水晶打印时提示：‘server has not yet been opened’，是什么原因引起的

XCX 2005-08-09 10:23:24

水晶打印时提示：‘server has not yet been opened’，是什么原因引起的

...全文

83 回复打赏收藏转发到动态举报

写回复

回复

切换为时间正序

请发表友善的回复…

发表回复

S7A驱动，7.20版本， Version history S7A OPC server and FIX driver for Siemens S7/S5 communication via Ethernet TCP/IP, MPI or Profibus Version 7.20 Date: 26.02.2007 Version history ================================================================================================================== Build 224 New Functions: Solved problems: - The setup program created a wrong e-doc link in Windows start menu. Now the proper link to S7A.CHM will be installed - From the S7A power tool it wasn't possible to open the online help file. Now it can be opened and also the context-sensitive help works properly. - When the signal conditioning function "S&M" was used for a byte value, the driver wrote wrong values to the PLC. Build 223 : New Functions: - The number of supported NetLinks has been increased from 4 to 64 - Communication error messages will be now displayed in the alarm history of iFIX. This helps to detect, trace and solve communication issues. - Support for MATRIX licence key added. This key does not require a special key driver. It simply needs a PC and a windows version which supports USB ports. - When using the driver in conjunction with NetLink adapter, the licensing now can take place by a license code in the NetLink. In this case a separate licence key (dongle) is not needed. Note: If the connection to the NetLink is disturbed, the driver starts in demo mode, because it can't read out the licence code. If afterwards the connection to the Netlink can be established, the driver automatically changes from demo mode to the licenced mode as long as the demo time has not yet expired. - Support for S7-200 with CP 243-1 was added. Solved problems: - Passing of invalid OPC Item IDs caused a memory leak of the driver's global memory. After the global memory was exhausted, the driver crashed. - Writing of string values failed when using hardware options NULLTERM or SPACEFILL - Wrong OPC timestamp occured when driver was running in NIO Simulation Mode - The driver could crash when the driver was stopped (either by hand or when the demo time was expired) and in the same moment the NetLink connection was disturbed. - Improvements on the seamless switching between primary and backup devices and vice versa. - The passing of invalid OPC Item IDs caused a memory leak in the driver's global memory. After the global memory was exhausted, the driver crashed. - The writing of strings failed when using the hardware options NULLTERM or SPACEFILL Build 222 : New Functions: - The driver can now be registered as a service with the local system account as the log-on account. For more information see the document "S7A_as_Service.PDF" on the installation CD . - The driver now switches "seamless" between primary and backup device and vice versa. Solved problems: - The field "Default Gateway" has been added to the NetLink's configuration property page - The string write from the OPC Server sometimes failed . This bug has been fixed - After driver uninstall, not all of the S7A-related registry entries were deleted. Now the registry will be completely cleaned from S7A registry entries after uninstall. - The Netlink module generated trace messages to debug window. These trace outputs has been removed. Build 221 : New Functions: - The information field "Runtime Mode" has been added to the system statistic display. This shows whether the driver is running in the demo mode or in the licenced mode (with a valid key). - New control tag: "!RuntimeMode:S7A". A digital (i.e. DI) or analog (i.e. AI) block with the I/O address "!RuntimeMode:S7A" delivers the value 0 (in the F_CV property), when the driver is running in the demo mode, or the value 1, when the driver has found a valid key, and therefore is licenced. This function is also available via the OPC server interface. The same item-ID address format has to be used as with the tag-I/O address, thus "!RuntimeMode:S7A" - The information field "Device in use" has been added to the device statistic display. This shows which device (primary or backup-device) is currently active. This state can also be requested via the "!switch" control tag. A digital (i.e. DI) or analog (i.e. AI) block with the I/O address "!switch: delivers the value 0, when the primary device is active and the value 1 when the backup device is active. For the place holder in the I/O address the logical name of the driver device, which is requested, must be entered . Example: "!switch:D11" delivers the state of the driver device with the name "D11". This function is also available via the OPC server interface. The same address format has to be used for the tag-I/O address as for the item-ID, thus "!switch:" - At the channel parameter dialog for the communicatiomn way "S7 