Merge branch 'errors-phase-2' into caveman-LOCs

18 files changed, 1507 insertions, 103 deletions

diff --git a/src/libutil/config.cc b/src/libutil/config.cc
index f03e444ec..8fc700a2b 100644
--- a/src/libutil/config.cc
+++ b/src/libutil/config.cc
@@ -65,60 +65,63 @@ void Config::getSettings(std::map<std::string, SettingInfo> & res, bool override
             res.emplace(opt.first, SettingInfo{opt.second.setting->to_string(), opt.second.setting->description});
 }
 
-void AbstractConfig::applyConfigFile(const Path & path)
-{
-    try {
-        string contents = readFile(path);
-
-        unsigned int pos = 0;
-
-        while (pos < contents.size()) {
-            string line;
-            while (pos < contents.size() && contents[pos] != '\n')
-                line += contents[pos++];
-            pos++;
-
-            string::size_type hash = line.find('#');
-            if (hash != string::npos)
-                line = string(line, 0, hash);
-
-            vector<string> tokens = tokenizeString<vector<string> >(line);
-            if (tokens.empty()) continue;
+void AbstractConfig::applyConfig(const std::string & contents, const std::string & path) {
+    unsigned int pos = 0;
+
+    while (pos < contents.size()) {
+        string line;
+        while (pos < contents.size() && contents[pos] != '\n')
+            line += contents[pos++];
+        pos++;
+
+        string::size_type hash = line.find('#');
+        if (hash != string::npos)
+            line = string(line, 0, hash);
+
+        vector<string> tokens = tokenizeString<vector<string> >(line);
+        if (tokens.empty()) continue;
+
+        if (tokens.size() < 2)
+            throw UsageError("illegal configuration line '%1%' in '%2%'", line, path);
+
+        auto include = false;
+        auto ignoreMissing = false;
+        if (tokens[0] == "include")
+            include = true;
+        else if (tokens[0] == "!include") {
+            include = true;
+            ignoreMissing = true;
+        }
 
-            if (tokens.size() < 2)
+        if (include) {
+            if (tokens.size() != 2)
                 throw UsageError("illegal configuration line '%1%' in '%2%'", line, path);
-
-            auto include = false;
-            auto ignoreMissing = false;
-            if (tokens[0] == "include")
-                include = true;
-            else if (tokens[0] == "!include") {
-                include = true;
-                ignoreMissing = true;
+            auto p = absPath(tokens[1], dirOf(path));
+            if (pathExists(p)) {
+                applyConfigFile(p);
+            } else if (!ignoreMissing) {
+                throw Error("file '%1%' included from '%2%' not found", p, path);
             }
+            continue;
+        }
 
-            if (include) {
-                if (tokens.size() != 2)
-                    throw UsageError("illegal configuration line '%1%' in '%2%'", line, path);
-                auto p = absPath(tokens[1], dirOf(path));
-                if (pathExists(p)) {
-                    applyConfigFile(p);
-                } else if (!ignoreMissing) {
-                    throw Error("file '%1%' included from '%2%' not found", p, path);
-                }
-                continue;
-            }
+        if (tokens[1] != "=")
+            throw UsageError("illegal configuration line '%1%' in '%2%'", line, path);
 
-            if (tokens[1] != "=")
-                throw UsageError("illegal configuration line '%1%' in '%2%'", line, path);
+        string name = tokens[0];
 
-            string name = tokens[0];
+        vector<string>::iterator i = tokens.begin();
+        advance(i, 2);
 
-            vector<string>::iterator i = tokens.begin();
-            advance(i, 2);
+        set(name, concatStringsSep(" ", Strings(i, tokens.end()))); // FIXME: slow
+    };
+}
 
