Not sure if it’s applicable to your requirements, but that’s the kind of thing I wrote my redirection class for on Clarionmag.
What is the output of fName.getValue()?
Might be useful to give an example how a filename looks like aka the whole img tag in html code.
1 Like
Hi Koen
Dirk’s earlier example was:
so based on that fName should hold ‘…/…/logo.jpg’
so after doing fName.replace(‘…/’,‘’) it should then hold just ‘logo.jpg’
To be more exact two examples
<img src=“…/…/…/…/Images/CC_LOG.JPG”
where the image is
C:\Clarion11|Images\CC_LOG.JPG
<img src=“…/…/…/MijnTaken/DATA_MYTASKS/RAPPORT/logo.jpg”
where the image is
C:\Clarion11\exes\MijnTaken\DATA_MYTASKS\RAPPORT\logo.jpg
So i need to replace / with \ (no problem) and then start from the root directory (where the html file is saved), count how many “…/” are found, remove them and then with the count# remove these subdirectories from the root to get the path to the image file. Any ideas?
I guess this api should help:
PathRelativePathToA function (shlwapi.h) - Win32 apps | Microsoft Learn
in DOS and Windows a parent directory is noted by two dots, so this might be worth a try:
fName.replace('…/','..\')
fName.replace('/','\')
and perhaps explicitly specify the path:
fName.replace('…/','..\')
fName.replace('/','\')
if fName.startsWith('..\') ! two dots indicates parent directory
fName.prepend(longPath() & '\')
elsif fName.startsWith('.\') ! one dot indicates current directory
fName.RemoveFromPosition(1,1) ! remove first character (or fName.top(1))
fName.prepend(longPath())
end
or alternatively
fName.replace('…/','..\')
fName.replace('/','\')
if fName.startsWith('..\') ! two dots indicates parent directory
fName.prepend(longPath() & '\')
elsif fName.startsWith('.\') ! one dot indicates current directory
fName.replace('.',longPath(),1) ! replace initial dot
end
I think that should work and be easier than splitting fName on ‘\’ and deleting lines then joining again - but let us know how you go…
cheers
Geoff R
but if you did want to do that it shouldn’t be too hard. Something like:
x long,auto
recs long,auto
code
fName.replace('…/','..\')
fName.replace('/','\')
if fName.startsWith('..\') ! two dots indicates parent directory
fName.prepend(longPath() & '\')
elsif fName.startsWith('.\') ! one dot indicates current directory
fName.replace('.',longPath(),1) ! replace initial dot
end
fName.split('\')
recs = fName.records()
loop x = 2 to recs
if fName.getLine(x) = '..' and x > 1
fName.deleteLine(x)
fName.deleteLine(x-1)
x -= 2
recs -= 2
end
end
fName.join('\')
as before, let us know how you go…
cheers
Geoff R
Why not just replace…
…/…/…/…/Images with C:\Clarion11|Images
and…
…/…/…/MijnTaken with C:\Clarion11\exes\MijnTaken
? If they are the only two paths?
is the 3 dots …/ correct? unix and dos/windows paths are 2 dots for the parent directory with *nix using a / as a seperator and dos a
So I would have thought …/…/file would have been the thing to use.
yes Geoff that is the question… I was giving a more general solution rather than “hard coding” in a path which may change down the track.
cheers
(the other) Geoff
no idea Sean - that is just what Dirk mentioned/specified.
I am replacing them with two dots to mean parent directory… as that is what Dirk’s examples implied.
fName.replace('…/','..\')
BTW stay safe from Cyclone Alfred heading your direction - hope it misses or peters out.
cheers
Geoff R
edit: I just realised the three dots on Dirk’s example is actually just one character (ellipsis ie. ‘<133>’) not three separate dot characters. Regardless, the replace statement above changes it to two dots.
1 Like
Hi all,
For some reason when i past …/ from html it is changed to …/ So i ment …/
The images can be inserted from any place from the local drive.
