Thanks for the reply!
Unfortunately copy pasting in a way which preserves the input from pdf's does create huge and slow files.
There are a few different ways to copy paste from pdf's,
1) selecting using the cursor and copy-pasting
2) selecting using the selection box and copy pasting
3) rendering to html, selecting using the cursor and copy-pasting
4) using the clip to tinderbox service
5) rendering the pdf to html and previewing the html note
I created a file(below) where I demonstrate the output from all of these ways.
so, 1) selecting text using the cursor and copying and pasting this, indeed does not create huge files, however it destroys almost all markup, formules and special characters. Also it detroys images and their location and is generally unstable.
2) Using the selection box and copying and pasting this creates an image, which creates huge files, takes away the ability to edit any text/formulas and generally looks very incongruent(however use an image adornment creates possibilities but severs all ties to the original content and has the same issues regarding editing the content of formulas. and it makes export almost impossible, since I will end up with a mish mash of picture cutout's from different pdf's with looks really unpleasant.
3) Copy pasting rendered html files creates the largest discrepancy between input and intended output.
4) clipping to tinderbox only copies pure text, no markup no images nothing, almost unreadable output.
5) Where rendering the output from PDF2HTMLEX faithfully recreates the PDF, using some scripts this could also be linked to the original file(references) and using some updating protocol I could change formulas using other software.
and the export actually looks nice.
Basically I want tinderbox to render MathML(tex would be even better, however this is difficult using HTML) in the note itself. The map view is really crucial in this case.
However, the html preview does not allow me to use the map view, since the text pane will only show me the html code, which leaves me none the wiser.
I have uploaded the files mentioned here: https://we.tl/Lbxa94LcmV
Since new users are not allowed to upload.
Would it be possible to store the html separately and let tinderbox store the rendered html as an image, which it would update when I change the html file?
And preferable use an an image adornment, but only when I open the actual file/parent node(such that tinderbox doesn't store bitmapped images of thousands of pdf files).
I looked through some of the action code, but I could not find an export command.