Source Code 1994 March

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

/ Source Code 1994 March / Source_Code_CD-ROM_Walnut_Creek_March_1994.iso / compsrcs / misc / volume33 / mbase / part01 < prev next >

Wrap

Text File | 1992-11-23 | 54.2 KB | 1,323 lines

Newsgroups: comp.sources.misc From: richid@owlnet.rice.edu (Richard Parvin Jernigan) Subject: v33i119: mbase - MetalBase 5.0, Portable database engine, Part01/08 Message-ID: <csm-v33i119=mbase.165613@sparky.IMD.Sterling.COM> X-Md4-Signature: 4e160b28912fbbc8564f40f7508addad Date: Mon, 23 Nov 1992 22:57:13 GMT Approved: kent@sparky.imd.sterling.com Submitted-by: richid@owlnet.rice.edu (Richard Parvin Jernigan) Posting-number: Volume 33, Issue 119 Archive-name: mbase/part01 Environment: AMIGA, MS-DOS, HP-UX, XENIX, UNIX, ULTRIX, SGI, SU, Curses Supersedes: mbase: Volume 28, Issue 40-44 [ This version of mbase has been released as donateware. Donate if you wish ] [ but you are not required to. See the file readme.too for more specifics. ] [ For archiving purposes, I do not consider this shareware. -Kent+ ] MetalBase 5.0 is a simple database engine, designed to be portable between almost any platform; Unix, MS-DOS, HP, Mac PW, NeXT, sgi, and many more, small- model and up. This latest release has more field types, three-phase locking to ensure multi-user access without any system-based IPC, and internal caching to provide twice the speed and up for rebalancing indices. A conversion utility for MetalBase 4.0 and 4.1a relations is provided. Other features include run-time encryption on all platforms, creation of relations on-the-fly and a flexible report writer. Richid ------------------------------------------------------------------------------- #! /bin/sh # This is a shell archive. Remove anything before this line, then feed it # into a shell via "sh file" or similar. To overwrite existing files, # type "sh file -c". # Contents: advert dox dox/mbase.dox sample src src/util1.c # Wrapped by kent@sparky on Mon Nov 23 16:33:11 1992 PATH=/bin:/usr/bin:/usr/ucb:/usr/local/bin:/usr/lbin ; export PATH echo If this archive is complete, you will see the following message: echo ' "shar: End of archive 1 (of 8)."' if test -f 'advert' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'advert'\" else echo shar: Extracting \"'advert'\" $1397 characters$ sed "s/^X//" >'advert' <<'END_OF_FILE' XAdvertisement :) MetalBase 5.0 X------------------------------------------------------------------------------- X XI'm looking for either of two things: X X - An internet newsfeed (ANY feed, stable or not) local to central Iowa, X around Ames/Des Moines preferably. X X - A job located somewhere in Iowa, again, preferably near Des Moines. I X know, I know--"Iowa???" That's what everyone says. But my SO is starting X school at Grinnell this fall, and I wanna be around for it. X X Microsoft's being nice to me, but my internship will be over around X January of 1993; and 'cause of regulations, they can't hire an intern X again until X months after the last internship ended. So I've gotta find X somebody else instead... X X If your company is in need of a very experienced Unix/MSDOS/VMS X C/C++/Pascal/FORTRAN/680x0/80x86/6502 programmer, let me know and I'll X float you a resume. I'm unsure of my internet connection for a while, so X contact me through: X Richard P Jernigan IV X PO Box 827 X Norris, TN 37828 X 615-494-0445 X Okay. So what are the odds I'll get a response to that? :) X XPostScript: I actually DID get a response! Didn't take 'em up 'cause X I'm doin that summer thang at Microsoft. But as I said, that's X almost over. X END_OF_FILE if test 1397 -ne `wc -c <'advert'`; then echo shar: \"'advert'\" unpacked with wrong size! fi # end of 'advert' fi if test ! -d 'dox' ; then echo shar: Creating directory \"'dox'\" mkdir 'dox' fi if test -f 'dox/mbase.dox' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'dox/mbase.dox'\" else echo shar: Extracting \"'dox/mbase.dox'\" $33032 characters$ sed "s/^X//" >'dox/mbase.dox' <<'END_OF_FILE' X /\ X / \ X../ \....METALBASE Version 5.0.............................................. X / \ X/ \ Written From October 1990 by Huan-Ti X\ / X \ / Revision 3.1 released December 10th 1990 X \ / Revision 3.2 released December 10th 1991 X \ / Revision 4.0 released Febuary 14th 1992 X \/ Revision 4.1 released April 1st 1992 X Revision 4.1a released May 10th 1992 X Revision 5.0 released October 1st 1992 X X X X X X X "And Calvin ate. And Food Lion went out of business. Amen." X X --1st Calvin 4:8 X X XWHY_METALBASE_?________________________________________________________________ X X The Amiga 500 multi-tasks. Wow. It has great graphics. Yeah. Phenomenal Xsound. Yippee. It costs $500. I bought one. X A DBase package costs more than the computer. So does Paradox. And XInformix. And almost every other database package on the market. I didn't Xbuy one. X But I needed a database for my system--anybody who's ever worked with a Xdatabase knows they're addictive; code that's written without a flexible Xdatabase backing it up lacks something. And DBase wants $600 to put that Xsomething back into programs. No thanks. X That's why MetalBase was written... its forerunners (versions 1.0 and X2.0, not available) were designed by the same author with the Amiga line's Xneeds specifically in mind: relatively slow floppy drives (Compared to hard Xdrives), so sparse read-head maneuvering was essential. Relatively small Xmemory, when compared with the amounts of memory required by most C compilers X(Lattice needs roughly 300k just to run, for example), so the code had to be Xconsise, pure, and use few variables. Expansion was always available; hard Xdrives, huge memory capacities, extra processors, ram drives -- so the system Xhad to be readily expandible and flexible enough to meet the requirements of Xboth a 512k, 1-drive machine and a 