Acorn RISC PD-CD 1

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

/ Acorn RISC PD-CD 1 / Acorn RISC PD-CD 1.iso / utilities / powerbase / pbasedoc / _tutorial < prev next >

Wrap

Text File | 1993-05-22 | 24.0 KB | 460 lines

Tutorial ======== This document is intended to acquaint the new user as painlessly as possible with the most commonly-used facilities of Powerbase. It does not explain how to create a database from scratch (the file 'Create' does that) but it does explain how to:- (a) Start up Powerbase and open an existing database. (b) Browse through the database and search for individual records. (d) Add new records and alter or delete existing ones. (d) Query the database to create simple reports. (e) Close the database. The supplied sample database 'Elements' is used to illustrate most of these activities. If you have a hard disc you should copy 'Elements' and Powerbase itself onto it before proceeding further. It is possible to use Powerbase using floppy discs only but some operations, such as creating reports, will be quite slow. Unless memory is very tight therefore, users of floppy discs are recommended to allocate at least 112K to the RAM disc and copy 'Elements' onto it. Everything will then run much faster. Powerbase can provide context-sensitive help via Acorn's !Help application and you are advised to load this, resizing the help window and placing it at the top right of the screen. Further help can be had by choosing the Help entry on Powerbase's main menu or on the icon-bar menu. Starting Powerbase ================== Double-click on the Powerbase icon. After a few seconds it will appear on the icon-bar with the words "No data" beneath it to show that no database is open. Dragging the 'Elements' icon onto the Powerbase icon on the icon-bar will open 'Elements' for use and the words "No data" will be replaced by the name of the database. Double-clicking on the 'Elements' icon has the same effect, in fact if Powerbase isn't already on the icon-bar a double-click on 'Elements' will install it for you provided that the Powerbase icon has been seen by the filer. Two windows will open. One is the RECORD WINDOW and displays the first record in the database. The other is the KEYPAD which allows you to control the browsing, searching and editing functions of Powerbase. The keypad is rather large and obscures quite a big area of the record window. The most-used buttons, however, are on the top two rows and the keypad can be moved part way off screen to get it out of the way. Browsing through the Records ============================ The top row of buttons on the keypad allows you to browse through the file. If you have !Help running, as previously suggested, it will tell you what each button does. Once you get the general idea you will find the symbols on the buttons logical and self-explanatory. The most commonly used ones act as follows: * Single arrows move you one record at a time. From ACTINIUM, clicking the right arrow takes you to ALUMINIUM. The left arrow takes you back to ACTINIUM. * Double arrows move you through the file by long hops. Thus the right double arrow takes you from ACTINIUM to BISMUTH and the left one back to ACTINIUM again. The "fast forward/rewind" value is shown in the little box between the "fast wind" buttons. It is initially set to 10 records but can be altered by clicking in it with SELECT and entering a new value. * The buttons at the far left and right take you to the beginning and end of the file. The rightmost button therefore takes you to ZIRCONIUM and the leftmost one to ACTINIUM. One further feature of these buttons should be noted; the file wraps around. If you are at the last record (ZIRCONIUM) one record "forward" takes you to ACTINIUM again. Similarly, one record "back" from ACTINIUM gives ZIRCONIUM. The second row has buttons marked like the 'play' and 'stop' controls of a cassette player. 'Play' allows an automatic 'flip-through' of the records without the need to touch any other buttons. When you get to the place you want, click on 'stop'. If the records go past too quickly you can slow the process down by increasing the number in the central icon of the top row of buttons. (This is normally used to set the 'fast-wind' interval; see above.) Key fields, Keys and Indices ============================ When using the six buttons on the top row of the keypad you will have noticed that the records appear to be arranged alphabetically by the name of the element. The actual order in which records occur in the file is in fact nothing like alphabetical, but the order in which they are ACCESSED is determined by an INDEX which forces them to appear in alphabetical order. Every database must have at least one index, but you can have as many extra ones as you wish (subject to memory and disc limitations). Changing index -------------- The second row of the keypad has two buttons featuring a blue arrow with a key symbol. Click on the right one and the record for HYDROGEN is displayed! If you now experiment with the top row of buttons as before you will find that the records no longer appear in alphabetical order of name but in order of ATOMIC NUMBER instead. What we have done is change to a different index. If a database has multiple indices the "key" buttons allow you to cycle through them in opposite directions. (Since Elements has only two indices the two buttons have exactly the same effect.) When you change to a different index the screen always displays the first record as determined by that index. Keys and key fields ------------------- An index is built from KEYS which are derived from a particular KEY FIELD. The two indices used by Elements are based on the NAME and ATOMIC NUMBER fields. The way in which the key stored in the index is derived from the key field is different in the two cases. For the NAME index the first four letters of the name are used. Since no two elements have names beginning with the same four letters this gives a unique "value" for the key. (Two letters