order by后为什么不能参数化原因

注:sql注入的防御手段主要为预编译,但是预编译对orderby,表名、列名都不生效,所以在orderby的位置测试sql注入往往成功率比较高,原因参考下面文章,这里给出结论"order by后不能参数化的本质是:一方面预编译又只有自动加引号的setString()方法,没有不加引号的方法;而另一方面order by后接的字段名不能有引号。"

感兴趣的可以看完整的分析文章SQL预编译中order by后为什么不能参数化原因 - 诸子流 - 博客园

  • 0
    点赞
  • 2
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
VCLZip Native Delphi Zip/UnZip Component! (VCLZip Lite: Version 2.23 April 14th, 2002) (VCLZip Pro: Version 3.10 Buid 1 - November 25th, 2007) IMPORTANT: If installing the registered version, please be sure to always re-install/rebuild the components (VCLZip and VCLUnZip) to the component pallette (or rebuild the design time package) so that the ThisVersion property and any other new properties will be properly updated. If your application still does not run without the IDE, open up VCLZip's package, click on options and look at the Directories/Conditionals tab. If KPDEMO is defined, remove it and recompile the package. ***IMPORTANT: Please remember do not install these components into a package by the name of either VCLZip or VCLUnZip. You will receive an error if you do. PLEASE TAKE A LOOK AT THE "WHAT's NEW IN THIS VERSION" LINK IN THE HELP FILE AS IT HAS CONVENIENT LINKS TO ALL OF THE NEW TOPICS. ==================== Version 3.10 Build 1 - Several bug fixes. - Added support for Delphi 2006, 2007 - Added support for BCB 2006, 2007 - Improved memory performance when working with archives containing extremely high number of compressed files. ==================== Version 3.06 Build 2 Made Delphi 2005 compatible Other assorted fixes ==================== Version 3.05 Build 1 Fixed a lot of incompatabilities between VCLZip and WinZip Other assorted fixes ==================== Version 3.04 Build 1 New ZLib methods for optimized compression and decompression of single entities of data in standard ZLib format, without the overhead of the PKZip format. This is excellent for compression of data to be sent across the net, compressing web pages (http compliant compression), blobs, etc. - ZLibCompressStream - ZLibDecompressStream - ZLibCompressBuffer - ZLibDecompressBuffer - ZLibCompressString - ZLibDecompressString Overloaded TStream Methods for Delphi 4,5, BCB 4, and 5 - UnZipToStream - UnZipToStreamByIndex - ZipFromStream Special OnGetNextTStream Event for Delphi 4,5, BCB 4, and 5 - Allows zipping multiple TStreams in one process - More efficient than calling ZipFromStream multiple times Capability to use the latest version of ZLib 1.2.1. - VCLZip currently uses 1.4.1 by default. - By defining ZLIB121, VCLZip will use the latest version of ZLib which is included with the registered version. Some optimization improvements which should show some improvement in zipping and unzipping speed when using TkpStreams with D4, D5, BCB4, and BCB5. ============ Version 3.03 (VCLZip Pro) - Please test your application thoroughly with this new version of VCLZip Pro. While it has been tested and has even been used in at least two production applications for several months now prior to initial release, there are so many combinations of property settings, environment differences, and ways to use VCLZip that you should always test VCLZip completely in your application before deploying. *** New Zip64 capabilities, properties, methods and events: - Uncompressed, Compressed, and Archive file sizes can be up to 2^63-1 bytes in length. - You can compress up to 2147483647 files into an archive. This is compatible with PKZip's Zip64 format. - If a file does not extend beyond any of the original limitations (filesizes of 4 gig or 65535 files) then no Zip64 format information is included in the archive. - property isZip64 - tells you when you are working with a zip file that is using Zip64 format. Much faster processing due to linking to Zlib object files for compression and decompression routines. Blocked Zip Files (spanned zip archives split onto hard drive) - Now completely compatible with PKZip and WinZip split archives file naming format. - For backwards compatability you can tell VCLZip to use the old VCLZip filenaming format by using the BlockMode property. - New method OnFileNameForSplitPart called just before each split filepart is created. VCLZip supplies a default implementation of this method so for most purposes you won't need your own. - method DefaultFileNameForSplitPart - VCLZip calls this internally if you don't define your own OnFileNameForSplitPart. You can also call it from your own OnFileNameForSplitPart if you wish to add some processing to the default behavior. - property BlockMode - determines whether VCLZip uses PKZip/WinZip standard naming convention or VCLZip classic method. - method DefaultGetNextDisk - VCLZip calls this internally if you don't define your own OnGetNextDisk. You can also call it from your own OnGetNextDisk event if you wish to add some processing to the default behavior. - Properties for controlling which files are zipped... - IncludeHiddenFiles - default False; - IncludeSysFiles: - default False; - IncludeReadOnlyFiles: - default True; - IncludeArchiveFiles: - default True; - Event OnGetNextStream - Allows you to zip from multiple streams when using the ZipFromStream method. This improves performance since repeated calls to ZipFromStream causes the archive to be updated on each subsequent call. - property ThisBuild - Tells you the current build. See also ThisVersion - property OnHandleMessage - Handles interactive messages with VCLZip. There is a default, so you don't need to define your own unless you wish to eliminate interactive messages and handle them on your own. This is helpful if you are using VCLZip as a service or on a webserver for instance. ******** Upgrading existing applications that use VCLZip 2.X ********** For the most part, existing applications will work as-is. Just install VCLZip 3.X and recompile your code. Here are some things to be aware of though... 