Update index.md

Author: pythonJaRvis

index.md

@@ -20,8 +20,7 @@ scale to large programs. In our preliminary experiments, it ran out of memory or
 2,000 lines of code. This limitation stems from the costly global fixed-point iterations required during analysis. In addition, PyCG is
 flow-insensitive and does not fully support Python’s dynamic features, which further limits its accuracy.
 
-
-    To overcome these drawbacks, we propose a scalable and precise approach for constructing application-centered call graphs for
+To overcome these drawbacks, we propose a scalable and precise approach for constructing application-centered call graphs for
 Python programs, and implement it as a prototype Jarvis. Jarvis maintains a type graph (i.e., type relations of program identifiers) for
 each function in a program to allow type inference.x Taking one function as an input, Jarvis generates the call graph on-the-fly,
 where flowsensitive intraprocedural analysis and interprocedural analysis are conducted in turn and strong updates are conducted.