Add full shell, replaceable Omics adapter and bilingual interaction

Implement approved specification #32 and tickets #33-#37. Keep host authentication server-verified and pin session interaction language. Compile scoped base selectors for browser compatibility and retain full gutters during CSS pruning.

tools/tht/internal/config/shell.go

@@ -0,0 +1,90 @@
+package config
+
+import (
+	"errors"
+	"regexp"
+	"strings"
+)
+
+// Shell is the public presentation configuration, independent of deployment topology and auth.
+type Shell struct {
+	Mode          string `yaml:"mode" json:"mode"`
+	DefaultLocale string `yaml:"defaultLocale" json:"defaultLocale"`
+	Adapter       string `yaml:"adapter,omitempty" json:"adapter,omitempty"`
+}
+
+// NormalizeDefaults resolves omitted installation settings before projecting them to the browser.
+func (s *Shell) NormalizeDefaults() error {
+	if s.Mode == "" {
+		s.Mode = "embedded"
+	}
+	if s.DefaultLocale == "" {
+		s.DefaultLocale = "en"
+	}
+	if s.Mode != "embedded" && s.Mode != "full" {
+		return errors.New("shell.mode must be full or embedded")
+	}
+	if !wellFormedLocale(s.DefaultLocale) {
+		return errors.New("shell.defaultLocale must be a well-formed BCP47 language tag")
+	}
+	if s.Adapter != "" && s.Adapter != "omics-portal" {
+		return errors.New("shell.adapter must be omics-portal")
+	}
+	if s.Mode == "full" {
+		s.Adapter = ""
+		return nil
+	}
+	if s.Adapter == "" {
+		s.Adapter = "omics-portal"
+	}
+	return nil
+}
+
+// RFC 5646 section 2.1 grammar. Capture variants and extensions to reject duplicates.
+// Validate syntax without consulting a language registry or the frontend's available catalogs.
+var localePattern = regexp.MustCompile(`^(?:` +
+	`(?:[a-z]{2,3}(?:-[a-z]{3}){0,3}|[a-z]{4}|[a-z]{5,8})` + // language / extlang
+	`(?:-[a-z]{4})?(?:-(?:[a-z]{2}|[0-9]{3}))?` + // script / region
+	`((?:-(?:[a-z0-9]{5,8}|[0-9][a-z0-9]{3}))*)` + // variants
+	`((?:-[0-9a-wy-z](?:-[a-z0-9]{2,8})+)*)` + // extensions
+	`(?:-x(?:-[a-z0-9]{1,8})+)?` + // private use after a language
+	`|x(?:-[a-z0-9]{1,8})+)$`) // private use alone
+
+func wellFormedLocale(locale string) bool {
+	// BCP47 is ASCII. Check before case folding, which could otherwise map Unicode to ASCII.
+	for _, character := range locale {
+		if character > 127 {
+			return false
+		}
+	}
+	tag := strings.ToLower(locale)
+	// These grandfathered irregular tags do not match the normal language-tag grammar.
+	switch tag {
+	case "en-gb-oed", "i-ami", "i-bnn", "i-default", "i-enochian", "i-hak", "i-klingon",
+		"i-lux", "i-mingo", "i-navajo", "i-pwn", "i-tao", "i-tay", "i-tsu",
+		"sgn-be-fr", "sgn-be-nl", "sgn-ch-de":
+		return true
+	}
+	parts := localePattern.FindStringSubmatch(tag)
+	if parts == nil {
+		return false
+	}
+	variants := make(map[string]bool)
+	for _, variant := range strings.Split(strings.TrimPrefix(parts[1], "-"), "-") {
+		if variants[variant] {
+			return false
+		}
+		variants[variant] = true
+	}
+	singletons := make(map[string]bool)
+	for _, subtag := range strings.Split(parts[2], "-") {
+		if len(subtag) != 1 {
+			continue
+		}
+		if singletons[subtag] {
+			return false
+		}
+		singletons[subtag] = true
+	}
+	return true
+}