High Voltage Shareware

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

/ High Voltage Shareware / high1.zip / high1 / DIR4 / ZDIFF14.ZIP / ZDIFF.DOC < prev next >

Wrap

Text File | 1993-12-24 | 25KB | 568 lines

ZDIFF v1.4 The Archive/Directory comparison Shareware program Copyright 1993, Christophe Dubourg Date : 24 December 1993 ---*--- Table of contents 1 ........................Purpose 2 ........................ Syntax 3 ...................... Features 4 .................. How it works 5 .................. Requirements 6 ........................ Limits 7 ....................... Remarks 8 .................... Disclaimer 9 ............ Distribution files 10 ............Registration/Author ---*--- 1. Purpose : ------------ ZDiff is a comparison utility program. You can compare a pair of archives, or a pair of directories, or an archive and a directory. (that's the new of this !) New in 1.4, you can also compare to a previously saved list. The result is a sorted list, showing the status of each file. ■ ZDIFF.EXE is the normal DOS version. ■ ZDIFFD.EXE is the DPMI hosted version. It permits a larger amount of files to be treated (up to 16380) and works with Windows or any other DPMI host. 2. Syntax : ----------- ZDIFF <New Archive/Directory/List> <Old Archive/Directory/List> or ZDIFF <New Arc/Dir/List> <Old Arc/Dir/List> [options] [wildcards] Note : To simplify the following, the archive or directory that is compared to an other archive or directory will be called a Set of files. There is a NEW set of files, and an OLD set of files. (to determine the one that is Newer and the one that is Older, for instance). The <New archive/directory> and <Old archive/directory> parameters are needed (not optional). The order in which you specify the New/Old pair is important. It will make the difference between a 'Deleted' file or 'Inserted' file, or a 'Newer' and 'Older', or 'Bigger' and 'Smaller'. (globally, it will determine the output result) To specify that a set of files is a LIST (not an archive or a directory), just add an '@' in front of the LIST filename. Like this : ZDIFF @MyList.LST MyNewDir /R This way, I compare MYLIST.LST (which is a list), and MYNEWDIR (which is a directory) The NEW set of files must be specified first. --------------------------------------------- Then, you may specify wildcards if you need. a. Wildcards The wildcards may be used to restrict the files compared inside the directory or archive. (for both the new and old sets, because it would be unlogical to have two different wildcards). The DOS standard wildcards are allowed. Note : wildcards are not used in LISTs. b. Options The options may be specified anywhere on the command line, and are not case sensitive. The options are : /b : Batch output. Only the file names are sent to standard output. You may re-use the list as input with an other program. (to update an archive with new files, or delete files that are already in an archive or on a backup disk, for example) /c : Country flag override. If the dates in the ZIP files are not well recognized, you may use this option to force the date format to use while reading ZIP files : /C0 - USA date format. (MM/DD/YY) /C1 - EUROPEAN date format. (DD/MM/YY) /C2 - JAPANESE date format. (YY/MM/DD) /d : Differences only. Once the two sets of files are compared, the output will only show the files that are different between the 2 sets. (ANY difference : date, size, crc, name...) By default, all the files of the two sets are displayed. /h : Help ! This display more help screens, and an information screen. /l : Generate a list. With this option turned on, you don't compare two sets of files, you just generate a list from ONE set of files, that can later be used to compare with... an other listing, or an other set of files. This allows you to gain a lot of time if you want to compare one set of files with several sets of files. (ZDiff, when using a list, don't have to scan a disk for names, sizes, dates, and CRC so it's incredibly FAST !) When you generate a list, ZDiff try to obtain the maximum informations from the source files. If the source is a directory, both 16-bit and 32-bit CRC are calculated and displayed in the output list. This allows you to later either compare with a 16-bit CRC archiver output, or a 32-bit CRC one. If the source is an archive, only the CRC used by the unarchiver output can be obtained and sent to the list output. As usual, the display is sent to standard output, so to save the list, just re-direct it, like this : ZDIFF EXAMPLES /L >EXAMPLES.LST or ZDIFF MYSAMPLES.ARJ /L >SAMPLSET.1 or ZDIFF /L DUMM.ZOO >DIR1.LST The first example outputs a list from the EXAMPLES directory, calculating both 16 and 32-bit CRC, to EXAMPLES.LST. The second example outputs a list of the files contained in MYSAMPLES.ARJ - the CRC is 32-bit, since ARJ use 32-bit CRC. The