Add tests for parse errors, will need to be updated as parser is developed (#75)

* Add tests for parse errors, will need to be updated as parser is developed

* go get -u ./...

Author: FollowTheProcess

.rumdl.toml

@@ -30,16 +30,11 @@ tasks:
 
   fmt:
     desc: Run go fmt on all source files
-    preconditions:
-      - sh: command -v rumdl
-        msg: rumdl not installed, see https://github.com/rvben/rumdl#installation
     sources:
       - "**/*.go"
       - .golangci.yml
-      - "**/*.md"
     cmds:
       - golangci-lint fmt ./...
-      - rumdl fmt --fix .
 
   test:
     desc: Run the test suite
@@ -100,17 +95,12 @@ tasks:
     sources:
       - "**/*.go"
       - .golangci.yml
-      - "**/*.md"
     preconditions:
-      - sh: command -v rumdl
-        msg: rumdl not installed, see https://github.com/rvben/rumdl#installation
-
       - sh: command -v typos
         msg: requires typos-cli, run `brew install typos-cli`
     cmds:
       - golangci-lint run --fix
       - typos
-      - rumdl check --fix .
 
   doc:
     desc: Render the pkg docs locally

Taskfile.yml

@@ -28,7 +28,7 @@ require (
 	github.com/charmbracelet/lipgloss v1.1.0 // indirect
 	github.com/charmbracelet/x/ansi v0.11.1 // indirect
 	github.com/charmbracelet/x/cellbuf v0.0.14 // indirect
-	github.com/charmbracelet/x/exp/strings v0.0.0-20251117104201-863d1fef9cfb // indirect
+	github.com/charmbracelet/x/exp/strings v0.0.0-20251118172736-77d017256798 // indirect
 	github.com/charmbracelet/x/term v0.2.2 // indirect
 	github.com/clipperhouse/displaywidth v0.6.0 // indirect
 	github.com/clipperhouse/stringish v0.1.1 // indirect

go.mod

@@ -30,8 +30,8 @@ github.com/charmbracelet/x/errors v0.0.0-20240508181413-e8d8b6e2de86 h1:JSt3B+U9
 github.com/charmbracelet/x/errors v0.0.0-20240508181413-e8d8b6e2de86/go.mod h1:2P0UgXMEa6TsToMSuFqKFQR+fZTO9CNGUNokkPatT/0=
 github.com/charmbracelet/x/exp/golden v0.0.0-20241011142426-46044092ad91 h1:payRxjMjKgx2PaCWLZ4p3ro9y97+TVLZNaRZgJwSVDQ=
 github.com/charmbracelet/x/exp/golden v0.0.0-20241011142426-46044092ad91/go.mod h1:wDlXFlCrmJ8J+swcL/MnGUuYnqgQdW9rhSD61oNMb6U=
-github.com/charmbracelet/x/exp/strings v0.0.0-20251117104201-863d1fef9cfb h1:bKbwRftwpZ3sN1karNah6eIjlaMeE1FKv7INGRFSv1g=
-github.com/charmbracelet/x/exp/strings v0.0.0-20251117104201-863d1fef9cfb/go.mod h1:/ehtMPNh9K4odGFkqYJKpIYyePhdp1hLBRvyY4bWkH8=
+github.com/charmbracelet/x/exp/strings v0.0.0-20251118172736-77d017256798 h1:g0RVaqkUdTikWLqrBdk2ZvJ9oTQOS0HZlYjYE8Tu7yg=
+github.com/charmbracelet/x/exp/strings v0.0.0-20251118172736-77d017256798/go.mod h1:/ehtMPNh9K4odGFkqYJKpIYyePhdp1hLBRvyY4bWkH8=
 github.com/charmbracelet/x/term v0.2.2 h1:xVRT/S2ZcKdhhOuSP4t5cLi5o+JxklsoEObBSgfgZRk=
 github.com/charmbracelet/x/term v0.2.2/go.mod h1:kF8CY5RddLWrsgVwpw4kAa6TESp6EB5y3uxGLeCqzAI=
 github.com/charmbracelet/x/termios v0.1.1 h1:o3Q2bT8eqzGnGPOYheoYS8eEleT5ZVNYNy8JawjaNZY=

go.sum

@@ -2,15 +2,22 @@
 package parser_test
 
 import (
+	"bytes"
 	"flag"
+	"fmt"
 	"os"
 	"path/filepath"
+	"slices"
+	"strings"
+	"sync"
 	"testing"
 
+	"github.com/rogpeppe/go-internal/txtar"
 	"go.followtheprocess.codes/snapshot"
 	"go.followtheprocess.codes/test"
 	"go.followtheprocess.codes/zap/internal/syntax"
 	"go.followtheprocess.codes/zap/internal/syntax/parser/v2"
+	"go.uber.org/goleak"
 )
 
 var (
@@ -19,7 +26,7 @@ var (
 )
 
 func TestParse(t *testing.T) {
-	pattern := filepath.Join("testdata", "src", "*.http")
+	pattern := filepath.Join("testdata", "valid", "*.http")
 	files, err := filepath.Glob(pattern)
 	test.Ok(t, err)
 