would not have been enough because of such names as CHLORINE and CHROMIUM.) The ATOMIC NUMBER index is based on the whole 3-digit maximum length and the ordering is NUMERIC. Whenever a record is displayed the key is shown in the window title (# characters are for padding and should be ignored.) The record screen clearly shows which fields are keyfields and which are not. The descriptors, or labels, of fields are normally black. Where the fields are keyfields they are shown in dark blue instead and the field used by the current index has its descriptor in red. Note also that the background colour of the NAME field is pale yellow whereas the rest are white . This indicates that NAME is the PRIMARY KEY field. The primary key and the index based on it is more important than the others. It is the one index which a database MUST have and is the one selected when a database is first opened. Searching for a record ====================== The rightmost button on the third row of the keypad shows a question mark followed by a "record card" icon We will call it the Search button. When clicked with SELECT the keypad window grows an extension on the right-hand side. This new area contains a writable icon in which to enter the key of the record you wish to display. Suppose you want to see the record for the element NITROGEN. If the current index is the NAME then type in NITR and either type RETURN or click the 'OK' button. The required record quickly appears. If you enter the key of a non-existent record an error box is displayed. You don't necessarily need to type the whole key; NIT would do in this case, but NI would be ambiguous (there are elements called NICKEL and NIOBIUM). To find NITROGEN when the ATOMIC NUMBER index is active you would need to know that the atomic number of nitrogen is 7. Type this in and type RETURN or click on 'OK' and the record appears. Abbreviated keys are NOT allowed in this case for obvious reasons; if you want element 103 (LAWRENCIUM) it's no good entering 10 (NEON)! Editing the Database ==================== This part of the tutorial involves changes to the data records. Make a working copy of the disc and put the original in a safe place. If you are using a copy of Elements on the RAM disc then the floppy disc won't be altered, but it's still not a good idea to work with the original disc. Modifying records ----------------- Pointing the mouse at any field and clicking SELECT places the caret in that field. The contents of the field may then be edited in the standard way for RISC OS writable icons. i.e. CTRL U clears the field. DELETE removes the character to the left of the caret. COPY removes the character to the right of the caret. SHIFT/COPY deletes the word to the right of the caret. CTRL/COPY deletes everything from the caret to the end of the line. ⇦ and ⇨ move the caret left/right by one character. SHIFT ⇦ and SHIFT ⇨ move the caret left/right by one word. CRTL ⇦ and CTRL ⇨ move the caret to start and end of the field. Changes made to a record cause the record to be written back to the database whenever you display a different record or close the database. There is a button on the keypad called 'Force update' but you don't normally need to use it. At the ends of the third row of keypad buttons are two which contain a vertical blue arrow. Clicking on the left button stores the contents of the field which has the caret in a scratchpad. Moving the caret to another field and clicking the right button copies the data from the scratchpad into that field, over-writing its previous contents. This provides a quick method of copying data from one field to another, or from one record to another. You may continue to "paste" the data into new fields as often as you like. Only when you click the left button again are the contents of the scratchpad changed. Data "pasted" into a field cannot "overflow"; if too long for the field it will be cut to the permitted length. As an added precaution you can't "paste" a null string into a field. Recovering data after a mistake ------------------------------- You will sometimes edit the wrong field by mistake and want to restore the previous contents. Click MENU over the field. The second entry on the main menu contains the "tag" of the field. Follow the right-pointing arrow to the 'Field' sub-menu and choose the last entry: 'Restore field'. Similarly, if you make a mess of a complete record and want to go back to the way it was you can choose 'Restore record' from the main menu. It is important to realise that these 'restore' facilities will not work if you have looked at another record before using them (or clicked on 'Force update'). In other words you can recover data which has been altered in or removed from the currently-displayed record only. Adding new records ------------------ First we will add a new record for a fictitious element as described below:- NAME TASMANIUM SYMBOL Ts ATOMIC NUMBER 108 We could invent a lot more, but this will be enough to get the idea across. At the far left of the second row on the keypad is a button with a plus sign and a "record card" icon. This is the 'Insert' button. A blank record will be displayed with the caret in the NAME field. Enter TASMANIUM here and press RETURN. The caret will jump to the SYMBOL field. Enter Ts. You will now need to press RETURN twice to get to ATOMIC NUMBER (One press takes you to the un-named field after NAME where the origin of the element's name is given.) Alternatively you could point the mouse at the ATOMIC NUMBER box and click SELECT. There is no need to press RETURN after entering a field in order to "accept" the data; it's simply the most convenient way of getting to the next field if that's what you want to do. When entering data into widely- scattered fields it is more convenient to use the mouse. Having entered your data click any of the first row of buttons so that a different record is displayed. Now use 'Search' to find TASMANIUM by name. Change to the other index and find it by atomic