Brain has stopped - help pleas - ASPImage

[bigmentor]

Good Morning Folks

I have bought and own a copy of ASPImage.
Installed on my Internet server.
Works like a Dream writing and saving image files containing text.


My brain seems to have stopped working.
I have tried to write a script that I expect cannot be more than ten lines long.
The towel has been thrown in !!


Then I thought some kind person in this forum could produce it in five minutes.


So please will anyone help


The task is simple.

I have a variable called
hColor
It contains (for example) "FFFF00" (for Yellow) (Note - no hash)



I want to create and save a "one pixel" by "one pixel" image of the color held in 'hColor'

The image file name is (in this case)
FFFF00_dot.gif


I will be forever in your debt
Thanks (in anticipation)
Pete (Northolt UK)

[jhardman]

Pete,

I've been thinking about this problem for a while, unfortunately, the best solution I can come up with is to save a single-pixel-image using paint or some such and then opening the bytes to figure what they say. I know for example that the beginning of the file says "GIF89a" and then the dimensions of the picture, but after that the data is compressed, I understand how the compression works, at least the theory behind it, but I don't think I could count backwards enough to read the data file. I had an old example of how to write a small 15 X 15 gif image, and sure enough it starts out "71 73 70 56 57 97 15 0 15 0", but after that I'm lost.

To make matters worse, when I tried to make a one pixel image in paint and then look at the bytes, the image starts out the same "71 73 70 56 55 97 1 0 1 0" as expected, but the file is actually much longer than the 15 X 15 image, ( I suspect that paint pads the file a bit, and it uses the old '87a' version, notice the difference of the 5th character) and I can see no pattern, either declaration of the compression library or anything. If you could elucidate the idea of what the file should look like, I could show you how to save the file, it's just coming up with the right sequence of numbers that is beyond me.

Jared

[jhardman]

after thinking some more, I muscled through it. Here's the solution, a 32-byte one pixel image:

<%
option explicit
 
'set gifHeader
dim gifHeader, fileBytes(31), x
 
gifHeader = "GIF87a" 'there's also a version "89a", definition and syntax is
    ' almost identical, I'm not sure what the differences are.
 
x = 0
do until len(gifHeader) = 0
    fileBytes(x) = AscB(left(gifHeader, 1))
    gifHeader = right(gifHeader, len(gifHeader)-1)
    x = x + 1
loop
 
'set image size
dim imgWidth, imgHeight
imgWidth = 1
imgHeight = 1
 
if imgWidth > 255 then
    'need to shorten, first send "lsb"
    'i'm glad I didn't need to calculate this.
else
    fileBytes(x) = imgWidth
    x = x + 1
    fileBytes(x) = 0
    x = x + 1
end if
 
if imgHeight > 255 then
    'need to shorten, first send "lsb"
    'i'm glad I didn't need to calculate this.
else
    fileBytes(x) = imgheight
    x = x + 1
    fileBytes(x) = 0
    x = x + 1
end if
 
'set color map, color resolution, and bits per pixel
dim M, CR, bpp
 
M = 1
CR = 0
bpp = 0
 
fileBytes(x) = bpp
fileBytes(x) = fileBytes(x) + (CR*16)
fileBytes(x) = fileBytes(x) + (M*128)
x = x + 1
 
'set background color
fileBytes(x) = 0 'first color in the dictionary
x = x + 1
 
'end screen descriptor
fileBytes(x) = 0
x = x + 1
 
'start color dictionary
fileBytes(x) = 255 'red index of first color
x = x + 1
fileBytes(x) = 255 'green index of first color
x = x + 1
fileBytes(x) = 0 'blue index of first color
x = x + 1
 
fileBytes(x) = 255 'red index of 2nd color
x = x + 1
fileBytes(x) = 255 'green index of 2nd color
x = x + 1
fileBytes(x) = 0 'blue index of 2nd color
x = x + 1
 
'start first image
fileBytes(x) = 44 ' comma is the first character of an image
x = x + 1
 
'positioning of image within total file
fileBytes(x) = 0 ' start in the corner, horiz (lsb format)
x = x + 1
fileBytes(x) = 0 ' more significant byte is 0 (number < 255)
x = x + 1
fileBytes(x) = 0 ' start in the corner, vert
x = x + 1
fileBytes(x) = 0 ' more significant byte
x = x + 1
 
'dimensions of this image
fileBytes(x) = 1 ' image width (lsb format)
x = x + 1
fileBytes(x) = 0 ' more significant byte is 0 (number < 255)
x = x + 1
fileBytes(x) = 1 ' image height
x = x + 1
fileBytes(x) = 0 ' more significant byte
x = x + 1
 
'use global color map
fileBytes(x) = 0 ' no local color map, just use global
x = x + 1
 
'start raster data, I'm still not sure about compression, but for one pixel:
fileBytes(x) = 0 'first color in color map
x = x + 1
 
fileBytes(x) = 1
x = x + 1
 
'end the file
fileBytes(x) = 59 'last byte is a semi colon, ends each image
 
for each x in fileBytes
    if x <> "" then
        response.write x & " "
    end if
next 
response.write "<br>" & vbNewLine
 
 
dim output
for each x in fileBytes
    if x <> "" then
        response.write chr(x) & " "
        output = output & chr(x)
    end if
next
 
'save file
dim objFSO, objTXT, filePath, fileName
filePath = request.serverVariables("appl_physical_path") & "temp"
fileName = "ffff00.gif"
filePath = filePath & "\" & fileName
response.write "<!-- path: " & filePath & " -->" & vbNewLine
set objFSO = server.createObject("scripting.fileSystemObject")
set objTXT = objFSO.openTextFile (filePath, 2, True)
objTXT.write output
response.write "<br>(File saved as: " & filePath & ")<br>" & vbNewLine
response.write "<img src='" & fileName & "'><br>" & vbNewLine
 
'spacer gif: 71 73 70 56 55 97 1 0 1 0 128 0 0 255 255 0 0 0 0 44 0 0
' 0 0 1 0 1 0 0 0 0 59 
'yellow gif: 71 73 70 56 55 97 1 0 1 0 128 0 0 255 255 0 255 255 0 44
' 0 0 0 0 1 0 1 0 0 0 1 59
%>
<div style="background-color: blue;">   
<img src='<%= fileName %>' height='30' width='30'></div>

There are a few tricky parts. First, the image size is defined more than once, because a gif file can hold more than one image (for animations, etc.). The first time is the total, all subsequent times are for individual pictures.

Second, the color map, color resolution and bits per pixel is combined into one byte. That makes that byte a little tricky to calculate. You can also define a different color map for each image, I just defined a single "global" color map and on the actual image left it blank.

Third, you need at least two colors defined. If you only use a single color, it doesn't matter what the second color is.

Fourth, image size and image placement numbers are in 2 byte LSB format, which means if the number is less than 255, you just use the first byte and leave the second byte at 0. (that makes a 1 x 1 pixel picture say "1 0 1 0") If the number is greater than 255 you need to use both bytes, and I'm glad I didn't need to calculate what the second byte was.

Fifth, (the really tricky one) is the compression formula. Fortunately this only applies to the raster data, so if you only have one pixel, the raster data should say (0 1) for the first color in your color map or (0 0) for transparent.

Jared