Lesson 02 · Rejecting a malformed core - Build a Semantic Version Parser and Range Matcher

Rejecting a malformed core

A parser is only trustworthy if it says no to garbage. Today you add the SemVer rules for a valid numeric core - exactly three parts, each a non-negative integer with no leading zeros - and return an error when they are broken.

The goal

Return an error for a core that has the wrong number of parts, a non-numeric part, or a leading zero.

Start here - the target

TO DO

Scenario: A malformed core is rejected

Givenmalformed version strings

Wheneach is parsed

ThenParse("1.2"), Parse("1.2.3.4"), Parse("1.2.x") and Parse("01.2.3") each return an error

AndParse("1.2.3") and Parse("0.0.0") still succeed, and Parse("1.02.3") returns an error

Background

The SemVer spec is strict about the core so that version strings are unambiguous. There must be exactly three dot-separated parts - 1.2 and 1.2.3.4 are not versions. Each part must be a non-negative integer, so 1.2.x is out. And crucially, a numeric part may not have a leading zero: 01.2.3 and 1.02.3 are invalid, because allowing them would make 01 and 1 two spellings of the same release. A lone 0 is fine - it is 0, not a leading zero.

That leading-zero rule is easy to miss and easy to get subtly wrong, so pin it now: a valid numeric identifier is either the single character 0 or a non-zero digit followed by any digits. Everything the rest of the project does assumes the core is trustworthy, so this small gate earns its place. Return an error rather than guessing at a fix - a parser that silently repairs input is worse than one that rejects it.

Make it work

// A valid numeric identifier: a single "0", or a non-zero digit followed by more digits.
// So "0" is ok, "10" is ok, but "01" and "" and "x" are not.
func validNum(s string) bool { /* ... */ }
// In Parse: require exactly three parts, each validNum, before converting.

CheckpointDONE

Parse now accepts only a well-formed numeric core and errors on the rest. Commit and stop here.