Oakland CPM Archive

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

/ Oakland CPM Archive / oakcpm.iso / sigm / vol194 / plot33.dqc / PLOT33.DOC

Wrap

Text File | 1985-02-09 | 75KB | 1,613 lines

PLOT33.DOC July 31, 1984 PLOT Version 3.3 (PLOT33) DOCUMENTATION PART 1. PLOT USER'S GUIDE General Information Running PLOT Application Programming PLOT Commands Pitfalls and Limitations Error Messages PART 2. APPLICATION PROGRAMS BASIC Programs TURBO PASCAL Library FORTRAN Library PART 3. INSTALLATION INSTRUCTIONS Sizing the Memory Map Printer Settings 1 PLOT33.DOC PLOT User's Guide July 31, 1984 PART 1 USER'S GUIDE GENERAL INFORMATION This collection of programs contains a graphics system that can produce high resolution graphics on dot matrix printers and other graphic devices. This system consists of three parts: - high level language subprograms that you include in your application programs, - a disk file created by the application program which is used to describe the pictures to be printed, - and low level device drivers that read the file and produce the pictures on whatever sort of hardware you are using. The application program "plots" by writing simple plot commands to a disk file, which is later read and printed by the low level driver. This file is the key to the whole system, because it makes the application program independent of the plotting device and vice versa. It also makes it possible to plot pictures that were created on different computers or to plot the same picture on different types of hardware, such as CRT's and printers. The system contained in this collection has high level subprogram libraries written in BASIC (GRAPH.BAS), Turbo Pascal(GRAF1.PAS and GRAF2.PAS), and FORTRAN (GRAF.FOR). A C library is being debugged. It contains one assembly language device driver, PLOT33.ASM, which will plot on a variety of dot matrix printers. These printers include the Epson MX-80 etc., CItoh Prowriter, Okidata 92, NEC, Gemini-10, Apple Dot Matrix, and IDS printers. PLOT33 lets you treat your printer as if it were a plotter capable of true vector graphics. All future versions of PLOT (3.x) will support the same basic file format and will be compatible with each other. This documentation describes how to get PLOT33 installed, specifies the format for the disk file of plotting vectors (".VEC" files), and gives an overview of the high level language packages. You will probably want to briefly skim over the descriptions below and go right to the Installation section. If you intend to mainly use the programs and subprogram libraries provided, the exact details of the format of the commands in the file are not important. 2 PLOT33.DOC PLOT User's Guide July 31, 1984 Capabilities PLOT's command set includes instructions for device independent graphics and commands that provide access to the unique capabilities of specific devices. The device independent commands are patterned after a pen plotter's characteristics, with a rudimentary capability for filling areas on raster scan devices. These device independent commands include the capability to move to a point without plotting, to plot points and straight line segments, to change the current color, and fill in a trapezoidal area. More complex areas can be decomposed into simpler regions for filling. The device dependent commands include the ability to plot/print a string using hardware character capabilities. The Upload command can be used to upload the red-green-blue values used for the color table on color graphics terminals (or stipple patterns to be used for "colors" on monochrome devices as in PLOT33). The Erase command will clear the screen, establish the background color, or advance the page, as appropriate for the type of plotter or display. An Output command is used to produce hardcopy output or print the memory map. A general instruction that can be used to extend the command set for any other non-standard command required rounds out the device dependent commands. Bug Fixes and Future Versions The assembly language and BASIC programs in this system have been tested so far on four different computers and six different types of printers. I would appreciate hearing about any problems you have or changes that you make to the program. I am particularly interested in versions for different printers or drivers for other plotting devices such as graphic terminals and plotters. Please send your comments and questions to the author at: Tom Speer 887 Briddlewood Ln. Dayton, Ohio 45430 (513)429-2781 It is important that a common version be maintained that incorporates fixes to all of the bugs and implements the improvements that are of general interest. As mentioned above, all future updates to this version of PLOT will be compatible with the existing plot files and high level software. Future drivers are also planned for creating and interpreting the vector command file on mainframe computers using CALCOMP or Tektronics plotters and terminals. As much as possible, the eXtend command should be used for adding neat new features for your printer or taking advantage of the capabilities of more advanced devices. This will make it 3 PLOT33.DOC PLOT User's Guide July 31, 1984 possible to maintain the file format as a means of exchanging information without a fatal proliferation of exclusive versions. Acknowledgments I owe a special debt of thanks to all the people that have been my guinea pigs in adding new printers and discovering bugs. Especially Hal Carter for his help and encouragement, Kirk Horton for really wringing out the system in his VLSI applications, and Mike Gingrich for his critiques of the system and documentation. RUNNING PLOT To use PLOT, first create a sequential disk file of plot commands. These are described fully in the next section. Then execute PLOT by typing: PLOT33 ifn [ofn] ifn is the input file name. A file type of .VEC is assumed if no file type is given. [ofn] is an optional output file name. If no output file is given, the output from PLOT will go directly to the printer (CP/M LST: device). If an output file is specified, all of the information normally sent to the printer will be stored in the disk file instead. In this case, no printed output will be produced. PIP or some other utility may be used to print the picture. This option is useful for making multiple copies or for including graphics in the middle of text files. The default output file type is .PLT . Note: Output files are large- typically 40 to 60k. Be sure to leave enough space on the disk for the entire file. APPLICATION PROGRAMMING Application programs "plot" by writing the appropriate plot command to a disk file. This file later becomes PLOT's input file. The plotting area is a square, with the coordinates (0.0 , 