Add set text analyser for MySQL lexer

Author: gandarez

Diff for: lexers/lexers_test.go

@@ -5,6 +5,7 @@ import (
 	"io/ioutil"
 	"os"
 	"path/filepath"
+	"strconv"
 	"strings"
 	"testing"
 
@@ -79,6 +80,11 @@ func TestLexers(t *testing.T) {
 	assert.NoError(t, err)
 
 	for _, file := range files {
+		// skip text analysis test files
+		if file.Name() == "analysis" {
+			continue
+		}
+
 		if file.IsDir() {
 			dirname := filepath.Join("testdata", file.Name())
 			lexer := lexers.Get(file.Name())
@@ -117,3 +123,52 @@ func TestLexers(t *testing.T) {
 		}
 	}
 }
+
+func FileTestAnalysis(t *testing.T, lexer chroma.Lexer, actualFilepath, expectedFilepath string) {
+	t.Helper()
+	t.Run(lexer.Config().Name+"/"+actualFilepath, func(t *testing.T) {
+		expectedData, err := ioutil.ReadFile(expectedFilepath)
+		assert.NoError(t, err)
+
+		analyser, ok := lexer.(chroma.Analyser)
+		assert.True(t, ok, "lexer %q does not set analyser", lexer.Config().Name)
+
+		data, err := ioutil.ReadFile(actualFilepath)
+		assert.NoError(t, err)
+
+		actual := analyser.AnalyseText(string(data))
+
+		if os.Getenv("RECORD") == "true" {
+			// Update the expected file with the generated output of this lexer
+			f, err := os.Create(expectedFilepath)
+			defer f.Close() // nolint: gosec
+			assert.NoError(t, err)
+
+			_, err = f.WriteString(strconv.FormatFloat(float64(actual), 'f', -1, 32))
+			assert.NoError(t, err)
+		} else {
+			expected, err := strconv.ParseFloat(strings.TrimSpace(string(expectedData)), 32)
+			assert.NoError(t, err)
+
+			assert.Equal(t, float32(expected), actual)
+		}
+	})
+}
+
+func TestLexersTextAnalyser(t *testing.T) {
+	files, err := filepath.Glob("testdata/analysis/*.actual")
+	assert.NoError(t, err)
+
+	for _, actualFilepath := range files {
+		filename := filepath.Base(actualFilepath)
+		baseFilename := strings.TrimSuffix(filename, filepath.Ext(filename))
+		lexerName := strings.Split(baseFilename, ".")[0]
+
+		lexer := lexers.Get(lexerName)
+		assert.NotNil(t, lexer, "no lexer found for name %q", lexerName)
+
+		expectedFilepath := "testdata/analysis/" + baseFilename + ".expected"
+
+		FileTestAnalysis(t, lexer, actualFilepath, expectedFilepath)
+	}
+}

Diff for: lexers/m/mysql.go

@@ -1,10 +1,17 @@
 package m
 
 import (
+	"regexp"
+
 	. "github.com/alecthomas/chroma" // nolint
 	"github.com/alecthomas/chroma/lexers/internal"
 )
 
+var (
+	mysqlAnalyserNameBetweenBacktickRe = regexp.MustCompile("`[a-zA-Z_]\\w*`")
+	mysqlAnalyserNameBetweenBracketRe  = regexp.MustCompile(`\[[a-zA-Z_]\w*\]`)
+)
+
 // MySQL lexer.
 var MySQL = internal.Register(MustNewLazyLexer(
 	&Config{
@@ -16,7 +23,25 @@ var MySQL = internal.Register(MustNewLazyLexer(
 		CaseInsensitive: true,
 	},
 	mySQLRules,
-))
+).SetAnalyser(func(text string) float32 {
+	nameBetweenBacktickCount := len(mysqlAnalyserNameBetweenBacktickRe.FindAllString(text, -1))
+	nameBetweenBracketCount := len(mysqlAnalyserNameBetweenBracketRe.FindAllString(text, -1))
+
+	var result float32
+
+	// Same logic as above in the TSQL analysis.
+	dialectNameCount := nameBetweenBacktickCount + nameBetweenBracketCount
+	if dialectNameCount >= 1 && nameBetweenBacktickCount >= (2*nameBetweenBracketCount) {
+		// Found at least twice as many `name` as [name].
+		result += 0.5
+	} else if nameBetweenBacktickCount > nameBetweenBracketCount {
+		result += 0.2
+	} else if nameBetweenBacktickCount > 0 {
+		result += 0.1
+	}
+
+	return result
+}))
 
 func mySQLRules() Rules {
 	return Rules{

Diff for: lexers/testdata/analysis/mysql.backtick.actual

@@ -0,0 +1 @@
+CREATE TABLE `my_table` (id INT);

Diff for: lexers/testdata/analysis/mysql.backtick.expected

@@ -0,0 +1 @@
+0.5