32-megabyte, 200-meg hard drive-driven Xsetup. And they were. X The problem was, Amigas can multi-task. So what is the rest of the Xcomputer doing while a database runs around looking up records on one drive? XNothing. Why not let it look up records on another drive? Fine... that works. XWhat about on the same drive... using, say, a ram disk, so there are no real Xread-heads to worry about positioning? That would mean each relation would Xhave to support multi-user access... a requirement around which MetalBase Xversions 3.0 and up focus. Not only can any given terminal access many files Xat once, but on multi-terminal systems, each terminal can work simultaneously Xon any given file. Terminals can each use as many relations as a system will Xallow, and each relation can be used by some sixty-four users at once. The Xcode used to create MetalBase 3.X and up is completely compatible with the XUnix operating system and all of its derivatives, so its inherent multi-user Xability places it in firm working footing on the most advanced systems on the Xmarket. X But the IBM users complained. And the unix users complained. And lo, Xeven the Amiga people complained--the code didn't work all the time, and Xwas generally crude. So MetalBase 4.0 solved these troubles and more--a Xcomplete rewrite of the engine has provided even faster and more efficient Xalgorithms, which keep MetalBase indices in AVL balance at all times. More Xconvenient functions and operations (such as build's ability to create header Xfiles for your code which reflect a schema's structure) have been added, Xalong with several utility programs for maintaining your database. X Relations are designed by a user to match his/her exact requirements, Xindices placed as needed on single or composite fields, and records can be Xadded, retrieved, changed, and deleted as any C programmer so desires. The Xinterface is as simple as possible, very fast in execution, and provides a wide Xrange of error types for exact specification of problems. A user on any Xterminal can choose to lock out other users from any number of relations, so Xsweeping changes in data will not be read by others before they can take Xplace. Automatic temporary locks are placed on a relation which is about Xto undergo a deletion or addition, strategically delaying other users' access Xto files for less than seconds, so that no data will be read until all indices Xhave been updated correctly. This process is entirely system-driven; users Xnever need worry about it at all. If one user is adding a record to a Xrelation, any terminal requesting a record lookup will simply wait until the Xaddition is complete before proceeding. As records are deleted and added Xthrough normal use, the algorithms used to update indices automatically keep Xthem at top operating speeds, so any given relation's number of records can Xgrow exponentially while causing only a linear increase in look-up time. Any Xrelation can contain over four billion records (Nearly four billion more than Xmost systems will ever need), and each record's length can be as large as its Xsystem's memory can handle. In perspective, MetalBase is infinitely Xexpandible, while requiring very little of the system on which it operates. X X[ The actual number of records allowable is 4,294,967,296... if you were to X add one record a second, it would take over 136 years to reach this limit ] X XRELATION_DESIGN________________________________________________________________ X X A sample relation. Data Field # -> 1 2 3 4 5 X is read by selecting one Record # _____ ___ ____ ___ ___________ X record from the others by 1 |_____|___|____|___|___________| X various criteria, and 2 |_____|___|____|___|___________| X returning it to the user. 3 |Smith| 12| 114| A2|Houston, TX| X Record 3 (In this example-- 4 |-----|---|----|---|-----------| X actual MetalBase records need X no such numbering system), if returned, would be placed in a structure X containing Smith,12,114,A2, and Houston,TX in the appropriate elements. X X A relation is a collection of records, each record consisting of the same Xnumber of fields--blocks of data whose individual lengths remain constant from Xrecord to record. Each field need not have the same length as the others in Xany given record, but from record to record, a field's length will remain the Xsame. The data the fields contain is, of course, the main purpose of a Xdatabase--when thousands of records exist, a system without a database would Xneed vast amounts of time to find any given record. Databases decrease Xdramatically record retrieval time, while requiring as little additional time Xto add new and change old records as possible. X The key to a database is its indices--pointers arranged in such a fashion Xas to show the system where to look for the record for which it is searching. XThe more of these collections of pointers, or indices, a database has, the Xgreater the number of different fields a system can search through in order to Xfind a record meeting certain criteria for a field (Or fields... a single index Xcan be used to first sort records by one field, and also sort duplicates with Xrespect to a second field, etc. MetalBase allows these composite indices to Xkey off of as many as 999 different fields). However, databases require time Xto update each index every time you add, delete, or change a field which has an Xindex placed upon it. So although retrieval criteria possibilites are expanded Xby using many indices, the amount of time required to change the existing Xrecords in a database will also increase. For this reason, the number of Xindices a relation will have should be considered carefully--use what is Xneeded, but no more. X X Relations are built from schema files, which decribe the relation in a Xmore human-friendly syntax. See build.dox for a more complete description of Xthis process. X XIMPLEMENTAION__________________________________________________________________ X X A sample program implementing MetalBase follows: X X #include <mbase.h> X #include "equipment.h" X X main () X { X relation *rel; X X if ((rel = mb_inc ("/usr/joe/equipment", strtokey("key"))) == RNULL) { X printf ("Can't open database - %s.