0.0) located at the bottom left corner, and the coordinates (1.0 , 1.0) located at the upper right corner. These coordinate values are independent of the resolution or aspect ratio of the plotting device. PLOT converts these values to the actual number of rasters (dots) used by the printer to produce a square plot approximately eight inches on a side. 4 PLOT33.DOC PLOT User's Guide July 31, 1984 All plot commands are written to disk one after the other, with no spaces or delimiting characters between them. Each command consists of an ASCII character followed by the binary data which it requires. Many languages, such as BASIC, insert extraneous characters, such as carriage returns or line feeds, at the end of each block written. Because of this, plotting commands should not be split between blocks, as these extra characters will be interpreted as data in a command. If the extraneous character occupies the position of a command character, however, it can be detected and ignored. The current application programs output plot commands in 128 byte blocks. If a command will not fit in a given block, the block is written to disk, and the next command written in a new block. This ensures that the extraneous characters will fall between commands where PLOT can deal with them. When using fixed length blocks or records, each can be padded with "N"'s to preserve its length before it is written. The recommended command sequence to produce a picture is: 1. Set the color to black or white (color values 127 or 0) for a background. Usually white is used. 2. Erase the picture to initialize the memory map area and establish the background. 3. Set the color value for plotting. Usually black (127) is used. 4. Establish a starting point by using the Move command, plotting a single point, or plotting a line segment using the Draw command. 5. Proceed to create the remainder of the plot. The Increment command is recommended when one line segment is connected to the previous one, as in a curve. 6. Be sure to include an Output command to print the final plot. 7. Use the Quit command to stop plotting. PLOT will terminate automatically at an end-of-file, but with an error message, and without printing the picture. PLOT COMMANDS Each command is of the form: Adata "A" represents the single ASCII character as explained below. "data" represents 0 to 86 bytes of data required for the particular command. Usually this consists of fixed point 5 PLOT33.DOC PLOT User's Guide July 31, 1984 numbers giving the coordinates of points or ends of line segments. No spaces or delimitating characters are used between fields within a command or between commands. The individual commands are described in Table 1. The data values required for the plot commands are not ASCII character strings. Thus, they cannot be made using normal formatted output of the coordinate variables in the application program. The coordinate values are 16 bit fixed point numbers. These are created in a high level program by multiplying the floating point value (between 0.0 and 1.0) by 32767 and truncating the result to an integer. In BASIC, the MKI$ function can be used to store the data in a string variable, which is then written to the disk file. FORTRAN programmers can simply use unformatted files or convert everything to characters and use A formats. Under PASCAL, the file can be declared to be a FILE OF CHAR and the binary values converted the characters using the CHR function. This does not work well for Turbo Pascal, however, since Turbo places the file size and record length at the head of a FILE OF CHAR. Turbo Pascal users should use TEXT files instead. The different word lengths, internal storage formats, and file structures have to be considered when constructing the command file on mainframes and microcomputers other than CP/M based machines. The ablility to structure a file as a byte stream is the basic requirement. The CP/M Dump utility can be used to ensure that the contents of the file are what you expected. The 16 bit data values are stored with the low order byte first (swapped format). This is the standard way of storing words for the 8080 and Z80 processors. In the case of the one byte integer required by the Color command, the first element of the string created by MKI$ should be used, as MKI$ automatically places the bytes in swapped format. Note that because the data values are not ASCII strings, the CP/M TYPE command will not properly display the file at the console. In addition, the [O] option must be specified when using PIP to transfer either the vector command file (.VEC) or the printer (.PLT) file. 6 PLOT33.DOC PLOT User's Guide July 31, 1984 TABLE 1 PLOT COMMANDS Command Length Name Character Data (bytes) Command Data Description ---------------------------------------------------------------------- Color C b 2 8 bit integer specifying one of a range of colors. 0 indicates white, 127 indicates black. Draw D X1,Y1, 9 Coordinate pairs for ends of line segments. X2,Y2 Erase E 1 Sets the entire ploting area to the currently selected Color value. Fill F X1,Y1, 11 Coordinate pairs for ends of line segment X2,Y2,Yf and horizontal level (Yf) defines area to be filled in with currently selected color. Incre- I X,Y 5 Coordinate pair for end of line segment. ment Starting coordinate is last point plotted. Move M X,Y 5 Coordinate pair for new "pen" position. No-op N 1 Ignored. Used as space filler in sequence of commands to exactly fill a disk record. Output O 1 Causes current picture to be printed. Point P X,Y 5 Coordinate pair for point to be plotted. Quit Q 1 Commands normal termination of program. String S X,Y, 6 Coordinate pair for start of character string to string to be printed. String 86 MUST end with a Carriage Return. Text T string 2+ String of arbitrary length. String MUST end with a <NULL> (00H) character. Upload U N, 3 N is a 16 bit unsigned integer giving the values to number of values to follow. Values N+3 specify basis for creating new "colors". eXtend X N, 3 N is a 16 bit unsigned integer giving the values to number of values to follow. Values are N N+3 bytes of whatever data is necessary. 