hscript: Implement Language, add tests

3 files changed, 123 insertions, 21 deletions

diff --git a/hscript/meta.cc b/hscript/meta.cc
index 7e8cfe8..e5e0ba4 100644
--- a/hscript/meta.cc
+++ b/hscript/meta.cc
@@ -13,6 +13,7 @@
 #include <assert.h>
 #include <fstream>
 #include <regex>
+#include <set>
 #include <sstream>
 #ifdef HAS_INSTALL_ENV
 #   include <unistd.h>
@@ -179,6 +180,77 @@ Key *PkgInstall::parseFromData(const std::string &data, int lineno, int *errors,
 }
 
 
+/* All ISO 639-1 language codes.
+ * Source: https://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
+ * Python to construct table:
+ * >>> f = open('ISO-639-2_utf-8.txt')
+ * >>> x = csv.reader(f, delimiter='|')
+ * >>> langs = [lang[2] for lang in iter(x) if lang != '']
+ * >>> print('"' + '", "'.join(langs) + '", "C."')
+ */
+const std::set<std::string> valid_langs = {
+    "aa", "ab", "af", "ak", "sq", "am", "ar", "an", "hy", "as", "av", "ae",
+    "ay", "az", "ba", "bm", "eu", "be", "bn", "bh", "bi", "bs", "br", "bg",
+    "my", "ca", "ch", "ce", "zh", "cu", "cv", "kw", "co", "cr", "cs", "da",
+    "dv", "nl", "dz", "en", "eo", "et", "ee", "fo", "fj", "fi", "fr", "fy",
+    "ff", "ka", "de", "gd", "ga", "gl", "gv", "el", "gn", "gu", "ht", "ha",
+    "he", "hz", "hi", "ho", "hr", "hu", "ig", "is", "io", "ii", "iu", "ie",
+    "ia", "id", "ik", "it", "jv", "ja", "kl", "kn", "ks", "kr", "kk", "km",
+    "ki", "rw", "ky", "kv", "kg", "ko", "kj", "ku", "lo", "la", "lv", "li",
+    "ln", "lt", "lb", "lu", "lg", "mk", "mh", "ml", "mi", "mr", "ms", "mg",
+    "mt", "mn", "na", "nv", "nr", "nd", "ng", "ne", "nn", "nb", "no", "ny",
+    "oc", "oj", "or", "om", "os", "pa", "fa", "pi", "pl", "pt", "ps", "qu",
+    "rm", "ro", "rn", "ru", "sg", "sa", "si", "sk", "sl", "se", "sm", "sn",
+    "sd", "so", "st", "es", "sc", "sr", "ss", "su", "sw", "sv", "ty", "ta",
+    "tt", "te", "tg", "tl", "th", "bo", "ti", "to", "tn", "ts", "tk", "tr",
+    "tw", "ug", "uk", "ur", "uz", "ve", "vi", "vo", "cy", "wa", "wo", "xh",
+    "yi", "yo", "za", "zu", "C."
+};
+
+
+Key *Language::parseFromData(const std::string &data, int lineno, int *errors,
+                             int *warnings) {
+    if(data.length() < 2 ||
+       valid_langs.find(data.substr(0, 2)) == valid_langs.end()) {
+        if(errors) *errors += 1;
+        output_error("installfile:" + std::to_string(lineno),
+                     "language: invalid language specified",
+                     "language must be a valid ISO 639-1 language code");
+        return nullptr;
+    }
+
+    /* We know a valid language appears, but is it real? */
+    if(data.length() > 2) {
+        /* data[1] is . if language is C.UTF-8 */
+        if(data[2] != '_' && data[1] != '.') {
+            if(errors) *errors += 1;
+            output_error("installfile:" + std::to_string(lineno),
+                         "language: invalid language specified",
+                         "language must be a valid ISO 639-1 language code, "
+                         "optionally followed by '_' and a country code");
+            return nullptr;
+        }
+        /* we don't really care about the country code, but we do care about
+         * codeset - we (via musl) *only* support UTF-8. */
+        std::string::size_type dot = data.find_first_of('.');
+        if(dot != std::string::npos && data.substr(dot+1, 5) != "UTF-8") {
+            if(errors) *errors += 1;
+            output_error("installfile:" + std::to_string(lineno),
+                         "language: invalid language specified",
+                         "you cannot specify a non-UTF-8 codeset");
+            return nullptr;
+        }
+    }
+
+    return new Language(lineno, data);
+}
+
+
+bool Language::execute(ScriptOptions) const {
+    return false;
+}
+
+
 Key *Firmware::parseFromData(const std::string &data, int lineno, int *errors,
                              int *warnings) {
     bool value;
diff --git a/hscript/meta.hh b/hscript/meta.hh
index 8a50576..7dc3f68 100644
--- a/hscript/meta.hh
+++ b/hscript/meta.hh
@@ -45,6 +45,13 @@ public:
 };
 
 class Language : public StringKey {
+private:
+    Language(int _line, const std::string &my_lang) :
+        StringKey(_line, my_lang) {}
+public:
+    static Key *parseFromData(const std::string &data, int lineno, int *errors,
+                              int *warnings);
+    bool execute(ScriptOptions) const override;
 };
 
 class Keymap : public StringKey {
@@ -63,7 +70,7 @@ class Timezone : public StringKey {
 
 class Repository : public StringKey {
 private:
-    Repository(int _line, const std::string my_url) :
+    Repository(int _line, const std::string &my_url) :
         StringKey(_line, my_url) {}
 public:
     static Key *parseFromData(const std::string &data, int lineno, int *errors,
diff --git a/hscript/script.cc b/hscript/script.cc
index 33003bf..69a46cc 100644
--- a/hscript/script.cc
+++ b/hscript/script.cc
@@ -88,8 +88,8 @@ struct Script::ScriptPrivate {
     std::set<std::string> packages;
     /*! The root shadow line. */
     std::unique_ptr<RootPassphrase> rootpw;
-    /*! Target system's mountpoints. */
-    std::vector< std::unique_ptr<Mount> > mounts;
+    /*! The system language. */
+    std::unique_ptr<Language> lang;
 