\n", mb_error); X mb_exit (1); X } X X equipment_rec.price_code = 114.20; /* Set up initial values */ X strcpy (equipment_rec.part_numb, "117"); /* to search for... */ X X if (mb_sel (rel, 0, &equipment_rec, EQUAL, NULL) != 0) { X printf ("%s.\n", mb_error); X mb_exit (1); X } X X /* Now that we've found a record, change it to something else... */ X X equipment_rec.num_purch = 14; X strcpy (equipment_rec.customer, "Su"); X strcpy (equipment_rec.part_numb, "112"); X strcpy (equipment_rec.shop_addr, "Dallas, TX"); X equipment_rec.price_code = 12; X X mb_upd (rel, &equipment_rec); /* Update the current selection */ X X mb_rmv (rel); // Unnecessary before mb_exit(), but good practice. X mb_exit (0); // Always a good idea to use mb_exit() with MetalBase. X } X X Manipulation of relations using MetalBase is simple... by including the XMetalBase header file in your programs and compiling using the option "-lmb" X(Or "library ..., mbase.lib;", depending on your compiler), you add extra Xcommands to C. A summary of these follows: X X - mb_inc (filename, int) -- this command includes a relation on your desktop, X so that it becomes accessable. Any relation you wish to work with X must be included using this call. If mb_inc returns RNULL ( defined X as (relation *)0 ), an error has occurred (The most likely one is that X the relation cannot be found under the given directory--mb_errno will X be set to an error number, and mb_error (global char*) will contain X text regarding the problem). Note that "filename" does NOT include the X ".rel" suffix. BUILD, in analog, does not require that you include the X ".s" suffix on the name of the schema you are working with, but the X schema file itself must end in ".s", as relations must end in ".rel", X forms in ".frm" and reports in ".rpt". X X The integer passed in is the encryption key for the relation--see the X section on encryption for more information. Note that MetalBase 3.0, X as it does not support encryption, does not require this integer; all X versions 3.1 and up do. Sorry for the inconvenience. X X I also STRONGLY recommend the use of the function strtokey() to X obtain this integer from an ascii string... all utility programs X shipped with MetalBase ask for "Encryption Key"s, and they expect X a string to hash up with strtokey(). So if you use encryption but X not this method of getting the integer, you're SOL when trying to X use vr or report. In fact, I encourage this so strongly that, in X future releases, I may replace the integer argument with a string X argument and do away with strtokey() entirely. X X - mb_tst (filename) -- This function goes through all the steps needed to X make sure a given relation can be opened, but doesn't open it (thus X it doesn't need an encryption key). vr uses this one to make sure X it can open a relation before asking for a key; that's pretty much X its only use. X X - mb_sel (file, index, buffer, action, comparison_buffer) -- this command X searches a given relation for the record you desire, described by X "action" and "comparison_value". "file" is a pointer to the structure X returned by mb_inc... this structure is used in all other functions to X indicate which relation you intend to work with. "index" is the index X number by which you wish to search... numbered sequentially by the X order in which they were defined in the schema, starting with the X number 0. To save trouble when changing schema, it is recommended you X use the routine idxnum() to look up an index number by name. "buffer" X is merely a structure (like &equipment_rec above) into which MetalBase X can return the record, when found. A list of valid actions follows: X FIRST (or FRST) -- searches for the first record alphabetically. X LAST -- similarly, seeks the last record alphabetically. X NEXT -- searches for the next sequential record alphabetically. X PREVIOUS (PREV) -- as NEXT, searches for the previous record. X EQUAL (EQUL) -- seeks a record containing fields exactly those in X "comparison_buffer". If the index used allows duplicates and X some exist, EQUAL returns the first of these: all duplicates X can be found by subsequent "NEXT" searches. Note that the X comparison_buffer and the return buffer need not be the same X space, although they may. X GTEQ -- EQUAL returns a NOT_FOUND error if a record cannot be found X to precisely match the comparison. GTEQ acts as EQUAL in all X respects but this; GTEQ will find the record greater X alphabetically, or precisely equal to (if one exists), the X comparison. In the event that duplicates of the record exist X ("Duplicate" being defined logically as a record whose fields X used by the current index are precisely equal to another such X record's fields for the same index), GTEQ returns the first X of these: all duplicates can be found by subsequent "NEXT" X searches. X GTHAN (GTHN) -- Acts as GTEQ, except GTHAN will ignore precise X matches. Useful if you wish to begin searching names, say, X after the "Smith"'s. Another way to accomplish this would be X to use LTEQ + "Smith", followed by a NEXT... X LTEQ -- Searches for a record less alphabetically, or equal to (if X one exists), the comparison. In the event that duplicates of X the record about to be returned exist, LTEQ returns the last of X these: all duplicates can be found by subsequent "PREVIOUS" X searches. X LTHN (LTHAN) -- Similar to LTEQ in all but one regard: LTHAN will X not return a precise duplicate, instead searching to its X left for the record before it. X CURR (CURRENT) -- After a record is