7 PLOT33.DOC PLOT User's Guide July 31, 1984 The following sections describe the way in which the commands have been implemented in PLOT Version 3.3 (PLOT33). Other drivers may not implement them in exactly the same way, but the general sense should be preserved. Colors are simulated in PLOT33 by using stipple patterns. These patterns apply to all points, lines, and filled areas. These patterns are based on 8 dot by 8 dot cells that tile the entire picture area. Imagine cutting each figure out of a sheet entirely covered with the desired pattern and then pasting it on the plot in progress. This gives a general idea of the way in which the patterns work. Three different methods of generating the patterns are used. Each method trades off the memory required to store the basic patterns with the ability to specify exactly what the pattern will look like. When a Color command is given, all subsequent plotting will be done with that "color", or pattern, until it is changed with another Color command. The Color command simply acts as a selector to pick one of a number of colors or patterns that are currently available. A color code of 0 corresponds to white. Lines and filled areas plotted with this color will absolutely erase any dots that were previously plotted. Color codes from 1 through 48 will produce cross hatched patterns that generate checkered areas and other distictive patterns. Codes from 49 to 63 correspond to 15 special patterns that can be programmed to exactly match any desired pattern based on the 8 x 8 dot cells. Codes from 64 through 127 are generated using an ordered dither matrix, in which each succeeding code is the same as the previous one, except for one more dot per cell being plotted. These are currently set to generate a range of shades of gray. All of these patterns can be changed under the control of the application program by using the Upload command. Each dot is added to the memory map by OR'ing it with the dots previously plotted. Each subsequent figure does not totally obscure the existing picture. For example, suppose that a gray shade is being used that consists of every other dot being plotted. The dots that are not plotted will not cause the existing dots to be set to white if they were previously plotted. Negative color values (-1 to -127) will use the same pattern as their positive counterpart, but will be XOR'ed with the existing picture. This means that the figure will be plotted black on white or white on black, depending on what the existing background is. Plotting the same figure twice with a negative color value will have no net effect on the picture. Filling the whole picture with -127 (negative black) will have an effect like making a photographic negative. The String command is used to mix printed characters and graphics. The characters are printed using the printers normal character set. No facility is provided in PLOT33 to send <ESC> 9 PLOT33.DOC PLOT User's Guide July 31, 1984 codes or control characters to the printer to change character sizes or font styles. The string is printed starting near the coordinates given in the command. The exact position will vary, but the first character will cover the coordinate location. The ambiguity comes from the fact that most printers can only print characters at their normal location on the page, and can't use graphics commands to position the characters at any arbitrary place on the line. In addition, a printed character can't be part on one line of print and part on another. Therefore, PLOT33 computes the closest character location to the desired coordinates and begins the string there. The Text command will immediately send a text string to the printer when the command is encountered. This is quite different from the String command, because the text is not put in the memory map and is not part of the picture. This command is useful for positioning the plot on the page, for printing titles and headers, and for advancing to the next page after the plot is printed. Any ASCII characters may be included in the string, including control codes, escape sequences, carriage returns, and linefeeds. The only exception is the null character, 00H, which is used to signal the end of the string. The string may be of any length. This makes it a convenient way of imbedding a figure in the middle of a document, since all of the text above the figure can be included in one Text command, as can all of the text after the plot. Any text to be included in the plotting area itself must be entered one line at a time with the String command. Upload Command The Upload command deserves some special attention, because it is the least standard of all of the commands currently implemented. Most dot matrix printers can, of course, only print in black. To give these printers a "color" capability, different patterns have been programmed to produce shades of gray and distinctive cross hatched patterns. The cross hatched patterns are intended to be used for bar charts, dotted and dashed lines, etc. The shades are the closest that PLOT33 can come to true colors. The special patterns are very useful for matching patterns produced by other systems, and for anything that just can't be done with the other two methods. The Color command acts as a selector to choose between the pre-defined patterns. The Upload command is only needed if the default patterns are not capable of producing the desired effect. It will redefine the patterns from which the Color command can select. The cross hatched patterns are based on seven bytes that provide the basic patterns in both the vertical and horizontal directions for each cell. One byte from the seven is selected as the X pattern and one byte from the seven is selected as the Y pattern. The X pattern is extended vertically throughout the 10 PLOT33.DOC PLOT User's Guide July 31, 1984 cell, and the Y pattern is extended horizontally throughout the cell. The two patterns are reversed (XOR'ed) where they cross. The effect is much like weaving a plaid material. The selection of the bytes for the X and Y patterns is made by the Color code. The lower three bytes of the Color value select the X byte and the next three bytes select the Y byte. The best way to determine the exact apperance of each pattern is to experiment by filling blocks with different patterns. If the existing patterns are not acceptable, new ones can be Uploaded. Set the number field, N, to 07H. The next seven bytes will be the bytes that form the basis for the patterns. Each bit corresponds to a dot in the pattern. A byte of 07H for the X pattern would produce the pattern .....*** . If an F0H is used for the Y pattern, the final cell will look like: *****... *****... *****... *****... .....*** .....*** .....*** .....*** The second type of pattern, used for the 15 special patterns, is specified by giving the 8 bytes that produce the pattern. Thus, the sequence 00H 00H 99H 00H 00H 99H 00H 00H will produce the pattern: ..*..*.. ........ ........ ..*..*.. ..*..*.. ........ ........ ..*..