Chip 1997 October

home *** CD-ROM | disk | FTP | other *** search

/ Chip 1997 October / Chip_1997-10_cd.bin / sharewar / jar / jar101dx.exe / JAR.DOC < prev next >

Wrap

Text File | 1997-02-03 | 120KB | 2,990 lines

USER'S MANUAL FOR THE JAR ARCHIVER PROGRAM January 1997 TOPICS COVERED IN THIS DOCUMENT ------------------------------- MAJOR FEATURES OF JAR ARCHIVER BENCHMARKING INSTALLATION QUICK START TO USING JAR CONVERTING OTHER ARCHIVE FILES TO JAR FORMAT HOW TO USE JAR IMPORTANT NOTES USING JAR AS A BACKUP PROGRAM JAR ERROR SITUATIONS HOW JAR USES AVAILABLE MEMORY ARCHIVE CHAPTERS JAR USER ACTION PROMPTS JAR COMMAND LINE SYNTAX JAR COMMANDS JAR SWITCH OPTIONS CONFIGURATION FILE CONFIGURATION FILE KEYWORDS IMPORTANT DIFFERENCES BETWEEN JAR AND ARJ JAR LIMITATIONS JAR ARCHIVE ACCESS MODES JAR ARCHIVE FLAGS JAR ERROR LEVELS FINAL COMMENTS MAJOR FEATURES OF JAR: JAR comes as a 16-bit executable (JAR16.EXE) for DOS and a 32-bit executable (JAR32.EXE) for Windows 95 and Windows NT (version 3.50 or higher). The latter takes full advantage of 32-bit instructions and the "flat" memory model. Sometimes this 32-bit executable is referred to as the "Win32 JAR version". Currently, JAR ranks as one of the best in compression in terms of size reduction of the currently available archivers. JAR compresses significantly better than such popular archivers as PKZIP 2.04, UC2 rev 3, RAR 1.55, RAR 2, and LHA. JAR uses a modified form of SOLID archive technology to provide excellent compression, fast updates and fast extraction. The ability to process and archive more than 50,000 files at one time. JAR provides the CHAPTERS concept which allows one to store multiple backups in the same archive using incremental compression. Archive individual file or chapter comments with the option of inputting comments from a file. Support of ANSI style comments. JAR has MS-DOS international language support for the proper casing of filenames. The ability to put the entire JAR command line in a response type file as in "JAR @command.rsp". Supports nested response files. 32 bit CRC file integrity check. Total percentage milestone display for the entire archive build. Option to test new archive before overwriting the original archive including the option to do a byte for byte file compare during a build. Archives that can span diskettes. This allows the user to backup a full hard disk drive to multiple floppies. Recovery of individual files is convenient because the JAR archive header keeps information about the disk location of each file. Built-in facility to recover files from broken archives. Also, the capability to add recovery records to important archives which allow one to repair damaged archives. Internal integrity check in JAR to resist hacking or virus attacks. Archive security envelope "seal" feature to resist tampering with secured archives. This feature disallows ANY changes to a secured archive. Archive and chapters locking feature. Password option to protect archived files both using authentication or encryption. Specification of the files to be added to or exclude from an archive via one or more list files. In addition, JAR can generate a list file. Specification of files to be excluded from processing by JAR. Special command to test JAR reliability. "testjar.bat" is a ready-to-use batch file for testing on specified drive. Optional long filenames translation into short names when extracting files in DOS. Configuration file which allows the customization of default JAR behavior. Includes a set of default options for each command. Breaks 640K DOS barrier using EMS and/or XMS. Query option when wildcard selection is not suitable. Files sorting in list commands. 32-bit JAR version allows the use or storage of both long and/or short filenames. It uses filenames compatible with DOS OEM encoding. ARCHIVER BENCHMARKING: Benchmarking results are very dependent upon the data used for the tests. We compared JAR with some of the leading commercial archivers: PKZIP 2.04g, UltraCompressor II Version 3, RAR 2.00 for Win32, and RAR 2.00 for DOS. LHA 2.X and ARJ 2.X are not included. Their compression size numbers are very similar to those of PKZIP 2.04g. COMPUTER: Pentium-100, 24MB memory, no external cache OPERATING SYSTEM: Windows 95 CONFIG.SYS: device=c:\win95\himem.sys dos=high files=40 devicehigh=c:\dos\power.exe JAR.CFG: 32MemoryLimit=10240 CheckAvailableMemory=No The only exception is JAR16.EXE which was tested in DOS 6.22 with 16M of EMS memory. This was due to the slow Windows 95 EMS implementation which decreases JAR's performance relative to the amount of EMS used (up to 1.5 times). We recommend the use of JAR32.EXE in Windows 95 and Windows NT. FILE SET: The Calgary Corpus (18 files totaling 3,251,493 bytes, popular file set used to compression rates). Archiver Method Compression time Compressed size JAR32 Maximum 31s 903,863 (100%) JAR32 Default 21s 910,676 (101%) JAR16 under DOS 6.22 Maximum 41s 935,675 (104%) with 16MB EMS JAR16 under DOS 6.22 Default 29s 945,690 (105%) with 16MB EMS RAR 2.00 -m5 -mde -s 261s 951,992 (105%) RAR 2.00 Default solid 48s 997,418 (110%) RAR 2.00 Default 41s 1,008,455 (112%) RAR 2.00 for DOS Default solid 25s 1,030,894 (114%) UC2 rev 3 Default 30s 1,051,516 (116%) PKZIP 2.04 Default 17s 1.074,550 (119%) FILE SET: Microsoft Visual C++ 1.52 (462 files and directories totaling 18,806,153 bytes, mix of executables, text, DLLs, and object files). Archiver Method Compression time Compressed size JAR32 Maximum 264s 5,984,311 (100%) JAR16 under DOS 6.22 Maximum 341s 6,279,415 (105%) with 16MB EMS RAR 2.00 -m5 -mde -s 1004s 6,351,777 (106%) JAR32 Default 173s 7,050,008 (118%) JAR16 under DOS 6.22 Default 250s 7,267,567 (121%) with 16MB EMS RAR 2.00 Default solid 273s 7,856,362 (131%) RAR 2.00 for DOS Default solid 151s 8,314,078 (139%) UC2 rev 3 Default 190s 8,556,558 (143%) PKZIP 2.04 -ex 140s 8,906,367 (149%) PKZIP 2.04 Default 96s 8,945,476 (149%) JAR16/JAR32 shows its best compression with large directories of files. INSTALLATION: This assumes that you have already executed the self-extracting distribution archive and extracted its archived files into a directory. To install the JAR software, simply copy JAR16.EXE, JAR32.EXE, JAR.CFG, REARJ.EXE, REARJ.CFG and optionally JAR_C.COM, JAR_C.DLL (available in non-export version only) to one of the directories named in your DOS PATH statement found in your AUTOEXEC.BAT. On many PCs, this directory may be C:\DOS or C:\BIN. In addition, for a DOS machine or for a DOS environment one MB of available XMS strongly recommended. Four MB of EMS preferred for outstanding compression. More memory is better. Detailed information about memory usage is available in section "HOW JAR USES AVAILABLE MEMORY". QUICK START TO USING JAR: See the document INTRO.DOC. CONVERTING OTHER ARCHIVE FILES TO JAR FORMAT: Included with this software is the program REARJ. This program can be used to individually or collectively convert archive files from other formats to the JAR format. Read REARJ.DOC for more details. HOW TO USE JAR: If you type JAR16 [return], you will see a simple help screen. If you type JAR16 -? [return], you will see more detailed help information. Please note that unless specifically indicated, references to JAR16 also include JAR32.EXE. IMPORTANT NOTES: When using the "-w" working directory switch, JAR does not check on space availability before overwriting the original archive if it exists. Be sure that you have enough disk space for the new archive before using the "-w" switch. If JAR aborts in this situation because of disk space, JAR will keep the temporary archive. Like ARJ and PKZIP, JAR requires extra disk space to UPDATE an archive file. JAR will backup the original archive while it creates the new archive, so enough room must be available for both archives at the same time. Also JAR requires additional space for a swap file which is used to store the archive header and a list of files to add during processing. By default during an add command, JAR will ALSO add to the archive all directories scanned while searching for matching filenames. Even if NO files were found, JAR will build an archive if any directories were scanned during the operation. See the "-hb" option for more information. Currently, JAR will not extract overwriting a readonly file unless the "-ha" option is specified. JAR uses a configuration file to store global user settings. By default the configuration file is named "JAR.CFG" and must be placed into the same directory as the JAR16.EXE and JAR32.EXE directory. USING JAR AS A BACKUP PROGRAM: JAR can be used as a substitute for a backup program. The following partial command lines illustrate a full backup command, an incremental backup command, and a restore command. The only parts missing are the names of the files to backup/restore. JAR16 a backup1 -jt -r -m4 -b0 JAR16 a backup2 -hba1 -b0 -jt -r -m4 JAR16 x backup2 You may then copy archive into diskettes: JAR16 y backup -oa:backup -vvas -hk Or you may create backup directly on diskettes: JAR16 a a:backup -jt -r -m4 -b0 -hk -wc:\ -vvas You should familiarize yourself with the above switches so that you can modify the above command lines as needed. Another possibility for incremental backups is to use chapter archives: JAR16 ac chapters -ht1 -r -js1 -hl -hz To restore the last chapter use: JAR16 x chapters To restore any other chapter (in our example #5) use: JAR16 x chapters -jb5 It may be useful to review chapter comments: JAR16 lc chapters We also recommend you test JAR reliability with your computer and software using the JAR "test" command. See TESTJAR.BAT. JAR ERROR SITUATIONS: ADD: If a user specified file is not found during an add, JAR will continue processing. Note that files specified within an JAR listfile that are not found during an add will NOT trigger an error unless the "-hl" option is also specified. In a disk full condition or any other file i/o error, JAR will promptly terminate with an error condition and delete the temporary archive. MOVE: JAR will only delete files after all them have been successfully added to the archive. EXTRACT: In a disk full condition or any other file i/o error, JAR will promptly terminate with an error condition and delete the current output file. CRC ERRORS OR BAD FILE DATA: In the case where an JAR archive has been corrupted, JAR will report a CRC error or a Bad file data error. These corruptions can be the result of an unreliable diskette, a computer memory problem, a file transfer glitch, or incompatible CACHING software or memory manager. Most of these errors are the result of file transfer glitches and bad diskettes. CRITICAL ERROR HANDLER: JAR sets up an interactive critical error handler to handle DOS critical errors like "sector not found" and "drive not ready". When a critical error occurs, JAR will prompt the user with the message "Retry Y/N?". The user may retry the failed operation by pressing "Y". Pressing "N" will fail the operation. The user can press Control BREAK to abort to DOS. There is an option to disable the critical error handler in the JAR configuration file. JAR32.EXE doesn't have an interactive hardware error handler. HOW JAR USES AVAILABLE MEMORY: One of JAR's advantages is its ability to use additional memory for better compression. Unlike most other archivers the JAR DOS version can use EMS or XMS to improve compression. JAR DOS version JAR uses a fixed amount of XMS memory to cache its overlay code (about 300K) and requires an additional 500K of XMS memory for better performance. When JAR compresses data it first looks at available EMS memory. If there are more than about 640 Kbytes EMS and the compression method is -m2..