found by a search, MetalBase X remembers the location of the record. Searching with the X action CURRENT re-reads this most recent record, and returns X it in its entirety. As with FIRST, LAST, NEXT and PREVIOUS, X the value of "comparison_value" is irrelevant in CURRENT -- X however, CURRENT is also oblivious to index used (As no X actual searching takes place... MetalBase merely recalls a X known record). X Please note that the above-described action names merely stand for X integers, and MetalBase will not understand if you use quotes around X the names. These actions must also be written in upper-case. X "comparison_buffer" is a structure identical to "buffer", but any X data in any of the fields is used for comparison when searching for X EQUAL, GTEQ, LTEQ, GTHAN or LTHAN. There are three ways to use that-- X give it a new record... X sample_str sample_rec, target_rec; X target_rec.num = 15; X mb_sel (rel, idxnum("ix_num"), &sample_rec, EQUAL, &target_rec) X or, the same one twice... X sample_str sample_rec; X sample_rec.num = 15; X mb_sel (rel, idxnum("ix_num"), &sample_rec, EQUAL, &sample_rec) X or, what I do most, pass in NULL as the comparison (works just like X the one above): X sample_str sample_rec; X sample_rec.num = 15; X mb_sel (rel, idxnum("ix_num"), &sample_rec, EQUAL, NULL) X All three of these would have the same effect. X X - mb_upd (file, new_rec) -- After a record is found, it may be updated X through mb_upd. MetalBase simply replaces the old record's fields with X those indicated in "new_rec". This process usually requires more time X than any other, as the record must be disconnected from each index in X turn, changed, and reconnected. Occasionally, changing a record will X not change certain indices--when this occurs, the index is simply X skipped. An update, like an add, will be aborted if the changes would X violate a no-duplicates-allowed index. X X - mb_del (file) -- After a record is found, it may be removed completely by X use of mb_del. Once a record is deleted thus, it CANNOT be retrieved. X Pre-5.0 versions of MetalBase used a second argument for tracking the X current pointer; this code was broken and I removed it. :) So now X there's no second argument--again, sorry for changing things on ya. X X - mb_add (file, new_rec) -- This command adds a new record to a relation. If X adding the record described by "new_rec" would violate a non-duplicate X index, the addition to the relation will be aborted and an error X returned (See the section on Errors, below). X X - mb_lck (file) -- If, as an example, a program reads a record from a relation X used by many terminals simulaneously, waits for a user to enter a new X value for one of its fields, and updates the record, it could prove X disastrous if the record was read by one terminal, then another... X updated on one & re-written, and re-updated by the second terminal. To X prevent this type of scenario, mb_lck, when used, excludes other users X from reading or writing to a relation. Every operation they try X (including mb_inc()) will return MB_LOCKED until the process which X called mb_lck() calls mb_unl(). X X - mb_unl (file) -- Removes a lock placed on a relation by mv_lck. X X - mb_rmv (file) -- This command removes a relation from your desktop. If the X file has been locked previously (via mb_lck), the lock is removed. X X - mb_die () -- This command removes all open relations from your desktop, as X if you'd called each one in turn with mb_rmv(file). X X - mb_exit (ret_code) -- Calls mb_die(), followed by exit(ret_code). X X - mb_num (file) -- Returns the number of records in a relation. Wheeee... X X XERRORS_________________________________________________________________________ X XIn addition, see the separate documentation on troubleshooting, under the name Xtrouble.dox. There's not much, but I tried... X XBuild / Errors : X X Cannot open <file.s>............The schema indicated to Build cannot X be found. X File <file.s> holds no schema...The schema indicated cannot be used X definition as a definition of a relation. The X most likely cause is that the file is X empty. X Field <field> declared after....All fields must be declared before any X indices indices are declared. X Field <field> declared twice....Field names must be unique. Spaces are X not allowed in field names. X No fields declared before.......A relation must contain fields, and these X end reached must be declared before any indices. X Incorrect syntax................The definition of a field's length has X been entered incorrectly. X Field <field> undefined.........An index has been placed on a non- X existant field. Check spelling. X No fields declared before end...The schema definition is incomplete: it X reached must contain at least one field and X one index. X No indices declared before.....See the above error description. X end reached X Identifier <?> not recognized...The only valid words for the beginning X of a line are "relation" or a relation- X name (as the first line), "field", X "index", and (optionally) "end". X Cannot open relation............The file <file.rel> cannot be written... X possible causes are a media error or X write-protection on the file. X Relation is busy................Relations cannot be re-built until it is X not being worked with by any users. X The file about to be created....If you build a relation where a relation X already exists of the same name already exists, any X data in the existing relation will be X erased in favor of the new, empty X relation. X XMetalBase / Errors : X XMetalBase routines either return necessary information (such as, mb_inc() Xreturns a relation-pointer, and mb_num() the number of records in a relation), Xor 0 to indicate success. The global variable mb_errno will be set to MB_OKAY X(0) on success, and another error number (in the 1000's) on failure; the Xvariable