*.. These patterns are uploaded by specifying 120 for N followed by the 120 bytes that define all 15 of the special patterns. No facility is provided for just uploading one set of 8 bytes. The other method for producing patterns for "colors" is an eight by eight byte dither matrix. In this method, each byte in the matrix corresponds to one dot in the eight by eight dot cell. These cells tile the picture plane, as before. Color code values from 64 to 127 are used to select the 64 patterns created by the dither matrix. Each element of the matrix is assigned a value from 63 to 126. If the Color value is greater than the value in the matrix corresponding to a given dot, that dot is plotted. Thus, 127 is greater than all of the elements, and all dots in the matrix will be plotted, producing black. If each element of the matrix is assigned a different value, 11 PLOT33.DOC PLOT User's Guide July 31, 1984 then a unique pattern will be associated with each Color value. For example, if the Color code is 64, only the element with a value of 63 will produce a dot. If the Color value is 65, those elements with 63 and 64 will produce dots. As the Color code is increased, and figures are plotted, another dot in the "tile" has the potential for being plotted. The existing arrangement of values in the dither matrix is designed to produce shades of gray with a minimum appearance of organized patterns. This matrix came from Foley and Van Dam's "Fundamentals of Interactive Computer Graphics". A common use of the Upload command is to change the matrix to provide diagonal patterns for shading areas. This would produce a pattern that looks like (for four cells): //// //// //// //// To design an Upload sequence for a given pattern, lay out a matrix representing a single cell. Start with 63 and place the numbers in the matrix in the pattern desired. For example, a cell from the diagonal pattern above could be created from the matrix: 63 111 85 103 75 115 93 95 112 64 104 86 116 76 96 94 79 105 65 117 87 97 77 121 106 80 118 66 98 88 122 78 71 119 81 99 67 123 89 107 120 72 100 82 124 68 108 90 91 101 73 125 83 109 69 113 102 92 126 74 110 84 114 70 To see the pattern, look at the matrix part way through its construction: 63 75 64 76 65 77 66 78 71 67 72 68 73 69 74 70 For this matrix, the Color code 79 will result in all of the above points being plotted in any filled in area, and the diagonal pattern will emerge. Once the matrix is constructed, the Upload command to load it is formed by an ASCII "U" followed by 40H, 00H to indicate 64 bytes to come, and finally 03H, 6FH, etc., the bytes of the 12 PLOT33.DOC PLOT User's Guide July 31, 1984 matrix. The matrix is read in one row after another ( the [1,1] element followed by the [1,2] element, ..., to the [1,8] element, then the [2,1] element, ...). All of the matrix values are binary 8 bit values. In BASIC, use the MKI$ function to convert the integers to strings, then use the first character. To summarize the Upload command: The Color codes from 1 to 48 are produced by cross-hatched patterns based on eight pattern bytes. These can by Uploaded by a "U" 08H 00H followed by eight new pattern bytes. The Color codes from 49 throuth 63 are Uploaded by a "U" 78H 00H and the 120 bytes that define the 15 cell patterns. The Color codes from 64 through 127 are produced by a dither matrix. This is created by assigning elements of an eight by eight matrix the values 63 through 126. The matrix is Uploaded by a "U" 40H 00H followed by the 64 elements in row major order. PITFALLS AND LIMITATIONS This section amplifies the more subtle points and limitaions of PLOT33's commands. The valid range for coordinate values is 0 to 1. This corresponds to binary integer values from 0 to 32767. As each point is plotted, a check is made to insure that the coordinate is within the memory map. If it is not, the point is not plotted. This is mainly to keep the program from running amok and writing over parts of memory that are not part of the memory map of the picture. This is not intended to be a form of windowing. If you want to zoom in on a part of a larger picture, the picture should be clipped by the application program. Note also, that most high level languages will not deal with positive integers that are larger than 32767, as these larger numbers represent negative integers. In contrast to the coordinate values used for all of the graphic commands, any coordinate value outside the picture used to indicate the start of printing a string will be set to zero. This is intended to make the String command act as much as possible like a normal printer. When the string goes off of the right hand side of the plot, the X coordinate for the next character will be set to zero and the string will wrap around- just like a carriage return without a linefeed. All characters that attempt to be printed above or below the plot will be printed on the bottom line. String has a few other points of interest. Only printable characters should be used. Control characters and escape codes will occupy their respective positions in the memory map but will not be printed. This will cause the rest of the line to be shifted to the left and destroy its alignment with the rest of the page. Also, the character pitch must agree with the 13 PLOT33.DOC PLOT User's Guide July 31, 1984 character width set at assembly time, or the alignment will be affected. The coordinates for the start of the string are truncated to the nearest multiple of the size of the character. Thus the specified point will be within the first character, but the character will not be centered or aligned with the given coordinate position. This was done because most printers insist that the printed characters occupy these locations, and PLOT33 has to respect this or the aligment of the rest of the line will again be affected. PLOT33 prints the strings, it does not plot them as line segments. Because of this, a printed character will absolutely occupy its location on the plot, and any overlapping lines or points will not appear. If a character is printed on a patterned background, it will appear as if it were in a white box the size of the character. If true graphic characters are desired, they must be drawn by the application program as line segments. An additional implication of the printed characters is is their size relative to the graphic bit images that are sent to the printer. The graphics are printed in "strokes" that are one dot wide and seven dots high. Many printed characters are higher than seven dots especially those that have descenders. A printed character occupies a space that is seven dots high in the memory map. Thus, the top line of capital letters and the lower case descenders can overlap the graphics area above and below the nominal area of a printed character. The distance that the paper is advanced between lines is dictated by the height of a graphic "stroke", and is nearly always smaller than the normal single spaced distance. Printed lines must therefore be "double spaced" by specifying starting coordinates that are well separated vertically. This is a device