Comparison between Bottom-up and Top-down parsing techniques

Aspect	Top-Down Parser	Bottom-Up Parser
Definition	Starts parsing from the start symbol and tries to derive the input string.	Starts from the input string and reduces it to the start symbol.
Parsing Direction	Left to right on the input; derivation is leftmost.	Left to right on input; derivation is rightmost in reverse.
Parse Tree Construction	From root to leaves (top-down).	From leaves to root (bottom-up).
Grammar Used	LL(k) grammar (subset of CFG).	LR(k) grammar (superset of LL grammars).
Backtracking	May require backtracking (unless predictive).	No backtracking required.
Error Detection	Detects errors early (near the root of the parse tree).	Detects errors late (closer to the leaves of the tree).
Complexity of Implementation	Easier to implement manually (e.g., recursive descent).	Harder to implement manually; generally uses tools.
Examples	Recursive Descent, Predictive Parser (LL Parser)	Shift-Reduce, LR, SLR, LALR parsers

Example

For an expression like a + b * c

• Top-Down Parser:
  Starts from expression rule and predicts what production to apply based on lookahead symbols.

• Bottom-Up Parser:
  Starts by shifting a, then +, then reduces b * c as a term, then combines with a.

Summary

• Use Top-down when grammar is simple and recursive.

• Use Bottom-up for more powerful grammar support and better error handling.