| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
ago in 1dedadcaba2cd474a251a24d893dc3b7b2a719a7)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* Remove a redundant ivar (and while we are at it, restructure some code)
* And then there was this gem:
(...)
if (fileMode == O_RDONLY) {
int i, c;
char bzbuf[4];
const char *charFileMode = fileMode == O_WRONLY ? "wb" : "rb";
(...)
|
| |
|
|
| |
* Move a lock() that could become unbalanced and theoretically cause havoc
|
| | |
|
| |
|
|
|
| |
* Avoid a malloc(4) / free() and let the compiler figure that out
* sizeof(pointer) != sizeof(memory area pointed to by pointer)
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
improve memory usage during large exports, particularly improving memory which appeared to never be reclaimed
- Fix some minor memory leaks throughout the application
|
| |
|
|
|
| |
- Tweak README
|
| |
|
|
|
|
|
|
|
| |
- The ability to easily index multiple columns
- The ability to specify the storage type (HASH or BTREE) that should be used when available and permitted
- The ability to specify a length prefix when indexing a specific column (required on TEXT columns)
Requires lots of testing.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
- A new SPExportFile class, providing an abstract interface to the handling and creation of export files.
- Enables the centralisation of all file/file handle creation logic as well as better support for handling situations where files fail to be created, including files that already exist at the export location.
- New SPExportFileHandleStatus constants to support the reporting of file handle creation.
- Update SPExporter to use the new file class instead of directly using an instance of SPFileHandle.
- Add the necessary logic to deal with files that already exist on disk, by providing the user with 3 options: cancel the export, ignore the files in question or overwrite them. We might want to enhance this to make new files sequential in name to prevent overwriting. Fixes issue #742.
- New SPExportFileUtilities category, which centralises all the logic relating to writing export type headers as well as dealing with problems occurred during file/file handle creation.
- Improve feedback given on the export progress sheet during export initialisation.
- Tidy up and improve comments.
|
| |
|
|
| |
when reading uncompressed files
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ability to use Bzip2 compression.
Other changes include:
+ Enable the use of export compression (Gzip and Bzip2) for all export formats.
+ Move the compression options in the export dialog to the 'Advanced' export settings view.
+ Simplify the setting of common exporter properties (e.g. the connection, use of compression).
+ Fix a potential memory leak in the dot exporter.
+ Update the data importer to recognise Bzip2 compressed files.
+ Fix several display issues on export dialog.
+ Restore the default .csv file extension of CSV exports.
+ Correctly update the default export filename when selecting a output compression type.
The addition of Bzip2 compression support implements issue #688.
|
| |
|
|
| |
increased separation of the writing thread. This results in a up-to 1.2x faster MySQL dump for fast servers, and makes SPFileHandle faster than NSFileHandle for writing data (either directly or GZIP compressed).
|
| |
|
|
|
|
|
| |
function documentation
- Tweak subfolder ordering
|
|
|
background thread, and integrate for SQL import:
- Implement streaming reading of gzip-compressed files for SQL import
- Support exporting SQL dumps into a gzip-compressed file
- SPFileHandle supports the most-used subset of NSFileHandle commands for easy integration
- Integrate zlib 1.2.4 for improved gzip streaming performance (and support for custom buffer sizes and file offset positions)
This implements Issue #571 .
|