dependent aspect, and a litle experimentation may be needed to achieve a good appearance. The last caution about the String command is that it MUST end with a carriage return character. This signals the end of the string, and without it, the program will continue to read the rest of the input file as part of the string until it either encounters a carriage return or the end of the file. The String and Text commands are the only commands with this potential, as all of the others have a definite number of bytes of data. All these caveats regarding the String command may sound a little gruesome, but it is really not much different from normal printing. If its limitations are too inhibiting, you must draw the characters instead of printing them. The Text command is output to the printer at a different time than any of the other commands. Its contents are sent immediately when it is read, whereas all other commands cause their actions to be stored in the memory map and only printed when an Output command is received. Thus, it cannot be used to send escape sequences to the printer in the middle of the 14 PLOT33.DOC PLOT User's Guide July 31, 1984 plot. Whether graphics commands (such as Draw, Fill, etc.) preceed or follow a Text command is immaterial, since these commands affect the memory map at the time that they are read and not the printer. What really matters is the position of a Text command relative to the Output commands in the file. Like the carriage return at the end of the String command, omitting the 0 byte at the end of the Text command can have disasterous consequences. The line spacing used in the Text command is different from the line spacing in the plot itself. The Output command sets the line spacing to be appropriate for graphics, and then resets it to the normal text single spacing when done. If a uniform appearance is required throughout, each Text command should contain the control characters or escape sequences required to set the line spacing to match that used in the plot itself. The final consequence of the line spacing is that the form length and the top of form position set in the printer will be wrong. The Output command will result in the printer being positioned at the bottom of the plot. This is to allow several plots to be printed sequentially as one continuous picture without any gaps between them. In order to advance to the top of the next page, the Text command must be used to issue the correct number of line feeds and then tell the printer to reset the top of form. The Output command does not erase the existing picture. Thus several copies of the same plot can be produced by repeating the command. In addition, this allows building up a picture in stages and printing it at each stage. To print one plot and start anew, you must Output the first, set the Color to white, Erase the page, and then set the Color back to the plotting color or pattern. The program does not automatically execute an Output command at the end-of-file or when Quitting. You must include an Output command to get any output at all. This philosophy is to accomodate those devices, like graphic terminals, that display the plot as it is being made, and only need an Output command for hardcopy. ERROR MESSAGES Disk is Full. Fatal error. No more room exists on the designated disk for writing the output file. Ensure that enough room is available and run PLOT again. Note that a different disk drive may be specified for the output file than the one used for the input file. If disk space is not available, run PLOT in the direct printing mode. 15 PLOT33.DOC PLOT User's Guide July 31, 1984 Any remnant of the output file may be printed, but the last record will probably contain an incomplete graphic output sequence. The printer will expect to receive additional bytes of graphics. This will produce a swath of dots with a random apperance, unless nulls are sent. In addition, the printer line spacing will remain set for the size of the graphic output. End of File. Fatal error. The end of the input file was encountered when PLOT attempted to read the next record. Add a Quit command for the normal termination of the program. File not Found. Fatal error. The file name given for the input file could not be found on the disk. Make sure that the spelling was correct, that the proper drive is specified in the file name, or the right drive is logged in as the default drive. No Directory Space Available. Fatal error. When PLOT tried to make a new output file, there was no more room in the disk directory. Erase a file, use a different diskette, or run PLOT in the direct printing mode, without making an output file. No File Specified. Fatal error. No input file name was given. PLOT will not prompt you for a file name, and there is no interactive mode for entering commands. Create an input file and try again. Undefined Command Character Encountered. Warning error. When PLOT attempted to interpret the ASCII letter portion of a command, a character was found which did not correspond to one of the established commands. PLOT ignores these characters and attempts to interpret the next character as a command. If a command does not have the proper number of data bytes, or if an extraneous character is inserted in the middle of the data portion of a command, the rest of the file will be out of sequence. Eventually, PLOT should get back in step with the command stream, but the overall effects are unpredictable. 16 PLOT33.DOC Application Programs July 31, 1984 PART 2. APPLICATION PROGRAMS BASIC PROGRAMS Several sample BASIC application programs are included as examples of possible implementations. BASIC was selected because it is almost universally available on CP/M systems and is a good language for quick one-of-a-kind plots. The first file is GRAPH.BAS. This file contains a plotting "library" of routines that give direct access to the low level commands used by PLOT. It also includes routines for scaling, and for drawing a complete graph with labeled axes. This file is used by the other programs. For each command that can be written to the .VEC file, there is a corresponding subroutine in GRAPH.BAS. To use these subroutines, first load the appropriate variables with the necessary data. For example, to draw a line segment, set X1 and Y1 to the starting coordinates, and set X2 and Y2 to the ending coordinates (remember that all coordinate values are between 0.0 and 1.0). Then GOSUB to the subroutine for that function, in this case the DRAW LINE SEGMENT routine, which is currently line 177. That's all that is required. Drawing axes or graphs is done the same way. Set the required values and GOSUB to the appropriate routine. All of the variables are documented in the REM statements at the top of the file, and