number. In both cases try moving to the record just before and just after the new one, noting that it has been placed in its correct position in both indices: between TANTALUM and TECHNETIUM in the NAME index and at the very end of the ATOMIC NUMBER index. Data Input Validation --------------------- When entering the above data you may have noticed that the NAME field will only accept capital letters and the ATOMIC NUMBER field will only accept numerals. This is because the fields have an associated TYPE which restricts the characters which the field will accept. Any others are simply ignored. We will call this feature TYPE VALIDATION. Another level of data validation, which we will now illustrate, is called TABLE VALIDATION Make an entry in the GROUP field. Type X, for example, then try to move to another field by either pressing RETURN or clicking the mouse. You'll find that you can't do so - you get an error message instead. This is because the GROUP field is linked to a VALIDATION TABLE which means that only entries which appear in the table may be used in that field. Fields linked to validation tables are distinguished by using dark green, instead of black, as the foreground colour. To see what the validation table contains, place the caret in the GROUP field then click on the List values button on the keypad. The validation table will then be displayed in its own window. Only those items in the left column of the table may be entered in the GROUP field. Element 108 would be a transition metal, so enter T. You will now find that Powerbase accepts your data. Both type validation and table validation may be turned off and on by means of an option on the Validation submenu. The item is normally ticked, showing that validation is turned on. Deleting records ---------------- Since "tasmanium" isn't a real element we want to remove it from our database. There is a "dustbin" button which will do just that: but don't use it yet. Make sure that the unwanted record is on the screen and click the button to the RIGHT of the bin. You will now find that "tasmanium" has vanished; the names run straight from TANTALUM to TECHNETIUM and the atomic numbers end with 103, just as they did before you added the extra record. Wasn't that rather TOO easy? Suppose you clicked the button by mistake and deleted an important record! Powerbase handles this very nicely. The record hasn't really been deleted at all. To prove this, click the button above the one you used for "deleting". The button shows a green right-pointing arrow and a "card index" icon. There's TASMANIUM again! You will soon find that it appears to be the ONLY record in the file! Don't worry; nothing has been lost. A Powerbase database is treated as six separate SUBFILES numbered 0-5. When you first open the database you are looking at Subfile 0 and the heading of the record window shows this. If you look at the heading now you will see that it has changed to Subfile 1. The buttons on either side of the bin carry pictures which suggest their true functions. They do NOT delete records; they merely MOVE records from one subfile to another. The right button moves the displayed record to the next subfile in sequence. Thus, in the above case the TASMANIUM record has been moved from subfile 0 to subfile 1. A record in subfile 1 would be moved to subfile 2 and so on. What happens when it gets to 5? The next click shoves it into subfile 0 where it was originally. As you might have guessed, the left button moves records in the opposite direction, i.e. from subfile 1 to 0 or from subfile 0 to 5. What happens when you DO want to lose a record for good? That's when you consign it to the dustbin! Since this button's action is more drastic than that of the other two you are asked for confirmation before deletion occurs. Printing Lists ============== Before Powerbase can produce a list it must be told: * which FIELDS should appear on the list * which RECORDS should appear on the list. Making the field selection -------------------------- Specifying the fields you want is extremely easy. To include a field simply point the mouse at it and click with ADJUST. The selected field will then appear in inverse video (i.e. normally white on black). If you change your mind about a field you can de-select it by clicking on it again with ADJUST. The order in which the fields will appear in the printed list is the order in which you select them. Take care with this since there is nothing on the screen to indicate what order the fields will be in. The fields chosen for inclusion in a printout make up a SELECTION which may be saved and reloaded for future use. Just for now select NAME, SYMBOL and GROUP. Choosing which records to print ------------------------------- Specifying which records to print is somewhat more complex, although efforts have been made to render it as simple as possible. With the mouse pointer in the record window click MENU. This displays the main Powerbase menu from which you should choose Print. A small window appears in which you can specify the criteria which must be met by records for inclusion in the list. If you just click OK (or press RETURN) without entering anything here you will obtain a list of the whole database (or, more accurately, of the the whole of the current subfile ). Try this first. Nothing will be printed until the search is complete. If the mouse pointer is over the main window the hourglass will tell you how far things have progressed. Pressing Escape causes the process to be aborted. When the process is complete the list will be displayed in a window. It may be saved as a text file and printed using the usual RISC OS drivers . If you point to a record in the displayed list and double-click with SELECT Powerbase will display the chosen record in the main record window. Note that the caret is placed in the field you were pointing to on the list. This is very useful if you spot errors in the data whilst examining a list. A double click on the offending item offers the record