third example outputs a list from DUMM.ZOO, which use the 16-bit CRC. /m : Matching files only. The output will only show the files that are BOTH in the NEW and the OLD set of files. Files that are only in the NEW set or only in the OLD set (i.e., Deleted or Inserted files) will not be shown. Useful if you want to compare a set which is only a subset of the other. /n : No Path in names With this option, you can compare a full directory tree with a single directory. Useful to compare a ZIP file that have a directory tree stored inside of it, and a directory that contains all the files without subdirectories. Example : to compare to directories, one containing the 'cut' product, and one containing the 'development' version, like : New directory : BUILD\*.* Old directory : CUT\DISK1\*.* CUT\DISK2\*.* CUT\DISK3\*.* etc... You can compare these 2 directories with the following command : ZDIFF /r /n BUILD CUT /p : Pause. With this option, ZDiff will prompt you before reading each set of file. The purpose of this is essentially while using floppies, or removeable media. For example, if you need to compare two sets that are on two different floppies that have the same format, you'll have to insert one disk, read the files from it, and then the second disk. Maybe, you could compare 2 removable hard disks, or a hard disk and an archive, or whatever you want. /q : Quick scan. If a directory is involved in comparison, this permits to skip the CRC calculation on each file of this directory, in order to speed up the process. Of course, use this option with caution, since the comparison is then only made with TimeStamp and Size ! (two files can have the same date and time of last modification, the same size, and be binary different : using this option, they will be marked as 'Same'). However, if the timestamp or size is different, 'Differs' will be displayed. /r : Recurse directories. By default, the subdirectories are not searched, either in the archive or the directory file structure. (to save time). This switch allows the comparison of the full tree of directories. Note : if you don't specify the -r option, the filenames that contains a path in the archives are not treated at all. A warning is displayed if a path is found in an archive, if you have not used the -r option. /v : Verbose output. This option causes a status to be output at the end of comparison. This status report shows the number of files that are : - Identical - Different - Newer - Older - Bigger - Smaller - Inserted - Deleted It may give an idea of the amount of differences between the two sets, without having to analyze the whole output. /z : Zpecial messages. Troubleshooting mode option - display more information while running, like Country code detected, date used, Zip version... It may help if something seems to go wrong (with date format detection, for example). Please note that using the DPMI version of ZDiff with the Borland RTM DPMI server, the Country Code number MAY be inaccurate. However this doesn't affect the DATE FORMAT detection. (See the /C option to override the Date format detection) 3. Features : ------------- Since this is a comparison program, you will probably want to know how it compares the two sets of files : The keys of comparison are : Name : If two names of the sets don't match, either one is new or one is deleted, depending on the order you specified for the comparison. if the -r option is specified, the full filename must match. Once two files are found in both the 2 sets, a more precise comparison occurs. The keys for comparison between two files that have the same name are : CRC : The most accurate CRC method is used to know if files are absolutely identical or not. CRC is either taken from the archive, or calculated from the files in the case of a directory. 32-bit CRC is the default. Note : if one of the sets is an archive of type LZH, LHA, ARC or ZOO, then a 16-bit calculation is done on files, to match the CRC that these archives use. -> If you compare one archive which use 16-bit CRC and an other which use 32-bit CRC, then they can't match : the Quick mode is then automatically set to ON. (CRC are not compared). In this case, the only way to compare these archives with a CRC is to extract one of them to a directory, and then compare this directory to the unextracted archive. If the Quick option is set, and at least one of the 2 sets is a directory, no CRC calculation occurs, and files are assumed binary identical. Date : To know which one is the oldest file, or the newest. (Time is used, too) Size : To know which one is the biggest file, or the smallest. 