a sample plot is depicted, followed by the settings that would produce it. The BASIC package automatically issues the commands to initialize the plotting area, so you can start plotting immediately when the the application part of the program starts executing. End the plot by calling the QUIT PLOTTING routine, at line 200. This will automatically add the commands to print the picture and close the file. Four statement functions are provided to make it easy to scale your data and compute plotting coordinates. Functions RX and RY will convert your "real world" values to X and Y coordinates for plotting. Functions UX and UY do just the opposite, converting coordinates on the page to the equivalent coordinates in the user defined plane. These functions require that you have established the margins, etc. as if you were plotting a graph. The default values for these settings will result in the conversion functions simply returning the same value as their input. Note that although the margins are useful for confining the plotting to one section of the page, no clipping is done to insure that all of the points plotted are indeed confined to that area. The next file is called TEST.BAS. It is a sample program that illustrates how to use the routines in GRAPH.BAS. To use TEST, enter BASIC-80, then LOAD "GRAPH.BAS". MERGE "TEST.BAS", and RUN it. The program can produce one or both of two plots. 17 PLOT33.DOC Application Programs July 31, 1984 The first is a polygon with all of the vertices connected to each other. This provides a good test of the line drawing capabilities of PLOT. The second plot is a graph of a damped sine wave. It illustrates nearly the complete range of PLOT's capabilities. Two programs, FILLS and UPLOAD, are an aid to experimenting with the patterns used to represent colors by PLOT33. Both are to be used with GRAPH.BAS in the same way as TEST.BAS. FILLS.BAS will produce an array of blocks showing the stipple patterns used to represent each "color". This makes an easy reference for future plotting. UPLOAD.BAS is similar to FILLS.BAS, except that before plotting the blocks, it Uploads new definitions for the patterns. The dither matrix used is like the example used in the previous section. The DATA statements of this program can be changed to experiment with different definitions and then incorporated into your applications. The final program is HANDPLOT.BAS. This is a general purpose program for manual plotting of data. The program uses a menu to display the BASIC plot package settings and allow them to be changed. Data can be typed in from the keyboard, saved in a disk file, and plotted. A limited number of symbols are avalilable for identifying data points and the points may optionally be connected by straight lines. Plotting scales and axes may be specified manually, or determined automatically from the range of values in the data. To make it easy to enter this program directly from the operating system command level, the routines in GRAPH.BAS have already been included and no MERGE operation is required. When HANDPLOT is first entered, it prompts the user for the name of the .VEC file, and for the dimensions of the data arrays. It then displays the menu. The menu is divided into three parts which are concerned with the data, the format of the graph, and the activities that can be performed by the program. It will look like this: 18 PLOT33.DOC Application Programs July 31, 1984 DATA SETTINGS No. of Variables 0 No. of Points 0 1> Data Entry KEYBOARD 2> Input File 3> Output File 4> X Variable 5> Y Variable 6> Symbol NONE 7> Connect points FALSE GRAPHIC SETTINGS X Axis 8> Min 0 9> Max 1 10> DeltaX 1 11> Label Chars 0 12> Divisions 0 Y Axis 13> Min 0 14> Max 1 15> DeltaY 1 16> Label Chars 0 17> Divisions 0 Margins 18> Left 0 19> Right 0 20> Bottom 0 21> Top 0 22> Color Value 127 23> Auto-Scale TRUE 24> Clear Plot Area First FALSE 25> Plot Output File Name JUNK.VEC ACTIONS 26> Read New Data 27> Edit data 28> Write Data 29> Plot Graph Axes 30> Plot Data 31> Title Plot 32> Advance Page 33> End Program Enter Number of SETTING or ACTION? From here, the program is operated by giving all of the required settings their correct value, and then selecting the desired action. This is done by entering the number printed next to the entry on the menu. Each setting will then prompt you for the data that it needs, if any. Several settings, such as Data Entry, only toggle between two values; in this case, KEYBOARD or DISK. The settings may be entered in any order. The first thing you will probably want to do is to enter some data to be plotted. If the data is to be read from a disk file instead, then enter 1 to toggle to DISK. You will then have to enter 2 to have the program prompt you for the input file's name. If the data is to be typed in, you are already to go. Now enter 26, and the program will start to read the disk file, or prompt you for the data. Once the data has been read in, it can be changed using option 27, Edit Data. The data can also be stored on disk using option 28. The output file name is handled in the same way as the input file name. Next you will want to indicate, using settings 4 and 5, which variables are to be used for the X and Y axes and whether the points should be plotted with one of the available symbols and/or connected with straight lines. If you are simply plotting a figure, you are ready to plot it using option 30, Plot Data. If you are plotting a graph, then all of the settings in the next section should be given their correct value before plotting the axes with option 29. Several sets of data can be plotted on the same graph, simply 19 PLOT33.DOC Application Programs July 31, 1984 by changing the X and Y variables and symbol and plotting the data again. Strings can be printed anywhere on the graph using Option 31. The DeltaX and Divisions settings are linked to each other, as are the DeltaY and Divisions setting for the Y axis. When the DeltaX or DeltaY setting is changed, the Divisions settings will be reset to indicate the number of divisions between the Min and Max values. If the Divisions setting is changed, the DeltaX or DeltaY setting will be calculated to give a good engineering value (multiples of 1, 2, 5, or 10), and the Divisions setting will reflect the revised number of divisions. Thus the value you enter into the Divisions setting will only be approximately the value that will ultimately result. If your engineering esthetics differ from the program's, you will have use the DeltaX and DeltaY settings. In a similar fashion, the Auto-Scale option will reset the Max and Min values to include the actual range of the data, and will use the Divsions settings to indicate the approximate number of labels to use for each axis. Advance