mb_error is a character string which will always describe the most Xrecent error. Possible errors are: X X 0..MB_OKAY............This is the generic no-problem return code. All X functions except mb_inc should return 0 if X they have been able to function properly. X 1001..MB_NO_ROOM.........Too many relations are open at once -- mb_inc X cannot allocate enough space for the new X relation. This error can be prevented by X re-defining MAX_REL, as with build's warning X messages, to a value greater than its default X of 5. In addition, having many files open at X once may require that you warn <stdio.h> X (or config.sys, if you're using MSDOS)... X consult your system manuals. X 1002..MB_NO_MEMORY.......A malloc() has failed, and the task must be X aborted. This is a bad, bad thing to happen. X How d'ya fix it? Well, that's hard to say... X look around and make sure you're free()'ing any X memory you malloc(). X 1003..MB_NO_OPEN.........The file "file.rel", where "file" is the name X passed to mb_inc, cannot be opened. In all X probability, it does not exist. X 1004..MB_NO_READ.........The file <file.rel> cannot be read. This error X is returned only from mb_inc, and is probably X caused by read-protection placed on the file. X 1005..MB_FORMAT..........This is the generic this-shouldn't-have-happened X error. Check out trouble.dox for a quick X explanation of what could cause it. X 1006..MB_LOCKED..........This error is returned if another user has X locked the relation you're trying to access... X wait until they haven't, and you'll be fine. X 1007..MB_BUSY............Only a certain number of users is allowed to X access a relation at any given time. The X limits are: X MetalBase 3.1, 3.2.........120 X MetalBase 4.0...............63 X MetalBase 4.1, 4.1a, 5.0...255 X 1008..MB_BAD_REL.........This essentially means that the relation* you X passed to a function was not returned by X a call to mb_inc()... it's an invalid pointer. X 1009..MB_NO_WRITE........This error is produced when the relation you've X opened can't be written to... perhaps you don't X have permission. X 1010..MB_TIMEOUT.........A user has stopped a program during a record X addition, deletion, or update. These functions X place temporary locks on a file, which are X removed at the end of their respective X procedures. Normally other functions will X simply wait for the file to be unlocked, but if X the wait time becomes excessive, this error is X returned and the function aborted. Try the X function again, and if you recieve this error X a second time, see the section on Utilites to X remove the lock manually. X X Note that this error is no longer used, as of X revision 4.1. X 1011..MB_BAD_REC.........A null pointer to a record has been received. X So essentially, you've given NULL instead X of &rec. Stupid thing to do. :( X 1012..MB_CORRUPT.........This relation has an error on an index. See the X section on Utilites to repair it. Try to find X which index you were working with at the time X of the error, in order to speed repair. X 1013..MB_BAD_DUP.........The record that has been offered for addition/ X update would violate an index that has been X declared not to allow duplicates. X 1014..MB_NO_CURR.........In order to perform a delete or an update, X you must have already issued mb_sel() to X determine which record you wish to modify. X This error indicates you have not done so. X 1015..MB_BAD_IDX.........A bad index number has been passed to mb_sel. X Valid numbers range 0 through the number of X indices in a relation, minus 1. X 1016..MB_NO_SUCH.........Generic code returned when the record described X cannot be found in the relation searched. If X you feel it should have been found, check the X index number, action, and comparison value used X in the search. X 1017..MB_UNKNOWN.........A very unlikely error, receipt of this code X indicates you have passed a number greater than X 12 or less than 0 to mb_sel as an action. X 1018..MB_NO_FIELDS.......You're trying to create a new relation without X specifying any fields for it. Stupid you. X 1019..MB_NO_INDICES......You're trying to create a new relation without X specifying any inidces for it. X 1020..MB_BAD_INDEX.......An index you're trying to create has no fields. X Remember that the format for indices being X created is "nnn,nnn,nnn" --no whitespace X allowed without using BUILD. X 1021..MB_DISKFULL........This will only happen on mb_add() or when using X mbconv, the 4.0+ => 5.0 conversion utility, X and is returned whenever there isn't enough X free space on the drive to perform the X requested operation. X 1022..MB_BAD_SERIAL......This will only happen on mb_upd(), if you've X changed the serial number of the record X you're trying to update. You can't. It won't X let you. X 1023..MB_TMPDIR..........On some systems, MetalBase requires that you X have the environment variable "TMP" set to X an appropriate directory for temporary files; X for example, "TMP=/tmp; export TMP" or X "set TMP=C:\TMP". If you prefer, you can X use the variable "TEMP" instead. X 1024..MB_TMPERR..........MetalBase cannot work with the directory you've X defined as per error #1023 (see TROUBLE.DOX). X XUTILITIES______________________________________________________________________ X X Five utilities have been released with this package; one, called blast, Xremoves user- and system-placed locks on a relation and sets the number of Xusers the relation thinks are using it to zero. Impromper closing of a Xrelation (ie, not closing one) will cause the number of users to build up Xuntil it cannot be used any longer. X X The second utility included is called vr, for View Relation. Please see Xthe separate documentation under ../dox/vr.dox for its description and Xinstructions for its use. X X The third utility is called report... it generates reports of a relation's Xcontents. There is sparse