TCP/IP" and "MPI/PB CIF", the combo box "Conn.Type" has been added, which now enables the selection of the connection types "PG", "OP" and "other". Until now, a PG connection was used. - In the device dialog of the communication way "MPI/PB CIF", the input fields "Rack:" and "slot no. of CPU" have been added. This extension now allows the connection to "external" Profibus- CPs (CP 342/443-5 or 343/443-5). The Profibus station address of the Profibus CP is stated in the MPI/PB address field. The rack and slot number of the S7-CPU must be stated in the corresponding fields. Please make sure that slot number of the CPU is entered and not of the Profibus-CP!! - The driver now supports the different date and time data types of the S7 PLC. See Topic "Hardware Options for OPC String and TX-Blocks to support S7 date and time data types" in the S7A online help for detailed information. Build 220 : Solved problems: - The own address of CIF50 MPI/PB communication way was limited to 31. Now the limit is 126. - Bug-fix in PG-PC Interface (CP5611) communication way. Device could be disabled from polling, after CP5611 sent a specific sequence of communication errors. Build 219 : Solved problems: Improvements and new functions: - New Parameter "Max Gap (Bytes)" on Setup parameter property page "Advanced". This parameter defines the maximum address gap between an existing and a new block at which the driver automatically appends the new block to the existing. If the address gap is greater, then the driver creates a new data block. This functionality is active, when the data block auto create option is set on. - New Radio Button selection "Show Window" on Setup parameter property page "Advanced" The driver automatically opens its window which shows trace log messages useful for trouble shooting, when the "Yes" Radio button is on. - New HTML Help Build 218 : Solved problems: - Problems with datablocks which have a length between 180 and 200 bytes. - Unsufficient exception handling in S7AMemory.Dll caused blocking of driver and all other clients (WSACTASK, Workspace, etc.) which reads data from the driver. - Error in MIX Read function could cause blocking of data blocks Build 217 : Solved problems: - In configurations with data blocks which have a primary rate of zero (as fast as possible), it sometimes happened that write requests were "stuck" in the write queue. - Sporadic locks of data blocks after the driver was switched from the primary device to the backup device Improvements and new functions: - Improved error and timeout handling on communication faults. The reply timeout on the device configuration page can no longer be configured. Instead it will be derived from the TCP/IP timeout value, configured on the channel configuration page. The value (in seconds) of reply timeout always is 1 second higher than the TCP/IP timeout. This ensures, that the logical read/write request (time-limited by reply timeout value) will not be cancelled (by timeout) before the physical request (time-limited by TCP/IP timeout value) is finished regulary or through the timeout. - The NIO Simulation mode has been expanded. If it is set "ON" the polling of data blocks will be disabled, due to which no communication errors will arise, as long as a PLC hardware is not connected to the driver. In this mode the iFix data base blocks reads the value of zero or respectively, the last value written to the driver`s output data area. - "!SyncPoll" and "!SyncSend" have been added to allow e.g. an EDA application to read/write a data block synchronously Build 216 : Solved problems: (06.06.05) - Reset of statistic counters failed Build 215 : Solved problems: (31.05.05) - An error in the OLE interface methods ReadValue ()and WriteValue() was corrected. These functions can now be used for example,within the Excel program, to access data block values. - Datablocks stopped polling after an 'unknown error' was returned by CP5611 Build 214 : Solved problems: - Syncronous reads/writes from concurrent OPC clients now work properly New function: - Performance has been improved, especially in configuration with a lot of short data blocks. - system, channel, driver and data block static display has been extended Build 213 : Solved problems: - No reconnect after the MPI/Profibus connection between CP5611 and the PLC was aborted. - S7ADRV.EXE could crash on Multi Processor Systems or Systems with Hyperthreading when error messages (e.g. "Unable to connect to MPI station") were put to driver's log window from different threads concurrently. Build 212 : Solved problems: - The AGLink error code -1268 (Unknown error message of the PLC) caused a blocking of the device. Now the device will be closed and re-opened to recover the error. - under very seldom conditions a AGLink error caused a crash of S7ADRV.EXE Build 211 : Solved problems: - Setup of TCP/IP parameters failed with error -510. Problem has been