-m4 then JAR allocates EMS and uses it to improve compression. You may specify that JAR use up to 16384 Kbytes of this memory. The more EMS - the better compression achieved, but JAR will work slower. If there is not enough EMS then JAR uses XMS instead. But "XMS based" compression is significantly worse. JAR compression rate and speed does not depend significantly on the XMS amount used. The optimal value is less than 4 Mbytes. The JAR XMS engine consumes about 60K more DOS conventional memory compared to the EMS engine. When extracting, JAR uses available XMS or EMS for faster decompression. The recommended value for data that has been compressed using the "-m2" and "-m3" method is 1.5 Mbytes of free EMS and/or XMS, for the "-m4" method, use at least 3 Mbytes of free EMS/XMS. With less memory JAR may use a swap file instead (this results in worse performance). When both EMS and XMS are available JAR itself chooses the optimal memory amount and type for the selected command. The JAR DOS version requires EMS 4.0+ and XMS 2.0+ versions. Although the JAR DOS version achieves significantly better compression and uses less memory with EMS available, special care is required to ensure that the EMS driver (such as EMM386.EXE) is configured and works properly. The EMS driver works in a more "complex" way than XMS. It searches for a "hole" in DOS address space for the page frame. Usually it also works with 386+ memory paging enabled. As result some older DOS software or hardware may conflict with the EMS driver. Refer to your memory manager documentation for details. We strongly recommend that you ensure that DOS JAR works reliably with EMS enabled using the "test" command. If you detected problems then you must disable EMS usage using "EmsLimit=0" in the JAR configuration file. When running JAR in Windows 3.X or OS/2 be sure the PIF file/DOS session settings allow JAR to use EMS and XMS. By default XMS and EMS usage may be limited to 1 Mbyte. We recommend increasing these settings to 4 Mbytes or more. JAR Win32 version JAR for Win32 requires about 1 Mbyte of memory for executable and data. The compression engine may use additional free memory for better compression since more available memory allows the JAR compression engine to match strings at larger offset distances. Very good compression is achieved when JAR uses 3-4 Mbytes of memory. 8-10 Mbytes gives the best compression. JAR may use less memory (0.5-3 Mbytes) for good compression, too. The JAR Win32 version works significantly faster and compresses better than the DOS version. So in Windows NT and Windows 95, you should use the Win32 JAR version. In DOS, Windows 3.X and OS/2 - use the DOS JAR version. Both versions The simplest way to tune JAR memory usage is to set "CheckAvailableMemory=Yes" in the configuration file. Then JAR detects the free physical memory currently available and uses it. Note that JAR never uses more memory than the limits that are set in the configuration file ("XmsLimit", "EmsLimit", "32MemoryLimit") even if there is actually more free memory. If "CheckAvailableMemory=Yes" you may use very large limit settings (e.g., 8192 Kbytes each). Experienced users may set "CheckAvailableMemory=No" and then specify the "real" memory amounts in "XmsLimit", "EmsLimit", "32MemoryLimit". Be sure that the memory specified is really available; otherwise, in operating systems as Windows 95, Windows NT, OS/2, intensive memory swapping may occur which will decrease overall system performance. "CheckAvailableMemory=No" is also recommended for batch processing. But again SPECIFY MEMORY AMOUNTS that ARE REALLY AVAILABLE. It is better to specify a smaller memory amount than wait for hours due to swapping. The EMS implementation in Windows NT is slow. So for good compression do not use the JAR DOS version. Use the JAR Win32 version instead. ARCHIVE CHAPTERS The advantage of a chapter archive over a series of separate archives of the same set of files is that space can be saved when files to be archived do not change or change slightly. One chapter archive can produce the same results as one hundred standard archive backups. In other words, one can update a chapter archive one hundred times and fully recover any one of those updates. Technically speaking, each file in a chapter archive has a chapter lifespan range. When a file is backed up and has not changed since the last backup, only the lifespan range is changed. A copy of the file is NOT added to the archive since a copy already exists. If the backed up file is different from the previous version, that file is added to the archive with an appropriate chapter range. The previous version of the file is still kept in the chapter archive. When a particular chapter is restored, JAR selects only the files whose lifespan includes the chapter number requested. As an example, suppose one made a series of daily chapter backups of a document directory where only one document out of the one hundred documents changes on a daily basis. The chapter archive would contain the original one hundred documents plus a copy of each day's changed document. JAR provides a lot of services for chapter management: keeps chapter creation/modification time, allows commenting or locking, can extract a specified chapter and also, subsequent changes to it. EXAMPLE Suppose you started to develop a new programming project and decided to use a JAR chapter archive for backups. The project directory is "c:\prj\", so you have written a special "back.bat" with the command line: JAR16 ac archive c:\prj\*.* -r -x*.obj -x*.exe -jt You are not interested in object or exe file backups so they are excluded. "-jt" stands for temporary archive testing after each update. After some coding you have completed the "main.h" file. "back.bat" adds this file into first archive chapter. Next thing you have completed is the "main.c" file. The second run of "back.bat" detects that there is no change to "main.h" since the previous chapter as file size and modification time have not changed. So JAR compresses only the new "main.c" file. As noted before each file has a chapter lifespan range. In your archive main.h lifespan range became from 1 to 2 ("<1-2>" in JAR notation) main.c - " - " - from 2 to 2 ("<2-2>" in JAR notation) Obviously JAR saved additional space as it did not compress and save main.h twice. JAR16 x archive -jb1 extracts file(s) which are included in the first chapter. That is only "main.h". The "-jb1" option is used to select chapter #1. JAR16 x archive -jb2 extracts the second chapter: both "main.c" and "main.h". The same result occurs when extracting without the option "-jb2" since JAR by default extracts (or lists/updates) the last chapter in the archive. Next step is a bug fix in file "main.h" and another backup. JAR detects the change of "main.h" and no changes to "main.c". So now the archive contains 3 files: main.c <2-3> main.h <1-2> main.h <3-3> (can be listed using JAR16 ls archive -jb1-3 ) Yes, there are two files named "main.h". However, the first one belongs to chapters 1 and 2 while the second belongs to chapter 3 only. So with the "-jb" option you can fully reconstruct any of these chapters. If you used separate archives for each backup then they would contain now 1+2+2=5 files (instead of only 3 files in chapters case). Moreover, in the case of chapter archives JAR optimizes the compression as it detects that main.h <1-2> and main.h <3-3> do not differ very much, so it compresses only the difference between them. Suppose two days later your project contains 20 files. And chapter archive contains 30 updates. You can list or extract any of them. Archive is still very small as JAR compresses only changes between chapters. If you used 30 separate archives instead of a single chapter archive, they would occupy probably 10 times more disk space. Suppose you noted that somewhere in the previous backups that an important piece of code was deleted from "main.h". You don't need to search all 30 archives for this code! JAR16 e archive main.h -jb1-30 extracts "main.h" from chapters 1-30 each into separate directories. Chapters where "main.h" has not been changed are skipped. The directory name is generated by enclosing in brackets the lifespan start number. As a result JAR extracts like this: main.h <1-2> to (1)\main.h main.h <3-15> to (3)\main.h main.h <16-30> to (16)\main.h You need to review only these 3 files to find the required code as all other chapters contain duplicates. Upon a successful search you may decide to comment chapter 3 with a special note about the useful code using the: JAR16 c archive -jb3 -hz command. So the main advantage of chapters archive is compact and convenient storage of many updates within single archive. There is easy access to any chapters. And of course, there is major disk space savings compare to multiple separate backups. Even if you decide to start a new archive backup after the 100'th chapter, your savings in compressed size will be outstanding. END OF EXAMPLE Please read the "ac", "lc" commands, "-js" and "-jb" options for further information about chapters. JAR USER ACTION PROMPTS: JAR prompts the user for action at certain times. There are several types of prompts. One is for yes/no permission, another is for a new filename, another is for archive comments. The JAR yes/no user prompts provide a lot of flexibility. In addition to the normal yes and no responses, JAR also accepts the following responses: quit, always, skip, global, and command. "Global" sets JAR to inhibit all subsequent user prompts by assuming YES for all queries as if "-y" were specified. "Always" sets JAR to assume YES for subsequent queries of the same class. For example, answering ALWAYS to the overwrite query will assume YES for ALL subsequent overwrite queries. "Skip" sets JAR to assume NO for ALL subsequent queries of the same class. "Command" prompts for one DOS command and then executes it. JAR then returns to expect an answer to the current yes/no query. Do not install a TSR program using this command prompt as JAR may not have enough memory left to continue execution. Abbreviated responses such as "y", "n", "g", "a", "s", "c" are allowed. JAR COMMAND LINE SYNTAX: JAR{16|32} <command> [<switch>...] <archive_name>[.j] [<arguments>...] <arguments> ::= <switch> | <wildcard> | <filename> | <list_file> | <response_file> <switch> ::= {'/'|'-'}<switch_character>['-'|<option>] <list_file> ::= '!'<filename> <response_file> ::= '@'<filename> Commands and switches can be entered in upper or lower case. Command is the first non-switch argument. JAR supports the use of either "-" or "/" as the switch option character. You may mix them. -s turns on switch "s" -s- turns off switch "s" -sname provides the name argument for switch "-s". Wildcards or filenames with spaces must be quoted: "c:\Program files\*" or c:\"Program files"\* Wildcards or filenames that can be parsed as a switch option must be quoted, too. JAR parses a switch option until the first quote. The remaining part may only be a switch argument: -x is switch "-x" is filename -x- turns off "-x" option -x"-" excludes filename "-" -"x-" reports error as "-" is not followed by a switch character (instead argument 'x-' is supplied) If you need to specify a quote character as an argument then you must specify two consecutive quotes (and argument itself must be quoted too): JAR16 a test -g123""""4567 or JAR16 a test -g"123""4567" in both cases password 1234"4567 is specified. Switch options can be combined to save command line length. However, some switch options take optional string arguments and therefore, must be the last option in a combined argument token string because JAR accepts the rest of the argument token as the optional argument. You can combine "-r" and "-y" to make "-ry". You can combine "-wd:\" and "-r" with "-rwd:\", because the switch option taking the string argument is last. Please note that switches that only accept the "-", "0", "1", "2" modifiers can be combined in any order. WARNING: The "j" in "-jX" option switch is a SHIFT symbol. Therefore, a combined "-jftl" is a combination of "-jf", "-jt", and "-jl", since the "j" shifts the meaning of subsequent symbols in the argument token string. Examples: -ryjftl <=> -r -y -jf -jt -jl -b1ywd:\ <=> -b1 -y -wd:\ The "h" is another SHIFT symbol. The standard JAR file suffix is ".j". The JAR archive name must be the first filename on the command line after the command. The switches and other filenames can be in any order. Wild names follow the Windows 95 (NOT DOS!) convention. Examples: "*" or "*.*" means all files "*.DOC" means all files with an extension of ".DOC" "?B*" means all files with a second character of "B" "??" means all two character length filenames "*1*" means all names that contains character "1" "*.