1) If your app currently creates mmBlock archives (spanned directly to hard drive) and you define your own OnGetNextDisk in VCLZip 2.X, you should move your code from this event that handles mmBlock events to the new event OnFileNameForSplitPart. However, if you simply rely on VCLZip's default OnGetNextDisk then you don't have to worry about this. 2) If your app creates mmBlock archives, the default naming convention has changed to match the PKZip/WinZip standard. If you wish to keep the same naming convention then set BlockMode := mbClassic. 3) OnGetNextDisk and OnPrepareNextDisk events are called for the 1st disk now. VCLZip 2.X only calls these events starting with the 2nd disk. 4) properties CompressedSize[Index], UncompressedSize[Index], ZipSize are now Int64 types. 5) Delphi 4, Delphi 5, BCB 4, and BCB5 are all capable of using the Zip64 format. However they use the TkpHugeStream decendants which act just like TStreams except they handle files/stream sizes larger than 2gig. There is a TkpHugeFileStream and a TkpHugeMemoryStream which should handle 99% of all necessary actions. If you currently work with VCLZip 2.X with TBlobStreams or some other type of streams, you can either define your own TkpBlobStream for instance which inherits from TkpHugeStream, or use the TkpHugeStream.CopyFrom(TStream, Count) and the TkpHugeStream.GetStream: TStream methods to give VCLZip your stream and get it back. Ofcourse when using regular TStream decendants in D4,4,BCB4,and 5, you cannot create Zip64 archives. If you use Delphi 6, 7, or BCB 6, you don't have to worry about any of this as the normal TSTream is used by VCLZip and handles large file/stream sizes. ============ Version 2.23 (VCLZip Lite) Added the OEMConvert property. Filenames stored in a PKZip compatible archive normally go through an OEM conversion to make them ascii compatible. When opening the zip file the conversion is undone. If you do not plan on having other zip utilities opening up your archives this conversion process is not really necessary. Setting this property to False will eliminate this process. The default value for this property is True for normal PKZip compatability. Added OnEncrypt and OnDecrypt events. These allow you to replace the standard pkzip encryption with your own. Data is passed to these events a buffer at a time. Use this with care as this is still somewhat experimental and I'm not sure how useful it is yet. You must make all changes within the buffer sent in to you. Treat the entire file as a stream. Byte for byte replacement only. No additional keys can be saved. Added OnRecursingFile event. Sometimes when using wildcards and recursing directories, there was no reporting of progress. This will be fired each time a file matches as the file list is being built while recursing directories. Added the EncryptBeforeCompress boolean property. The default for this property is False and if left like this VCLZip will behave like normal. If set to True, VCLZip will encrypt each buffer prior to compressing it instead of afterwards. This will cause files to not be decryptable by normal zip utilities thereby adding a bit of extra security. Bugs Fixed: IMPORTANT!!! Behavior of freeing the ArchiveStream (compressed stream) has been modified. VCLZip will now no longer try to free ArchiveStream, you must free it yourself. This was due to a problem where it would be freed automatically if there was a problem with the ArchiveStream when trying to open it as a zip file (possibly corrupt). Best practice is that ArchiveStream should always point toward a TMemoryStream that you create anyway. Modified the SFX code (the code used to create the SFX stub distributed with VCLZip) so that it handles filenames that have been run through an OEM Conversion. The SFX was losing accented characters. This modification means that if you are creating zip files to be used as SFX's you will want to leave the OEMConvert property mentioned above, set to it's default value of True. Modified so that when cursor is changed to hourglass by VCLZip, previous cursor is saved correctly instead of just changing it back to default cursor. Now saves Central Directory Extra Fields correctly. Fixed the SFX code so that it works properly if you use Copy /B to concatenate a zip file to the stub. Due to a Delphi strange behavior sometimes path names for directory only entries would become corrupted. Removed reference to QConsts, replaced with RTLConsts. Sometimes a GPF would result if a corrupt zip file was opened. Using a wildcard in pathname added to FilesList did not work. Using '*.*' as a wildcard in files added to FilesList now is the same as using '*'. VCLZip will now check for CancelTheOperation during initial building of the fileslist instead of just during compression processing. Added a final call to OnTotalPercentDone with 100% because this didn't always happen. Attributes were not getting set correctly for directory-only entries. Fixed a problem that was not allowing ZipComment's to be added correctly to spanned or blocked zip files. Not the same fix as in 2.22. Directories (directory-only entries) were not being restored properly unless DoAll was True. You were unable to delete a directory from which files were recursively zipped until exiting your application. ============ Version 2.22 Now Delphi 6 compatible. New event called {link=93,OnRecursingFile} which gets called as VCLZip recurses directories searching for files that match a wildcard that is entered in the FilesList. This gets called each time a file matches the wildcard. Fixed a bug which kept diskettes from being labeled when creating spanned zip files on WIN31. Fixed a bug which sometimes did not allow zip comments to be added to blocked zip