@@ -48,6 +55,72 @@ func TestParse(t *testing.T) {
 	}
 }
 
+// TestInvalid is the primary test for invalid syntax. It does much the same as TestParse
+// but instead of failing tests if a syntax error is encounter, it fails if there is not any syntax errors.
+//
+// Additionally, the errors are compared against a reference.
+func TestInvalid(t *testing.T) {
+	// Force colour for diffs but only locally
+	test.ColorEnabled(os.Getenv("CI") == "")
+
+	pattern := filepath.Join("testdata", "invalid", "*.txtar")
+	files, err := filepath.Glob(pattern)
+	test.Ok(t, err)
+
+	for _, file := range files {
+		name := filepath.Base(file)
+		t.Run(name, func(t *testing.T) {
+			defer goleak.VerifyNone(t)
+
+			archive, err := txtar.ParseFile(file)
+			test.Ok(t, err)
+
+			test.Equal(
+				t,
+				len(archive.Files),
+				2,
+				test.Context("%s should contain 2 files, got %d", file, len(archive.Files)),
+			)
+			test.Equal(
+				t,
+				archive.Files[0].Name,
+				"src.http",
+				test.Context("first file should be named 'src.http', got %q", archive.Files[0].Name),
+			)
+			test.Equal(
+				t,
+				archive.Files[1].Name,
+				"want.txt",
+				test.Context("second file should be named 'want.txt', got %q", archive.Files[1].Name),
+			)
+
+			src := archive.Files[0].Data
+			want := archive.Files[1].Data
+
+			collector := &errorCollector{}
+
+			parser, err := parser.New(name, bytes.NewReader(src), collector.handler())
+			test.Ok(t, err)
+
+			_, err = parser.Parse()
+			test.Err(t, err, test.Context("Parse() failed to return an error given invalid syntax"))
+
+			got := collector.String()
+
+			if *update {
+				archive.Files[1].Data = []byte(got)
+
+				err := os.WriteFile(file, txtar.Format(archive), 0o644)
+				test.Ok(t, err)
+
+				return
+			}
+
+			test.DiffBytes(t, []byte(got), want)
+		})
+	}
+}
+
 // testFailHandler returns a [syntax.ErrorHandler] that handles syntax errors by failing
 // the enclosing test.
 func testFailHandler(tb testing.TB) syntax.ErrorHandler {
@@ -57,3 +130,40 @@ func testFailHandler(tb testing.TB) syntax.ErrorHandler {
 		tb.Fatalf("%s: %s", pos, msg)
 	}
 }
+
+// errorCollector is a helper struct that implements a [syntax.ErrorHandler] which
+// simply collects the scanning errors internally to be inspected later.
+type errorCollector struct {
+	errs []string
+	mu   sync.RWMutex
+}
+
+func (e *errorCollector) String() string {
+	e.mu.RLock()
+	defer e.mu.RUnlock()
+
+	// Take a copy so as not to alter the original
+	errsCopy := slices.Clone(e.errs)
+
+	var s strings.Builder
+
+	slices.Sort(errsCopy) // Deterministic
+
+	for _, err := range errsCopy {
+		s.WriteString(err)
+	}
+
+	return s.String()
+}
+
+// handler returns the [syntax.ErrorHandler] to be plugged in to the scanning operation.
+func (e *errorCollector) handler() syntax.ErrorHandler {
+	return func(pos syntax.Position, msg string) {
+		// Because the scanner runs in it's own goroutine and also makes use of the
+		// handler
+		e.mu.Lock()
+		defer e.mu.Unlock()
+
+		e.errs = append(e.errs, fmt.Sprintf("%s: %s\n", pos, msg))
+	}
+}

internal/syntax/parser/v2/parser_test.go

@@ -0,0 +1,8 @@
+-- src.http --
+### Bad
+GET https://blowup.com
+
+< £$%^&
+-- want.txt --
+bad-body-file.txtar:1:1-4: parseStatement: unrecognised token: Separator
+bad-body-file.txtar:4:5: unrecognised character: '£'

internal/syntax/parser/v2/testdata/invalid/bad-body-file.txtar

@@ -0,0 +1,9 @@
+# Content-Type is missing a colon
+
+-- src.http --
+### Bad
+GET https://api.something.com/v1
+Content-Type application/json
+-- want.txt --
+bad-headers.txtar:1:1-4: parseStatement: unrecognised token: Separator
+bad-headers.txtar:3:13: expected ':', got ' '

internal/syntax/parser/v2/testdata/invalid/bad-headers.txtar

@@ -0,0 +1,14 @@
+# This one tests our parser synchronisation method, the request after BadMethod
+# is fine but if the parser is cascading errors (not synchronising) then it will
+# report a bunch of errors in this one too as it will be confused and expecting
+# different tokens
+
+-- src.http --
+### BadMethod
+DESTROY https://api.something.com/v1
+
+### Okay
+GET https://api.somethingelse.com/v1
+-- want.txt --
+bad-method.txtar:1:1-4: parseStatement: unrecognised token: Separator
+bad-method.txtar:4:1-4: parseStatement: unrecognised token: Separator

internal/syntax/parser/v2/testdata/invalid/bad-method.txtar

@@ -0,0 +1,12 @@
+-- src.http --
+### Bad
+GET https://blowup.com
+
+{
+  "hello": "body"
+}
+
+> )(*&^%
+-- want.txt --
+bad-response-file.txtar:1:1-4: parseStatement: unrecognised token: Separator
+bad-response-file.txtar:8:4: unrecognised character: ')'

internal/syntax/parser/v2/testdata/invalid/bad-response-file.txtar

@@ -0,0 +1,12 @@
+-- src.http --
+### Bad
+GET https://blowup.com
+
+{
+  "hello": "body"
+}
+
+<> )(*&^%
+-- want.txt --
+bad-response-ref.txtar:1:1-4: parseStatement: unrecognised token: Separator
+bad-response-ref.txtar:8:5: unrecognised character: ')'