tried for sanskrit corpus but output was not manageable because sanskrit has riches in inflections, sandhi, samaas and all. Big sandhi words were majorly problematic. Therefore i changed to hindi corpus. Sorry, i did not inform you for this and also for selection of list30.txt, i didn't check email and started working with list30.txt!

Corpus selected Hi_blogs1.txt
Unique Words file: wordlist.hi
segmented output: segmentedll.hi
300 words file:list30.txt 
from these words present in finalsegmentedall.hi : segUniqs.txt ----- after removing english words also urdu and less frequent words.
getting 300 words of list30 in: segUniq300.hi
manually segmented words file: manualSeg.txt
matching segmented words file: matched1.txt
this gives exact accuracy of (92/134*100): 68%
non matching from segmented words file : nonmatchingMn.txt
non matching from manually segmented words file : NMmanSeg.txt

command for tokenization:

cat Hi_Blogs.txt|sed 's/[[:punct:]]//g;s/[0-9A-Za-z]//g;s/[०-९]//g'|sed 's/\t//g;s/\///g;s/ /\n/g'|grep -e "^$" -v|sort|uniq -c|sed -e 's/^[ \t]*//' -e 's/[ \t]*$//' > wordlist.hi

for searching the words of list30.txt command used grep -'....word sequence...' "wordlist.hi" > "seg300.hi" 
unique words from seg300.hi in: segUniq300.hi

everything done with terminal commands.