sets. Fixed a bug which caused VCLZip to not properly handle the IncompleteZip exception on spanned zip sets unless you called ReadZip prior to calling UnZip. Version 2.21 (Changes are shown in the build stages as they were implemented) Pre-Release Build 5: When working with temporary files, VCLZip will now rename, instead of copy, the temp file if the destination is on the same drive. This will speed up the adding of files to an existing zip file when the resulting zip file is very large. Pre-Release Build 4: New event called OnPrepareNextDisk which is an event that will allow you, when creating spanned zip files across diskettes, to do things like format a diskette that has just been inserted, or to add or delete files from the diskette before continuing with the zipping process. Fixed a problem that was causing the CancelTheOperation Method to not work properly. Pre-Release Build 3: Fixed bug which caused VCLZip to miscalculate space needed for zfc file if wildcards are put into the FilesList. Fixed bug so you could have FilePercentDone without needing TotalPercentDone when creating spanned zip files Fixed so relative_offset set correctly for spanned zips. Side effect of removing needless write of header. Added code to read local fileheaders if exception thrown when reading a central fileheader. Fixed problem where directories couldn't be created from directory entries because the fullpath wasn't known yet. Result of having moved this code to earlier. Fixed typo in creation of LOC header values which could cause error if reading local headers. Changed so Zip Comment starting position is calculated based on end of central record instead of end of file. Pre-Release Build 2: IMPORTANT: Changed default for FileOpenMode back to fmShareDenyNone as it had been for all but version 2.20. Fixed a problem where drivepart (i.e. C:\) was not being stripped when saving relative paths. Added a BufferedStreamSize property which can increase the speed of creating zips to floppy (and other slow media) dramatically. The new default for this should increase the speed by as much as 3 times, but you can now tweak this especially for your application! Added an ImproperZip property which gets set when VCLZip detects an inconsistency with the zip. This can be useful for detecting when VCLZip was able to open the zip in spite of an inconsistency found. There was no way to know this in the past. Fixed a problem where zip comments in zfc files were not being read correctly. Added a setZipSignatures procedure which allows you to modify the signatures of your zip file. This will cause other zip utilities to not be able to recognize or read your zip files created with VCLZip. Useful if you want to add further security to your zip files. Pre-Release Build 1: Some zip files would not open correctly, throwing an incomplete zip file exception due to an erroneous "extra field length" identifier in headers of some compressed files. These zip files are rare, but a very few people seemed to have several of them. This problem would not affect zip files created by VCLZip, and this problem should only occur in VCLZip 2.20, not in any previous version. If you had Range Checking turned on, VCLZip would get a range check error when using a wildcard that ended with a * as in 'somefile.*'. Under certain circumstances, drive information would not be stripped from path information if zipping recursively (including subdirectories) "Retrying" to zip a file that could not be opened using the OnSkippingFile event would not always work correctly. Creating spanned zip set to floppy should be faster now due to removing a needless header write to disk for each file. VCLZip would not compile correctly with MAKESMALL defined. Added code to make VCLZip work with BCB5. Haven't tested this yet though since I don't have BCB5 myself yet. Added readonly boolean ImproperZip property which will be set to True when some sort of problem is found when opening the zip file, even if recoverable. This property will be enhanced and refined in the future. If KeepZipOpen is set to True, when putting in the wrong disk in a spanned zip set, VCLZip would not always properly close the file on the old diskette before trying to open the file on the next diskette. Added ECantWriteUCF exception which will be thrown if VCLZip runs out of room to write the uncompressed file when unzipping. Timestamp was not being set properly when unzipping readonly files. Moved setting of the timestamp to before the attributes get set. ============ Version 2.20 Changes have been made in the following areas: --Performance There are a few code optimizations that should speed up the zipping process slightly. --Spanned Zip Files A new feature, turned on with the SaveZipInfoOnFirstDisk allows VCLZip to create and read spanned zip files starting with the first disk instead of the normally required last disk of the spanned disk set by saving a Zip Configuration File on the first disk. This feature can be used even if creating the spanned zip file directly to your hard drive. A new property, SaveOnFirstDisk, allows you to save room on the first disk when creating a spanned zip file, to allow room for other files, such as setup programs, data files, or a Zip Configuration File. Spanned zip files can now be directed toward disks greater than 2 gig in size as long as you are using Delphi 5 or BCB 4. --UnZipping The new Selected indexed property offers another way to flag files to be unzipped. Files that have the Selected property set to True can be unzipped using the UnZipSelected method. The Selected property will be cleared (set to False) for each file as it is unzipped, but you can also call the ClearSelected method to clear them all. At anytime the NumSelected property can be checked to see how many files have been selected. Also, the