4. How it works --------------- To compare two directories, or a directory and a saved list, or two lists, you won't need nothing but ZDIFF.EXE. But to compare two archives, or an archive and a directory, you'll need the correct unarchiver in your PATH, to access the archives you have specified. The program supports all the current archive formats that are commonly used and that I am aware of : ZIP, ARJ, LHA, LZH, PAK, ZOO, ARC. The needed unarchiver, for each extension is : *.ZIP PKUNZIP.EXE (v1.10 or v2.04+) *.ARJ ARJ.EXE (v2.30) *.LHA \ *.LZH - LHA.EXE (v2.12) *.PAK / *.ZOO ZOO.EXE (v2.1) *.ARC PKUNPAK.EXE (v3.61) (each of these can be easily found on many BBS or Compuserve) Now, here is how ZDiff works : The program first checks if the archive or directory or list that you specified exists. (to prevent an error after some precious seconds of calculation). At this point, it decides if a 32-bit or 16-bit calculation will occur (or no crc, if Quick mode is set). Then, it analyzes the NEW set, and then the OLD set. Three choices, concerning each set of file : a. If it's an archive: The unarchiver is called, to obtain a list of each file in the archive, along with the date, crc, size, and so on. This list is saved in a temporary file. (Using the TEMP environment variable if available) Then, the temporary file is opened, and scanned for each file in it. Since all the informations are already calculated by the unarchiver, there's nothing to compute (the crc is already there). Each file entry is then compared or stored to a list in memory. b. If it's a directory: The directory is scanned, and all the files are opened, to obtain their date, time, size, and a CRC calculation occurs. (except if the Quick mode is set). Each file entry is then treated like archive files entries. c. If it's a list: The list is opened, and scanned like an archiver output. The only difference is that since BOTH the 16 and 32 bits CRC can be presents, the CRC choosen is the one that correspond the best to the other set of file to compare to. (the most accurate is always taken) Each file entry is then treated like archives file entries. After all the files have been read in memory (the NEW set and the OLD set of files), the sorted list is sent to standard output. Since you can redirect standard output, you could use this facility to pipe the output to a finder program (FIND, GREP), or a viewer program (VIEW, LIST), or any filter program of your choice. Here are some examples : 1. ZDIFF NEWARC.ZIP OLDARC.ARJ | more This outputs the result of the comparison via the DOS's MORE filter command. (if the archive contains sub-dirs, they are not included, since no -r parameter is there - instead, you'll obtain a warning while running ZDiff). 2. ZDIFF /r NEWARC.ZIP OLDDIR *.DOC | FIND "Newer" | LIST /s This displays, via the LIST full-screen viewer, all the NEW files in NEWARC.ZIP, compared to the OLDDIR directory files. Subdirectories are scanned, too. Only the .DOC files are processed. 3. ZDIFF NEWZIP.ZIP . -r -q -c1 This outputs on screen the differences between the archive NEWZIP.ZIP and the current directory. The subdirectories are scanned, too, (-r) both in the archive and the directory. A Quick scan (-q) of directories files will be made (no CRC), so the comparison will be based on date, time and size only. The date format used by PKUNZIP on this machine (-c1) is the EUROPEAN date format (DD/MM/YY). 4. ZDIFF -v . -r OLDARC.LHA /d /q -m This output on screen the differences between the current directory and the archive OLDARC.LHA. Subdirs (-r) are scanned also. A Verbose (-v) status is displayed at end of comparison. Only the differences are shown (/d), and no CRC is calculated (/q). Only the matching files (-m) of the two sets are displayed. 5. ZDIFF -L \BIN >MASTER.LST This outputs in MASTER.LST the list of files in the \BIN directory, along with their date, time, 16-bit and 32-bit CRC, and size. For later comparison purpose. 6. ZDIFF LATEST.ZIP @MASTER.LST -d This outputs the differences between the latest product zipped in LATEST.ZIP, and the master list generated by the previous example. 5. Requirements : ---------------- ZDiff is compiled for 80286 and above systems. DOS version : a minimum of 200kb is required. DPMI version : 2MB or more memory would be good. You need a DPMI host, of course. (Windows, for instance). 6. Limits : ----------- ■ ZDIFF.EXE : All the files informations are stored in DOS memory. (no EMS) It allows between about 5500 to 9900 files to be stored, depending on the path lengths and your available system memory - this is much enough for everyday use, I guess. ■ ZDIFFD.EXE : All the files informations are stored in the GLOBAL memory. (DPMI) So, you can virtually store as many files that you have memory free, up to 16380 files at one time in memory. You need Windows or any other DPMI server, like Borland's RTM. Note : Borland DPMI server can be found on Compuserve : GO BORFORUMS, then go to 'Borland Pascal forum'. Then, choose the 'From Borland' library section, and download the file called NEWRTM.ZIP. ■ Note concerning the archives : if the -r