*.*" means all names that contain two "." characters NOTE: There are minor differences with Windows NT wildcard matching. JAR interprets a wildcard (?) as "any character" instead of "zero or one character". The default for <wild_name> for all commands except for "d" is "*". The argument "" may be used as the wildcard that never matches any files. You can supply one or more filenames for files containing lists of files to be added to an archive. The filenames must be listed one per line. The list filename(s) must be prefixed with "!" as in "!listfile". You can also build a long command line using response files. The response file is "included" into the command line at the specified place. Multiple and nested response files are allowed. The command line may be split over as many lines as necessary. Example: JAR16 a archive @switches.txt @wildcards.txt JAR COMMANDS: a: Add files to archive This is the basic command to add disk files to a JAR archive. If you specify the "-r" switch (recurse subdirectories), JAR will add all of the files in all of the subdirectories that match the specified wildname. Along with these files JAR also will add all of the scanned subdirectories. Note that this last action is similar to UC2 but different from ARJ behavior. See the "-hb" option for more information. Examples: JAR16 a archive subdir\* Archive all files in directory "subdir". JAR32 a archive "c:\Program files\*" -r Archive all files in directory "Program files" and subdirectories. NOTE: Both DOS and Win32 JAR versions ALWAYS use Windows 95 like wildcard matching style. This means that "subdir\*" works like "subdir\*.*". The wildname "subdir\*." selects files without extension. After add and listing commands with all files selected JAR shows the total archived data ratio. This ratio does not include JAR headers and recovery records. Only "pure" compressed data. JAR compresses archive header. As a result whole archive size differs time to time even when compressing the same data. ac: Add new Chapter The "ac" command works like the "a" command, but adds files into a newly created chapter. JAR manages an almost unlimited number of chapters. See the "-js" option for IMPORTANT DETAILS about methods on how files are added to a new chapter. Example: JAR16 ac archive subdir\* Add all files in the directory "subdir" into new chapter JAR uses two compression methods. The "solid" method is used to compress files in the usual manner. The "version" method is used to compress files with the same path but within different chapters. This method is optimized to compress only the DIFFERENCES between files while updating from chapter to chapter. NOTE: The current JAR version supports "version" compression method only for files with a size less than about 500 Kbytes. "Version" compression also allows faster updates to an archive. The only exception is the first time that a new file version is compressed. In this case JAR must change the compression method used for the file (from "solid" to "version") and this takes additional time. The next version update will be significantly faster. See the "-jb" option for more details how to extract or list chapters. c: Comment archive files This command allows you to comment the archive header or individual files or archive chapters. JAR will prompt you for each comment. The user will be prompted for up to 50 lines for each comment. A line containing only a [return] will terminate the comment. The user can choose to input comment data from a file by entering the comment filename preceded by an "!" as in "!archive.txt" starting in column 1 of the first comment line. This file is read as a text file. Only the first 4096 bytes of the file will be accepted by JAR. To erase a comment from an archive, type [space] [return] on the first comment line and [return] on the second comment line. To add file comments but not archive comment, use following command: JAR16 c archive To add the archive comment but not file comments, use the following command: JAR16 c archive -z To add only the archive comment at the command line, use the following command: JAR16 c archive -zcomment.txt To remove a comment via the command line, use predefined "NUL" filename: JAR16 c archive -zNUL You may also comment an archive with other update commands: JAR16 a -zcomment.txt archive *.txt JAR allows the storing of ANSI style comments. While displaying these comments JAR itself emulates ANSI.SYS driver behavior in a virtual screen and then dumps the complete picture to the output. NOTE: The virtual screen is two columns narrower than the current "real" screen. cfg: extract default configuration file This command accepts a single argument - destination filename. If it is omitted, JAR extracts the configuration file as "jar.cfg". This command is useful if the configuration file is missing or damaged. d: Delete files from archive This command allows you to delete files from the archive. When wildcard selection is not suitable, you can use the "-q" switch to set JAR to prompt you for deletion for each file selected. These deleted files are physically removed from the archive. Examples: JAR16 d archive *.c Delete all files in archive ending in ".c" JAR16 d -q archive *.c Prompt before deleting each file ending in ".c". Even if JAR removes all the files, the empty archive WILL NOT be deleted. You can specify the "-n" (new) option to delete the files in the archive that DO NOT EXIST on the disk. Then the "-o" option may be used to specify the directory to scan. Example: JAR16 d archive *.* -n da: Delete files in All chapters Delete the specified files in all chapters while showing the deleted files with the corresponding chapter lifespan ranges. The arguments are the same wildcards as used with the "d" command. Example: JAR16 da archive *.c Deletes all files in all chapters ending in ".c". dc: Delete Chapters Delete chapters in the specified range. The command arguments are chapter range "1-", "-2", "1-5" like. The "-x" option may be used to exclude a range. Each file's chapter lifespan range is modified with regard to the existing chapters. If there are no remaining chapters then the file is removed. Examples: JAR16 dc arch 5 6 8 deletes chapters 5, 6 and 8 JAR16 dc arch 5-8 11-12 -x6-7 deletes chapters 5, 8, 11, 12 JAR16 dc arch "-10" deletes first ten chapters JAR16 dc arch 10- deletes all chapters from 10 and above e: Extract files from archive This command will extract one or more files from the archive to the current directory or to the output directory if specified. JAR will prompt the user before overwriting existing files unless the user specifies the "-y" switch. If the user gives a "no" answer, JAR will prompt must file be extracted to other filename. Then a "no" answer skips file and "yes" prompts for new filename. JAR does not ask overwrite permission while extracting existing directories unless the archive directory and existing directory attributes do not match. The "-o" option can be used to specify the destination directory. Example: JAR16 e archive -osoft\ *.c Extract all files ending in ".c" to subdirectory "soft". If JAR encounters a CRC or Bad file data error during extraction, JAR will delete the corrupted extracted file unless the "-jr" option is used. You can extract files from a specified chapter using the "-jb" option: JAR16 e archive *.txt -jb20 Using same option you can also extract files which span a specified chapter range. Example: JAR16 e archive file.txt -jb5-10 Extracts the "file.txt" changes in chapters 5-10. For convenience when extracting a chapter range, JAR places files with different starting chapters into different directories "(5)","(6)"... (brackets enclose starting chapter number). NOTE: During extraction JAR DOES NOT EXTRACT DUPLICATE files into all chapter directories from the file lifespan range. A file is extracted ONLY into the starting chapter directory of the file. This feature saves space and lets one easily find chapters where the selected file has been changed. So "JAR16 x archive -jb5-10" will reconstruct the full chapter 5, BUT EXTRACT ONLY THE CHANGES in chapters 6-10. f: Freshen files in archive Update matching files in the archive that are OLDER than the selected disk files. Example: JAR16 f archive *.c *.h l: List contents of archive List contents of archive. You can get sorted output by using the "-ju" option. Example: JAR16 l archive *.c *.h You can specify the "-n" (new) option to list the files in the archive that DO NOT EXIST on the disk. The "-o" option may be used to specify directory to scan. Example: JAR16 l -n -oc:\source archive *.c *.h When listing a few chapters JAR appends the corresponding file's chapter lifespan range in the format "<from-to>". Example: JAR16 l archive *.c *.h -jb1-10 lc: List Chapters List chapter information: comment, creation and modification time, and lock flag. The command arguments are sets of chapter ranges as in the "dc" command. Example: JAR16 lc archive List all chapter information in archive ls: List contents of archive (Short form) List contents of archive in the shortest form. Example: JAR16 ls archive *.c *.h lt: List Technical information for shells List technical information for archiver shell programs. This command accepts a single parameter - a file that must not exist on disk. JAR dumps the archive header information to this file. Example: JAR16 lt archive header.txt See TECHNOTE.DOC for details. m: Move files to archive The "m" command adds the selected files to the archive. If the adds are successful, then the added files and directories are deleted. The move command does not ask permission before deleting the files. Example: JAR16 m archive soft\*.* It is STRONGLY suggested that you always use the "-jt" option with move commands to set JAR to verify the archive before deleting the input files. reg: register program The "reg" registration command is available only in the JAR DOS version. When you order JAR, you will get a registration key which matches the OWNER and security envelope SIGNATURE specified on the order form. The "reg" command simultaneously registers both JAR16.EXE and JAR32.EXE. Both executables must be in the same directory before processing. Since JAR will modify both executables, JAR16.EXE and JAR32.EXE must not be in use. If the "reg" command is executed without a filename argument, JAR will prompt for the registration key interactively. Otherwise, JAR will search for the registration key within the file specified. Examples: JAR16 reg key.txt JAR16 reg rh: Recover archive Header This command is available only in Win32 JAR version. If your archive is damaged, the remaining undamaged files may be recovered in the following ways: - you can try to extract files using the "-hx" option. This option uses the recovery records (if available in archive) to repair slightly damaged archives. - you can try to extract files using the "-jr" option (this prevents JAR from aborting on damaged files). The "-hx" and "-jr" options can be combined. Both listed options do not help if the archive header is damaged. The "rh" command may recover useful archive data (if such) and the header from a seriously damaged archive. It uses signatures to search for valid compressed data chains. You must place all damaged archive volumes into same directory and then run: JAR32 rh archive JAR will output the recovered archive into the current directory as "$recover.j". The archive may still contain partially damaged data so it may be necessary to use the "-jr" option to extract data from the archive. NOTE: The "rh" command recovers only the most important archive data: file contents, file CRC, file modification date but not time. The file's chapter lifespan range information is lost, too, but JAR tries to group different versions of the same file into different chapters. Be sure you specify the correct password and method for the protected damaged file; otherwise, JAR will report that there is no recoverable data. After repair of archive with multiple chapters, "$recover.j" may contain a smaller number of files than the original archive (even