solved. - Problems in configurations with more than 8 devices within a channel have been solved - Import of CSV file from S7A Version 7.16 failed - Failed to read the AG State - Configuration of a backup IP address caused a driver crash when it tried to switch from primary to backup device Build 210 : New functions: - Communication way MPI/Profibus using Siemens CPs (CP5611, 5613) added - Communication way MPI/Profibus using Hilscher CIF50 PB added - Communication to "old" S5 series via TCP/IP added - The old Sentinel Keys will not supported any longer - Check Box "Disable comm alarming" in datablock configuration allows to suppress the COMM alarms of Tags which addresses a driver datablock which is in communication error state. - Now the OPC server is OPC DA 2.05 complient Build 210 : Solved problems: - Under Windows XP the driver crashed when the network cable was not connected to ethernet card and the driver started. - A communication error on a single device caused comm errors on all other devices of the channel Build 201 : Solved problems: - CSV Import/export works now even for the new communication ways. Build 200 : Solved problems: - Handle leakage when using asynchron read mode of OPC server. This problem was fixed by migration to OPC server toolkit version 7.20 New functions: - The driver now supports additional communication ways (serial adapter, NetLink adapter and Siemens CPs (eg. CP5611). It's no longer a pure TCP/IP driver but a general driver for Siemens S7 which allows to connect via TCP/IP, MPI bus or Profibus using different interface hardware (CPs). Build 128 : Solved problems: - In big configurations with a lot of data blocks when a data block has been written it took too long until this data block was read back. Thus between the write and the read the clients (OPC or FIX) read the old data. Now after a write the driver reads the data block priorized. Build 127 : Solved problems: - The addressing range for PLC's input area and output area has been extended from 0..255 bytes to 0..4095 bytes. Build 126 : Solved problems: - Errors in manual and automatic block write functions has been fixed Build 125 : Solved problems: - Very fast consecutive analog writes to the same data block caused a crash of the S7ADRV.EXE process. This problem has been solved. Build 124 : Solved problems: - Different communication problems has been solved. - Design of message passing to driver window has changed to prevent the driver dead locks which occored under a very seldom conditions. Build 123 : Solved problems: - Changing of the IP-Address while the driver was running had no effect. This problem has been solved. Build 122 : New functions: - Now the devices within a channel processing in parallel mode - Added "Bytes per Second" to device and channel statistics - Logging of read/write requests/responses into a log file is now configurable for individual data blocks. File logging can basically be switched on/off at register page "advanced" in power tool's setup screen. - This version is already prepared to support the S7T loadable database block which will allow to work with the S7 TIME and TIME_OF_DAY data types in iFIX database. - Shared memory management has been changed to fix problems with large configurations and multi processor systems. Solved problems: - In previous versions the S7 rack number had no effect. Now it will be mapped into the right bit positions of the "called TSAP" field when opening the ISO connection within the TCP connection. - An error in output simulations mode was fixed. - An error in auto creation of data blocks was fixed. Build 121 : Solved problems: - When the slot and rack number in the device setings has been changed, the new values wasn't used when the connection has been re-established. Now the change of these parameters take place immediately. Fixed wrong display of Transmitted and Received counters of mission control's driver statistics Build 120 : Solved problems: - The OPC Server now supports read/write of Control-Tags (eg. !Send or !PollRate). Address syntax is equal to Tag I/O Address within FIX-Database. - Driver doesn't work when a channel with 8 devices was configured Build 119 : Solved problems: - Timeout value of channel template was not stored, thus it had no effect when a new channel was added. - IP-Address of device template was not stored, thus it had no effect when a new device was added. - To avoid duplicate Connection Numbers in Driver Device config, now these Connection Numbers are assigned automatically. The "Connection No" Field in Powertool's device configuration dialog is no longer enabled for input but for displaying the assigned Connection Number. - The driver devices are now limited to 8 per channel. - When running S7A driver on W2K without hardlock key (Sentinel or HASP) the Power Tool displayed an exception message box and the driver didn't start in demo mode. This problem has been