-            set(name, concatStringsSep(" ", Strings(i, tokens.end()))); // FIXME: slow
-        };
+void AbstractConfig::applyConfigFile(const Path & path)
+{
+    try {
+        string contents = readFile(path);
+        applyConfig(contents, path);
     } catch (SysError &) { }
 }
 
diff --git a/src/libutil/config.hh b/src/libutil/config.hh
index 7ea78fdaf..5c7a70a2e 100644
--- a/src/libutil/config.hh
+++ b/src/libutil/config.hh
@@ -7,6 +7,38 @@
 
 namespace nix {
 
+/**
+ * The Config class provides Nix runtime configurations.
+ *
+ * What is a Configuration?
+ *   A collection of uniquely named Settings.
+ *
+ * What is a Setting?
+ *   Each property that you can set in a configuration corresponds to a
+ *   `Setting`. A setting records value and description of a property
+ *   with a default and optional aliases.
+ *
+ * A valid configuration consists of settings that are registered to a
+ * `Config` object instance:
+ *
+ *   Config config;
+ *   Setting<std::string> systemSetting{&config, "x86_64-linux", "system", "the current system"};
+ *
+ * The above creates a `Config` object and registers a setting called "system"
+ * via the variable `systemSetting` with it. The setting defaults to the string
+ * "x86_64-linux", it's description is "the current system". All of the
+ * registered settings can then be accessed as shown below:
+ *
+ *   std::map<std::string, Config::SettingInfo> settings;
+ *   config.getSettings(settings);
+ *   config["system"].description == "the current system"
+ *   config["system"].value == "x86_64-linux"
+ *
+ *
+ * The above retrieves all currently known settings from the `Config` object
+ * and adds them to the `settings` map.
+ */
+
 class Args;
 class AbstractSetting;
 class JSONPlaceholder;
@@ -23,6 +55,10 @@ protected:
 
 public:
 
+    /**
+     * Sets the value referenced by `name` to `value`. Returns true if the
+     * setting is known, false otherwise.
+     */
     virtual bool set(const std::string & name, const std::string & value) = 0;
 
     struct SettingInfo
@@ -31,18 +67,52 @@ public:
         std::string description;
     };
 
+    /**
+     * Adds the currently known settings to the given result map `res`.
+     * - res: map to store settings in
+     * - overridenOnly: when set to true only overridden settings will be added to `res`
+     */
     virtual void getSettings(std::map<std::string, SettingInfo> & res, bool overridenOnly = false) = 0;
 
+    /**
+     * Parses the configuration in `contents` and applies it
+     * - contents: configuration contents to be parsed and applied
+     * - path: location of the configuration file
+     */
+    void applyConfig(const std::string & contents, const std::string & path = "<unknown>");
+
+    /**
+     * Applies a nix configuration file
+     * - path: the location of the config file to apply
+     */
     void applyConfigFile(const Path & path);
 
+    /**
+     * Resets the `overridden` flag of all Settings
+     */
     virtual void resetOverriden() = 0;
 
+    /**
+     * Outputs all settings to JSON
+     * - out: JSONObject to write the configuration to
+     */
     virtual void toJSON(JSONObject & out) = 0;
 
+    /**
+     * Converts settings to `Args` to be used on the command line interface
+     * - args: args to write to
+     * - category: category of the settings
+     */
     virtual void convertToArgs(Args & args, const std::string & category) = 0;
 
+    /**
+     * Logs a warning for each unregistered setting
+     */
     void warnUnknownSettings();
 
+    /**
+     * Re-applies all previously attempted changes to unknown settings
+     */
     void reapplyUnknownSettings();
 };
 
diff --git a/src/libutil/error.cc b/src/libutil/error.cc
index 585059f22..a713fe9db 100644
--- a/src/libutil/error.cc
+++ b/src/libutil/error.cc
@@ -10,13 +10,16 @@ namespace nix {
 
 const std::string nativeSystem = SYSTEM;
 
-
+// addPrefix is used for show-trace.  Strings added with addPrefix
+// will print ahead of the error itself.
 BaseError & BaseError::addPrefix(const FormatOrString & fs)
 {
     prefix_ = fs.s + prefix_;
     return *this;
 }
 
+// c++ std::exception descendants must have a 'const char* what()' function.
+// This stringifies the error and caches it for use by what(), or similarly by msg().
 const string& BaseError::calcWhat() const
 {
     if (what_.has_value())
@@ -124,25 +127,25 @@ void getCodeLines(NixCode &nixCode)
     }
 }
 
