rbock
abf4bb8e9a
Added serializer to interpreter
...
The serializer has partial specializations for all templates. It is a
good basis if the connector/database requires strings close to the
standard.
The interpreter is unspecialized (and uses a static assert to say so).
It is a good basis if the connector/database requires a different
interpretation, e.g. re-writing the expression tree.
The interpretable_t implements three methods for interpretation:
serializing with the sqlpp::serializer_context, serializing with the
database's serialization context and interpretation with the database's
interpretation context.
2014-02-18 17:45:08 +01:00
rbock
61a4797d0b
Simplified syntax for multi_columns a lot
...
Instead of
multi_column(alias, columns...)
it is now
multi_column(columns...).as(alias)
Also, you can now write:
all_of(table).as(alias)
2014-02-13 10:07:51 +01:00
rbock
9e8979c6f0
Fixed a bunch of small errors and shorted compile error messages
2014-02-13 08:44:09 +01:00
rbock
7c928ae6d0
Added table sets to several more expressions
2014-02-10 19:55:48 +01:00
rbock
54687c5e69
Added a bunch of static asserts to prevent misuse of insert and update
...
There are quite a few more to be inserted
2014-02-10 17:56:16 +01:00
rbock
e4c7b5235f
Added several checks for correct types in insert columns/values
2014-02-10 09:46:31 +01:00
rbock
22c43358f2
All statements back to composite design
...
The point is that even though the CRTP code looked nice, it was too hard
to comprehend.
2014-02-09 23:30:30 +01:00
rbock
cc76820870
Migrated insert and remove back to composite design
2014-02-09 18:14:52 +01:00
rbock
ae2f1948ac
Added checks to prevent non-matching columns from being inserted
2014-02-09 15:06:42 +01:00
rbock
61ed1d9f37
Corrected inconsistency in select value_type
2014-02-08 23:48:35 +01:00
rbock
88fc0959b5
Added value type back to select
2014-02-08 22:50:26 +01:00
rbock
3c6e7cb89a
Fixed a bunch of smaller migration errors
2014-02-08 22:24:05 +01:00
rbock
68750aac80
Refactored insert to use policies
...
Much cleaner now.
2014-02-08 21:06:23 +01:00
rbock
5b67061fca
Fixed multi_column to handle all_of again
2014-02-08 17:12:40 +01:00
rbock
b82a0295cf
Fixed all_of use in select
2014-02-08 16:29:21 +01:00
rbock
0a744455ec
basic select methods work again
...
I've let go of the select(flags, columns) monster methods, these are now
separated
2014-02-07 23:43:26 +01:00
rbock
9c4832df0f
Switched remove to policy based design.
...
A lot of sweating to figure out how to do it, but now it looks MUCH
cleaner and it is probably a lot easier to extend :-)
2014-02-07 21:25:23 +01:00
rbock
b907e3482d
Using explicit call to python now, this might work on Windows even ;-)
2014-02-02 19:10:14 +01:00
rbock
28fc5900f2
Added missing overload for prepared statement parameters
2014-02-01 14:28:23 +01:00
rbock
ec2a8587d4
Added python code generator as an example
2014-01-28 20:53:22 +01:00
rbock
dfcd372611
tvin works with const char*, too
2014-01-26 15:44:44 +01:00
rbock
b0e7395f47
Using assignments in add_values now
...
This makes mistakes much less likely
2014-01-26 15:09:30 +01:00
rbock
5ef5259988
Added columns and add_values to insert for multi row inserts
...
Need to change add_values method to use assignments, though.
2014-01-26 13:24:42 +01:00
rbock
f802300685
Added sqlpp::null and sql::default_value for insert and update
2014-01-25 22:38:16 +01:00
rbock
019a920739
Added a .where(bool) to choose all or no row
2014-01-22 22:53:23 +01:00
rbock
abce8592e6
Added DISTINCT flag to avg and sum
2014-01-22 09:31:36 +01:00
rbock
34f6c7a2eb
Added optional DISTINCT flag to count
2014-01-21 23:42:27 +01:00
rbock
370c61870b
Updated tests a bit
2014-01-21 23:19:03 +01:00
rbock
a22b07aa03
Added more tests (and fixes) for dynamic columns and flags
2014-01-21 09:50:58 +01:00
rbock
075e63c486
Added test for (empty) dynamic select flag list
2014-01-20 23:17:07 +01:00
rbock
ee7cda0a5d
Started to change the way dynamic columns and flags are specified
2014-01-20 21:26:16 +01:00
Mateusz Loskot
2734675750
Add headers to sources to enable file browsing in IDEs
...
