Refuse an exclude pattern that does not compile at config load (closes #455)

An exclude pattern that did not compile was skipped with a warning during the backup, which then stored what the user meant to leave out. Config.Validate now refuses it, naming the setting and quoting the pattern, for example exclude[2]: "build[" is not a valid exclude pattern: ...

The compile step moves to config.CompileExcludePattern, which Validate and the scanner both call. The scanner's skip-and-warn path is gone; a pattern that does not compile there now fails the scanner.

Model: opus-5-5

internal/config/exclude.go

@@ -1,11 +1,79 @@
 package config
 
 import (
+	"errors"
 	"fmt"
+	"strings"
 
-	"sneak.berlin/go/vaultik/internal/exclude"
+	"github.com/gobwas/glob"
 )
 
+var errBadExcludePattern = errors.New("is not a valid exclude pattern")
+
+// CompileExcludePattern compiles one exclude pattern into the globs that
+// match it. anchored reports whether the pattern matches only from the root
+// of the source directory. An empty pattern yields no globs. Config.Validate
+// and the scanner both compile patterns with it, so a pattern that loads is
+// one the scanner can use.
+func CompileExcludePattern(pattern string) ([]glob.Glob, bool, error) {
+	if pattern == "" {
+		return nil, false, nil
+	}
+
+	// A leading "/" anchors the pattern to the source directory.
+	trimmed, anchored := strings.CutPrefix(pattern, "/")
+
+	// A trailing "/" marks a directory and is not matched.
+	trimmed = strings.TrimSuffix(trimmed, "/")
+
+	// A leading "**/" demands a "/" before the name and so misses an
+	// entry directly under the source path. It means "at every depth",
+	// which is what an unanchored pattern already does, so drop the
+	// prefix and match unanchored: as in gitignore, "**/x" and
+	// "/**/x" then also match a top-level "x".
+	if rest, found := strings.CutPrefix(trimmed, "**/"); found {
+		trimmed = rest
+		anchored = false
+	}
+
+	// Likewise "/**/" inside a pattern demands two slashes, so
+	// "a/**/*.log" misses "a/x.log". As in gitignore it stands for
+	// zero or more directories, so also compile the pattern with it
+	// made a single "/". The glob library's {a,b} is no help: with a
+	// "**" inside it in mid-pattern, it misses paths it should match.
+	var globs []glob.Glob
+
+	for _, variant := range doubleStarVariants(trimmed) {
+		g, err := glob.Compile(variant, '/')
+		if err != nil {
+			return nil, false, fmt.Errorf("%w: %w", errBadExcludePattern, err)
+		}
+
+		globs = append(globs, g)
+	}
+
+	return globs, anchored, nil
+}
+
+// doubleStarVariants returns pattern once for every way of keeping each
+// "/**/" in it or making it a single "/".
+func doubleStarVariants(pattern string) []string {
+	const doubleStar = "/**/"
+
+	before, after, found := strings.Cut(pattern, doubleStar)
+	if !found {
+		return []string{pattern}
+	}
+
+	var variants []string
+
+	for _, rest := range doubleStarVariants(after) {
+		variants = append(variants, before+doubleStar+rest, before+"/"+rest)
+	}
+
+	return variants
+}
+
 // validateExcludes refuses an exclude pattern, global or in a snapshot,
 // that does not compile, naming the setting and quoting the pattern.
 func (c *Config) validateExcludes() error {
@@ -29,7 +97,7 @@ func (c *Config) validateExcludes() error {
 // compile. setting is the name of the list, as in "exclude".
 func validateExcludeList(setting string, patterns []string) error {
 	for i, pattern := range patterns {
-		_, _, err := exclude.Compile(pattern)
+		_, _, err := CompileExcludePattern(pattern)
 		if err != nil {
 			return fmt.Errorf("%s[%d]: %q %w", setting, i, pattern, err)
 		}

internal/exclude/exclude.go

@@ -1,76 +0,0 @@
-// Package exclude compiles exclude patterns into globs.
-package exclude
-
-import (
-	"errors"
-	"fmt"
-	"strings"
-
-	"github.com/gobwas/glob"
-)
-
-var errBadExcludePattern = errors.New("is not a valid exclude pattern")
-
-// Compile compiles one exclude pattern into the globs that match it.
-// anchored reports whether the pattern matches only from the root of the
-// source directory. An empty pattern yields no globs. Config.Validate and
-// the scanner both compile patterns with it, so a pattern that loads is one
-// the scanner can use.
-func Compile(pattern string) ([]glob.Glob, bool, error) {
-	if pattern == "" {
-		return nil, false, nil
-	}
-
-	// A leading "/" anchors the pattern to the source directory.
-	trimmed, anchored := strings.CutPrefix(pattern, "/")
-
-	// A trailing "/" marks a directory and is not matched.
-	trimmed = strings.TrimSuffix(trimmed, "/")
-
-	// A leading "**/" demands a "/" before the name and so misses an
-	// entry directly under the source path. It means "at every depth",
-	// which is what an unanchored pattern already does, so drop the
-	// prefix and match unanchored: as in gitignore, "**/x" and
-	// "/**/x" then also match a top-level "x".
-	if rest, found := strings.CutPrefix(trimmed, "**/"); found {
-		trimmed = rest
-		anchored = false
-	}
-
-	// Likewise "/**/" inside a pattern demands two slashes, so
-	// "a/**/*.log" misses "a/x.log". As in gitignore it stands for
-	// zero or more directories, so also compile the pattern with it
-	// made a single "/". The glob library's {a,b} is no help: with a
-	// "**" inside it in mid-pattern, it misses paths it should match.
-	var globs []glob.Glob
-
-	for _, variant := range doubleStarVariants(trimmed) {
-		g, err := glob.Compile(variant, '/')
-		if err != nil {
-			return nil, false, fmt.Errorf("%w: %w", errBadExcludePattern, err)
-		}
-
-		globs = append(globs, g)
-	}
-
-	return globs, anchored, nil
-}
-
-// doubleStarVariants returns pattern once for every way of keeping each
-// "/**/" in it or making it a single "/".
-func doubleStarVariants(pattern string) []string {
-	const doubleStar = "/**/"
-
-	before, after, found := strings.Cut(pattern, doubleStar)
-	if !found {
-		return []string{pattern}
-	}
-
-	var variants []string
-
-	for _, rest := range doubleStarVariants(after) {
-		variants = append(variants, before+doubleStar+rest, before+"/"+rest)
-	}
-
-	return variants
-}

internal/snapshot/scanner.go

@@ -19,8 +19,8 @@ import (
 	"github.com/spf13/afero"
 	"sneak.berlin/go/vaultik/internal/blob"
 	"sneak.berlin/go/vaultik/internal/chunker"
+	"sneak.berlin/go/vaultik/internal/config"
 	"sneak.berlin/go/vaultik/internal/database"
-	"sneak.berlin/go/vaultik/internal/exclude"
 	"sneak.berlin/go/vaultik/internal/log"
 	"sneak.berlin/go/vaultik/internal/storage"
 	"sneak.berlin/go/vaultik/internal/types"
@@ -2329,7 +2329,7 @@ func compileExcludePatterns(patterns []string) ([]compiledPattern, error) {
 	var compiled []compiledPattern
 
 	for _, p := range patterns {
-		globs, anchored, err := exclude.Compile(p)
+		globs, anchored, err := config.CompileExcludePattern(p)
 		if err != nil {
 			return nil, fmt.Errorf("%q %w", p, err)
 		}