-
+// if nixCode contains lines of code, print them to the ostream, indicating the error column.
 void printCodeLines(std::ostream &out, const string &prefix, const NixCode &nixCode)
 {
     // previous line of code.
     if (nixCode.prevLineOfCode.has_value()) {
-        out << fmt("%1% %|2$5d|| %3%",
-            prefix,
-            (nixCode.errPos.line - 1),
-            *nixCode.prevLineOfCode)
-            << std::endl;
+        out << std::endl 
+            << fmt("%1% %|2$5d|| %3%",
+                prefix,
+                (nixCode.errPos.line - 1),
+                *nixCode.prevLineOfCode);
     }
 
     if (nixCode.errLineOfCode.has_value()) {
         // line of code containing the error.
-        out << fmt("%1% %|2$5d|| %3%",
-            prefix,
-            (nixCode.errPos.line),
-            *nixCode.errLineOfCode)
-            << std::endl;
+        out << std::endl
+            << fmt("%1% %|2$5d|| %3%",
+                prefix,
+                (nixCode.errPos.line),
+                *nixCode.errLineOfCode);
         // error arrows for the column range.
         if (nixCode.errPos.column > 0) {
             int start = nixCode.errPos.column;
@@ -153,20 +156,21 @@ void printCodeLines(std::ostream &out, const string &prefix, const NixCode &nixC
 
             std::string arrows("^");
 
-            out << fmt("%1%      |%2%" ANSI_RED "%3%" ANSI_NORMAL,
-                prefix,
-                spaces,
-                arrows) << std::endl;
+            out << std::endl
+                << fmt("%1%      |%2%" ANSI_RED "%3%" ANSI_NORMAL,
+                    prefix,
+                    spaces,
+                    arrows);
         }
     }
 
     // next line of code.
     if (nixCode.nextLineOfCode.has_value()) {
-        out << fmt("%1% %|2$5d|| %3%",
-            prefix,
-            (nixCode.errPos.line + 1),
-            *nixCode.nextLineOfCode)
-            << std::endl;
+        out << std::endl
+            << fmt("%1% %|2$5d|| %3%",
+                prefix,
+                (nixCode.errPos.line + 1),
+                *nixCode.nextLineOfCode);
     }
 }
 
@@ -239,17 +243,15 @@ std::ostream& operator<<(std::ostream &out, const ErrorInfo &einfo)
             levelString,
             einfo.name,
             dashes,
-            einfo.programName.value_or(""))
-            << std::endl;
+            einfo.programName.value_or(""));
     else
         out << fmt("%1%%2%" ANSI_BLUE " -----%3% %4%" ANSI_NORMAL,
             prefix,
             levelString,
             dashes,
-            einfo.programName.value_or(""))
-            << std::endl;
+            einfo.programName.value_or(""));
 
-    // filename, line, column.
+    bool nl = false;  // intersperse newline between sections.
     if (einfo.nixCode.has_value()) {
         switch (einfo.nixCode->errPos.origin) {
             case foFile: {
@@ -273,30 +275,36 @@ std::ostream& operator<<(std::ostream &out, const ErrorInfo &einfo)
             default:
                 throw Error("invalid FileOrigin in errPos");
         }
+        nl = true;
     }
 
     // description
     if (einfo.description != "") {
-        out << prefix << einfo.description << std::endl;
-        out << prefix << std::endl;
+        if (nl)
+            out << std::endl << prefix;
+        out << std::endl << prefix << einfo.description;
+        nl = true;
     }
 
-
     if (einfo.nixCode.has_value()) {
         NixCode nixcode = *einfo.nixCode;
         getCodeLines(nixcode);
       
         // lines of code.
         if (nixcode.errLineOfCode.has_value()) {
+            if (nl)
+                out << std::endl << prefix;
             printCodeLines(out, prefix, nixcode);
-            out << prefix << std::endl;
+            nl = true;
         }
     }
 