if it has not been damaged). This is due to the loss of "CHAPTER range" information and as a result two files with the SAME DATA in neighboring chapters may be recovered as a single file. If you have lost one or more archive volumes, then after all existing volumes placed into single directory, before running "rh" command, you must substitute lost volumes with any other file which size greater than 64 bytes. E.g., with file which contains 65 space characters. Obviously JAR will not find recoverable data in such volume substitution, however such file will allow JAR "rh" command to process through all volumes. sl: Scan Locked files After command scans selected file set it tries to open each file for reading. Files which JAR failed to open are listed. You can copy this list into file using "-l" option. Later this list file may be used to exclude locked files. Command does not accept archive name. Example: JAR32 sl c:\* -r -l"To be excluded" NOTE: When wildcard used with "-x" or "-x!" does not contain path specified (only filename) then this wildcard is used to match to ALL filenames (path is ignored). You must keep this in mind when using exclude lists generated by "sl" command. Read "-x" option for details. t: Test integrity of archive Test the contents of ALL files for the correct CRC value. JAR uses a 32 bit CRC to validate the contents of the files. test: test JAR reliability This is the command to test the reliability of JAR compressing and decompressing selected files. When executing this command JAR works in similar way like "add" command. But after each compressed file portion JAR decompresses and tests files using comparison and then removes compressed data. This saves disk space and allows JAR to test gigabytes of files with limited free disk space. This command requires a temporary archive filename which will be used for testing. The needed free disk space can be found from the following table (in megabytes): Compression method: -m1,-m2,-m3 -m4 1,000 files to test 6M 20M 5,000 files to test 10M 30M 20,000 files to test 20M 40M 50,000+ files 60M 60M This table assumes that the JAR swap file is located on the same drive as the temporary archive. Otherwise, use following table (space for temporary archive/swap file): Compression method: -m1,-m2,-m3 -m4 1,000 files to test 4M/3M 20M/5M 5,000 files to test 4M/5M 20M/7M 20,000 files to test 4M/10M 20M/15M 50,000+ files 4M/30M 20M/30M Depending on the free disk space JAR will also skip files too large to test. JAR must have enough space to store whole compressed file before testing decompression and comparing contents. When running this command JAR also uses additional methods to detect conflicts with hardware or software, memory or disk corruption. If JAR detects such problems then it will report the following errors: Wrong program flow detected at (<module>,<line>) Possibly conflict with used software or hardware or Internal at (<module>,<line>) Possibly conflict with hardware We strongly recommend that you run such tests before using JAR, especially for backup purposes. It is also useful to retest JAR after hardware upgrades or installation of new memory managers or other "critical" software. TESTJAR.BAT contains a ready-to-run command line. You can modify this batch file to test other compression methods (-m1,-m3, or -m4) or other drives. JAR16.EXE and JAR32.EXE should be tested separately in the intended operating environments. Example: JAR16 test archive c:\* d:\* -r -m1 -ho -jf -xwin386.swp Test compression/decompression of the files on few drives. "win386.swp" is excluded as it may be locked. Be sure you exclude files which may be locked for reading during testing; otherwise, JAR will abort on the locked file. Testing may take a long time (hours or days). You can interrupt or shorten any test by pressing CTL BREAK. With the "-ht0" option JAR will test the decompression using CRC32 only instead of a file contents comparison. This may work faster. It may be useful also to test encryption/protection using "-g" and "-hg" options together with "test" command. u: Update files to archive Update older files in the archive and add files that are new to the archive. Example: JAR16 u software v: Verbosely list contents of archive This command lists the full pathname and comments of the archive files as well as the same information as the "l" command. It also shows the compression method, file access and creation time, and host OS. This command may be used with the "-hf" options to look at long filename translations. x: eXtract files with full pathname This command extracts one or more files from the archive to their full paths in the current directory or to the output directory if specified. Example: JAR16 x archive *.c Refer to the description of the "e" command for more information about extraction. y: copY archive with new options The JAR "y" command provides an easy interface to the options that change archive settings. Examples: JAR16 y archive -jl locks archive JAR16 y archive -g? -jg? changes archive password JAR16 y archive -hk adds recovery records to archive JAR16 y archive -hk- removes recovery records JAR16 y archive -v20k -osplit splits archive into 20K volumes JAR16 y split -oarchive concatenates archive volumes back JAR16 y archive -hx repairs damaged archive using recovery records NOTE: When changing the password JAR always tests the archive. JAR SWITCH OPTIONS: b0: reset archive bits The "-b0" switch does NOT affect file selection. It causes JAR to reset the archive bits of added files. If the "f", "u", or "ac" command has been selected, JAR will also reset the archive bits of files that are already duplicated in the archive. Example: JAR16 a e:archive c:\*.* -b0 -r archives all files on the C drive and resets all archive bits. This option also causes JAR to reset the archive bit of extracted files. NOTE: JAR ignores archive bit set/reset errors. If error occurs then added file archive bit value is not changed. For extracted file archive bit remains set to 1. b1: set archive bits Works like "-b0" but sets the archive bit instead. d: with Delete (move) This switch with the "a" command is similar to the standard MOVE command. Successfully added files will be deleted. When used with the "f", "u", or "ac" command, JAR will also delete files that are already duplicated in the archive. JAR will prompt the user before deleting the files unless the "-y" or "-jyd" switch is specified. Also, you can use the "m" command which does not prompt before deleting the files. Example: JAR16 a archive filename -d -y is equivalent to JAR16 m archive filename This option works with "add" type commands only. f: Freshen existing files This switch works only with the "e" and "x" commands. "-f" specifies JAR to extract newer files from the archive. "-f1" specifies to extract older files. "-f2" specifies to files with a changed date/time or size. NOTE: JAR ignores date/time comparison between directories (always assumed are the same). Example: JAR16 x archive -osource -f Extract newer files g: specify archive password This switch will specify the archive password. An archive can be protected from unauthorized access by using a password with either of two different protection methods. The protection method may be specified using the "-hg" option. The "Authenticate" method establishes an archive password WITHOUT the actual encryption of its contents. The "GOST v1.01" method uses GOST 28147-89 encryption in CBC mode to encrypt archive files (but NOT archive header which contains comments and filenames). With the default configuration file and without specifying the protection method ("-hg") JAR uses the "Authenticate" method. The "DefaultAccessProtection" setting in the CFG file allows one to change this behavior. NOTE: Encryption is not available in the export JAR version. Once an archive is protected, JAR requires the password when extracting or updating the archive. JAR stores within the archive header a secure password hash. This allows verification of the specified password before the actual update/extract operation. But there is a very small (smaller than 1e-30) probability that the hash function will not detect an incorrect password during an update operation. In this case, different files may be encrypted using different passwords. So JAR may report "CRC error" or "Bad data" when using the wrong password for update. Short and trivial passwords allow brute-force attacks against protected archive. We recommend to use passwords longer than 16 characters. Special characters and digits in password also recommended. Example: JAR16 e archive -gpassword If the "-g" option is followed by a question mark "?", JAR will prompt the user for the password without displaying the password. The backspace key can be used to erase characters. Example: JAR16 a archive -g? Use the "y" command with "-jg" option to remove/set password for an already existing JAR archive. Please save your password in a safe place. There is no trick, trapdoor, or shortcut to recovering a lost password. NOTE: The first JAR 1.01 BETA used the now obsolete "Authentication" method to protect an archive. Please read WHATSNEW.DOC for details. h: selects (second) alternate set of switch characters. Works like the "-j" shift option but selects another set of switch characters. See the "-j" switch for more details. j: selects alternate set of switch characters. This switch toggles the set of switch characters. The toggle is reset at end at of each separate switch sequence back to the main set of switch characters. For example, "-jf" is not the same function as "-f". However, "-jjf" is the same as "-f" because of the double toggle. Also, "-jljf" is NOT the same as "-jl -jf". The switch sequence "-jlf" is the same as "-jl -jf". It is not recommended that you use these combined shift sequences. l: create List_name file This switch will set JAR to dump to the filename specified after the "-l" switch all of the filenames to be processed by this JAR command. For add type commands, this list contains all files that matched the file wildnames given on the command line. For "sl" command, this list contains files which JAR failed to open for reading. For other commands, the file will contain only the original filenames and/or wildcard names. Other options do NOT affect the output of filenames to this list file. This list file can be used as a listfile on the command line. Example: JAR16 a -lname.lst archive *.exe This example will create a file named "name.lst" with all scanned *.exe files. m: with Method 1, 2, 3, 4 Method 1 = fastest compression. This method uses a small fixed memory amount for compression in comparison with -m2..4 . Method 2 = normal compression (default). JAR uses available memory to improve compression. Method 3 = better compression. This method works slower but may improve compression a few percent. Method 4 = best compression. This method may compress better with large amounts of data (>2 Mbytes). "-m4" is not recommended for archives that may be modified in future. JAR requires up to 12 Mbytes of additional disk space for a swap file when updating these archives. In addition, updates are much slower compared to the other methods. Examples: JAR16 a archive *.exe -m3 Better compression JAR16 a archive *.exe -m4 Uses the best compression Recommended values of free memory for decompression depend on the method used: "-m1" - does not require additional memory "-m2" and "-m3" - 1.5 Mbytes "-m4" - 3.0 Mbytes These values are RECOMMENDED for optimal operation; however, JAR will decompress files with less memory. Usually methods "-m3" and "-m4" work 1.5 times slower than default "-m2". What is difference between these methods? JAR uses a greatly improved LZ family algorithm. "-m1" matches strings at short distances (comparable with PKZIP - about 32Kbytes). "-m2" searches for strings at large distances (about 500K-1000Kbytes and sometimes more) depending on the memory amount used. "-m3" method works like "-m2" but does a more exhaustive slower search. "-m4" matches strings at even larger distances and like "-m3" uses an exhaustive search. The second difference between methods is data block size. For "-m1", "-m2", "-m3", JAR concatenates files into 1000K-1500Kbytes pieces to obtain good compression and reasonable update speed. This also makes a JAR archive more robust than a "FULLY" solid archive (a damaged byte affects at most a single block, not the whole archive). The only drawback is a slight lose in compression rate. "-m4" compresses better by concatenating files into significantly larger blocks - up to 12Mbytes. Therefore, we recommend the use of "-m2" for archives which may be updated frequently and "-m4" for the "best and solid" compression. n: only New files (not exist) With the "a" command, add files that do not exist in the target archive. With the "d" command, delete files that do not exist in the target directory. With the list type command, list files that do not exist. Target directory may be specified using "-o" option. o: Output directory/file For commands which scan files this option specifies the destination directory within the archive for scanned files. Example: JAR16 a archive dir1\a.txt -oversion1 -jf- File will be added to archive as "version1\dir1\a.txt" For extraction commands this option specifies the destination directory where files are to be extracted. Example: JAR16 e archive dir1\a.txt -oc:\version1 -jf- File will be extracted to "c:\version1\a.txt" For the "y" command, this option specifies the updated archive name. Currently, the "y" command together with "-o" option is the only way to update a multi-volume JAR archive. Examples: JAR16 y archive -oa:archive Copies archive to floppy JAR16 y archive -oa:archive -va Copies archive to multiple floppies and automatically splits to volumes See the "y" command description for more details. For the "d", "l", "v", "ls" commands with the "-n" (new files) option, the "-o" option specifies the directory to scan. q: Query on each file This switch causes JAR to prompt the user prior to acting upon each archived file for all but the "l", "lc", "ls", "lt", "rh", "t", "v" and "y" commands. This allows you to selectively extract, delete, add, etc. r: Recurse subdirectories This switch will set JAR to recurse any wildcards specified on the command line by traversing all subdirectories scanning for matches. JAR will also recurse non-wildcard filenames as in: JAR16 a archive FILE.BBS -r By default this option is ON for all commands that matches files within archive ("c", "d", "e", "da", "l", "ls", "v", "x"). By default this option is toggled OFF for all commands that scan files ("a", "ac", "f", "m", "u", "test"). Examples: JAR16 l archive -r- *.doc List *.doc files only in root archive directory. JAR16 d archive *.doc Delete ALL *.doc files from archive JAR16 d archive test\*.doc -r- Delete *.doc files from test\ directory, but not subdirectories s: set archive time-Stamp This switch causes JAR to set the date-time stamp of the archive to that of the newest file in the archive. This option works only with update operations. The "-s1" switch is used to keep the old archive date-time stamp. The new archive will have the same date-time stamp as the original archive. u: Update files This switch works only with the "e" and "x" commands. "-u" specifies JAR to extract only newer and non-existing files from the archive. "-u1" specifies to extract only older and non-existing files. "-u2" specifies to extract non-existing files and files that date/time or size changed. NOTE: JAR ignores date/time comparison between directories (always assumed are the same). Example: JAR16 x archive -osource -u2 Extracts new or changed files v: enable multiple Volumes While creating a NEW archive this option specifies the new archive volume size. This option cannot be used while updating an existing archive. The only exception is the "y" command with the "-o" option. As such, the "y" command allows one to split an archive, to change the existing archive volume size, or to concatenate all the volumes into one volume. The "-v" option is unnecessary for non-update operations since the JAR archive header contains information about each volume created and its size. JAR will report an error when "-v" is used together with unsupported command. Thus while extracting a single file JAR will only prompt for the required volume(s). To list the archive contents JAR usually requires only the last volume (".j" default extension). In rare cases, a few last volumes are required as the archive header may not fit the last volume. JAR volumes are suffixed in following order: .j01, .j02, ... .j99, .100, .101 ... .990. The LAST VOLUME by default is suffixed .j and it must be used in the command line for future access to the archive. JAR supports up to 990 volumes. When creating new volumes on a removable drive JAR pauses and checks that the diskette has been really changed. However, this check is limited to 3 retries for the same volume. This test can be inhibited with the "i" modifier. JAR does not do this check when accessing volumes for reading or testing. Example: JAR16 a a:archive -vi -va The "-v" switch will accept the abbreviations 360, 720, 1200, and 1440. These will be translated to 362,000, 730,000, 1,213,000, and 1,457,000 bytes, respectively. Please note that if the available disk space is less than the requested amount, JAR will prompt the user for permission to continue. You can use the "K" modifier as a shortcut for "000". For example, 100K means 100000. "K" does NOT mean 1024. The "-va" switch sets the disk space auto-detect option. JAR will check for the disk space available on the target drive and try to use all or most of it. This option is aimed at diskette usage. Please note that this switch option detects free disk space. It does NOT detect formatted disk space. If the space available is less than 10,000 bytes, JAR will prompt the user for permission to continue. Examples: JAR16 a a:backup -va JAR16 a backup -v360 The "v" modifier turns on the next volume beep option. When you select this option, JAR will sound a beep prior to the next volume. The "r" modifier allows the reservation of disk space on the first volume. The number after the "r" specifies the space to reserve. Example: JAR16 a install -v360r50K Make first volume 50000 bytes smaller than 360K. The switch modifier "s" can be used to make JAR execute one specified system command prior creation of new volume or make JAR pause for manual execution of system commands. This is useful for purging target diskettes before JAR writes to them. The switch modifier "z" is identical to the "s" modifier except that the system command is not echoed to the screen. Optionally, after the "s" modifier, you can specify a system command or batch filename. JAR will automatically execute the command or batch file before creating each new volume. If the command has embedded blanks, then the switch option must be surrounded by double quotes. The system command is executed before JAR executes the auto-detect space option. Examples: JAR16 a A:backup -vas JAR16 a A:backup -vvas JAR16 a A:backup -v360s JAR16 a A:backup -vv360s JAR16 a A:backup -vaspurge.bat JAR16 a A:backup -v360sdelete.bat JAR16 a A:backup -vas"FORMAT A:" JAR16 a A:backup -vas"DIR C:\" The "-v" switch accepts its modifiers in any order. The "s" and "z" modifiers must be last because any succeeding text is considered the system command to execute. You can suppress the next volume prompt with the "-y" or the "jyv" option. However, when using the "-y" option with multiple volume JAR archives on removable media drives, JAR will ALWAYS PROMPT for the next diskette. The "-jyv" option WILL ALWAYS suppress the prompt even for removable media. During JAR operation in non-update commands, JAR will not abort when it cannot open the correct sequential JAR volume archive. It will report the open error and proceed to the prompt. At this point it is possible to swap diskettes to put the correct volume in place. This feature is disabled if the "next volume" prompt has been disabled. w: assign Work directory This switch option is HIGHLY recommended if you are archiving directly to diskettes. By default, JAR builds a new JAR archive file in the same directory as the old archive file. By specifying the "-w" switch, you can specify the working directory where the temporary archive file will be built. After the temporary archive file is built, it is copied over the original archive and deleted. Normally JAR requires enough disk space for the original archive and the new temporary archive. Specifying the "-w" switch allows you to move some of that disk space requirement to another directory. When using the "-w" switch while archiving to multiple volumes, JAR will check the available disk space in the work directory. If there is not enough space, JAR will prompt for permission to continue. This prompt can be suppressed with the "-y" and "-jyk" switches. If the copy of the temporary archive on top of the original archive aborts, you must do the copy manually. JAR will not delete the temporary archive in this error situation. Example: JAR16 a -we:\temp\ archive *.c For speed reasons, you should always use the "-w" option when creating or modifying archives on diskettes. You should specify a work directory on your hard disk or a sufficiently sized RAMDISK. Example: JAR16 a -we:\ b:archive *.c x: Exclude filenames This switch is used to exclude filenames or wildnames from the list of filenames to be processed. Example: JAR16 a archive soft\* -r -x*.exe -x*.obj -xtest This example will archive all files in the "soft" directory and subdirectories except for any files named "test" or ending in ".exe" and ".obj". You can also specify an exclude file list by preceding the filename with the list character "!". The exclude file list must contain a list of filenames/wildnames specified one per line. Example: JAR16 a archive soft\* -r -x!exclude.lst You can specify a pathname to exclude as in: JAR16 a archive \dir\*.* -r -x\dir\*.* In the above example, JAR will exclude all files in \dir but not the subdirectories in \dir. The "-x" option also allows you to exclude an entire path including subdirectories from being selected for processing. Syntax: JAR16 a archive C:\*.* -xC:\TEMP\ The "\" after C:\TEMP indicates to JAR that everything below C:\TEMP\ is to be excluded. Note that the exclude directory option requires that the directory be specified in the same relative path that the selected files use. Example: JAR16 a archive \temp\*.* -x\temp\firstdir is correct usage JAR16 a archive \temp\*.