linter和验证器有什么区别？

Question

I'm working on organizing some things and wanted to combined these two sets. In fact I don't know of any other validator besides the W3 - http://validator.w3.org/

Linters - jslint | jshint | eslint
Validators - w3

Is there any difference between these 2 types of tools. Both just look at the code statically and test it against a set syntax correct ?

Answer 1

A lint program checks for suspicious or dangerous code, and checks the code against styling guidelines. A linter must be able to parse the code, so it also validates against a language spec. Therefore, to a degree, a linter is a validator plus extra stuff.

A validator program verifies the code adheres to language specifications. It doesn't care about style or logic.

Examples of things caught by a linter, but not by a validator:

• Division by zero (its valid syntax, so it passes the validator).
• Using a variable before its been declared or populated (again, valid syntax).
• Detecting out-of-bounds problems with array indexing (again, valid syntax).
• Style issues (same thing)
• Code like if (x = 3) { } is valid syntax in many languages, and will pass a validator. A linter will say "hey, what are you doing, bro? this doesn't look right to me." (an example of suspicious code)

Answer 2

A lint looks for syntax and other problems. Validators look for compliance with particular specifications.

Answer 3

Linters make sure that code is properly formed and syntaxed as to prevent bugs - usually in realtime or upon save/compile.

Validators (in the case of W3) make sure that the markup adheres to the rules defined for that specific doctype.

There is overlap between the two (as they are both intended to help preempt bugs, but a Linter is typically far more exhaustive.)