     // hint
     if (einfo.hint.has_value()) {
-        out << prefix << *einfo.hint << std::endl;
-        out << prefix << std::endl;
+        if (nl)
+            out << std::endl << prefix;
+        out << std::endl << prefix << *einfo.hint;
+        nl = true;
     }
 
     return out;
diff --git a/src/libutil/error.hh b/src/libutil/error.hh
index d60c0fefe..bd17a47c1 100644
--- a/src/libutil/error.hh
+++ b/src/libutil/error.hh
@@ -22,6 +22,23 @@
 
 namespace nix {
 
+/* 
+
+This file defines two main structs/classes used in nix error handling.
+
+ErrorInfo provides a standard payload of error information, with conversion to string
+happening in the logger rather than at the call site.
+
+BaseError is the ancestor of nix specific exceptions (and Interrupted), and contains
+an ErrorInfo.
+
+ErrorInfo structs are sent to the logger as part of an exception, or directly with the
+logError or logWarning macros.
+
+See the error-demo.cc program for usage examples.
+
+*/
+
 typedef enum {
     lvlError = 0,
     lvlWarn,
@@ -38,7 +55,7 @@ typedef enum {
     foString
 } FileOrigin;
 
-
+// ErrPos indicates the location of an error in a nix file.
 struct ErrPos {
     int line = 0;
     int column = 0;
@@ -50,6 +67,7 @@ struct ErrPos {
         return line != 0;
     }
 
+    // convert from the Pos struct, found in libexpr.
     template <class P>
     ErrPos& operator=(const P &pos)
     {
@@ -74,8 +92,6 @@ struct NixCode {
     std::optional<string> nextLineOfCode;
 };
 
-// -------------------------------------------------
-// ErrorInfo.
 struct ErrorInfo {
     Verbosity level;
     string name;
@@ -164,7 +180,7 @@ public:
     {
         errNo = errno;
         auto hf = hintfmt(args...);
-        err.hint = hintfmt("%1% : %2%", normaltxt(hf.str()), strerror(errNo));
+        err.hint = hintfmt("%1%: %2%", normaltxt(hf.str()), strerror(errNo));
     }
 
     virtual const char* sname() const override { return "SysError"; }
diff --git a/src/libutil/logging.cc b/src/libutil/logging.cc
index 071a7ec7e..41378b0db 100644
--- a/src/libutil/logging.cc
+++ b/src/libutil/logging.cc
@@ -251,10 +251,9 @@ bool handleJSONLogMessage(const std::string & msg,
         }
 
     } catch (std::exception & e) {
-        logError(
-            ErrorInfo { 
-                .name = "Json log message",
-                .hint = hintfmt("bad log message from builder: %s", e.what())
+        logError({ 
+            .name = "Json log message",
+            .hint = hintfmt("bad log message from builder: %s", e.what())
         });
     }
 
diff --git a/src/libutil/logging.hh b/src/libutil/logging.hh
index 39692a291..eeb7233e9 100644
--- a/src/libutil/logging.hh
+++ b/src/libutil/logging.hh
@@ -150,9 +150,23 @@ bool handleJSONLogMessage(const std::string & msg,
 
 extern Verbosity verbosity; /* suppress msgs > this */
 
-/* Print a message if the current log level is at least the specified
-   level. Note that this has to be implemented as a macro to ensure
-   that the arguments are evaluated lazily. */
+/* Print a message with the standard ErrorInfo format.
+   In general, use these 'log' macros for reporting problems that may require user
+   intervention or that need more explanation.  Use the 'print' macros for more
+   lightweight status messages. */
+#define logErrorInfo(level, errorInfo...) \
+    do { \
+        if (level <= nix::verbosity) { \
+            logger->logEI(level, errorInfo); \
+        } \
+    } while (0)
+
+#define logError(errorInfo...) logErrorInfo(lvlError, errorInfo)
+#define logWarning(errorInfo...) logErrorInfo(lvlWarn, errorInfo)
+
+/* Print a string message if the current log level is at least the specified
+   level. Note that this has to be implemented as a macro to ensure that the
+   arguments are evaluated lazily. */
 #define printMsg(level, args...) \
     do { \
         if (level <= nix::verbosity) { \
@@ -166,18 +180,7 @@ extern Verbosity verbosity; /* suppress msgs > this */
 #define debug(args...) printMsg(lvlDebug, args)
 #define vomit(args...) printMsg(lvlVomit, args)
 