* -xC:\temp\firstdir is NOT correct because of the C: The "-x" switch also works for non-update commands like "e" and "l". Examples: JAR16 e archive -x*.exe Extract all files except *.exe files JAR16 l archive -x*.exe -x*.obj List all files except *.exe and *.obj JAR16 e archive \ -xdir\*.c Extract all files except *.c in directory "dir" The "-x" switch with "lc" and "dc" allows the exclusion of chapter ranges: JAR16 lc archive -x1-50 List all chapters except first 50. y: assume Yes on all queries Use this switch for batch type uses of JAR. This switch disables most of the normal user queries during JAR execution. Use this switch to suppress overwrite queries in the "e" and "x" commands, to suppress the make new directory query in the "e" and "x" commands, to suppress the new filename prompt when an open file operation fails in the "e" and "x" commands and to suppress the next volume pause using the "-v" option to create volumes on the hard disk. The "-y" switch will NOT suppress the "-v" next volume prompt when the archive is located on a drive with removable media such as a diskette drive. Use this option with due caution, especially during extraction as this sets JAR to overwrite files. NOTE: JAR does not overwrite readonly files unless the "-ha" option is used. As an alternative, you can use the "-jy" option that lets you suppress individual types of user queries. z: supply file for archive comment The "-z" switch has been provided to facilitate batch commenting of JAR archives with the "c" command. When the "-z" option has been specified with or without a comment file, JAR will only process the archive header comment. A comment file containing only a single blank can be used to erase comments from JAR archives. Syntax: JAR16 c archive -z JAR16 c archive -zcomment.txt JAR16 a archive -zcomment.txt Another method to erase the comment is to specify the predefined "NUL" filename: JAR16 c all -zNUL The "-z" option also allows one to dump the JAR archive header comment to a file. This will extract only the main archive comment and not the individual file comments. Syntax: JAR16 e archive "" -zcomment.txt dumps only the comment JAR16 x archive -zcomment.txt dumps comment and files jb: select chapter or chapter range Syntax: -jb10 Selects chapter 10 -jb10- Selects all chapters from 10 and above -jb-10 Selects chapters from 1 to 10 -jb10-20 Selects chapters from 10 to 20 -jb1- Selects all chapters -jb"-" Also selects all chapters For listing commands ("v", "l", "ls"), this option specifies the file chapter range to list. Examples: JAR16 l archive -jb10 Lists chapter 10 JAR16 l archive -jb1- Lists all files in all chapters JAR16 l archive Lists LAST chapter For the comment command ("c"), this option specifies the range of chapters to comment or files in the chapter range to comment. Examples: JAR16 c archive -hz -jb1-10 Comment a few chapters JAR16 c archive -jb1-10 Comments all files in chapters from 1 to 10 JAR16 c archive -hz"Ansi comment".txt Comments last chapter from archive Commenting will fail if the affected chapter is locked. For extract commands ("e", "x"), this option specifies the chapter to extract. Examples: JAR16 x archive -jb5 -oc:\dest Extracts whole chapter 5 into "c:\dest" JAR16 x archive *.c Extracts selected files from LAST chapter JAR can also extract one chapter and subsequent changes to it in selected chapter range. Example: JAR16 x archive -jb5-10 text.txt Will extract the full chapter 5 and ONLY the changed "text.txt" versions in chapters from 6 to 10. JAR uses the corresponding chapter number in destination path in the form "(5)","(6)"... See also the "e" command for more details. If selected operation affects few chapters at once then JAR shows each file lifespan range in the format "<from-to>". The switch "-jb" can not be used with other commands. With update commands "a", "f", "m", "u", "d" you can access only the LAST chapter. You can delete from a few chapters at once using "da" or "dc" commands. jf: store Full matched path JAR uses two methods to store the scanned file path within an archive. If the "-jf" option is ON then the full scanned file path is stored. Only the drive letter and root slash is removed. If the "-jf" option is OFF then the RELATIVE scanned file path with respect to the wildcard is stored. Examples: JAR16 a all c:\dir1\dir2\a.txt -jf file will be stored within archive as "dir1\dir2\a.txt" JAR16 a all c:\dir1\dir2\a.txt -jf- file will be stored within archive as "a.txt" JAR does not support storing an archive path which contains ".\" and "..\". Only the path back to the ".\" or "..\" is stored (-jf is ON). Examples: JAR16 a all c:\dir1\..\dir2\a.txt -jf file will be stored within archive as "dir2\a.txt" JAR16 a all c:\dir1\..\dir2\a.txt -jf- file will be stored within archive as "a.txt" This option affects only the commands that scan files ("a", "ac", "u", "f", "m"). The default value for the "-jf" option may be changed within the configuration file. See the "FullPath" keyword for additional information. jg: specify new password This option with the "y" command can be used to remove or set or change the archive password. JAR16 y archive -goldpassword -jgnewpassword Changes archive password JAR16 y archive -gpassword -jg Removes archive password JAR16 y archive -jgpassword -hg1 Sets archive password (GOST encryption method) With "-jg?" JAR will prompt for new password. Example: JAR32 y archive -jg? -v20k -o"Split and garbled" splits and garbles archive into new "Split and garbled.j" archive NOTE: After the password is changed JAR always tests the temporary archive before overwriting the original archive. jl: Lock archive This option sets a global archive flag for update operations. An archive with this flag set does not allow any archive modifications. There is no way to clear this "Lock" flag. You can only decompress or list files from a locked archive. Example: JAR16 y archive -jl Lock existing archive jl1: Lock existing chapters This option locks all existing chapters. Any update operation that attempts to modify locked chapters, files in locked chapters, or their comments will fail. There is no way to "unlock" the chapter. You can only decompress and list this chapter. Unlike the "-jl" option, new chapters may be added to the archive and they will not be locked. Example: JAR16 ac archive c:\* -jl1 Add new chapter and lock all chapters jl2: make auto-Lockable archive This option sets the global archive flag "auto-lock". Now after each "ac" command, JAR immediately locks the newly created chapter. So you can only list files, extract files or add new chapters. No other modifications are allowed. Example: JAR16 y archive -jl2 Makes existing archive "auto-lockable" There is no way to clear "auto-lock" flag. jr: Recover broken archive files This option is available only in the Win32 JAR version. This switch is used to access files in an archive corrupted either with bad data or missing data. JAR will continue processing despite these errors. NOTE: JAR does not scan the whole archive during extraction as it knows the exact location of each compressed file. So JAR can "skip around" the damaged files if you exclude them from the requested files set. This switch also will allow the extraction of files with CRC or Bad file data errors. Normally, JAR will delete these extracted files. Example: JAR32 e archive -jr With this option JAR will NOT check that the supplied volume is correct. Substituting another volume is one possible workaround in the case of a lost volume. js: Select chapter creation mode Each file in archive has a chapter lifespan range. When a file is backed up using the "ac" command and has not changed since the last chapter in archive, only the lifespan range is changed. A copy of the file is NOT added to the archive since a copy already exists in the archive. If the backed up file is different from the previous version, then that file is added to the archive with an appropriate chapter range. The "-js" option selects method used to detect changed (different) files. If "-js0" is set, then JAR considers a file different if its size or modification date/time or creation date/time do not equal the corresponding values of the file stored within the archive in the last chapter. If "-js1" is specified JAR works like "-js0" but IGNORES a changed creation time. So the file may have different creation date/time but still will be considered as the same as the file in archive. This option is useful when working with DOS tools. DOS programs do not know about file creation time and so do not store/restore it properly (e.g., DOS archivers). JAR for DOS assumes the scanned file creation time is equal to the file modification time. So the "-js1" option is also useful when adding new chapters into the same archive with both JAR for Win32 and JAR for DOS. The "-js1" option is default. With "-js0" or "-js1" JAR does NOT consider a file different if the file attributes (read only/system/hidden/archive) are the only file properties changed. With "-js0" or "-js1" JAR NEVER adds copy of a directory. Instead, JAR modifies the directory's chapter lifespan range. You can also specify the "-js-" option. Then JAR will assume ALL files and directories are different in comparison with the last archive chapter. This causes JAR to work slower as ALL files must be compressed again. NOTE: JAR uses a special compression method that compresses only the difference between different file versions in subsequent chapters. This produces a major compression improvement for a multiple-chapter archive. Currently JAR does not use this compression method if the file size is more than about 500 Kbytes. The "-js" options may be used only with the "ac" command. During this command JAR reports all unchanged files as "marked". This message may be turned off using a configuration file setting. See the "HideMarkedFiles" keyword. See the "ac" command for other details. jt: Test temporary archive The "-jt" switch can be used to set JAR to execute an archive integrity check on an intermediate temporary archive before overwriting any pre-existing original archive. If any error occurs, JAR will not overwrite the original archive. With the "-w" option JAR also will verify that the temporary archive is successfully copied. Example: JAR16 m archive *.c -jt ju0: sort files by name during listing This option sorts files with the "l", "ls" and "v" commands. This option is set by default. To prevent JAR from sorting, use the "-ju-" option. This causes JAR to work a bit faster. Example: JAR16 ls archive -ju0 ju1: sort files by extension during listing This is like "-ju0" but it first sorts files by extension, then by name. Example: JAR16 ls archive -ju1 jx: hide archive contents Together with any update command this option sets the "HIDDEN" archive flag. When this flag is ON, JAR asks for the archive access password for ALL operations. By default (HIDDEN flag is OFF) JAR asks for the archive password only when it is necessary. This means that the password is NOT asked for when using list commands and some "y" and "c" commands. The "-jx" option specifies JAR to ask for the password before any operation with the archive. The option is ignored for archives that are not protected/encrypted. NOTE: This option only enables an internal JAR check, it does NOT encrypt the JAR archive header. If you need a more secure method for hiding archive header information then you need to compress and encrypt this archive into another archive. NOTE: Encryption is not available in export JAR version. Example: JAR16 y archive -jx If "archive.j" is protected then this command "hides" archive contents JAR16 y archive -gpassword -jx- Now archive contents can be listed without specifying password jy: suppress queries assuming Yes This switch is used to suppress individual JAR user prompts. Syntax: JAR16 e archive -jy<flags> where <flags> is one or more of the following characters: c, d, k, n, o, t, v. For each flag selected, JAR will act as if you had answered YES to the suppressed question. C - skip create directory query D - skip delete files query K - skip disk space availability query N - skip new filename prompt O - skip overwriting existing file query T - skip overwriting directory attributes query V - skip ok to proceed to next volume query Example: JAR16 x volume -v -jyco Suppress create, and overwrite prompts. JAR will allow you to specify individual "jy" options in separate switch strings as in "-jyo -jyc". You may inhibit a switch as in "-jyc-". jz: supply file for file comment The "-jz" switch has been provided to facilitate batch commenting of one or more files within an JAR archive. Only one comment file can be specified. That file will be used to comment each file specified on the command line. A comment file containing only a single blank can be used to erase comments from JAR archives. Another method to erase the comment is to specify the predefined filename "NUL". Example: JAR16 c archive -jzNUL ha: ignore readonly Attribute The "-ha" switch sets JAR to reset the attributes of a file to be overwritten or a file to be deleted. This allows the overwrite or delete operation to proceed. Examples: JAR16 e archive -ha JAR16 m archive -ha dir\*.* hb: select files by attributes Option requires parameter in the form: [<bits which must be equal 0>'0'][<bits which must be equal 1>'1'] Bits encoded as: a - archive d - directory h - hidden r - readonly s - system To emulate the directory handling of archivers such as ARJ, you should use the "-hbd0" option. Normally, JAR will scan and add all directories that are encountered as files are searched via the wildcards. With a command like "JAR a test c:\*.doc -r", JAR by will search the C drive for all .DOC files and also add ALL directories found on the C drive unless the option "-hbd0" is specified. Examples: JAR16 a archive c:\* -r -hba1 -b0 Add all files with archive bit set and then reset this bit JAR16 a archive *.