a user using jdbc can plug in with
(set-query-deref-behaviour! #(... jdbc-query-code-here ...))
if they want, but if a user wants to use the code for something else then they
can do so without needing to import jdbc stuff. (I'll admit I'm not sure what
else they'd do with it, but that's slightly beside the point.) Still left to do
is to provide a helper for the common case (performing an SQL query with jdbc).
|
|
The parameters should be passed in the `rest` of the vector, not as a separate
seq. My bad.
|
|
The string parameters are now put in the query as a '?' and the string which
should go in their place is now placed in an auxiliary list when the query is
constructed. This should make it easier to avoid SQL injection stuff. (Although
table/column names are still vulnerable to SQL injection, they should not be
dynamic so the issue should be minimal.)
There was also another issue where some things were used before they were
declared (as a result of repl development) which has now been corrected.
|
|
Need to fix up the compilation to return a query ready for value substitution
by the jdbc stuff, or something (rather than embedding parameters in the
query).
Also need to add insert!, update! and delete! functions.
|
|
|