UnZipToBufferByIndex and UnZipToStreamByIndex methods allow you to unzip files specified by their index instead of by name or wildcard. The BufferLength property allows buffered output (buffer smaller than the total uncompressed filesize) when unzipping directly to memory (see UnZipToBuffer and UnZipToBufferByIndex). This will cause the OnGetNextBuffer Event to be called everytime BufferLength bytes have been output by VCLZip. Modified to work in all ways with zip files that have "extra fields" in their headers. These tend to be quite rare, but they do show up from time to time. --Zipping Added a property called FileOpenMode which allows you to define the file open mode for files when they are opened to be zipped. Added a Retry parameter to the OnSkippingFile Event that can be used to re-attempt to open a file for zipping that is open by another process. This gives the chance to close the file and continue with the zipping process rather than having to start over again. Added a ENotEnoughRoom exception which will be thrown if there is not enough room to write to the archive, i.e. out of disk space. The new OnUpdate Event gets fired when updating or freshening an existing archive. It is triggered for each file that already exists in the archive as it is either replaced or kept in the updated archive. The AddDirEntriesOnRecurse will cause separate directory entries to be included in archives when doing recursive zips through subdirectories. --Integrity Checking A new method, CheckArchive, will perform an integrity check on all files in an archive. This is much faster than using FileIsOK on each file if testing all files in an archive with VERY MANY files. Further improved checking for corrupted zip files when opening zip files. --Encryption The following new properties and methods allow lower level work with password encrypted archives: DecryptHeader Gets the decryption header for a particular compressed file in an archive GetDecryptHeaderPtr Same as DecryptHeader but easier to use in BCB. DecryptHeaderByte Method Tests a password against the decryption header found in the DecryptHeader property. GetDecryptHeaderByteByPtr Same as DecryptHeaderByte but easier to use in BCB. --Self Extracting Executables Changes were made to the ZIPSFX32.BIN stub itself: - Modified to work with zip files containing "extra fields" in their headers. - Modified to change mouse cursor to an hour glass during processing. - Check for correct file size is now done automatically - Now uses the end of central and central headers to find the first local header. - Added a progress meter - Better checking for corrupted zip files. - Added an information window that can optionally be shown when the sfx is initially started up. - Added an AutoRun option to make the sfx stub run automatially when double clicked with no other interaction from the user. For the new modified sfx stub, ZIPSFX32.BIN, instead of using kpSFXOpt, you should now use the TSfxConfig component to set the options for the sfx stub. The new sfx can be found in the sfx\ subdirectory as usual and is called ZIPSFX32.BIN and the original sfx can be found in the same subdirectory except it is now called ORGSFX32.bin. Just rename it if you prefer that one (use KPSFXOPT instead of TSfxConfig with the old stub). --Miscellaneous The installation is now easier, atleast for first time installers of the source code. The .DPK files for Delphi and .CPP files for BCB are now included. Now these files simply have to be compiled and that's it. There is a separate option in the installation for installing to the different versions of Delphi and BCB. Added a property called FlushFilesOnClose which will cause all files opened for write by VCLZip to have their disk buffers flushed to disk when closed. Added the capability to delete Selected files from an archive using the DeleteEntries Method. The behavior of the OnInCompleteZip Event has been greatly improved. You can now use this event to ask the user to insert the last disk of a spanned disk set rather than having to handle this situation from outside VCLZip. The register procedures were changed so that the components now get installed to the "VCLZip" tab on the palette. I found that for all but Delphi 1 I had to actually manually move the components to the "VCLZip" tab. You may find that you have to do this too if you have already installed VCLZip before. The components now use new bitmaps in place of the old ones on the component palette. Separated many compiler defines into a new file called KPDEFS.INC. ==================================== Version 2.18: 1) Thanks to the hard work of a fellow registered user, added the capability to remove all dependencies on the Dialogs, Forms, Controls, and FileCtrl units by defining the conditional MAKESMALL, which results in a smaller footprint. This can be quite useful when putting VCLZip into a DLL for instance. In order to make this work, go into your Project | Options and select the Directories/Conditionals tab and enter MAKESMALL in the conditional defines text box. In Delphi you can add this conditinal define to the project options of your application that uses VCLZip and then do a "build all". In BCB you will have to add this to the project options of the package that contains VCLZip and then rebuild the package. If you define MAKESMALL, the only things you lose are: a) ZIP file open dialog box that appears when the ZipName is set to "?" b) Select Directory dialog box that appears when the DestDir is set to "?" c) Changing the cursor to an hour glass during some operations. d) No long filename support in Delphi 1 2) Made VCLZip completely BCB4 compatible. 3) Added some exception handling to KPUNZIPP and KPINFLT, mainly to handle unexpected situations when wrong passwords are entered. This fixes the problem with PRP, the password recovery program. 