with the caret correctly positioned for editing. Most database queries, however, don't involve printing all the records! You will usually need to enter a SEARCH FORMULA before clicking OK. Search formulae can be quite complex and a more detailed treatment is given in a later section. Only a couple of simple examples will be shown here. Suppose we want a list of all the transitional elements. i.e. all those which have T in the GROUP field. We can do this by entering the search formula GP=T. Why GP? Shouldn't it be GROUP=T? The latter seems more obvious, but what happens if the field has no descriptor? The record screen of !Elements contains many such fields such as the one on the first line which gives the origin of the element's name. Every field which can hold data has to have a "handle" by which we can refer to it and that handle is called the field TAG. Tags are very short; never more than four characters. You have seen some tags already. Look at the list you printed earlier. The three columns are headed NAME, SYM and GP not NAME, SYMBOL and GROUP. The user therefore meets tags in two contexts: * They are used to specify fields in search formulae * They are the default column headings in lists. The Print options window allows you to change the column headings to the field descriptors if you so wish. Type GP=T in the space provided and click OK. This time the list will be a much shorter one and you will see that the GP column contains T in every case. Our second example involves printing all elements whose names end in "IUM". What we need to specify is that "NAME contains IUM" and we do this by entering the search formula NAME{IUM. Finally we will combine the previous two examples into a two-pronged search for all the TRANSITIONAL elements whose names end in "IUM". The search formula this time is NAME{IUM AND GP=T. Note the spaces before and after AND. These are essential. You might like to try an alternative way of entering this search formula. Proceed as follows: * Click the "toggle size" icon of the matching records window. This makes it expand to show more controls. * NAME should already be visible as the required field, but you need to select the radio button for Contains. * Click in the space next to Value: and enter IUM. * Click the AND button to select it. * Click Add to formula. The search formula should now read NAME{IUM AND. * Click on the field tag (currently NAME) until it reads GP * Select Is equal to. * Enter T in the Value field. * Click Add to formula. The complete search formula should now be visible so click OK and the list will be produced. The above takes far longer to describe than it does to do and can be a great help to the infrequent Powerbase user who has trouble remembering the syntax of search formulae. If you use certain complex queries often you may find it convenient to save them for future use . The biggest problem you are likely to encounter is not knowing the tags for the various fields. The method just described gets round the problem by letting you cycle through the tags by clicking on an icon rather than typing them directly. You will also find that clicking with SELECT on a field while holding down ALT causes the tag of that field to appear in the search formula icon. Closing down ============ A database may be closed down by choosing the Close database entry from the icon bar menu. When you do so, all the indices stored in memory are saved to disc (the Save indices button on the keypad gives a clear indication of this) and the database is closed. You do not have to do anything to save the database itself to disc since it has been there all along. Only the current record is in memory at any one time and this is automatically written back to disc when it has been changed in any way. Also saved automatically is the link file which holds the relational links between the main database and the validation tables. The tables themselves are NOT saved automatically. If you have edited a table, or created a new one, you will need to click MENU over the displayed table and choose the Save option from it. If you are going to use another database straight away just double click on it or drag it to the Powerbase icon on the icon bar. In fact you don't even need to close the previous database before doing this; Powerbase will do it for you. When working with a database on the RAM disc Powerbase will issue a warning that the data is not secure and that you need to drag it from the RAM disc to a floppy or hard disc before switching off. To terminate Powerbase itself use the Quit option on the icon bar menu. Merging data into an Impression document ======================================== The distribution disc contains a simple Impression document called 'MergeTest'. It will be used in conjunction with the 'Elements' database to illustrate direct data-merging with Impression. (1) Make sure that Impression has been "seen" by the filer. (You need not actually load it.) (2) Open the 'Elements' database and drop 'MergeTest' onto the record window. (3) Click on 'Merge' in the 'Merge data' dialogue box. You will see that data relating to the record for ACTINIUM has been inserted into the Impression document. (4) The arrow and double-arrow buttons (identical to those on the keypad) may be used to move through the database causing different records to be merged into the document. (5) A search formula may be used to restrict the operation to records matching certain criteria. (6) 'Print' may be used to print the document with the currently-merged data. To print a set of documents, one for each matching element, select the 'Print all' switch then click on 'Merge'. During data-merging Impression editing is disabled - you can't place the caret in the document and the mouse pointer appears as a red and blue cross. Clicking on 'Cancel' or closing the 'Merge data' dialogue box restores normal editing. For more information on data-merging, including instructions on how to prepare the Impression document, see 'MailMerge'.