-#define logErrorInfo(level, errorInfo...) \
-    do { \
-        if (level <= nix::verbosity) { \
-            logger->logEI(level, errorInfo); \
-        } \
-    } while (0)
-
-#define logError(errorInfo...) logErrorInfo(lvlError, errorInfo)
-#define logWarning(errorInfo...) logErrorInfo(lvlWarn, errorInfo)
-
-
-
+/* if verbosity >= lvlWarn, print a message with a yellow 'warning:' prefix. */
 template<typename... Args>
 inline void warn(const std::string & fs, const Args & ... args)
 {
diff --git a/src/libutil/lru-cache.hh b/src/libutil/lru-cache.hh
index 8b83f842c..6ef4a3e06 100644
--- a/src/libutil/lru-cache.hh
+++ b/src/libutil/lru-cache.hh
@@ -1,5 +1,6 @@
 #pragma once
 
+#include <cassert>
 #include <map>
 #include <list>
 #include <optional>
diff --git a/src/libutil/tests/config.cc b/src/libutil/tests/config.cc
new file mode 100644
index 000000000..74c59fd31
--- /dev/null
+++ b/src/libutil/tests/config.cc
@@ -0,0 +1,264 @@
+#include "json.hh"
+#include "config.hh"
+#include "args.hh"
+
+#include <sstream>
+#include <gtest/gtest.h>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * Config
+     * --------------------------------------------------------------------------*/
+
+    TEST(Config, setUndefinedSetting) {
+        Config config;
+        ASSERT_EQ(config.set("undefined-key", "value"), false);
+    }
+
+    TEST(Config, setDefinedSetting) {
+        Config config;
+        std::string value;
+        Setting<std::string> foo{&config, value, "name-of-the-setting", "description"};
+        ASSERT_EQ(config.set("name-of-the-setting", "value"), true);
+    }
+
+    TEST(Config, getDefinedSetting) {
+        Config config;
+        std::string value;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> foo{&config, value, "name-of-the-setting", "description"};
+
+        config.getSettings(settings, /* overridenOnly = */ false);
+        const auto iter = settings.find("name-of-the-setting");
+        ASSERT_NE(iter, settings.end());
+        ASSERT_EQ(iter->second.value, "");
+        ASSERT_EQ(iter->second.description, "description");
+    }
+
+    TEST(Config, getDefinedOverridenSettingNotSet) {
+        Config config;
+        std::string value;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> foo{&config, value, "name-of-the-setting", "description"};
+
+        config.getSettings(settings, /* overridenOnly = */ true);
+        const auto e = settings.find("name-of-the-setting");
+        ASSERT_EQ(e, settings.end());
+    }
+
+    TEST(Config, getDefinedSettingSet1) {
+        Config config;
+        std::string value;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> setting{&config, value, "name-of-the-setting", "description"};
+
+        setting.assign("value");
+
+        config.getSettings(settings, /* overridenOnly = */ false);
+        const auto iter = settings.find("name-of-the-setting");
+        ASSERT_NE(iter, settings.end());
+        ASSERT_EQ(iter->second.value, "value");
+        ASSERT_EQ(iter->second.description, "description");
+    }
+
+    TEST(Config, getDefinedSettingSet2) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+
+        ASSERT_TRUE(config.set("name-of-the-setting", "value"));
+
+        config.getSettings(settings, /* overridenOnly = */ false);
+        const auto e = settings.find("name-of-the-setting");
+        ASSERT_NE(e, settings.end());
+        ASSERT_EQ(e->second.value, "value");
+        ASSERT_EQ(e->second.description, "description");
+    }
+
+    TEST(Config, addSetting) {
+        class TestSetting : public AbstractSetting {
+            public:
+            TestSetting() : AbstractSetting("test", "test", {}) {}
+            void set(const std::string & value) {}
+            std::string to_string() const { return {}; }
+        };
+
+        Config config;
+        TestSetting setting;
+
+        ASSERT_FALSE(config.set("test", "value"));
+        config.addSetting(&setting);
+        ASSERT_TRUE(config.set("test", "value"));
+    }
+
+    TEST(Config, withInitialValue) {
+        const StringMap initials = {
+            { "key", "value" },