4) For Borland C++ Builder, changed any COMP types to double, getting rid of the compiler warnings for unsupported comp type. This affects the OnStartZipInfo and OnStartUnZipInfo events, so you'll have to change the comp parameter to double in these events if you use them (in both your header files and in the CPP files). 5) Modified OnStartUnZip event so that FName (the filename of the file that is about to be unzipped along with complete path) is now a VAR parameter and can be modified. This allows you to change the path and name of a file that is about to be unzipped. This is especially helpfull in applications like Install Programs. NOTE: You will need to change your current code to add the VAR to the event definition and implementation if you already use this event in your application. (In BCB, add a & just before the parameter instead of VAR) 6) Moved many type definitions to VCLUNZIP.PAS so that kpZipObj won't have to be included in your USES list. 7) Fixed bug that caused GPF when setting Zip Comment to '' (empty string). 8) Moved strings in VCLZip/VCLUnZip into a string table, making the code size a little smaller as well as making it much easier to localize string information. However you have the option of not using the new string table, for whatever reason, by defining NO_RES in your project options (in the conditional defines text box on the Directories/Conditionals tab). 9) Removed the need for several files. No longer included are kpstrm.res, kpstrm.rc, kpsconst.res, kpsconst.rc, kpstres.pas, and for Delphi 1, kpdrvs.pas. In some cases the need for these files was eliminated and in other cases just rolled into the newly included kpzcnst.rc, kpzcnst.pas, and kpzcnst.res. Definining NO_RES in your project options will elimiate the need for these new files but will make your code size slightly larger and you won't be able to localize your application without changing VCLZip source code. 10) Modified the OnFilePercentDone and OnTotalPercentDone progress events to work better when creating spanned disk sets and blocked zip sets. They no longer report 100% when the compressed file still has to be copied to disk. 11) Added the ReplaceReadOnly property. Setting this to true will allow files with the ReadOnly attribute to be replaced during the unzip process. 12) Added the ifNewer and ifOlder options to the OverwriteMode property. (This had somehow made it into the help file but not into VCLUnZip) 13) Added the SFXToZip method which will convert an SFX file to a regular zip file. The header pointers will be properly adjusted during the conversion. 14) Fixed a problem where the OnGetNextDisk event would always revert to the DefaultGetNextDisk method instead of what you entered into the Object Inspector each time your project was re-opened. 15) Fixed a bug that caused CRC errors when unzipping files from spanned disk sets if they were STORED (no compression) and spanned across disks. 16) Added the OnZipComplete and OnUnZipComplete events. If defined, these will fire at the very end of a zip or unzip operation (after all files have been processed, not after each file). These events will rarely be used since, normally you will be able to do the same thing at the point that the call to Zip or UnZip returns, but these events can be useful when using VCLZip in threads where in certain circumstances the return from the Zip or UnZip methods are not seen. 17) Creation of SFX files has never been easier!!! The addition of the MakeNewSFX method allows you to create Self Extracting Executables without the need to create a zip file first. The files that you specify in the FilesList property will be zipped, using all the normal VCLZip property settings, and the SFX will be created, all in one step! In addition, you can create configurable SFX files using this method, and you can do this especially easy by adding the new unit kpSFXOpt to your application's USES list and using the new 32bit SFX stub that is now distributed with VCLZip. This allows you to easily set things like SFX Dialog caption, default target extraction directory, file to launch after extraction, etc. 18) Fixed a memory leak that only affects applications using VCLZip that are compiled with Delphi 2, and that use wildcard specifications in the FilesList property. Version 2.17a: 1) Fixed a bug that was keeping VCLZip from reading truncated zip files or sfx files that did not have their headers adjusted. 2) Fixed a bug that was causing a directory to be created on the C drive when doing integrity checking with the FileIsOK property. 3) Added {$V-} to kpZipObj.PAS 4) Moved two AssignTo methods to public instead of private in kpZipObj.PAS Version 2.17: 1) Added Memory zipping and unzipping capabilities through the UnZipToBuffer and ZipFromBuffer methods. See the documentation for these methods in the Help File for more information. 2) New FileIsOK Property allows you to check for the integrity of individual files within an archive without actually unzipping the file. 3) Fixed a bug that kept checking of volume labels from working on WIN31 when working with spanned disk sets. 4) Removed all references to ChDirectory so that VCLZip will be more thread safe allowing separate instances of VCLZip in separate threads to be performing zip/unzip operations at the same time. 5) A new public property PreserveStubs allows you to make modifications to sfx archives and have the archive remain an SFX rather than revert back to a normal zip file. 6) Added a default OnGetNextDisk event. If one is not defined, then the default event will be called when the situation arises that a new disk is needed when zipping or unzipping a spanned or blocked zip archive. 