solved. Build 118 : New function: - Additionaly to the Rainbow Sentinel Key the driver now supports new Aladdin HASP Key. It detects either Sentinel or HASP Key. Solved problems: - The OPC Server now supports String addresses Build 117 : Solved problems: - Now the Input Data Area of S7 can be read. That was not possible in build version 115 and 116. - The I/O address in data base builder now accepts inputs of adresses without blank between data area identifier and adress index (e.g. D11:Q1.7 or D11:I10.0). - Under a certain condidtion the S7ADrv process runs into a deadlock situation This problem has been solved. Build 116 : Solved problems: - Now the switching between primary and backup device can be configured and used. If the communication to the primary device fails, the driver automatically switches to the device configured as backup device. If the connection to the backup device fails, it switches back to the primary device. The switching even can be done by "hand" using a DI Block with I/O address "!switch:. A F_CV value of 0 means primary device is active, F_CV value 1 means backup device is active. To change the active device just assign the corresponding value to blocks F_CV property. Build 115 : Solved problems: none New functions: - To make troubleshooting easier, the driver now allowes to log messages to a log file. To enable file logging, set the following registry key value to 1: "HKLM\Software\Intellution\Drivers\S7A\DriverDefaults\WriteLogToFile" By default this value is set to 0 which disables the file logging. The log file is named S7ADrv.log and located in the drivers default config file path. All messages dispatched to the driver server window will be logged in the file as well. Caution: The log file will "not" be deleted and re-created when driver starts. It's in the users responsibility to watch the size of the log file. If you enable the dispatching of debug level 1 to 3 messages, the log file can become a huge size in a short time!! Build 114 : Solved problems: - If SAC writes faster to the driver's image table than the driver can write to the PLC, after a while the write queue filled up to 100%. This version of the driver no longer queues write requests to PLC addresses which already in the queue, but uses the already queued request by just replacing the old write value by the new value. - In configurations with more than 1 channel, the driver sometimes crashed since This problem has been solved. - The Send and Receive counts in mission control was swapped. New functions: - Driver now runs with dongle RN-4BDNBH-B (S7A) and RN-4CKEBH-B (S7A+S7B) Build 113 : Solved problems: - Error in S7AMemory.dll caused problem in applications with very large configurations (where the shared memory "S7A_InputArea" exceeded 64k). The data type information stored in the Input Area was overwritten to WORD. Thus the SAC Task interpreted the block values always as an integer no matter what data type was set for the data block in power tool. New functions: - none Build 112 : Solved problems: - CSV File import now works fine - When choosed S7M 1.x compatible address syntax, the data type combo box was been set to a data type other that saved before. New functions: - The S7-TIME format now is supported. Use a TX-blocks and the data type identifier 'T' in the Tag's I/O address to address a double word which contains a value in S7-TIME format (e.g. D11.DB100.DBT 10). Write is also possible. Build 111 : Solved problems: - Sometimes the CPU load raised to 100% when writing values. - Read/write of the last byte of a data block failed. - Block write in manaul or auto mode failed - In data block statistics sreen sometimes "stats error windows" occured because some error codes could not be found in AGLink_Error.txt file. New functions: - none Build 110 : Solved problems: - none New functions: - Added hardware options NULLTERM and SPACEFILL to control string write. Build 109 : Solved problems: - Sometimes driver hangs after communication failed. Occured when a lot of data blocks with a very short poll time has been configured and processed. New functions: - none

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 vclzip@bigfoot.com for further information Thanks! Kevin Boylan

47 结构化绑定 48 条件分支语句初始化 49 聚合初始化 50 嵌套命名空间 52 lambda表达式捕获*this的值 53 改写/继承构造函数 54 用auto作为非类型模板参数 55 __has_include 56 fallthrough 57 nodiscard 57 maybe_...

在创建session时需要写入registration 方法,这样就可以解决这个问题了。....Session s = NotesFactory.createSessionWithFullAccess();Registration r= s.createRegistration()；r.switchToID(“c:\\notesIDFile.id”,“notespasswor...

1、前言页面端通常有需求想要准实时知道后台数据的一个变化情况，比如扫码登录场景，或者跳转到网银支付场景，在旧有的短轮训实现下，通常造成大量的不必要请求和查询，这里基于spring websocket+sockjs来解决该问题。2、websocketwebsocket是html5的一个新特性，目前oracle已经统一java websocket api，只要容器支持JSR356（tomcat7以上支持

数据库报表

6,108

社区成员

4,750

社区内容

发帖

与我相关

我的任务

社区管理员

加入社区

近7日
近30日
至今

加载中

查看更多榜单

社区公告

暂无公告

试试用AI创作助手写篇文章吧

+ 用AI写文章