     /*! Network addressing configuration */
     std::vector< std::unique_ptr<NetAddress> > addresses;
@@ -103,6 +103,8 @@ struct Script::ScriptPrivate {
 
     /*! Disk identification keys */
     std::vector< std::unique_ptr<DiskId> > diskids;
+    /*! Target system's mountpoints. */
+    std::vector< std::unique_ptr<Mount> > mounts;
 
 #ifdef NON_LIBRE_FIRMWARE
     std::unique_ptr<Firmware> firmware;
@@ -119,18 +121,6 @@ struct Script::ScriptPrivate {
                    int *errors, int *warnings, const ScriptOptions &opts) {
         if(key_name == "network") {
             return store_network(obj, lineno, errors, warnings, opts);
-        } else if(key_name == "hostname") {
-            return store_hostname(obj, lineno, errors, warnings, opts);
-        } else if(key_name == "pkginstall") {
-            return store_pkginstall(obj, lineno, errors, warnings, opts);
-        } else if(key_name == "rootpw") {
-            return store_rootpw(obj, lineno, errors, warnings, opts);
-        } else if(key_name == "firmware") {
-            return store_firmware(obj, lineno, errors, warnings, opts);
-        } else if(key_name == "mount") {
-            std::unique_ptr<Mount> mount(dynamic_cast<Mount *>(obj));
-            this->mounts.push_back(std::move(mount));
-            return true;
         } else if(key_name == "netaddress") {
             std::unique_ptr<NetAddress> addr(dynamic_cast<NetAddress *>(obj));
             this->addresses.push_back(std::move(addr));
@@ -139,14 +129,20 @@ struct Script::ScriptPrivate {
             std::unique_ptr<NetSSID> ssid(dynamic_cast<NetSSID *>(obj));
             this->ssids.push_back(std::move(ssid));
             return true;
+        } else if(key_name == "hostname") {
+            return store_hostname(obj, lineno, errors, warnings, opts);
+        } else if(key_name == "pkginstall") {
+            return store_pkginstall(obj, lineno, errors, warnings, opts);
+        } else if(key_name == "rootpw") {
+            return store_rootpw(obj, lineno, errors, warnings, opts);
+        } else if(key_name == "language") {
+            return store_lang(obj, lineno, errors, warnings, opts);
+        } else if(key_name == "firmware") {
+            return store_firmware(obj, lineno, errors, warnings, opts);
         } else if(key_name == "repository") {
             std::unique_ptr<Repository> repo(dynamic_cast<Repository *>(obj));
             this->repos.push_back(std::move(repo));
             return true;
-        } else if(key_name == "diskid") {
-            std::unique_ptr<DiskId> diskid(dynamic_cast<DiskId *>(obj));
-            this->diskids.push_back(std::move(diskid));
-            return true;
         } else if(key_name == "username") {
             return store_username(obj, lineno, errors, warnings, opts);
         } else if(key_name == "useralias") {
@@ -157,6 +153,14 @@ struct Script::ScriptPrivate {
             return store_usericon(obj, lineno, errors, warnings, opts);
         } else if(key_name == "usergroups") {
             return store_usergroups(obj, lineno, errors, warnings, opts);
+        } else if(key_name == "diskid") {
+            std::unique_ptr<DiskId> diskid(dynamic_cast<DiskId *>(obj));
+            this->diskids.push_back(std::move(diskid));
+            return true;
+        } else if(key_name == "mount") {
+            std::unique_ptr<Mount> mount(dynamic_cast<Mount *>(obj));
+            this->mounts.push_back(std::move(mount));
+            return true;
         } else {
             return false;
         }
@@ -241,6 +245,18 @@ struct Script::ScriptPrivate {
 #endif
     }
 
+    bool store_lang(Keys::Key *obj, int lineno, int *errors, int *,
+                    ScriptOptions) {
+        if(this->lang) {
+            DUPLICATE_ERROR(this->lang, std::string("language"),
+                            this->lang->value())
+            return false;
+        }
+        std::unique_ptr<Language> l(dynamic_cast<Language *>(obj));
+        this->lang = std::move(l);
+        return true;
+    }
+
     bool store_username(Keys::Key *obj, int lineno, int *errors, int *,
                         ScriptOptions) {
         if(accounts.size() >= 255) {
@@ -527,9 +543,16 @@ bool Script::validate() const {
     /* REQ: Runner.Validate.rootpw */
     if(!this->internal->rootpw->validate(this->opts)) failures++;
 
-    /* language */
+    /* REQ: Runner.Validate.language */
+    if(internal->lang && !internal->lang->validate(this->opts)) failures++;
+
     /* keymap */
-    /* firmware */
+
+#ifdef NON_LIBRE_FIRMWARE
+    /* REQ: Runner.Validate.firmware */
+    if(!this->internal->firmware->validate(this->opts)) failures++;
+#endif
+
     /* timezone */
 
     /* REQ: Script.repository */