7) Added more power to the wildcard capabilities. Now you can qualify the * wildcard character, for instance: * would satisfy any number of contiguous characters as long as they are all a thru e. * would satisfy any number of contiguous characters as long as none of them were a thru e. This allows you to do things like include files in specific direcories into your ExcludeList. For instance: VCLZip1.ExcludeList.Add('c:\test\*.txt') would exclude the zipping of all .txt files in the test directory but not in any subdirectories. 8) Fixed other minor bugs and made other code enhancements. Version 2.16: ***Please be aware that if you currently use the OnSkippingFile event in any of your applications, version 2.16 will require a small modification as this event has an added parameter and one of the current parameters is used a little differently when being called by the zip operation. Please see the help file for more information. 1) The OnSkippingFile Event has been changed slightly, adding a parameter for the filename. 2) OnSkippingFile is now called when a file to be zipped is skipped because it is locked by another application. See the Help File for more information. 3) Fixed a bug with the Exclude and NoCompressList where they were ignoring entries with anything before the extention (i.e. 'somefile.*' as opposed to '*.zip') if you were saving directory information. 4) Fixed a bug that caused an error if you added a wildcard with a non-existent directory to the FilesList. 5) A few other minor bug fixes. Modifications for 2.15 include: 1) PackLevel can now be set to 0 (zero) which means no compression at all (STORED only). 2) New property ExcludeList is a new stringlist that you can add filenames and wildcards to in order to specify files that you do not wish to be included in an archive. 3) New property NoCompressList is a new stringlist that you can add filenames and wildcards to in order to specify files that you wish to be STORED with a PackLevel of 0 (zero), no compression. 4) All compiler warnings and hints were removed. Modifications for 2.14 include: 1) Delphi 4 compatability. 2) Added ability to use complex wildcards when specifying which files are to be zipped. This includes wildcard characters not only in the filename but also in the pathname. This allows you to specify directories using wildcards, for instance: VCLZip1.FilesList.add('c:\test\w*\mycode*.pas'); would get all PAS files beginning with mycode in subdirectories under TEST that begin with the letter w. Wilcards may be much more complex than this. Please see the help file for more information. 3) Added the ability to override the RECURSE property setting when specifying files to be zipped. By adding the following characters to the beginning of the filenames being added, you can override whatever the current setting is for the RECURSE property: '>' will force recursion into subdirectories '|' will force NO-recursion For instance: VCLZip1.FilesList.add('>c:\windows\*.ini'); will get all .ini files in and below the windows directory reguardless of what the recurse property setting is. and: VCLZip1.FilesList.add('|c:\windows\sys*\*.dll'); will get all .dll files in subdirectories of the windows directories that start with 'sys' but will not recurse into any directories below the sys* directories. 4) The [ and ] characters previously used as special wildcard characters have been changed to since [ and ] are valid filename characters. If you still need to use the previous characters for backward compatability, I can show registered users how to easily modify a couple of constants in the source code in order to go back to the old style. See "Using Wildcards" in the help file for more information. 5) A few bug fixes. Modifications for 2.13 include: 1) New property ResetArchiveBitOnZip causes each file's archive bit to be turned off after being zipped. 2) New Property SkipIfArchiveBitNotSet causes files who's archive bit is not set to be skipped during zipping operations. 3) A few modifications were made to allow more compatibility with BCB 1. 4) Cleaned up the Help File some. 5) KWF file now works for Delphi 1 and Delphi 2 again. Still can't get context sensitive help in Delphi 3. 6) Cleaned up some of the code that was causing compiler warnings and hints. Modifications for 2.12 include: 1) Added a TempPath property to allow the temporary files path to be different from the Windows default. 2) Modified VCLZip so that any temporary files that are created receive a unique temporary filename so as not to clash with any other files in the temporary directory. This also allows working with zip files residing in the temporary directory. 3) Fixed a bug in the relative path feature. 4) Fixed a bug that caused a "list out of bounds" error if a file in the FilesList did not actually exist. Modifications for 2.11 include: 1) Fixed password encryption bug for 16 bit. 2) Fixed "invalid pointer operation" when closing application bug. 3) Fixed path device truncation bug which caused inability to modify existing archives in 16 bit. 4) Fixed inability to cancel during wilcard expansion bug. 5) Added capability to better handle corrupted timestamps. 6) Added capability to open and work with SFX files that were created with the COPY/B method (header files not adjusted). 7) Other small bug fixes. I'm still working on a bug which causes a GPF when continually unzipping the same file thousands to millions of times. This mainly affects programs like the Password Recovery Program (PRP) which uses the brute force method of searching for an archive's password. Modifications for 2.10 include: 1) Capability for 16bit VCLZip to store long file/path names when running on a 32bit OS. 2) New property (Store83Names) which allows you to force DOS 8.3 file and path names to be stored. 3) Better UNC path support. 