There are several options for you to consider. I think the splicing method you mentioned is one that I had to do for a paper many decades ago and some call it re-basing (i.e., converting nominal data that transends multiple base years). I can still recall how I did it and will share it if the rest of this answer does not illuminate a better solution for you.

Now it is much easier with a variety of price indices available just to deflate whatever nominal series you wish; however, extra work is necessary for GDP because it combines all (or almost all) areas of economic activity.  

The actual adjustment of GDP is rather complex because it uses multiple price indices; for example, there is personal consumption deflator for GDP that suggests price movements for personal consumption expenditures may change at a different rate than capital goods or other elements not in consumption). In general, this approach transforms the nominal data int to real data (i.e., adjusted for inflation or deflation) for each of the components of GDP.  These more specific price indices should create more precise values for real GDP, which is also called constant dollar GDP.  From this we can get an implicit deflator for GDP, which is like a composite price index. Fortunately, real GDP is readily available and so it is rare for us to emulate the process from scratch for GDP.

If you are just trying to transform a a single data series (other than GDP or these complex composites) for inflation or deflation, the quantitative task is much easier. One can simply divide the nominal amount of the variable by decimal equivalent of the appropriate price index.  

Splicing or re-basing is rarely necessary with the robust inclusion of price indices available going back decades.  For example, look at Federal Reserve Economic Data (FRED) and you will find several dozen price indices that go back decades without going through the arduous task of re-basing. If you have not used FRED, consider adding the  InExcel Add-in, which makes it really easy to mach time periods to deflate nominal data.  In addition, the US Bureau of Labor Statistics (BLS) and Bureau of Economic Analysis at the US Department of Commerce (BEA) have hundreds of price indices that will make your task much easier and the result can be highly focused data transformations.

I hope this helps. Let me know if you want more details.

http://research.stlouisfed.org/fred2/

I suspect you are working with data from India where my familiarity with the data and institutions is limited. Regardless, I would urge caution if you are trying to convert Gross State Domestic Product (GSDP) if it originally was calculated using specific price indices for each sector.  For example, if a price index was used in each sector to deflate the nominal data, then you may introduce error or bias into your data by using the implicit price deflator because prices in the different sector likely changed by different amounts.

If not and you are focused on a single non GDP variable , then some relatively simple solutions are to do one of the following:

If you have both nominal (i.e., the observed values that have not had ANY adjustment for inflation or deflation) and real data (i.e., data that has been adjusted for inflation or deflation which is sometimes called constant  or adjusted data) for each time period (e.g., monthly quarterly, annually), then ignore the real data for a moment and divide the nominal data by the decimal equivalent appropriate price index. For example, if  you are working with a broad-based series and you think the Implicit Price Deflator for GDP in India is the appropriate price index to use, then go to a reputable source for the price index.  I obtained the following OECD data from FRED:

Year              IPD for GDP- India

01/01/2005      71.5

01/01/2006      75.6

01/01/2007      80.0

01/01/2008      87.2

01/01/2009      91.1

01/01/2010    100.0

01/01/2011    108.9

01/01/2012    116.7

01/01/2013    125.2

The decimal equivalents are easy:

Year                 IPD      Dec Eq

01/01/2005      71.5    .715

01/01/2006      75.6    .756 

01/01/2007      80.0    .800 

01/01/2008      87.2    .872 

01/01/2009      91.1    .911  

01/01/2010    100.0   1.000  

01/01/2011    108.9   1.089

01/01/2012    116.7   1.167 

01/01/2013    125.2   1.252 

Then you divide the nominal amounts (e.g., in this case I created a fictitious variable called nominal xx) by the decimal equivalents as follows:   So if your series called nomianal xx was  ₹1000 in 2013, then the real value of nominal xx in 2013 would be ₹798.6 in 2010 ₹; in other words, in constant 2010 ₹ the real value of xx in 2013 is ₹798.6 (see below)  

Time                 IPD        dec Eq    Nominal xx   Real xx where 2010=100"

01/01/2005     71.5         .715             400          559.8

01/01/2006     75.6         .756             475          627.9

01/01/2007     80.0         .800             550          687.4

01/01/2008     87.2         .872             625          716.6

01/01/2009     91.1         .911             700          768.0

01/01/2010   100.0        1.000            775          775.0

01/01/2011   108.9        1.089            850          780.8

01/01/2012   116.7        1.167            925          792.4

01/01/2013   125.2        1.252          1000          798.6  (i.e., 1000/1.252 = 798.6)

So what happens if you do not have the nominal values? If you have the price index that was used for the adjustment, then just reverse the process to convert back to nominal values and then you could divide these nominal values by the appropriate price index that spans all of your time periods.

Please be aware of some challenges here.  Among the reasons we re-base a price index is due to the changes that that take place in the quality of goods over time. For example, a  vehicle produced in 2015 is likely to be significantly different than one manufactured in 1975. The more often we re-base a price index the less this makes a difference because a vehicle manufactured in 2012 is probably roughly similar in comparison with on made in 2015. 

Finally, when one finds overlapping real values without nominal values, then converting (i.e., slicing in your terms) can be achieved by algebraically converting the overlapping values. I have a class soon and do not have time to illustrate it now.

I hope I did not misunderstand your question and hope this helps.

Rob

sir thank you very much for your suggestion. i will workout.