package org.shedlang.compiler.tokeniser;

import org.junit.jupiter.api.Test;
import org.zwobble.sourcetext.SourceText;

import java.util.List;

import static org.zwobble.precisely.AssertThat.assertThat;
import static org.zwobble.precisely.Matchers.equalTo;
import static org.zwobble.precisely.Matchers.isSequence;

public class RegexTokeniserTests {
    enum TokenType {
        UNKNOWN,
        KEYWORD,
        IDENTIFIER,
        WHITESPACE
    }

    @Test
    public void emptyStringIsTokenisedToNoTokens() {
        var tokeniser = new RegexTokeniser<>(TokenType.UNKNOWN, List.of());
        var sourceText = sourceText("");

        var result = tokeniser.tokenise(sourceText);

        assertThat(result, isSequence());
    }

    @Test
    public void regexIsAppliedToGenerateToken() {
        var tokeniser = new RegexTokeniser<>(TokenType.UNKNOWN, List.of(
            RegexTokeniser.rule(TokenType.IDENTIFIER, "[A-Za-z0-9]+")
        ));
        var sourceText = sourceText("count");

        var result = tokeniser.tokenise(sourceText);

        assertThat(result, isSequence(
            equalTo(new Token<>(
                sourceText.characterPosition(0).to(sourceText.characterPosition(5)),
                TokenType.IDENTIFIER
            ))
        ));
    }

    @Test
    public void firstMatchingRegexIsUsed() {
        var tokeniser = new RegexTokeniser<>(TokenType.UNKNOWN, List.of(
            RegexTokeniser.rule(TokenType.KEYWORD, "if"),
            RegexTokeniser.rule(TokenType.IDENTIFIER, "[A-Za-z0-9]+")
        ));
        var sourceText = sourceText("if");

        var result = tokeniser.tokenise(sourceText);

        assertThat(result, isSequence(
            equalTo(new Token<>(
                sourceText.characterPosition(0).to(sourceText.characterPosition(2)),
                TokenType.KEYWORD
            ))
        ));
    }

    @Test
    public void sequenceOfTokensCanBeGenerated() {
        var tokeniser = new RegexTokeniser<>(TokenType.UNKNOWN, List.of(
            RegexTokeniser.rule(TokenType.IDENTIFIER, "[A-Za-z0-9]+"),
            RegexTokeniser.rule(TokenType.WHITESPACE, "\\s")
        ));
        var sourceText = sourceText("one two");

        var result = tokeniser.tokenise(sourceText);

        assertThat(result, isSequence(
            equalTo(new Token<>(
                sourceText.characterPosition(0).to(sourceText.characterPosition(3)),
                TokenType.IDENTIFIER
            )),
            equalTo(new Token<>(
                sourceText.characterPosition(3).to(sourceText.characterPosition(4)),
                TokenType.WHITESPACE
            )),
            equalTo(new Token<>(
                sourceText.characterPosition(4).to(sourceText.characterPosition(7)),
                TokenType.IDENTIFIER
            ))
        ));
    }

    @Test
    public void whenNoRulesMatchThenUnknownTokensAreGenerated() {
        var tokeniser = new RegexTokeniser<>(TokenType.UNKNOWN, List.of(
            RegexTokeniser.rule(TokenType.IDENTIFIER, "[A-Za-z0-9]+")
        ));
        var sourceText = sourceText("!!a!");

        var result = tokeniser.tokenise(sourceText);

        assertThat(result, isSequence(
            equalTo(new Token<>(
                sourceText.characterPosition(0).to(sourceText.characterPosition(1)),
                TokenType.UNKNOWN
            )),
            equalTo(new Token<>(
                sourceText.characterPosition(1).to(sourceText.characterPosition(2)),
                TokenType.UNKNOWN
            )),
            equalTo(new Token<>(
                sourceText.characterPosition(2).to(sourceText.characterPosition(3)),
                TokenType.IDENTIFIER
            )),
            equalTo(new Token<>(
                sourceText.characterPosition(3).to(sourceText.characterPosition(4)),
                TokenType.UNKNOWN
            ))
        ));
    }

    private SourceText sourceText(String sourceContents) {
        return SourceText.fromString("<filename>", sourceContents);
    }
}