4) Fixed a bug to allow files to be added to an empty archive. Modifications for 2.03 include: 1) Volume labels now get written correctly to spanned disk sets in Delphi 1 for all versions of Windows. 2) Delphi 1 VCLZip now correctly recognizes when it is running on Windows NT. 3) Fixed a problem with zipping files in the root directory when StorePaths = True. 4) File and Zip Comments are now read correctly from spanned/blocked zip archives. 5) Fixed a buf that was causing "Duplicate Object" errors. Modifications for 2.02 include: 1) Fix for file comments which were supposed to be fixed in version 2.01 but weren't. 2) Fix for stream zipping. Version 2.01 would not create a new archive if using a stream. (The Stream Demo now allows creating new zip files to streams too) 3) A few other minor modifications to further solidify the code. 4) A modification to the Zip Utility Demo which allows unzipping from Blocked zip files as if they were single zip files. 5) Added a read-only, published ThisVersion property which reflects the version of the VCLZip/VCLUnZip that you are currently working with. Modifications for 2.01 include: 1) Fixes for exceptions that were caused when CANCELING a zip or unzip of a spanned zip file. 2) Fix for a possible problem when zipping or unzipping a spanned zip file when one or more of the compressed files resided on more than 2 of the spanned parts. 3) Fix for file comments which were broken in version 2.00. Additional features for version 2.00 include: 1) Modify/Add internal file details (filename, pathname, timestamp, comment) for any file while zipping, in the OnStartZip event. 2) Add an Archive Comment while zipping in the OnStartZipInfo event. 3) Delphi 1 compatiblity for VCLZip. 4) Stream to Stream Zipping - Archives themselves can now be TStreams! 5) New Relative Path Information option. 6) Unzip archives that weren't zipped with the Relative Path option turned on as if they had been by determining how much path information to use with the Rootpath property. 7) Modify timestamps for files in existing archives (you could already modify filenames and pathnames for files in existing archives) 8) The OnBadPassword event now allows you to supply a new password and try the same file again when unzipping. 9) Source code has been cleaned up so that it will compile under Borland C++ Builder with no modifications. Also some bugs were fixed, most importantly: 1) An empty file, that had been compressed into an archive would cause any file added to the archive to cause the archive to approximately double in size. Any archives containing empty files are not corrupted, they are OK. This was simply a fix to the way the archive was processed. 2) After creating an SFX file, you had to close the zip file before you could modify it in any way, otherwise a stream read error was encountered. See the Help file for more information on new features. This zip file is part of a self contained installation program. Just run it and the installation program will begin. Contact [email protected] for further information Thanks! Kevin Boylan
第1章 数据库与sql基础 1 1.1 数据库的基本概念 1 1.1.1 数据库的由来 1 1.1.2 数据库系统的概念 3 1.2 数据库系统的结构、组成及工作流程 3 1.2.1 数据库的体系结构 3 1.2.2 数据库系统的组成 4 1.2.3 数据库的工作流程 5 1.3 数据库的发展 6 1.3.1 第一代数据库 7 1.3.2 第二代数据库 7 1.3.3 新一代数据库技术的研究和发展 7 1.4 关系数据库 8 1.4.1 关系模型 8 1.4.2 codd十二法则 9 1.4.3 范式 10 1.5 sql语言基础 11 1.5.1 sql的历史 11 1.5.2 sql语言的组成 12 1.5.3 sql语句的结构 13 .1.5.4 sql的优点 13 1.5.5 sql的执行 14 1.6 sql环境 15 1.6.1 环境 15 1.6.2 sql的层次结构 15 1.6.3 客户程序和服务程序系统 17 1.6.4 sql环境中对象的命名规则 18 第2章 主要的关系数据库与sql 21 2.1 sql server 21 2.1.1 sql server的结构 21 2.1.2 数据库访问标准化接口—odbc 22 2.1.3 使用查询分析器执行sql语句 22 2.2 transact-sql 24 2.2.1 transact-sql 概述 24 2.2.2 transact-sql的主要组成 25 2.2.3 transact-sql的一些重要命令 26 2.3 oracle数据库 29 2.3.1 oracle数据库软件组成 29 2.3.2 oracle数据库体系结构 29 2.3.3 oracle数据库系统结构 30 2.3.4 使用sql*plus执行sql语句 31 2.4 pl/sql简介 32 2.4.1 pl/sql的特点 32 2.4.2 pl/sql程序结构 33 第3章 创建、修改和删除表 37 3.1 表的基础知识 37 3.1.1 表的基本结构 37 3.1.2 表的种类 38 3.2 sql数据类型 39 3.2.1 字符型数据 39 3.2.2 数字型数据 40 3.2.3 日期数据类型 41 3.2.4 二进制数据类型 43 3.2.5 文本和图形数据类型 44 3.2.6 自定义数据类型 44 3.3 表的创建(create) 46 3.3.1 创建基本表 46 3.3.2 非空约束 47 3.3.3 default指定缺省值 49 3.4 表的修改 50 3.4.1 增加新列 50 3.4.2 删除列 51 3.4.3 修改列 53 3.5 表的删除与重命名 55 3.5.1 重命名表 55 3.5.2 删除表 56 3.6 创建、删除数据库 56 3.6.1 数据库的创建 56 3.6.2 sql server中数据库的创建 57 3.6.3 删除数据库 58 第4章 索引与视图的创建 61 4.1 索引的基础知识 61 4.1.1 索引的概念 61 4.1.2 索引的结构 61 4.2 索引的创建与销毁 63 4.2.1 基本创建语法 63 4.2.2 本章实例用到的实例表 64 4.2.3 创建简单的非簇索引 66 4.2.4 多字段非簇索引的创建 69 4.2.5 使用unique关键字创建惟一索引 70 4.2.6 使用clusterde关键字创建簇索引 71 4.2.7 索引的销毁 73 4.2.8 使用索引的几点原则 73 4.3 视图的基础知识 74 4.3.1 视图简介 74 4.3.2 视图的优缺点 74 4.4 视图的创建与销毁 75 4.4.1 基本创建语法 75 4.4.2 创建简单的视图 75 4.4.3 利用视图简化表的复杂连接 78 4.4.4 利用视图简化复杂查询 79 4.4.5 视图的销毁 81 4.4.6 使用视图的几点原则 82 第5章 简单的查询 83 5.1 查询的基本结构 83 5.1.1 select语句的结构 83 5.1.2 select语句的执行步骤 84 5.2 列的查询 84 5.2.1 本章用到的实例表 85 5.2.2 单列查询 85 5.2.3 使用distinct去除重复信息 86 5.2.4 多列查询 87 5.2.5 查询所有的列 88 5.3 排序查询结果 89 5.3.1 单列排序 89 5.3.2 多列排序 90 5.3.3 采用序号进行多列排序 91 5.3.4 反向排序 92 5.4 使用where子句定义搜索条件查询 93 5.4.1 where子句单条件查询 93 5.4.2 单值比较运算符 94 5.4.3 between运算符范围筛选 96 5.4.4 null值的判断 97 第6章 复杂搜索条件查询 99 6.1 本章用到的实例表 99 6.2 组合查询条件 100 6.2.1 and运算符 100 6.2.2 or运算符 101 6.2.3 and、or运算符的组合使用 103 6.3 in运算符 104 6.3.1 in运算符的使用 104 6.3.2 in运算符与or运算符 105 6.4 not运算符 106 6.4.1 使用not运算符 106 6.4.2 not运算符与[]运算符 108 6.5 使用like进行模糊查询 109 6.5.1 like运算符 109 6.5.2 “%”通配符 110 6.5.3 “_”通配符 112 6.5.4 “[]”通配符 114 6.5.5 使用escape定义转义符 115 第7章 连接符、数值运算与函数 117 7.1 本章实例用到的表 117 7.2 连接符 118 7.2.1 连接符的应用 118 7.2.2 使用别名 119 7.3 数值运算 121 7.3.1 数学运算符的种类 121 7.3.2 数学运算符的运用 122 7.3.3 使用cast表达式转换数据类型 123 7.3.4 使用case表达式 124 7.4 函数 126 7.4.1 有关函数的说明 126 7.4.2 字符处理函数 126 7.4.3 算术运算函数 131 7.4.4 日期时间函数 133 7.4.5 convert()函数转换日期、时间 136 第8章 聚合分析与分组 139 8.1 聚合分析的基本概念 139 8.1.1 聚合分析 139 8.1.2 聚合函数 139 8.2 聚合函数的应用 140 8.2.1 求和函数—sum() 140 8.2.2 计数函数—count() 140 8.2.3 最大/最小值函数—max()/min() 143 8.2.4 均值函数—avg() 145 8.2.5 