By adding headers to sources of targets, CMake adds them to list of project files generated for IDEs.
This is non-intrusive trick, which does not affect targets compilation.
2014-01-19 19:56:37 +01:00
rbock
31127812a0
Added some template aliases to ease specialization and added a few more operator overloads
2014-01-19 18:11:05 +01:00
rbock
24b44fae65
Re-animated all tests
2014-01-18 20:58:51 +01:00
rbock
312e735d6d
Moved a bunch of things into namespace vendor
...
This namespace is for things the developer will not get in touch with
(normally) but the vendor might need to use for specialization of the
interpreter.
This separation is not fully completed yet, some changes will have to
follow...
2014-01-18 15:50:16 +01:00
rbock
93129be748
Added missing interpreter specializations, especially for serializable
2014-01-17 09:56:35 +01:00
rbock
9826ef79e4
Added interpreters to is_null, multi_column, alias and select pseudo table
2014-01-16 22:31:43 +01:00
rbock
afcc62fc75
Added interpreters for a bunch of functions and table aliases
2014-01-15 19:42:14 +01:00
rbock
54d45e97ae
Added interpret support for remove()
2014-01-15 08:24:42 +01:00
rbock
54fa55e6a5
interpret works for in()
2014-01-15 07:38:53 +01:00
rbock
602f33726f
interpreter added for update()
2014-01-15 07:22:54 +01:00
rbock
9335a62087
Replaced implicit trivial_value_is_null tag for columns with explicit tvin method
...
This isn't database vendor specific, but library vendor specific, as I
used this implicit behaviour quite a lot, but it is utterly confusing
when mixed with prepared statements. Explicit is better here.
2014-01-14 22:54:07 +01:00
rbock
fa18ce5476
interpret works for insert now
...
Also added explicit .default_values() method
2014-01-13 23:05:48 +01:00
rbock
eb9f92543e
Interpret works for parameters now
...
Also, paramters inherit the operators they need
2014-01-13 21:00:20 +01:00
rbock
c03ef1b0b1
interpret handles order_by(), limit() and offset()
2014-01-13 06:24:27 +01:00
rbock
7fc5c34190
interpret works for group_by(), having() und .like()
2014-01-12 16:19:09 +01:00
rbock
d56d1422cd
interpret works for where()
2014-01-12 16:01:38 +01:00
rbock
1d3ea8516f
interpreter for from()
2014-01-12 15:45:50 +01:00
rbock
d957e8c0ae
Added interpreter to select, select_flag_list, select_expression_list
2014-01-12 12:22:15 +01:00
rbock
47ae6a2e76
Continued to implement standard serializing interpreters
...
Everything that is interpreted also has to move out of detail namespace,
because it might have to be specialized for a database connector
2014-01-11 21:51:57 +01:00
rbock
bef7cea6a6
Started to switch from member serialize -> non-member interpret
...
This will allow database connectors to specialize the interpretation of
the expression tree and interpret queries in vendor specific ways where
required.
2014-01-11 00:11:47 +01:00
Roland Bock
51e0db883f
Added static assert to prevent automatic rivial->null conversion to happen in where/having
2014-01-05 13:29:08 +01:00
Roland Bock
1360b1d9db
Turned value_type::parameter_t into a non-template
2014-01-02 13:11:19 +01:00
Roland Bock
b031bda5fc
Can now prepare a select (tested with mysql)
2013-12-28 22:52:54 +01:00
Roland Bock
b4baf38fab
Parameters of prepared statements can be null now
2013-12-26 19:05:05 +01:00
Roland Bock
a915bd9e50
Refactored of result and result row (clearer structure depending on connectors)
2013-12-17 22:12:13 +01:00
Roland Bock
9597c3712c
Moved result construction more towards connector.
...
