PureYAML/Tests/LibYAMLScannerFetchTests.swift

@testable import PureYAML
import Testing

/// End-to-end tokenization through the ported fetch loop. The headline case is
/// the original block-mapping bug: `- a:` followed by a same-column `b:` must
/// tokenize as siblings (one BLOCK-MAPPING-START), matching libyaml, while a
/// more-indented value nests (two BLOCK-MAPPING-STARTs).
@Suite("LibYAML Scanner Fetch")
struct LibYAMLScannerFetchTests {
    private typealias TokenType = PureYAML.Parsing.LibYAML.TokenType

    private func tokenize(_ input: String) throws -> [TokenType] {
        var scanner = PureYAML.Parsing.LibYAML.Scanner(input)
        return try scanner.scanAll().map(\.type)
    }

    @Test("a block mapping inside a sequence entry keeps same-column keys siblings")
    func test_sequenceEntryMappingSiblings() throws {
        // The original bug: `- a:` / `  b: c` is one mapping {a: null, b: c}.
        let tokens = try tokenize("- a:\n  b: c\n")
        #expect(tokens == [
            .streamStart,
            .blockSequenceStart,
            .blockEntry,
            .blockMappingStart,
            .key,
            .scalar(value: "a", style: .plain),
            .value,
            .key,
            .scalar(value: "b", style: .plain),
            .value,
            .scalar(value: "c", style: .plain),
            .blockEnd,
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a more-indented value nests a second mapping")
    func test_moreIndentedValueNests() throws {
        let tokens = try tokenize("a:\n  b: c\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "a", style: .plain),
            .value,
            .blockMappingStart,
            .key,
            .scalar(value: "b", style: .plain),
            .value,
            .scalar(value: "c", style: .plain),
            .blockEnd,
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a simple block mapping pair")
    func test_simplePair() throws {
        let tokens = try tokenize("key: value\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "key", style: .plain),
            .value,
            .scalar(value: "value", style: .plain),
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a multi-line plain scalar folds its line break to a space")
    func test_plainScalarFolding() throws {
        let tokens = try tokenize("a\nb\n")
        #expect(tokens == [
            .streamStart,
            .scalar(value: "a b", style: .plain),
            .streamEnd,
        ])
    }

    @Test("a flow sequence of plain scalars")
    func test_flowSequence() throws {
        let tokens = try tokenize("[a, b]\n")
        #expect(tokens == [
            .streamStart,
            .flowSequenceStart,
            .scalar(value: "a", style: .plain),
            .flowEntry,
            .scalar(value: "b", style: .plain),
            .flowSequenceEnd,
            .streamEnd,
        ])
    }

    @Test("a flow mapping with a key and value")
    func test_flowMapping() throws {
        let tokens = try tokenize("{k: v}\n")
        #expect(tokens == [
            .streamStart,
            .flowMappingStart,
            .key,
            .scalar(value: "k", style: .plain),
            .value,
            .scalar(value: "v", style: .plain),
            .flowMappingEnd,
            .streamEnd,
        ])
    }

    @Test("single-quoted scalar with an escaped quote")
    func test_singleQuoted() throws {
        let tokens = try tokenize("'it''s here'\n")
        #expect(tokens == [
            .streamStart,
            .scalar(value: "it's here", style: .singleQuoted),
            .streamEnd,
        ])
    }

    @Test("double-quoted scalar decodes escapes")
    func test_doubleQuotedEscapes() throws {
        let tokens = try tokenize("\"a\\nb\\tc\\u263A\"\n")
        #expect(tokens == [
            .streamStart,
            .scalar(value: "a\nb\tc\u{263A}", style: .doubleQuoted),
            .streamEnd,
        ])
    }

    @Test("a multi-line quoted scalar folds line breaks")
    func test_quotedFolding() throws {
        let tokens = try tokenize("\"a\nb\"\n")
        #expect(tokens == [
            .streamStart,
            .scalar(value: "a b", style: .doubleQuoted),
            .streamEnd,
        ])
    }

    @Test("an anchored scalar value and an alias")
    func test_anchorAndAlias() throws {
        let tokens = try tokenize("a: &x 1\nb: *x\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "a", style: .plain),
            .value,
            .anchor(name: "x"),
            .scalar(value: "1", style: .plain),
            .key,
            .scalar(value: "b", style: .plain),
            .value,
            .alias(name: "x"),
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a literal block scalar keeps line breaks and clips the trailing newline")
    func test_literalBlockScalar() throws {
        let tokens = try tokenize("key: |\n  a\n  b\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "key", style: .plain),
            .value,
            .scalar(value: "a\nb\n", style: .literal),
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a folded block scalar joins lines with a space")
    func test_foldedBlockScalar() throws {
        let tokens = try tokenize("key: >\n  a\n  b\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "key", style: .plain),
            .value,
            .scalar(value: "a b\n", style: .folded),
            .blockEnd,
            .streamEnd,
        ])
    }

    @Test("a strip chomping indicator drops the trailing newline")
    func test_literalStripChomping() throws {
        let tokens = try tokenize("key: |-\n  a\n")
        #expect(tokens == [
            .streamStart,
            .blockMappingStart,
            .key,
            .scalar(value: "key", style: .plain),
            .value,
            .scalar(value: "a", style: .literal),
            .blockEnd,
            .streamEnd,
        ])
    }
}