option is not specified, the files containing a path in their name are not included in the comparison. Only the files in the main directory of the archive are treated, and only the files in the specified directory are treated also. However, a warning is displayed if a directory path is found in an archive, and if you haven't specified the -r option. ■ Note concerning the LISTS generated by ZDiff : you can't use wildcards if a list is used in a comparison. -> But you can use wildcards to generate a list. 7. Remarks ---------- I made my possible to make this program free of bugs, but if ever something is wrong with ZDIFF, tell me. I'll fix it ASAP. ZDiff is written in Pascal and Assembly, compiled using Borland Pascal 7.01, (for DOS and DPMI targets). Please notice that ZDiff supports international information from DOS, since PKUNZIP does. (it concerns the date format, used for date comparison, that is different in US and in France, for instance). The DOS Country information is read for this purpose. The 3 date formats that exists are USA, European and Japanese. If ever the comparison of dates seems to be wrong in ZDiff, you can override (with the /C option) the automatic detection to set a date format to use on your machine. See the 'Syntax' section. Note : if you use the /Z option to display technical information, the country information reported by ZDIFFD.EXE (the DPMI hosted version) may be inaccurate, if you use the Borland DPMI RTM ; The date format, however, will be correctly detected and used. The correct country number will be correctly obtained, running ZDiffD under Windows DPMI. Note : The country number is for information only. Don't worry if it's wrong. Just check that the date format displayed match your PKUNZIP outputs. Since ZDiff writes a few temporary data on your hard disk (in your temp directory), you should not reboot while ZDiff is running, as this could result in lost sectors. (CHKDSK /F would corrects it anyway) Just use Ctrl-Break if you want to abort the program. 8. Disclaimer ------------- Even if there is really no possibility for this program to damage your data, here is the usual disclaimer message : As a shareware program, ZDIFF is distributed as-is. The author disclaims all warranties, expressed or implied. The author will assume no liability for damages either from the direct use of this product or as a consequence of it's use. PKUNZIP, PKUNPAK, ZIP, ARJ, ZOO, PAK, LZH, LHARC, LHA are copyrighted by their respective authors. Borland Pascal is Copyright Borland International. 9. Distribution files --------------------- Here is the file distribution list : Length Date Time CRC-32 Name Description ------ -------- ----- -------- ------------ ------------------- 3449 24-12-93 01:40 04f427ef WHATSNEW.DOC News in 1.4 version 58084 24-12-93 01:40 6e9abc7a ZDIFF.EXE The std DOS program 69080 24-12-93 01:40 74c281e0 ZDIFFD.EXE The DPMI program 23688 24-12-93 01:40 ZDIFF.DOC You read it! 1738 24-12-93 01:40 a5de6fa1 REGISTER.DOC Registration form 348 24-12-93 01:40 db44371a FILE_ID.DIZ Short description ------ -------- ----- -------- ------------ ------------------- 10. Registration / author ------------------------- If you like this program, and plan to use it, you may try it during a 21 days period, after what you should register. I develop many tools for my own use, (this one helps me a lot!) and it would be very pleasant to me, to know that these tools are useful for someone else, too ! If you register, it will be a proof that my tools are useful : in this case, I'll spend more time to update them, and adding features - according to your requests or suggestions. The registration fee is $10 for a single user. For multi-user registration, please look at the REGISTER.DOC form included in the ZDiff package. I made this registration form for your convenience ; you can print it, fill-in the blank fields, and send it to me along with your check or bank note(s). ** Thanks in advance. ** ------------------------------------------------- HERE IS MY ADDRESS Christophe Dubourg FOR (paper) MAIL, 23, rue Victor Hugo CHECKS, or 78370 PLAISIR CASH : FRANCE ------------------------------------------------- --> Even if you don't send money, please feel free to send me your remarks, suggestions, and specific problems encountered with this program. It will be a pleasure to read your mail and to answer it ! Don't hesitate to contact me, at the following addresses : --------------------------------------------------------------- Compuserve : 73233,1412 (Christophe Dubourg) or 73233,1412@compuserve.com Forum : IBMSYS MCI Mail : CDUBOURG (453-6013) or cdubourg@mcimail.com Internet : cdubourg@borland.com At home (paper Mail) : Christophe Dubourg 23, rue Victor Hugo 78370 PLAISIR FRANCE --------------------------------------------------------------- Thanks for using ZDiff, I hope it helps. //ChD