* -r -hbd0 Do not add directories (allows a bit smaller archives) JAR16 e archive -hbsh0 Extract all files except "system" or "hidden" JAR16 a archive *.* -hbd0r1 Add readonly files (but NOT directories) he: skip test of security Envelope The "-he" switch sets JAR to skip the verification step when processing a JAR-SECURED archive. If a JAR archive has a damaged JAR-SECURITY envelope, you will have to use the "-he" option to extract the archive. hf: match long and short Filenames This option affects only the Win32 JAR version and is only usable with commands that scan files. By default JAR matches wildcards using only long filenames. This option will match both short and long scanned filenames. This dual matching behavior is the default for the DIR command in Windows 95. Suppose there is a file "New document.doc" with the short name "NEWDOC~1.DOC". Then: JAR32 a archive NEWDOC~1.doc Will report "No such files" JAR32 a archive -hf NEWDOC~1.doc Will add file "New document.doc" to archive as its short name is matched by wildcard Example: JAR32 a archive *~* -x*2* -hf Add all files in current directory with a short or long filename that contains "~" but not "2" hf1: use only short Filenames This option affects only the Win32 JAR version and is only usable with commands that scan files. This switch disables long filenames usage. The Win32 JAR version will work like the DOS version during file scanning. Example: JAR32 a archive *~* -x*2* -hf1 Add all files in current directory with a short filename that contains "~" but not "2" hf2: convert long Filenames into short This option sets JAR to do long filename translation during extract and list commands. It is useful when extracting files in DOS that have been previously compressed by the Win32 JAR version. After reading the archive header JAR converts all long filenames into a shortened form. The process is similar to that done by Windows 95 to generate short filenames. During this conversion JAR substitutes characters that are not supported in DOS filenames using underscore "_". Then the filename is trimmed to 8.3 form. If there is another file with the same shortname, "~1","~2",... is appended until the generated shortname is unique. JAR guarantee that files from different directories will remain in different directories after name conversion. Suppose archive contains 3 files: "long name\Document1.doc" "long name2\Document2.doc" "long name2\Document3.doc" After conversion filenames will be: "LONG_NAM\DOCUME~1.DOC" "LONG_N~2\DOCUME~2.DOC" "LONG_N~2\DOCUME~3.DOC" or so. The exact generated short name also depends on the operating system in which the file is stored. This is because the Win32 JAR version always stores the SHORT filename (if available) along with the LONG filename as a HINT for future filename translations. This HINT takes precedence over the JAR filename trimming algorithm. Before extraction you can review conversion results using the "v" command and the "-hf2" option which will show both the original long filename with the shortened name. The "l" and "ls" commands with "-hf2" will show only the shortened filenames. Afterwards, you can extract the files using: JAR16 x archive -hf2 The "-hf2" option does not work with update commands. hf3: convert long Filenames into short (second method) This switch works exactly like "-hf2" but also converts those filenames that contain "international" characters with hex code greater than or equal to 0x80. These characters are substituted with the underscore "_". This option helps to solve problems when extracting filenames encoded using a different OEM code page. hg: access protection method This option accepts argument - method name. Currently, JAR can protect the information in an archive from unauthorized access by using either of two methods: -hg"Authenticate" or -hg"GOST v1.01" The first method establishes archive password WITHOUT actual encryption of its contents. The second method uses GOST 28147-89 encryption in CBC mode to encrypt archive files (but NOT archive header which contains comments and filename information). The "Authenticate" method sets JAR to make a password check before processing the archive. If the password is incorrect then JAR aborts. Thus the "Authenticate" method is very fast but the data in the archive is NOT encrypted. The "method" also can be chosen using shorthand notation "-hg0" for "Authenticate" and "-hg1" for "GOST v1.01" encryption. NOTE: Encryption is not available in export JAR version. Any time you are creating a new archive and using the "-g" option (specify password) you can select the protection method using the "-hg" option. Otherwise, JAR will use the "DefaultAccessProtection" setting in the CFG file (see description for more details). The default CFG file specifies to use "Authenticate" if no "-hg" option is specified. You can also specify the method when protecting an unprotected archive with the "y" command: JAR16 y archive -jgpassword -hg1 If you need to change the protection method then you must first remove the archive password: JAR16 y archive -gpassword -jg and then protect the archive as in the first example. Because these two protection methods give significantly different security levels, JAR prints different messages when accessing protected archives. JAR prints that an archive is "PROTECTED" when using the "Authenticate" method and "ENCRYPTED" when using the "GOST" method. To access protected archives, you need to specify the password for most operations. The "-jx" option establishes a password check for ALL operations. See "-g", "-jg", "-jx" options for more details. NOTE: The first JAR 1.01 BETA used the now obsolete "Authentication" method to protect an archive. Please read WHATSNEW.DOC for details. hk: add recovery records This switch sets the global archive flag that specifies adding of data recovery records to the archive after each update. These records allow the repair of a slightly damaged archive. See the "-hx" option on how to use previously stored recovery records. This option increases the size of a medium size archive about 2 percent with a minimum of a 2 Kbyte increase. Large files (more than 3 Mbytes) will expand about 0.5 percent in size. This additional space is used to store tables which will allow the repair of archive damage. With a very high probability JAR is able to recover about 28 Kbytes of damaged archive section(s) on a 1440K volume. If your archive is damaged and recovery records are present then JAR will report an error unless the "-hx" option is specified. This option lets JAR recover damaged archive sections "on the fly". The global archive "add recovery records" flag may be turned off on any update using the option "-hk-". Example: JAR16 y archive -hk Makes archive with "recovery records" JAR16 y archive -hk- Makes archive without "recovery records"" hl: warn of non-matching wildcards With "-hl" switch JAR reports non-matching wildcards/filespecs during an "add" type command ("a", "f", "m", "u", "ac"). A wildcard/filespec is considered "non-matching" if it does NOT match any files (NOT including directories). JAR will subsequently return a non-zero DOS errorlevel when a non-matching wildcard or filespec has been specified. Example: JAR16 a a:backup !listfile -va -hl -hbd0 hm: On or after date YYMMDDHHMMSS The switch "-hm" by itself will select files modified today. If "-hm" is followed by a date and optionally a time, JAR will select files modified on or after that date-time. Example: JAR16 a test -hm9601021700 means select files modified on or after Jan 2, 1996, 5:00 PM. Years less than "80" will be considered as 21st century years. There is no option for using other date-time formats. The switch "-hmb" selects files modified before today. If "-hmb" is followed by a date and optionally a time, JAR will select files modified before that date-time. JAR accepts the use of "-hm" and "-hmb" options simultaneously to specify a range of selected dates. This is useful for selecting files on a specific date. Example: JAR16 a archive *.* -hm910101 -hmb910102 The "-hm" option and/or the "-hmb" option will default to today's date at midnight (0000 hrs) when specified without dates. The "-hmd" switch sets JAR to select files no older than the specified number of days. The days are calculated from midnight. Examples: JAR16 a archive -hmd0 select files modified today JAR16 a archive -hmd1 select files modified yesterday and today "-hmd" and "-hmb" can be combined too. ho: Output directory translation For add type operations the "-ho" option allows one to specify the file destination directory within an archive depending on the file path on disk. This option accepts an argument in the form: -ho<archive path>:<disk path> During add type operations, all files with the specified <disk path> will be preceded with the corresponding <archive path> when storing within an archive. JAR allows one to specify a set of translations. In such a case, the longest match is chosen. The option "-ho-" disables all translations. Examples: JAR16 a archive c:\* d:\* -hodrive\c:c:\ -hodrive\d:d:\ files from disk "c:" will be stored into "drive\c" and files from disk "d:" will be stored into "drive\d" JAR16 a archive src\* src\lib\* -holib:src\lib -hofiles:src stores files from "src\lib" into "lib\" and other files from "src\" into "files\" NOTE: The destination directory for a file also depends on the "-o" and "-jf" options. JAR compares the <disk path> letter by letter, so you must specify path in the same form. During extraction JAR does a backward translation (replaces specified archive directory with destination directory). Example: JAR16 x archive -hodrive\c:c:\ -hodrive\d:d:\ files from "drive\c" will be extracted to "c:\" and from "drive\d" into "d:\". All other files extracted into current directory. JAR does not support the storing of two files with the same path within an archive. Also, JAR does not support storing drive letters within an archive. The "-ho" option provides a workaround for backing up and restoring data from/to different drives or network servers using a single archive. Example: JAR16 a archive \\server\user\* c:\* -hod:\\server\user -hoc:c: The "-ho" option without arguments selects the built-in default translation. The default option translates all drives "a:\", "b:\"... into archive directories "_a_\", "_b_\".... ht: Test new/updated files This switch tests all newly compressed data. The "-ht0" option tests data using the file's CRC32 value. The "-ht1" option tests data by COMPARING the actual file contents. Example: JAR16 a archive c:\* -r -ht1 Verifies all added files using comparison With the "-w" option JAR will verify that the temporary archive is successfully copied. When used with volume archive this option may cause JAR to re-read previous volumes. You can use "-jt" option to test the whole updated archive. See this option for details. hx: use recovery records Sets JAR to use previously stored records to recover slight archive damage. It may be used with any command. It is recommended that this option be used with an existing archive containing "recovery records" on update operations. JAR does not detect archive damage unless the "-hx" option is specified or the JAR decompression engine fails on damaged data. Updating a damaged archive with "recovery records" without the "-hx" option may result in JAR computing new recovery records using corrupted data. Examples: JAR16 e archive -hx Extract data from archive using recovery records JAR16 y archive -hx Repair damaged archive using recovery records hy: add security envelope JAR adds a security envelope to a JAR archive. This option is only available in the registered JAR version. The JAR-SECURITY ENVELOPE feature provides a facility similar to other archivers. This feature disallows any type of modification, even commenting, to a JAR-SECURE sealed archive by JAR. Moreover, there are additional internal checks to determine if the JAR-SECURE sealed archive has been modified in any way. This feature is intended for use by software developers who distribute their software in archived format or for businesses that distribute data files. However, there can be no guarantee that this mechanism cannot be defeated. When processing a JAR-SECURED archive, JAR will display a message indicating that JAR is verifying the security envelope. If the envelope is damaged, a message will be displayed indicating so. If the envelope is intact, then the JAR-SECURITY signature of the user "sealing" the archive will be displayed. hz: comment chapters This switch is used to comment a chapter during an update operation. It also may be used to specify a comment file. Examples: JAR16 ac archive c:\* -hz Add new chapter and prompt for the comment for it JAR16 c archive -hzcomment.txt -jb1-10 Comment chapters from 1 to 10 using file comment.txt JAR16 c archive -hznul -jb"-" Clears all chapter comments CONFIGURATION FILE: JAR uses a configuration file to store global user settings. By default the configuration file is named "JAR.CFG" and must be placed into the same directory as the JAR16.EXE and JAR32.EXE directory. The environment variable JAR_CFG_FILE is used to specify an alternate JAR configuration file location. Example: SET JAR_CFG_FILE=c:\user\jar.cfg The JAR configuration file format is very similar to Windows *.INI files. It contains strings of three types: 1. Comments. This string must start with ";" character 2. Keys with set values in the form: <Key name>=<value> Key name is case insensitive. 