documentation under ../dox/report.dox; it might Xhelp, it might not. Play with it and it'll work--honest. :) X X The fourth, and newest utility, is called mbconv--it converts pre-5.0 XMetalBase relations to 5.0 syntax. Note that the process is safe but Xirreversable; and your schema files may need semicolons added just to be Xcomplete. X X And, the final utility is called "form"--it compiles data-entry templates Xinto .h files your programs include for data-entry. There's practically Xno documentation; just the one example under ../sample. Good luck to you. X X Other supporting utilities to MetalBase have not been released with this Xpackage, primarily because I have not bothered to write them yet. See the XReadMe.Too file up one directory to see how to register yourself with me, as XI'll write any of the following upon request once you're registered (and Xmaybe before, if I have time). X X Future utilities: X X * Repair of bad indices and recover of unreclaimed space from X previous deletes X X * Utility to change a relation's schema without destroying its X contents (oh boy...) X X * Relation import/export from standard, Emailable text files. X X XENCRYPTION______________________________________________________________________ X X MetalBase 3.1 and up (not 3.0) automatically encrypt all files unless Xtold otherwise. When a relation is opened (using mb_inc()), an integer Xencryption-key is passed in to MetalBase. Any records which are added to that Xrelation are encrypted using this key, and any records which are retrieved are Xdecrypted similarly before they are returned. In this way, any information Xwhich is physically stored is encrypted beyond recognition. As long as this Xinteger key is the same every time the relation is opened, the encryption is Xtotally user-transparent: records are passed to mb_inc and mb_upd in normal Xtext, and are retrieved in normal text, but are stored garbled. If the key Xchanges, however, the records retreived in the second session will be decrypted Xunder the new key, whereas they were encrypted under the former... and the Xoutput will be meaningless, for numbers and strings alike. X X Note that the encryption is apparently working for ALL supported machines Xnow--however, this has always been the hardest part to port between machines. XIf necessary (or if you simply want to), compile mbase.c (and vr.c/sample.c/ Xwhatever) with -DNOENCRYPT, and it'll be turned off. X X For a more thorough explanation of MetalBase's encryption process, see the Xseparate documentation crypt.dox. X X-- X ___ X .-',' /\ X / (_______.------------------. / \ Huan-Ti X ( _______|))))))))))))))))))| \ / X )__( `------------------' \/ X `----' richid@owlnet.rice.edu X END_OF_FILE if test 33032 -ne `wc -c <'dox/mbase.dox'`; then echo shar: \"'dox/mbase.dox'\" unpacked with wrong size! fi # end of 'dox/mbase.dox' fi if test ! -d 'sample' ; then echo shar: Creating directory \"'sample'\" mkdir 'sample' fi if test ! -d 'src' ; then echo shar: Creating directory \"'src'\" mkdir 'src' fi if test -f 'src/util1.c' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'src/util1.c'\" else echo shar: Extracting \"'src/util1.c'\" $16120 characters$ sed "s/^X//" >'src/util1.c' <<'END_OF_FILE' X/* X * METALBASE 5.0 X * X * Released October 1st, 1992 by Huan-Ti [ richid@owlnet.rice.edu ] X * [ t-richj@microsoft.com ] X */ X X#define UTIL_C X#include "mbase.h" X#include "internal.h" X Xextern int _started; Xextern relation *_list[MAX_REL]; X X/****************************************************************************/ X X#ifdef LONGARGS X static void _dec_user (relation *); X#else X static void _dec_user(); X#endif X Xvoid X_seterr (err) Xmb_err err; X{ X switch (mb_errno = err) X { X case MB_OKAY: mb_error="No error"; break; X case MB_NO_ROOM: mb_error="MAX_REL is #defined to be too small"; break; X case MB_NO_MEMORY:mb_error="Not enough memory for requested task"; break; X case MB_NO_OPEN: mb_error="Cannot open given filename"; break; X case MB_NO_READ: mb_error="Cannot read given filename"; break; X case MB_FORMAT: mb_error="Relation is not in MB 5.0 format"; break; X case MB_LOCKED: mb_error="Relation is locked by another user"; break; X case MB_BUSY: mb_error="Relation is too busy"; break; X case MB_BAD_REL: mb_error="Function passed bad relation struct"; break; X case MB_NO_WRITE: mb_error="Cannot write given to relation"; break; X case MB_TIMEOUT: mb_error="Temporary lock has not been removed"; break; X case MB_BAD_REC: mb_error="A null rec pointer has been received"; break; X case MB_CORRUPT: mb_error="A corrupt index has been detected"; break; X case MB_BAD_DUP: mb_error="Addition would violate a nodups idx"; break; X case MB_NO_CURR: mb_error="Current record required for operation";break; X case MB_BAD_IDX: mb_error="A bad index number has been received"; break; X case MB_NO_SUCH: mb_error="The specified record can't be found"; break; X case MB_UNKNOWN: mb_error="Search command invalid"; break; X case MB_NO_FIELDS: mb_error="The new relation has no fields"; break; X case MB_NO_INDICES: mb_error="The new relation has no indices"; break; X case MB_BAD_INDEX: mb_error="The new index has no fields"; break; X case MB_DISKFULL: mb_error="There is not enough free space left"; break; X case MB_BAD_SERIAL: mb_error="The record's serial number changed"; break; X case MB_TMPDIR: mb_error="You must define a TMP directory"; break; X case MB_TMPERR: mb_error="Cannot work with TMP directory"; break; X default: mb_error="Undefined error--rebuild and pray"; break; X } X} X Xrelation * X_fill_info (rel, fld, idx) Xrelation *rel; Xlong fld, idx; X{ X register int i, j; X int s, e; X char temp[4], buf[128]; X short tshort; X#ifdef STRUCT_4 X int done = 0; X#endif X X rel->iser = rel->num_f; /* Until we find otherwise */ X X for (i=rel->rec_len=0; i<rel->num_f; i++) X { X if (lseek (rel->relcode, fld, 0) != fld) X { X _clr_lck (rel); free (rel); X relerr (MB_FORMAT, RNULL); X } X X readx (rel->relcode, buf, 1); rel->type[i] = (ftype)buf[0]; X readx (rel->relcode, &tshort, 2); X switch (rel->type[i]) X { X case T_SHORT: X case T_USHORT: rel->siz[i] = 2; /* sizeof(short) */ break; X case T_FLOAT: rel->siz[i] = 4; /* sizeof(float) */ break; X case T_DOUBLE: X case T_MONEY: rel->siz[i] = 8; /* sizeof(double) */ break; X case T_TIME: X case T_LONG: X case T_ULONG: rel->siz[i] = 4; /* sizeof(long) */ break; X case T_DATE: rel->siz[i] = 4; /* sizeof(long) */ break; X case T_PHONE: rel->siz[i] = 20; /* char[20] */ break; X case T_SERIAL: rel->siz[i] = 4; /* sizeof(long) */ X rel->iser = i; X break; X default: rel->siz[i] = (int)tshort; break; X } X X#ifdef STRUCT_1 X if (rel->type[i] != T_CHAR && rel->type[i] != T_PHONE) X { X if (rel->siz[i] == 8) rel->rec_len = round8(rel->rec_len); X else if (rel->siz[i] == 4) rel->rec_len = round4(rel->rec_len); X else rel->rec_len = round2(rel->rec_len); X } X#endif X X#ifdef STRUCT_3 X if (rel->type[i] != T_CHAR && rel->type[i] != T_PHONE) X { X if (rel->siz[i] == 2) rel->rec_len = round2(rel->rec_len); X else rel->rec_len = round4(rel->rec_len); X } X#endif X X#ifdef STRUCT_4 X if (rel->type[i] != T_CHAR && rel->type[i] != T_PHONE) X { X done=1; X rel->rec_len = round2(rel->rec_len); X } X#endif X X rel->start[i] = rel->rec_len; X rel->rec_len += rel->siz[i]; X X readx (rel->relcode, buf, 20); X X for (j=0; buf[j]!='|'; j++) X ; X fld += (long)j+4; /* Advance to next field */ X X buf[j] = 0; strcpy (rel->name[i], buf); X } X X#ifdef STRUCT_1 X rel->rec_len = round8(rel->rec_len); X#endif X X#ifdef STRUCT_3 X rel->rec_len = round4(rel->rec_len); X#endif X X#ifdef STRUCT_4 X if (done) X rel->rec_len = round2(rel->rec_len); X#endif X X if (lseek (rel->relcode, idx, 0) != idx) X { X _clr_lck (rel); X free (rel); X relerr (MB_FORMAT, RNULL); X } X X for (i=0; i<rel->num_i; i++) X { X readx (rel->relcode, buf, 1); e=(int)buf[0]; /* Dups/NoDups */ X readx (rel->relcode, buf, 1); s=(int)buf[0]; /* Num/Fields */ X readx (rel->relcode, buf, 20); X X for (j=0; buf[j]!=':'; j++) X ; X X idx += (long)j + 3L; X lseek (rel->relcode, idx, 0); /* Advance to index' fields */ X X buf[j] = 0; X strcpy (rel->iname[i], buf); X rel->itype[i]=e; sprintf (rel->idxs[i], "%03d", s); X X for (j=0; j<s; j++) X { X readx (rel->relcode, &tshort, 2); X sprintf (temp, "%03d", tshort); X strcat (rel->idxs[i], temp); X } X X idx = lseek (rel->relcode, 0L, 1); /* Get current position */ X } X X rel->hack = idx; /* Current position == reserved segment */ X X if (rel->ver == verCURRENT && ! rel->rdonly) X { X lseek (rel->lckcode, lckPOS_USERS, 0); X readx (rel->lckcode, buf, 1); X buf[0] = (char)((unsigned int)buf[0] +1); /* Increment the */ X lseek (rel->lckcode, lckPOS_USERS, 0); /* number of users */ X writx (rel->lckcode, buf, 1); X X _clr_lck (rel); X } X X relerr (MB_OKAY, rel); X} X Xvoid X_close_proc (rel) Xrelation *rel; X{ X if (rel->ver == verCURRENT) X { X (void)mb_unl (rel); X _dec_user (rel); X } X X close (rel->relcode); X free (rel); X} X Xvoid X_divine_mask (rel, key) Xrelation *rel; Xint key; X{ X rel->mask = (char)( X ((key & (int)0x04) << 5) | /* 00000100 -> 10000000 */ X ((key & (int)0x30) << 1) | /* 00110000 -> 01100000 */ X ((key & (int)0x80) >> 3) | /* 10000000 -> 00010000 */ X ((key & (int)0x03) << 2) | /* 00000011 -> 00001100 */ X ((key & (int)0x40) >> 5) | /* 01000000 -> 00000010 */ X ((key & (int)0x08) >> 3)); /* 00001000 -> 00000001 */ X} X Xint X_identify (rel) Xrelation *rel; X{ X register int i; X if (rel == RNULL || !_started) return -1; X for (i=0; i<MAX_REL; i++) X if (_list[i] == rel) break; X if (i == MAX_REL) return -1; X if (rel->ver != verCURRENT) return -2; X return i; X} X Xmb_err X_format (rel, rec, stage) Xrelation *rel; Xdataptr rec; Xint stage; X{ X register int i; X double tdbl; X long nexts, tlong; X int a; X long ac,pre,num,ext; X char *pch, temp[25]; X X if (! rec) baderr (MB_BAD_REC); X X if (stage == 2) X { GO_NEXTS (rel); X readx (rel->relcode, &nexts, 4); X } X X for (i=a=0; i<rel->num_f; i++) X { X if (stage == 1) X { X if (rel->type[i] == T_PHONE) X { X pch = (char *)rec + rel->start[i]; X scn_phone (&ac,&pre,&num,&ext, pch); X strcpy (temp, fmt_phone (ac,pre,num,ext, -1)); X strncpy (pch, temp, 20); X } X if (rel->type[i] == T_MONEY) X { X tdbl = *(double *)((char *)rec+rel->start[i]); X tlong = (long)(tdbl * 100.0); tdbl = (double)tlong / 100.0; X *(double *)((char *)rec+rel->start[i]) = tdbl; X } X } X else X { X if (rel->type[i] == T_SERIAL) X { X *(long *)((char *)rec+rel->start[i]) = nexts; X rel->serial = nexts; X _cryptf ((dataptr)((char *)rec+rel->start[i]), 4, rel->mask); X a = 1; X } X } X } X X if (a) X { X GO_NEXTS (rel); nexts++; X writx (rel->relcode, &nexts, 4); X } X X return MB_OKAY; X} X Xvoid X_crypt (rel, rec) Xrelation *rel; Xdataptr rec; X{ X#ifndef NOENCRYPT X register int i; X X if (rel->mask) X for (i=0; i<rel->num_f; i++) X _cryptf ((char *)rec+rel->start[i], rel->siz[i], rel->mask); X#else X ; /* Some compilers complain about null-functions */ X#endif X} X Xvoid X_cryptf (rec, siz, mask) Xdataptr rec; Xint siz, mask; X{ X#ifndef NOENCRYPT X register int i; X register char *c; X X if (mask != 0) X for (i=0,c=rec; i<siz; i++,c++) X { *c ^= mask; X mask = (mask + 1) & (int)0xFF; X } X#else X ; /* Some compilers complain about null-functions */ X#endif X} X Xmb_err X_check_dup (rel, rec, idx, ign) Xrelation *rel; Xdataptr rec; Xint idx; Xlong ign; X{ X dataptr mem; X long pos; X int dir; X X if (rel->itype[idx] == 1) reterr (MB_OKAY, 0); X X GO_TOP (rel, idx); X readx (rel->relcode, &pos, 4); X if (pos==0L) X reterr (MB_OKAY, 0); X X if ((mem = (dataptr)malloc (1+ rel->rec_len)) == NULL) X reterr (MB_NO_MEMORY, -1); X X for (;;) X { X dir = _compare (rel, rec, _memrec(rel,pos,mem), idx); X X if (dir == 0) X { X if (pos == ign) break; /* If we're about to change the rec, okay */ X free (mem); X X reterr (MB_BAD_DUP, -1); X } X X GO_POINT (rel, pos, idx, dir); X readx (rel->relcode, &pos, 4); X X if (pos == 0L) break; X } X free (mem); X X reterr (MB_OKAY, MB_OKAY); X} X Xlong X_append (rel, rec) Xrelation *rel; Xdataptr rec; X{ X long rcd, temp; X register int i; X char buf[2]; X X lseek (rel->relcode, POS_NUMREC, 0); X readx (rel->relcode, &rcd, 4); rcd++; X lseek (rel->relcode, POS_NUMREC, 0); X writx (rel->relcode, &rcd, 4); X X GO_START (rel, rcd); X X for (i=0,temp=0L,buf[0]=BAL_EV; i<rel->num_i; i++) X { X writx (rel->relcode, &temp, 4); X writx (rel->relcode, &temp, 4); X writx (rel->relcode, &temp, 4); X writx (rel->relcode, buf, 1); X } X writx (rel->relcode, rec, rel->rec_len); X X return rcd; X} X Xvoid X_remove (rel, bad) Xrelation *rel; Xlong bad; X{ X int i, dir; X long rep, tmp; X char temp[5]; X dataptr buf; X X lseek (rel->relcode, POS_NUMREC, 0); X readx (rel->relcode, &rep, 4); rep--; X lseek (rel->relcode, POS_NUMREC, 0); X writx (rel->relcode, &rep, 4); rep++; /* Find replacement record */ X X if (rep == bad) return; X X if ((buf=(dataptr)malloc (rel->rec_len +13*rel->num_i +1)) == NULL) return; X X GO_START (rel, rep); X readx (rel->relcode, buf, rel->rec_len +13*rel->num_i); X GO_START (rel, bad); X writx (rel->relcode, buf, rel->rec_len +13*rel->num_i); X X for (i = 0; i < rel->num_i; i++) X { X GO_POINT (rel, rep, i, 0); X readx (rel->relcode, &tmp, 4); X GO_BAL (rel, rep, i); X readx (rel->relcode, temp, 1); dir = (temp[0] & PARDIR); X X if (tmp == 0L) X GO_TOP (rel, i); X else X GO_POINT (rel, tmp, i, (dir ? 1 : -1)); X writx (rel->relcode, &bad, 4); X X GO_POINT (rel, rep, i, -1); X readx (rel->relcode, &tmp, 4); X if (tmp != 0L) X { X GO_POINT (rel, tmp, i, 0); X writx (rel->relcode, &bad, 4); X } X X GO_POINT (rel, rep, i, 1); X readx (rel->relcode, &tmp, 4); X if (tmp != 0L) X { X GO_POINT (rel, tmp, i, 0); X writx (rel->relcode, &bad, 4); X } X } X X free (buf); X} X Xmb_err X_link (rel, rcd) /* CACHED */ Xrelation *rel; Xlong rcd; X{ X register int i; X X _free_cache (); X X for (i=0; i<rel->num_i; i++) X { X _drop (rel, rcd, i, 0L); X if (_check (rel, rcd, 0L, i)) X { X _flush_cache (rel, i); X return mb_errno; X } X _flush_cache (rel, i); X } X return MB_OKAY; X} X Xvoid X_drop (rel, rcd, idx, top) /* CACHED */ Xrelation *rel; Xlong rcd, top; Xint idx; X{ X long pos, par; X int dir; X char temp[5]; X dataptr a, b; X cache *ptr; X X ptr = _read_cache (rel, top, idx); X if (top == 0L) X { X pos = ptr->num; X par = 0L; X } X else X { X par = ptr->parent; X pos = top; X } X X a = (dataptr)malloc (1+ rel->rec_len); X b = (dataptr)malloc (1+ rel->rec_len); X X for ( ; pos != 0L; ) X { X ptr = _read_cache (rel, pos, idx); X temp[0] = ptr->parbal; X X dir = _compare (rel, _memrec(rel,rcd,a), b=_memrec(rel,pos,b), idx ); X/* X * POSSIBLE ERROR POINT -- _memrec is being used twice in one X * function call. If it's going to be a problem, it'll show up after X * the second add, delete or update.... if it doesn't, this isn't X * the cause of your trouble. X * X */ X X if (dir == 0) X { X dir = ((temp[0] & BAL) == BAL_RT || (temp[0] & BAL) == BAL_FR) ? -1:1; X } X temp[0] += dir; X X _change_cache (ptr, parbal, temp[0]); X X par = pos; X pos = _cache_field (ptr, dir); X } X X free (a); free (b); X X if (par == 0L) X { X ptr = _read_cache (rel, 0L, idx); X _change_cache (ptr, num, rcd); X } X else X { X temp[0] = (char)(((dir==1) ? PARDIR:0) | BAL_EV); X X ptr = _read_cache (rel, rcd, idx); X _change_cache (ptr, parbal, temp[0]); X X ptr = _read_cache (rel, par, idx); X if (dir == -1) _changeqcache (ptr, left, rcd); X else if (dir == 0) _changeqcache (ptr, parent, rcd); X else _changeqcache (ptr, right, rcd); X ptr->changed = 1; X } X X ptr = _read_cache (rel, rcd, idx); X _change_cache (ptr, parent, par); X} X Xmb_err X_check (rel, st, ed, idx) /* CACHED */ Xrelation *rel; Xlong st, ed; Xint idx; X{ X long pos, tmp; X cache *ptr; X X for (pos = st; pos != ed; ) X { X if (pos == 0L) break; X X ptr = _read_cache (rel, pos, idx); X tmp = ptr->parent; X X if (! BALANCED ( (ptr->parbal & BAL) )) X if (_balance (rel, pos, idx, (int)(ptr->parbal & BAL))) X return mb_errno; X pos = tmp; X } X X return MB_OKAY; X} X X/****************************************************************************/ X Xint Xcompare (rel, ptra, ptrb, idx) /* -1 == ptra < ptrb */ Xrelation *rel; Xdataptr ptra, ptrb; Xint idx; X{ X int i; X if (_identify (rel) < 0) interr (MB_BAD_REL, -1); X if (idx < 0 || idx >= rel->num_i) interr (MB_BAD_IDX, -1); X if (ptra == ptrb) interr (MB_OKAY, 0); X X _crypt (rel, ptra); X _crypt (rel, ptrb); X i = _compare (rel, ptra, ptrb, idx); X _crypt (rel, ptra); X _crypt (rel, ptrb); X interr (MB_OKAY, i); X} X Xint Xidxnum (rel, name) Xrelation *rel; Xchar *name; X{ X int i; X X if (_identify (rel) < 0) interr (MB_BAD_REL, -1); X X for (i = 0; i < rel->num_i; i++) X if (! strcmp (name, rel->iname[i])) X break; X if (i == rel->num_i) interr (MB_BAD_IDX, -1); X X interr (MB_OKAY, i); X} X Xstatic void X_dec_user (rel) Xrelation *rel; X{ X char buf[2]; X X lseek (rel->lckcode, lckPOS_USERS, 0); X readx (rel->lckcode, buf, 1); X buf[0] = (char)((unsigned int)buf[0] -1); X if ((uchar)buf[0] > 254) buf[0] = 0; X lseek (rel->lckcode, lckPOS_USERS, 0); writx (rel->lckcode, buf, 1); X} X END_OF_FILE if test 16120 -ne `wc -c <'src/util1.c'`; then echo shar: \"'src/util1.c'\" unpacked with wrong size! fi # end of 'src/util1.c' fi echo shar: End of archive 1 $of 8$. cp /dev/null ark1isdone MISSING="" for I in 1 2 3 4 5 6 7 8 ; do if test ! -f ark${I}isdone ; then MISSING="${MISSING} ${I}" fi done if test "${MISSING}" = "" ; then echo You have unpacked all 8 archives. rm -f ark[1-9]isdone else echo You still must unpack the following archives: echo " " ${MISSING} fi exit 0 exit 0 # Just in case...