hi nripa,
writing a dummy issue.

Function findrootword is taking exorbitantly long, rather it kills the PC.
Have a look at it once again.
Not working properly.
It used to work reasonably well earlier.

@drdhaval2785

Is this under active development? I would like to reuse if possible.

Following issues are seen

1. upasargas = etree.parse('SL_upasargas.xml')
   throws an error. No such file exists. There is an SL_preverbs.txt
2. Cannot import this directly as a module simple mod would be to change the last line to

if __name__ == "__main__":
        convertfromfile('sanskritinput.txt','analysedoutput.txt')

If you're not intending to support this, I'll reimplement based on your logic.

यथात्रोक्तम् -
" एतदपि ननु pip क्षेत्रेऽभविष्यच् चेद् अज्ञासिषम्!​ सरलमेतत्। ततो यः कोऽपि सरलतया स्थापयित्वा तम् तन्त्रांशम् प्रायोक्ष्यत "s = sanskritmark.analyser(ot,split=False)" ​इत्यत्र यथा।​ अस्तु, issues क्षेत्रेऽनुवर्तिष्ये।"

अत्र यथापेक्षमहम् शक्नोमि सहकर्तुम्। स्वविचारः सूच्यताम्।

Running the script causes an error
Line 518, which is currently:
if i % 2 == 0 and i != len(dat): # Even members of datum are the words and odd members are word boundaries. Therefore, processing only even members.
should have:
... i != (len(dat)-1)
to skip the last element, which is empty. i never reaches len(dat)

(I could have forked, fixed and created a pull request, but it seems too much of an overhead for such a small fix. My apologies).

The script is painfully slow.
Results are good, but very slow to be of some use.
Check the time or memory leakage and fix.