This will make it easier to handle different formats
2013-12-15 13:02:41 +01:00
Roland Bock
221e20918b
select now collects parameters from Where and Having
2013-11-26 23:45:31 +01:00
Roland Bock
8f99a03359
Can extract parameters from expression
2013-11-22 09:42:28 +01:00
Roland Bock
8d12d23d67
Added verbatim table
2013-11-16 23:19:37 +01:00
Roland Bock
a0af42e57e
Added a few more type tests.
2013-11-10 18:03:39 +01:00
Roland Bock
81c77a58f4
Even moe function tests and a few corrections
2013-11-08 09:09:52 +01:00
Roland Bock
813549690e
More function tests and a few corrections for names and return types
2013-11-07 15:06:13 +01:00
Roland Bock
9d94f6770b
Added function tests and fixed a few return types
2013-11-07 13:36:54 +01:00
Roland Bock
0a57af0b3d
Added tests for integral and floating point
2013-11-07 12:23:49 +01:00
Roland Bock
d3963e04b6
Added more checks for database traits
2013-11-01 15:25:52 +01:00
Roland Bock
e63a798a82
Started to use flags to indicate database traits
...
e.g. support for any or outer join, or how string concatenation is
implemented
2013-11-01 12:23:50 +01:00
Roland Bock
902583a33b
Added support for dynamic columns to select_expression_list
...
Dynamic columns are not yet supported in results, though.
2013-10-09 10:36:38 +02:00
Roland Bock
c59c8d9ba9
Reduced likelihood of mistakes in CRTP for tables.
...
Renaming serialize -> serialize_impl helped tracking down an error in
the connection implementation examples
2013-10-03 22:25:23 +02:00
Roland Bock
4d8e448152
Disabled select(tab). Use select(all_of(tab)) instead.
2013-10-03 21:36:50 +02:00
Roland Bock
96c6e5d435
Dynamic methods of select, insert and co return *this now.
...
Not sure if chaining makes much sense, but it is possible now
2013-10-03 20:53:55 +02:00
Roland Bock
a2d23006f0
Fixed ambiguous assignment operator
2013-10-02 07:32:28 +02:00
Roland Bock
2e7d5478f6
Added dynamic functions to insert, remove and update
2013-09-30 07:46:50 +02:00
Roland Bock
d0d5fd2969
Replaced tag_yes and tag_no by std::true_type and std::false_type
2013-09-29 09:02:51 +02:00
Roland Bock
5604d5d663
Added dynamic limit and offset
2013-09-26 18:54:52 +02:00
Roland Bock
472833016f
Added dynamic versions of from and where to select
2013-09-22 20:42:19 +02:00
Roland Bock
5576df1775
Cleaned up select_pseudo_table and started to prepare for dynamic select parts
2013-09-22 12:16:28 +02:00
Roland Bock
0db405c28c
Rewrote like() member function, added more type control to concat()
2013-09-19 09:04:07 +02:00
Roland Bock
1596b33e52
Minor cleanup
2013-09-13 10:38:25 +02:00
Roland Bock
161da75723
Stripped a lot of superfluouos type information from result rows.
...
Moved member template into name class to utilize the fact the class
aliases are really just aliases.
This makes the code leaner in many cases and less complex for the compiler
(I guess). It also has the benefit, that the field name is available as
string in the result_rows. This might be useful for debugging one day.
2013-09-13 09:18:15 +02:00
Roland Bock
2defeff18e
Introduced field template to make result_row_t less specific.
...
It is sufficient to have name and type. There is no need for the result
"to know" what exact expression was used to define the column.
Surprisingly, template alias creates new templates (in contrast to
non-template using, which really just creates an alias of a type).
template<typename T> struct A{};
struct X
{
template<typename T>
using U = A<T>;
};
struct Y
{
template<typename T>
using U = A<T>;
template<>
using U<int> = X;
};
template<template<typename> class X>
struct Z{};
static_assert(std::is_same<X::U<int>, Y::U<int>>::value, "class aliases are really just aliases");
static_assert(not std::is_same<Z<X::U>, Z<Y::U>>::value, "template aliases are new templates");
int main()
{
}
2013-09-13 07:24:41 +02:00
Roland Bock
f39582082b
Minor cleanup
2013-08-17 11:27:41 +02:00
Roland Bock
712d3a5c53
Renamed examples -> tests
...
Will need to add reasonable examples
2013-08-17 11:22:38 +02:00