Page will issue an Output command, clear the frame, and set the color to black for further plotting. The frames will be plotted one after the other with no space in between them. End Program will issue an Output command, close out all files and, return to the BASIC command level. Several plots can be put on one page, simply by adjusting the margin values between plotting each graph. If Clear Plot Area First, setting 24, is TRUE, the area inside the margins will be erased to white using the Fill command before the axes are plotted. This is useful for plotting one graph as an inset within a larger figure. The data file has a simple format, and can be made using your editor or another program, rather than laboriously entering all the data with HANDPLOT. The first record contains the number of variables, or fields, in the data records to follow. The next entries are names, up to 20 characters long, for each of the variables. These can be on the same line, separated by commas, or each may be on a record by itself. The data values are next. One value for each variable is read in turn, until the end of the file is encountered. The values on any given line are separated by commas, and each logical record (one value for each variable) may be spread over several lines of data. A sample data file is shown below: 3 <--- Three Variables X, Y, Z <--- Names for each variable 123, 4.56, 78.9 <--- First record: X=123, Y=4.56, Z=78.9 10.11,1213,1516.17 18.0, 19.0, 20 <--- Last record 20 PLOT33.DOC Application Programs July 31, 1984 TURBO PASCAL LIBRARY Turbo Pascal support is provided in the form of two "include" files and one test program. GRAF1.PAS contains the global declarations and lowest level routines. These provide direct access to the commands in the .VEC file. It also provides functions to do scale conversions. GRAF2.PAS uses these routines to provide the additional capability to plot graphic characters, plot axes, and do whole Cartesian grids. Each routine and its calling arguements are summarized in Table 2. Plotting is initiated with the procedure GRINIT. This procedure opens the disk file and initializes the plot, as discussed previously. It also initializes all global variables. The GRFINI routine writes the Output and Quit commands and closes the plot file. The following routines correspond directly to commands in the plot file: COLOR, ERASE, FILL, GPRINT, GSTRNG, GMOVE, POINT, SEGMNT, and VECTOR. Scaling is accomplished by using RWINDO to set the real world limits and SWINDO to map the real world limits to a specific area on the page. In the interest of execution speed, no clipping is done to ensure that the plotting is confined to either window. The functions SX and SY are used to convert from the real world coordinates to the normalized device, or "screen" coordinates. The reverse conversion is performed by RX and RY. GRAPH plots a complete grid given the range of data and area on the page. AXIS can be used to make special graphs and DXDY will calculate good engineering values for AXIS. The graphic characters and strings are plotted by GCHAR and GWRITE, respectively. 21 PLOT33.DOC Application Programs July 31, 1984 TABLE 2 TURBO PASCAL ROUTINES Name Inputs Outputs Function ----------------------------------------------------------------- color Sets color/pattern- sends Color command. code integer which selects color or pattern erase Clears page- sends Erase command. fill Fills trapezoidal area- sends Fill command. x1 normalized X coordinate for start of segment y1 normalized Y coordinate for start of segment x2 normalized X coordinate for end of segment y2 normalized Y coordinate for end of segment yf normalized Y coordinate for fill level gmove "Pen up" move to point- sends Move command. x normalized X coordinate y normalized Y coordinate gprint Prints picture- sends Output command. grinit Initializes .VEC file name string specifying name of .VEC file grfini Quits plotting, closes .VEC file gstrng Prints string- sends String command. x normalized X coordinate for first character y normalized Y coordinate for first character strng string to be plotted (no <CR> at end) point Plots point- sends Point command. x normalized X coordinate y normalized Y coordinate segmnt Plots line segment- sends Draw command. x1 normalized X coordinate for start of segment y1 normalized Y coordinate for start of segment x2 normalized X coordiante for end of segment y2 normalized Y coordinate for end of segment vector "Pen down" move- sends Increment command. x normalized X coordinate y normalized Y coordinate 22 PLOT33.DOC Application Programs July 31, 1984 TURBO PASCAL ROUTINES (CONT) Name Inputs Outputs Function -------------------------------------------------------------------- rwindo Establishes real window for scaling. xmini real world value at left side of window xmaxi real world value at right side of window ymini real world value at bottom side of window ymaxi real world value at top side of window swindo Establishes window on page for scaling. sxlti normalized X coordinate for left side sxrti normalized X coordinate for right side syboti normalized Y coordinate for bottom side sytopi normalized Y coordinate for top side sx Converts from real world to normalized coord rxi real world X coordinate sx normalized X coordinate sy Converts from real world to normalized coord ryi real world Y coordinate sy normalized Y coordinate rx Converts from normalized to real world coord sxi normalized X coordinate rx real world X coordinate ry Converts from normalized to real world coord syi normalized Y coordinate ry real world Y coordinate axis Plots axis at arbitrary orientation. r1 real world value at start of axis r2 real world value at end of axis dri real world increment for tic mark intervals sx1 normalized X coordinate for start of axis sy1 normalized Y coordinate for start of axis sx2 normalized X coordinate for end of axis sy2 normalized Y coordinate for end of axis ticlen normalized length of tic mark ticang CCW angle, in deg, from horiz. to tic mark lblnum integer number of char's in tic mark label lbldec integer number of char's after decimal pt. lblang CCW angle, in deg, from hor. to lbl X axis 23 PLOT33.DOC Application Programs July 31, 1984 TURBO PASCAL ROUTINES (CONT) Name Inputs Outputs Function -------------------------------------------------------------------- graph Plots complete Cartesian grid. xmini real world value at left side of graph xmaxi real world value at right side of graph nx integer- approximate number of X intervals ymini real world value at bottom side of graph ymaxi real world value at top side of graph ny integer- approximate number of Y intervals sxl normalized X coordinate for left side syb normalized Y coordinate for bottom side sxr normalized