3. Section separators in the form: '['<Section name>']' Section name is case insensitive. Example: [OS/2 settings] ColoredOutput=No ; black and white output SwapFileDirectory=h:\tmp The given example contains one line with the section name, two lines with keys, and one line with a comment. On startup JAR reads predefined key values from the configuration file and uses these values as global settings. Sections are used to keep setting variants in a single file. You can set the JAR_CFG_SECTION environment variable to a section name which will take precedence over default values. Example: SwapFileDirectory=c:\tmp [OS/2 settings] SwapFileDirectory=d:\tmp [Windows NT settings] SwapFileDirectory=e:\tmp If JAR_CFG_SECTION has the value "OS/2 settings" then the JAR swap file will be located on drive D. If JAR_CFG_SECTION has the value "Windows NT settings" then the JAR swap file will be located on drive E. Otherwise, it will be located on drive C. If JAR does not find a key in the configuration file then it uses built-in default values. CONFIGURATION FILE KEYWORDS: The right side of "=" describes default key value (used if it is not set within configuration file). ColoredOutput=No Possible values: Yes/No. Specifies the kind of JAR screen output. SwapFileDirectory= JAR chooses the swap file valid directory in the following order: - from CFG file but not equal to "" - from TMP environment variable (DOS JAR version) or temporary directory (Win32 JAR version) - current directory JAR swap file usage depends on the total files in archive and on command type. JAR may need up to a 2 Mbytes swap file when updating an existing medium-size archive (with about 1,000 files). Example: SwapFileDirectory=c:\tmp FullPath=Yes Possible values: Yes/No. Specifies the "-jf" switch default for JAR. Users who want ARJ-like file path storing should set this key to "Yes". Users who want UC2-like file path storing should set this key to "No". DefaultAccessProtection=Authenticate Possible values: "Authenticate" or "GOST v1.01". Specifies default archive protection method. See "-hg" option for details. XmsLimit=1024 EmsLimit=0 These settings LIMIT MAXIMUM XMS and EMS usage (in kilobytes) by the program. These settings are used only by the DOS JAR version. You can set corresponding values to 0 so JAR will not use XMS/EMS at all. It is strongly recommended to let JAR use at least 1024 kilobytes of XMS for better performance. Recommended - 3072. JAR does not necessarily use all the specified memory; instead, it chooses an optimal value based upon need. The maximum practical value is 5120. JAR uses EMS for better compression. JAR conventional memory consumption also decreases by about 60 kilobytes when EMS is available. Optimal value - 4096. Maximum amount used by JAR - 16384. When more EMS is available, a better compression rate can be achieved. IMPORTANT NOTE: See the CheckAvailableMemory keyword for memory usage details. CheckAvailableMemory=Yes Possible values: Yes/No. If this key value is set to "Yes" then JAR empirically tests available memory before the compression engine starts and sets internal memory consumption limits that should not result in a decrease in overall system performance. Multitasking operating systems like OS/2 or Windows 95 report available virtual memory instead of real "physical" memory. A large allocation of virtual memory amount may drastically slow down the OS or even hang it due to a lack of resources. JAR uses memory access timings to allocate only "safe" memory amounts. If you are an experienced user you may toggle this feature off. Then JAR will not test for available memory, but will use as much as it decides but no more than the set current limits (see XmsLimit, EmsLimit, 32MemoryLimit). JAR also starts faster when the "No" value is set. This setting is ignored by list and extraction JAR commands. 32MemoryLimit=0 Limits memory usage by the COMPRESSION ENGINE. This setting is used only by the Win32 JAR version. Recommended value 4096. Maximum which may be used - 10240 kilobytes. IMPORTANT NOTE: See the CheckAvailableMemory keyword for memory usage details. DOS InteractiveCriticalErrorHandler=Yes Possible values: Yes/No. The interactive handler asks "Abort, Retry?" on such hardware errors like "Drive not ready". For batch processing you can disable these prompts. This setting is used only by the DOS JAR version. ErrorsToStdout=No If "Yes" then reports errors to stdout instead of stderr. So you redirect error messages to a file using ">" or ">>" as well as all other JAR output. EmulateANSIComments=Yes Possible values: Yes/No. This option is used to disable the JAR ANSI comments emulation. OptionsForCommand A= OptionsForCommand AC= ... This group of keys is used to specify SEPARATE default options for ANY JAR command. The specified options are parsed before the command line is parsed. So you always can overwrite this setting. Examples: OptionsForCommand L=-ju1 Now "l" command by default will sort files by extension OptionsForCommand V=-ju- -r- By default "v" command will not sort and recurse files HideMarkedFiles=Yes Possible values: Yes/No. This setting hides the "Marking file" message for "ac" command. DefaultColor=7 ErrorColor=12 SignalColor=15 CommentColor=3 HighlightColor=2 Possible values 0-15. It is used to customize colored JAR output. The following table must be used for color translation: BLACK (0) GREY (8) BLUE (1) LIGHT BLUE (9) GREEN (2) LIGHT GREEN (10) CYAN (3) LIGHT CYAN (11) RED (4) LIGHT RED (12) MAGENTA (5) LIGHT MAGENTA (13) BROWN (6) YELLOW (14) WHITE (7) LIGHT WHITE (15) HighIntensityColors=No Possible values: Yes/No. It uses only "light" colors for JAR output. This may be useful for windowed DOS sessions. IMPORTANT DIFFERENCES BETWEEN JAR AND ARJ: Most JAR commands and switches follow ARJ naming. This DOES NOT mean that all JAR options and commands work exactly as in ARJ. Please read the documentation for details. These are a list of useful options that work slightly differently from ARJ: -b0 -b1 -r -jb -jf -hf -hg These are a list of options that have different meanings: -o -jg -jl -ju -js -jx -hm -ho -ht -hx -hz JAR DOES NOT support the ARJ base/target directory option. JAR provides the "-o" option instead. JAR always adds files with "system" and "hidden" attributes. ARJ by default does not. By default during an add command, JAR adds all directories scanned while searching for files that match the specified wildcards. An archive may be created with directory names only even if no files were found. For example, "JAR16 a archive c:\abcdefgh.* -r" would create an archive containing every directory name on drive C and probably no files. Use the "-hbd0" option to not include the directories. By default for commands that scan an archive, JAR wildcarding uses directory recursion. This means that the specified directory and all sub-directories will be scanned by the command. To get ARJ-like scanning behavior, you need to specify the "-r-" option to turn off recursion. Example: JAR16 x archive dir\docs\*.* -r- Unlike ARJ, JAR cannot join archives. Instead, JAR has the option of using the "y" command to join volumes into a single "solid" archive. It is also possible to change archive volume size. JAR LIMITATIONS: JAR uses a "central directory" scheme to store information about the compressed files in an archive. Since such a directory may be very large, JAR uses built-in virtual memory (VM) which allows JAR to work in a limited amount of physical memory (then swap file is used). VM is also used to store other intermediate data: list of files to add, temporary compression buffers, etc. During archive UPDATE operations VM size is limited to 25 Mbytes. If JAR exceeds this then the update operation fails. As a result, the maximum number of files that can be stored within an archive is about 80,000 files (depends on files path length). An archive compressed with the "-m4" method can be updated if it contains less than about 40,000 files. DOS JAR version memory usage: EMS available No EMS, XMS only No EMS and XMS ------------- ---------------- -------------- Update 431K(*) 484K 517K commands Extract 368K 434K 467K commands List 351K 351K 384K commands (*) - 484K for "-m1" method "K" means kilobyte (1024 bytes) Processing an ENCRYPTED archive requires an additional 8 Kbytes of memory. JAR ARCHIVE ACCESS MODES: Below is a description of the archive file opening modes used by JAR. This information may be important when sharing archives on networks. JAR does the following steps: 1. Opens archive file (last volume), reads archive header and then closes archive file. Sometimes archive header does not fit in the last volume and is distributed within the last few volumes. 2. Using the archive header information, JAR decides which archive volume is required for future processing. 3. Opens necessary volume, reads data, opens another volume and so on... So even in case of a single volume, JAR opens an archive twice. First to read the header and second to read the actual compressed data. On update type operations both DOS and Win32 JAR open an archive using "DENY READ and WRITE" mode (exclusive mode). On non-update type operations DOS JAR opens an archive using "COMPATIBILITY" mode and Win32 JAR - using "DENY WRITE" mode. As a result you can simultaneously list or extract from archives on network. JAR ARCHIVE FLAGS: JAR shows the archive flags and archive comment any time the archive is accessed. Here is a list of possible flags with a short explanation: ENCRYPTED Archive is protected using encryption. PROTECTED Archive is protected using "Authenticate" method. PROTECTED(obsolete method) Archive is protected using "Authentication" method. This obsolete method was used by the first JAR 1.01 BETA. Please read WHATSNEW.DOC for details. HIDDEN Archive created using "-jx" option (requires password for any command). SECURED JAR_SECURED archive created using "-hy" option. RECOVERY RECORDS Archive contains recovery records ("-hk" option). LOCKED Archive is locked using "-jl" option. Archive cannot be updated. AUTO-LOCK Archive is auto-lockable ("-jl2" option). JAR ERROR LEVELS: 0 -> success 1 -> warning 2 -> fatal error 3 -> CRC error 5 -> disk full or write error 6 -> cannot open archive 7 -> simple user error (bad parameters) 8 -> not enough memory 9 -> not a JAR archive 10 -> internal error 11 -> Ctrl+C pressed FINAL COMMENTS: We do hope that you find this program as useful as we have. We would appreciate any suggestions to improve this archiver. You can reach us at: Internet address: robjung@world.std.com CompuServe userid: 72077,445 JAR users are encouraged to use the above addresses to submit questions, problems, and new features requests. ***** UPDATED MAILING ADDRESS ***** Robert and Susan Jung ARJ Software P.O. Box 249 Norwood MA 02062 USA Fax number: 617-769-4893 *********************************** end of document