+        };
+        Config config(initials);
+
+        {
+            std::map<std::string, Config::SettingInfo> settings;
+            config.getSettings(settings, /* overridenOnly = */ false);
+            ASSERT_EQ(settings.find("key"), settings.end());
+        }
+
+        Setting<std::string> setting{&config, "default-value", "key", "description"};
+
+        {
+            std::map<std::string, Config::SettingInfo> settings;
+            config.getSettings(settings, /* overridenOnly = */ false);
+            ASSERT_EQ(settings["key"].value, "value");
+        }
+    }
+
+    TEST(Config, resetOverriden) {
+        Config config;
+        config.resetOverriden();
+    }
+
+    TEST(Config, resetOverridenWithSetting) {
+        Config config;
+        Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+
+        {
+            std::map<std::string, Config::SettingInfo> settings;
+
+            setting.set("foo");
+            ASSERT_EQ(setting.get(), "foo");
+            config.getSettings(settings, /* overridenOnly = */ true);
+            ASSERT_TRUE(settings.empty());
+        }
+
+        {
+            std::map<std::string, Config::SettingInfo> settings;
+
+            setting.override("bar");
+            ASSERT_TRUE(setting.overriden);
+            ASSERT_EQ(setting.get(), "bar");
+            config.getSettings(settings, /* overridenOnly = */ true);
+            ASSERT_FALSE(settings.empty());
+        }
+
+        {
+            std::map<std::string, Config::SettingInfo> settings;
+
+            config.resetOverriden();
+            ASSERT_FALSE(setting.overriden);
+            config.getSettings(settings, /* overridenOnly = */ true);
+            ASSERT_TRUE(settings.empty());
+        }
+    }
+
+    TEST(Config, toJSONOnEmptyConfig) {
+        std::stringstream out;
+        { // Scoped to force the destructor of JSONObject to write the final `}`
+            JSONObject obj(out);
+            Config config;
+            config.toJSON(obj);
+        }
+
+        ASSERT_EQ(out.str(), "{}");
+    }
+
+    TEST(Config, toJSONOnNonEmptyConfig) {
+        std::stringstream out;
+        { // Scoped to force the destructor of JSONObject to write the final `}`
+            JSONObject obj(out);
+
+            Config config;
+            std::map<std::string, Config::SettingInfo> settings;
+            Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+            setting.assign("value");
+
+            config.toJSON(obj);
+        }
+        ASSERT_EQ(out.str(), R"#({"name-of-the-setting":{"description":"description","value":"value"}})#");
+    }
+
+    TEST(Config, setSettingAlias) {
+        Config config;
+        Setting<std::string> setting{&config, "", "some-int", "best number", { "another-int" }};
+        ASSERT_TRUE(config.set("some-int", "1"));
+        ASSERT_EQ(setting.get(), "1");
+        ASSERT_TRUE(config.set("another-int", "2"));
+        ASSERT_EQ(setting.get(), "2");
+        ASSERT_TRUE(config.set("some-int", "3"));
+        ASSERT_EQ(setting.get(), "3");
+    }
+
+    /* FIXME: The reapplyUnknownSettings method doesn't seem to do anything
+     * useful (these days).  Whenever we add a new setting to Config the
+     * unknown settings are always considered.  In which case is this function
+     * actually useful? Is there some way to register a Setting without calling
+     * addSetting? */
+    TEST(Config, DISABLED_reapplyUnknownSettings) {
+        Config config;
+        ASSERT_FALSE(config.set("name-of-the-setting", "unknownvalue"));
+        Setting<std::string> setting{&config, "default", "name-of-the-setting", "description"};
+        ASSERT_EQ(setting.get(), "default");
+        config.reapplyUnknownSettings();
+        ASSERT_EQ(setting.get(), "unknownvalue");
+    }
+
+    TEST(Config, applyConfigEmpty) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        config.applyConfig("");
+        config.getSettings(settings);
+        ASSERT_TRUE(settings.empty());
+    }
+
+    TEST(Config, applyConfigEmptyWithComment) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        config.applyConfig("# just a comment");
