Say you have an HTML file which contains more than 100,000 characters. How do you extract all the Phone Numbers from it

Given a website you have to find number of hits in the last second, last minute and last hour.

Given a signature, compute the lexicographically smallest permutation of [1,2,....n].

You are given an array of n elements [1,2,....n]. For example {3,2,1,6,7,4,5}.
Now we create a signature of this array by comparing every consecutive pir of elements. If they increase, write I else write D. For example for the above array, the signature would be "DDIIDI". The signature thus has a length of N-1. Now the question is given a signature, compute the lexicographically smallest permutation of [1,2,....n]. Write the below function in language of your choice.   vector* FindPermute(const string& signature);

This is another good one from google easy though :)

Find All subsubsequence of size r in given array of length n

Given a positive integer N, arrange the integers 1 to N such the position of the average of two numbers (if present) falls outside the two numbers. For example, for a[i] and a[j], if b = (a[i] + a[j])/2, then if a[k] = b, k < i or k > j. Hint :- If the average of two numbers is not an integer then we can assume that they do not have an average.

One possible solution  i can think of is  divide the array in two parts , odd and even , sort them and use be.low algo , like if we have the array say N=10 and array is 3 10 2 7 5 3 4 6 9 8 1 after dividing and sorting
array will be 1 3 5 7 9 2 4 6 8 10
now calculate length of even and odd list ,if length is odd,then divide it into 2 equal parts and align alone one middle value like
1 3 5 7 9 2 4 6 8 10
now swap all values of two list except first value based on middle if each sub-list
1 7 5 3 9 2 8 6 4 10

i have not coded it yet so feel free to comment if it fails for any test case :)

How would you implement trending topics for twitter. Think at large scale as well.

Design a system where you can reutrn top 20 queries made in last 24 hours to users. Think on the scale of Google and Yahoo. How would you store data. What will be your data structures, algorithm to get that data.Describe your assumptions etc. For simplicity, you can assume that every web server create a log file with query and timestamp.

A legend among a group of n people is a person who is top rated in all respects.The task is to identify a legend by asking a single question of the form "who deserves it?" Design an efficient algorithm to identify a legend or determine if the group has no such person.How many questions does your algorithm need in the worst case??

Given a string array， return all groups of strings that are anagrams

Convert Prefix expression to postfix expression