X coordinate for right side chset Sets size and orientation of graphic char's xsize normalized width of character ysize normalized width of character theta CCW angle, in deg, to char X axis from hor. gchar Plots graphic character. cx normalized X coord of bottom left corner cy normalized Y coord of bottom left corner charin ASCII character to be plotted gwrite Plots string of graphic characters. x normalized X coord of bottom left corner y normalized Y coord of bottom left corner chars string to be plotted nchar integer number of characters to be plot dxdy Calculates arguements for axis. x1 real world value at start of axis x2 real world value at end of axis nx integer- approximate number of intervals lblnum integer number of char's in tic mark label lbldec integer number of char's rt. of decimal posang Converts angle to range in 0 to 360 deg angle angle to be converted posang converted angle ticend Calculates values for ends of axis rmin raw real world value for start of axis rmax raw real world value for end of axis dr real world increment between intervals pr1 converted real world value at start of axis pr2 converted real world value at end of axis 24 PLOT33.DOC Application Programs July 31, 1984 TURBO PASCAL ROUTINES (CONT) Name Inputs Outputs Function ------------------------------------------------------------------ writecmd Buffered output of command string to file. cmd output string cmdlen integer number of characters to be output ne 4 buffout Flushes buffer. outunit file variable specifying output file outbuf string of characters forming buffer size number of characters to be sent out concat2 Concatenates portions of strings. strng1 first string to be concatenated nchar1 number of characters from strng1 strng2 second string to be concatenated nchar2 number of characters from strng2 strng3 string containing strng1 and strng2 FORTRAN LIBRARY A collection of FORTRAN routines is also included in GRAF.FOR. This file is the FORTRAN complement of the Turbo Pascal library. The input and output requirements for the FORTRAN routines are similar to those of the Pascal routines, and are fully documented in the comments within them. They communicate with each other through a labeled COMMON block. These routines have been run under Microsoft FORTRAN-80, but may require some modification for other compilers. 25 PLOT33.DOC Installation Instructions July 31, 1984 PART 3. INSTALLATION INSTRUCTIONS SIZING THE MEMORY MAP PLOT33 has been designed to keep the printer driver program small while providing the necessary primitive graphic operations for producing most pictures and plots. The reason for the simplicity of the of the command set is because the bit map for the picture is LARGE. It can use as much memory as your computer can give it. The size of the memory available for the bit-map, in conjunction with the resoultion of your printer, sets the size of the plot that can be made. The first step in installing the program is to decide how large the memory area for the bit map can be. The program itself occupies about 4K. Subtract this from the size of the TPA, and reduce the result by a little more to provide a conservative pad. The end result is the target size for the size of the memory map. Divide the target size by the printer resolution (dots per inch) in the horizontal direction, and divide again by the printer's vertical resolution. Multiply this result by 7 (7 dots are stored in each byte). You now have the number of square inches that can be plotted. Take the square root, and round down to a convenient size for each side of the plot. Now multiply the height of the plot by the printer's vertical resolution and round down to the nearest multiple of 7. The number of dots in the vertical direction should be a multiple of 7 because seven dots are stored in each byte of the memory map. PLOT33 has no provisions for handling a byte which is contains bits that are part in and part out of the map. You now have the total number of dots in the vertical direction, and dividing by 7 gives the number of lines that will be printed when the plot is made. Finally, divide the number of dots in the vertical direction by the vertical resolution to get the final size, in inches, of the plot. Multiply the final size of the plot by the horizontal resolution to get the final number of dots accross the page. The final map size is the product of the number of dots horizontally multiplied by the number of dots vertically, divided by 7. As an example, consider a 56K CP/M system. Subtracting 4K for the PLOT program and 100H for the start of the TPA: Target Size = 57344 - 4096 - 256 = 52992 bytes 26 PLOT33.DOC Installation Instructions July 31, 1984 The Epson MX-80 has 60 dots per inch horizontal and 72 dots per inch vertical resolution. Target Size/ Hor. Res./ Vert. Res. * 7 = 52992 / 60 / 72 * 7 = 85.9 sq.in ==> 9.2 in. per side Since the MX-80 can only plot 8 inches horizontally, the plot is limited by the physical limits of the printer, not by the memory avalilable. Calculating the vertical number of dots for an 8 inch plot: 8 * 72 = 576 dots ==> 82 lines @ 7 dots/line ==> 574 dots vertically Since only 2 dots were lost due to truncating to the nearest number of whole bytes, there is little point in trying to adjust the horizontal size to keep the plot square. The horizontal size is therefore: 8 * 60 = 480 dots The final map size is calculated as: 480 * 574 / 7 = 39360 PRINTER SETTINGS Once the plot has been sized, you are ready to modify the settings in the assembly language source file (PLOT33.ASM). This file is large, over 70K, so watch the disk space. If your editor cannot handle a file this large, the CP/M standard editor, ED, can. ED can also read from one disk and write to another, if necessary. For each type of printer there is an associated flag that controls the conditional assembly for that printer. Set the flag for your printer to TRUE and set the others to FALSE. If your printer is not listed, it may be similar to one that is. For example, the Gemini-10 operates identically (for the purposes of this program) with the Epson printers. The IDS printers use the same method as the Okidata printers for their graphics, and the NEC printers act the same as the C.Itoh printers. Next, the settings for your printer must be changed or verified. MAPSIZE should be set to the value calculated above. MAXX and MAXY must be set to 1 less than the number of dots in the horizontal and vertical directions. For the example above, MAPSIZE becomes 39360, MAXX is set to 479, and MAXY to 573. Once the settings are made for your printer, exit the editor and assemble the program using ASM or another 8080 assembler. Use LOAD to create the .COM file from the .HEX file, and the installation is finished. 27