* Move const.inc and compilers.inc back to main source directory
* Move remaining few lines of code from heidicomp.pas to helpers.pas
* Remove components/compilerdetection and components/heidisql
* Remove remainders of EDBImage component, unused since Aug 08.
Implement an optimized and universal way to safe and restore the column visibibility of ListTables (and all other VirtualStringTrees) by merging that code with the code for the column-resizing feature.
- Drops global var ListTablesColumnNames and proc SetupListTablesHeader.
- Change + move "RestoreColumnWidths" to private "RestoreListSetup", add code from SetupListTablesHeader
- Change + move "SafeColumnWidths" to private "SafeListSetup", add code from MenuTablelistColumnsClick
- Default visible columns are (must be) set at design time now
Pro:
- RestoreListSetup is called once in FormShow, while SetupListTablesHeader was called each time in LoadDatabaseProperties.
- SafeListSetup is only called once in FormClose
- Just safes the column indexes, not the column names in registry. Makes it safe to rename columns in the future
- Less registry read/write access in general.
- Safe/RestoreListSetup open an easy way to implement further list features (column moving, etc.)
- Makes it a cakewalk to implement the same column un/hiding feature for the remaining 4 Lists.
Contra:
- people/developers who used a build from 870 to this one on will see their activated columns in ListTables reset to the default (7 first columns) because the registry key and value has changed. (but: better do that now than after an official relase)
- Drop pagecontrol to access more than one selected table. Didn't seem helpful for any case, looked ugly and was a click hell to use.
- Implement changeable editors for table name, comment, engine, charset, collation and auto_increment value.
- Backward compatibility kept: A 3.23.58 server leads to disabling editors for engine, charset and collation while keeping the others working.
A table's charset and collation can now be provided at creation time and at altering time. Just as it's implemented for databases. Fullfills now most of the wishes in RFE #1693393.
- Synchronize height of all listviews after moving the splitter on one page.
- Drop maximum dimensions of form.
- Store/restore form and list dimensions via registry
- Documentation
- Simplify code
- Add useful popupmenu with items "Copy" + "Select all" to SynMemos which display the CREATE statements
- Rely on modalresult of form, drop OnClick handler of Close-button
- Format sum of table sizes using FormatByteNumber instead of always using KB as the unit of choice.
- Add a statusbar to the form to signalize that the form can be resized.
Pros:
- ShowDBProperties doesn't reset column layout
- Sort direction is remembered automatically
- Reading columnlist from registry is done once in a TMDIChild, not on each ShowDBProperties
- Hiding/unhiding columns doesn't call ShowDBProperties
- No more hassling with what is called a "default column" or not. popupDBGrid is usable like in Windows Explorer.
- Column layout is stored global, not per session.
Cons:
- Column layout from old settings logic needed to be discarded
Side effect:
- Fixes a potential AV in FormatNumber(str) with empty strings.
- Created an overloaded FormatByteNumber() which can take a string as input
1. When user opens a file which is bigger than LOAD_SIZE (currently 5M), ask what to do
2. User can normally open the file, cancel, or use the new mechanism:
3. Load a chunk of LOAD_SIZE of SQL into memory
4. Split chunk with parseSQL into single queries
5. Run queries and go on with 3.
parseSQL is the bottleneck here, very CPU consuming, as it has to take care of different comment-styles and delimiters. So, the above strategy effected a good compromise regarding overall performance on different tests with worst case SQL files:
- "Wide" table exports with many big sized fields => long lines
- "Narrow" table exports with only one mini-sized field, extended INSERTs => short lines
Especially in the latter case it avoids to cause a hellfire of parseSQL-calls
Still seems to have some memory leaks somewhere.