聚合分析的重值处理 147 8.2.6 聚合函数的组合使用 148 8.3 组合查询 148 8.3.1 group by子句创建分组 149 8.3.2 group by子句根据多列组合行 150 8.3.3 rollup运算符和cube运算符 151 8.3.4 group by子句中的null值处理 153 8.3.5 having子句 153 8.3.6 having子句与where子句 154 8.3.7 select语句各查询子句总结 156 第9章 多表查询 157 9.1 本章用到的实例表 157 9.2 表的基本连接 158 9.2.1 连接表的目的 158 9.2.2 简单的二表连接 159 9.2.3 多表连接 161 9.2.4 使用表别名 162 9.2.5 采用join关键字建立连接 163 9.3 表的连接类型 163 9.3.1 自连接 163 9.3.2 自然连接(natural join) 166 9.3.3 内连接(inner join) 167 9.3.4 外连接(outer join) 170 9.3.5 交叉连接(cross join) 176 9.4 union与union join 177 9.4.1 关系的集合运算 177 9.4.2 union运算符 178 9.4.3 order by子句排序union运算结果 180 9.4.4 对多表进行union运算 181 9.4.5 union join 连接表 183 9.5 表连接的其他应用及注意问题 183 9.5.1 连接表进行聚合运算 183 9.5.2 多表连接的综合运用 185 9.5.3 多表连接注意事项 186 第10章 子查询 187 10.1 创建和使用返回单值的子查询 187 10.1.1 在多表查询中使用子查询 187 10.1.2 在子查询中使用聚合函数 188 10.2 创建和使用返回多行的子查询 190 10.2.1 in子查询 190 10.2.2 in子查询实现集合交和集合差运算 191 10.2.3 exists子查询 192 10.2.4 exists子查询实现两表交集 194 10.2.5 some/all子查询 195 10.2.6 unique子查询 197 10.3 相关子查询 198 10.3.1 使用in引入相关子查询 198 10.3.2 比较运算符引入相关子查询 200 10.3.3 在having子句中使用相关子查询 201 10.4 嵌套子查询 203 10.5 使用子查询创建视图 204 10.6 树查询 205 第11章 数据插入操作 209 11.1 插入单行记录 209 11.1.1 基本语法 209 11.1.2 整行插入 209 11.1.3 null值的插入 211 11.1.4 惟一值的插入 212 11.1.5 特定字段数据插入 213 11.1.6 通过视图插入行 214 11.2 插入多行记录 216 11.2.1 由values关键字引入多行数据插入 217 11.2.2 使用select语句插入值 217 11.3 表中数据的复制 220 11.3.1 基本语法 221 11.3.2 应用实例 221 11.4 从外部数据源导入、导出数据 222 11.4.1 access数据库数据的导出 223 11.4.2 access数据库数据的导入 225 11.4.3 sql server数据库数据导出 227 11.4.4 sql server数据库数据导入 230 第12章 数据的更新和删除 233 12.1 更新表中的数据 233 12.1.1 update语句的基本语法 233 12.1.2 update语句更新列值 234 12.1.3 利用子查询更新多行的值 237 12.1.4 依据外表值更新数据 238 12.1.5 分步更新表 239 12.2 删除表中的数据 240 12.2.1 delete语句基本语法 240 12.2.2 delete语句删除单行数据 240 12.2.3 delete语句删除多行数据 241 12.2.4 delete语句删除所有行 242 12.2.5 truncate table语句 242 12.3 通过视图更新表 243 12.3.1 可更新视图的约束 243 12.3.2 通过视图更新表数据 243 12.3.3 通过视图删除表数据 245 第13章 安全性控制 249 13.1 sql安全模式 249 13.1.1 授权id 249 13.1.2 sql的安全对象和权限 250 13.1.3 授权图 251 13.2 角色管理 251 13.2.1 create语句创建角色 252 13.2.2 drop语句删除角色 252 13.2.3 grant语句授予角色 252 13.2.4 revoke语句取消角色 253 13.3 权限管理 254 13.3.1 grant语句授予权限 254 13.3.2 revoke语句取消权限 255 13.3.3 select权限控制 256 13.3.4 insert权限控制 258 13.3.5 update权限控制 258 13.3.6 delete权限控制 259 13.4 sql server安全管理 260 13.4.1 sql server登录认证 260 13.4.2 sql server用户id的管理 261 13.4.3 sql server权限管理 263 13.4.4 sql server角色管理 266 13.5 oracle安全管理 268 13.5.1 oracle中用户、资源、概要文件、模式的概念 269 13.5.2 oracle中的用户管理 269 13.5.3 oracle中的资源管理 274 13.5.4 oracle中的权限管理 277 13.5.5 oracle中的角色管理 278 第14章 完整性控制 281 14.1 完整性约束简介 281 14.1.1 数据的完整性 281 14.1.2 完整性约束的类型 281 14.2 与表有关的约束 282 14.2.1 列约束与表约束的创建 282 14.2.2 not null(非空)约束 282 14.2.3 unique(惟一)约束 284 14.2.4 primary key(主键)约束 285 14.2.5 foreign key(外键)约束 287 14.2.6 check(校验)约束 289 14.3 深入探讨外键与完整性检查 290 14.3.1 引用完整性检查 290 14.3.2 match子句 291 14.3.3 更新、删除操作规则 293 14.4 域约束与断言 295 14.4.1 域与域约束 295 14.4.2 利用断言创建多表约束 296 14.5 sql server中的完整性控制 296 14.5.1 创建规则(rule) 297 14.5.2 规则的绑定与松绑 298 14.5.3 创建缺省值(default) 299 14.5.4 缺省值的绑定与松绑 300 第15章 存储过程与函数 303 15.1 sql中的存储过程与函数 303 15.2 sql server 中的流控制语句 304 15.2.1 begin…end语句 304 15.2.2 if…else语句 304 15.2.3 while、break和continue语句 304 15.2.4 declare语句 306 15.2.5 goto label语句 306 15.2.6 return语句 307 15.2.7 waitfor语句 307 15.2.8 print语句 308 15.2.9 注释 308 15.3 sql server中的存储过程和函数 308 15.3.1 系统存储过程 308 15.3.2 使用create procedure创建存储过程 309 15.3.3 使用execute语句调用存储过程 310 15.3.4 使用create function创建函数 312 15.3.5 使用enterprise manager创建存储过程和函数 315 15.3.6 修改和删除存储过程和函数 317 15.4 oracle中的流控制语句 319 15.4.1 条件语句 319 15.4.2 循环语句 320 15.4.3 标号和goto 322 15.5 oracle数据库中的存储过程 322 15.5.1 存储过程的创建与调用 322 15.5.2 oracle中存储过程和函数的管理 324 第16章 sql触发器 325 16.1 触发器的基本概念 325 16.1.1 触发器简介 325 16.1.2 触发器执行环境 325 16.2 sql server中的触发器 326 16.2.1 sql server触发器的种类 326 16.2.2 使用create trigger命令创建触发器 326 16.2.3 insert触发器 328 16.2.4 delete触发器 329 16.2.5 update触发器 330 16.2.6 instead of触发器 332 16.2.7 嵌套触发器 334 16.2.8 递归触发器 336 16.2.9 sql server中触发器的管理 338 16.3 oracle数据库中触发器的操作 340 16.3.1 oracle触发器类型 340 16.3.2 触发器的创建 341 16.3.3 创建系统触发器 342 16.3.4 触发器的触发次序和触发谓词的使用 343 16.3.5 oracle触发器的管理 346 第17章 sql中游标的使用 349 17.1 sql游标的基本概念 349 17.1.1 游标的概念 349 17.1.2 游标的作用及其应用 350 17.2 sql游标的使用 351 17.2.1 使用declare cursor语句创建游标 351 17.2.2 使用open/close语句打开/关闭游标 352 17.2.3 使用fetch语句检索数据 352 17.2.4 基于游标的定位delete语句 354 17.2.5 基于游标的定位update语句 356 17.3 sql server中游标的扩展 357 17.3.1 transact_sql扩展declare cursor语法 357 17.3.2 @@cursor_rows全局变量确定游标的行数 359 17.3.3 @@fetch_status全局变量检测fetch操作的状态 360 17.3.4 游标的关闭与释放 361 17.3.5 游标变量 362 17.3.6 使用系统过程管理游标 363 17.4 oracle中游标的使用 365 17.4.1 显式游标与隐式游标 365 17.4.2 游标的属性 366 17.4.3 %type、%rowtype定义记录变量 367 17.4.4 参数化游标 368 17.4.5 游标中的循环 369 17.4.6 游标变量 371 17.5 小结 372 第18章 事务控制与并发处理 373 18.1 sql事务控制 373 18.1.1 事务控制的引入 373 18.1.2 事务的特性 373 18.1.3 sql中与事务有关的语句 374 18.2 事务控制的具体实现 376 18.2.1 开始事务 376 18.2.2 set constraints语句设置约束的延期执行 377 18.2.3 终止事务 378 18.3 并发控制 380 18.3.1 并发操作的问题 381 18.3.2 事务隔离级别 382 18.3.3 set transaction设置事务属性 383 18.4 sql server中的并发事务控制 384 18.4.1 锁的分类 384 18.4.2 sql server中表级锁的使用 385 18.4.3 设置隔离级别实现并发控制 387 18.4.4 死锁及其预防 391 18.5 oracle中的并发事务控制 393 18.5.1 通过加锁避免写数据丢失 393 18.5.2 设置只读事务(read only) 394 18.5.3 oracle中的隔离级别 395 第19章 嵌入式sql 397 19.1 sql的调用 397 19.1.1 直接调用sql 397 19.1.2 嵌入式sql 398 19.1.3 sql调用层接口(cli) 399 19.2 嵌入式sql的使用 401 19.2.1 创建嵌入式sql语句 401 19.2.2 sql通信区 402 19.2.3 主变量 404 19.2.4 嵌入式sql中使用游标 406 19.3 检索、操作sql数据 407 19.3.1 不需要游标的sql dml操作 407 19.3.2 使用游标的sql dml操作 410 19.3.3 动态sql技术 412 19.4 sql server中嵌入式sql的编译运行 413 19.4.1 嵌入式sql代码 413 19.4.2 预编译文件 415 19.4.3 设置visual c++ 6.0连接 417 19.4.4 编译运行程序 419 19.5 oracle中嵌入式sql的编译运行 420 19.5.1 嵌入式sql代码 420 19.5.2 预编译文件 421 19.5.3 设置visual c++ 6.0编译环境 423 19.5.4 编译运行程序 424 附录a sql保留字 427 附录b 常用的sql命令 431 附录c 关于运行环境的说明 435 c.1 sql server 2000 435 c.1.1 直接访问 435 c.1.2 从企业管理器访问 436 c.2 oracle系统 436

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值