So i need something (pseudcode)
fName.replace('/','\')
if fName.startsWith('..\')
count# '..\'
Loop Count#
delete from right to left fname '\' to '\'
delete from left to right fname '..\'
end
fname=longpath()&'\'&fname
If fname not exists
Message('Local Image(s) will not show correctly when sending html E-mail, Change the image(s) to embed data in the html editor' )
end
3/5/25 Carl Edited post to change to CODE with ```
For some reason the two dots are changed to 3 but it’s about two dots and a backslash
Dirk, unless I have missed something I think you already have the code you need above - try it and see.
cheers
Geoff R
seems weird but there are options to tell it to not mess with the text
try “quoting” or “formatting” the code/text etc by putting three tildas or backticks before and after text. or you can try putting a > at the start of a single line
like this
for clarion code you can put three tildas (~) or backticks (`) followed by the word “clarion” ie. ```clarion.
for more info see
hth
As Geoff said wrap your code with 3 backticks ``` plus “clarion” and it will be CODE with color syntax highlighting for Clarion.
I edited your post. Below is screen cap of Editor on Left with markdown and Preview on Right that shows CODE:
Without that it looked like below:
One thing to consider in your file name work is you may have a Root Path like C:\
Thanks all,
It’s working now
good to hear that Dirk, glad you got it working and thanks for letting us know.
in case anyone with a similar requirement refers to this thread in the future, would you like to post your final working code?
cheers
Geoff R
Hi Geoff,
fname.SetValue(filename)
fName.replace('/','\')
fName.split('..\')
recs = fName.records()-1
Rootfolder=glo:uncmap&'\HTML_MERGEFILES' ! here are the html files
loop recs times
Rootfolder=removelastdir(Rootfolder)
end
loop x = 1 to recs
if fName.getLine(x) = '..\' and x > 1
fName.deleteLine(x)
fName.deleteLine(x-2)
x -= 3
recs -= 3
end
end!loop
fName.join('')
if img.loadfile(rootfolder&'\'&fname.GetVal())
if img.len() > 0
img.base64Encode(st:noWrap)
replaces=st.replace(tagContents[tagStart+1:tagEnd - 1],'data:image/'&fName.extensionOnly()&';base64,'&img.getValue())
end
end
1 Like
thanks Dirk for posting your code and, as I said earlier, I am glad you have it working for you.
However, having said that, I do have some concerns about your code.
in order to get rid of the “..\” or “move up a level to parent directory” I earlier posted this split/join code:
fName.split('\')
recs = fName.records()
loop x = 2 to recs
if fName.getLine(x) = '..' and x > 1
fName.deleteLine(x)
fName.deleteLine(x-1)
x -= 2
recs -= 2
end
end
fName.join('\')
in your code this is changed to:
fName.split('..\')
recs = fName.records()-1
Rootfolder=glo:uncmap&'\HTML_MERGEFILES' ! here are the html files
loop recs times
Rootfolder=removelastdir(Rootfolder)
end
loop x = 1 to recs
if fName.getLine(x) = '..\' and x > 1
fName.deleteLine(x)
fName.deleteLine(x-2)
x -= 3
recs -= 3
end
end!loop
fName.join('')
initially I was concerned that you are deleting line (x-2) not (x-1) and then decrementing by 3 not 2 when you have deleted two lines.
then I noticed it was moot/academic as that code in the loop is never performed.
consider this:
st.setValue(‘1,2,3,4,5’)
st.split(‘,’)
there are now 5 lines in the lines queue:
1
2
3
4
5
there is no way in this example that a comma is in a line as that was used as the “boundary” to split on.
similary you have changed the split to be on ‘..\’ rather than ‘\’ and so:
if fName.getLine(x) = '..\' and x > 1
can never be true.
the unknown at the moment is the earlier loop
loop recs times
Rootfolder=removelastdir(Rootfolder)
end
maybe that is doing the actual work???
can you please show the code for removelastdir() so we can understand what is going on?
thanks and cheers and hth
Geoff R