+        config.getSettings(settings);
+        ASSERT_TRUE(settings.empty());
+    }
+
+    TEST(Config, applyConfigAssignment) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+        config.applyConfig(
+            "name-of-the-setting = value-from-file #useful comment\n"
+            "# name-of-the-setting = foo\n"
+        );
+        config.getSettings(settings);
+        ASSERT_FALSE(settings.empty());
+        ASSERT_EQ(settings["name-of-the-setting"].value, "value-from-file");
+    }
+
+    TEST(Config, applyConfigWithReassignedSetting) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+        config.applyConfig(
+            "name-of-the-setting = first-value\n"
+            "name-of-the-setting = second-value\n"
+        );
+        config.getSettings(settings);
+        ASSERT_FALSE(settings.empty());
+        ASSERT_EQ(settings["name-of-the-setting"].value, "second-value");
+    }
+
+    TEST(Config, applyConfigFailsOnMissingIncludes) {
+        Config config;
+        std::map<std::string, Config::SettingInfo> settings;
+        Setting<std::string> setting{&config, "", "name-of-the-setting", "description"};
+
+        ASSERT_THROW(config.applyConfig(
+            "name-of-the-setting = value-from-file\n"
+            "# name-of-the-setting = foo\n"
+            "include /nix/store/does/not/exist.nix"
+        ), Error);
+    }
+
+    TEST(Config, applyConfigInvalidThrows) {
+        Config config;
+        ASSERT_THROW(config.applyConfig("value == key"), UsageError);
+        ASSERT_THROW(config.applyConfig("value "), UsageError);
+    }
+}
diff --git a/src/libutil/tests/hash.cc b/src/libutil/tests/hash.cc
new file mode 100644
index 000000000..7cb439817
--- /dev/null
+++ b/src/libutil/tests/hash.cc
@@ -0,0 +1,80 @@
+#include "hash.hh"
+#include <gtest/gtest.h>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * hashString
+     * --------------------------------------------------------------------------*/
+
+    TEST(hashString, testKnownMD5Hashes1) {
+        // values taken from: https://tools.ietf.org/html/rfc1321
+        auto s1 = "";
+        auto hash = hashString(HashType::htMD5, s1);
+        ASSERT_EQ(hash.to_string(Base::Base16), "md5:d41d8cd98f00b204e9800998ecf8427e");
+    }
+
+    TEST(hashString, testKnownMD5Hashes2) {
+        // values taken from: https://tools.ietf.org/html/rfc1321
+        auto s2 = "abc";
+        auto hash = hashString(HashType::htMD5, s2);
+        ASSERT_EQ(hash.to_string(Base::Base16), "md5:900150983cd24fb0d6963f7d28e17f72");
+    }
+
+    TEST(hashString, testKnownSHA1Hashes1) {
+        // values taken from: https://tools.ietf.org/html/rfc3174
+        auto s = "abc";
+        auto hash = hashString(HashType::htSHA1, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),"sha1:a9993e364706816aba3e25717850c26c9cd0d89d");
+    }
+
+    TEST(hashString, testKnownSHA1Hashes2) {
+        // values taken from: https://tools.ietf.org/html/rfc3174
+        auto s = "abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq";
+        auto hash = hashString(HashType::htSHA1, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),"sha1:84983e441c3bd26ebaae4aa1f95129e5e54670f1");
+    }
+
+    TEST(hashString, testKnownSHA256Hashes1) {
+        // values taken from: https://tools.ietf.org/html/rfc4634
+        auto s = "abc";
+
+        auto hash = hashString(HashType::htSHA256, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),
+                "sha256:ba7816bf8f01cfea414140de5dae2223b00361a396177a9cb410ff61f20015ad");
+    }
+
+    TEST(hashString, testKnownSHA256Hashes2) {
+        // values taken from: https://tools.ietf.org/html/rfc4634
+        auto s = "abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq";
+        auto hash = hashString(HashType::htSHA256, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),
+                "sha256:248d6a61d20638b8e5c026930c3e6039a33ce45964ff2167f6ecedd419db06c1");
+    }
+
+    TEST(hashString, testKnownSHA512Hashes1) {
+        // values taken from: https://tools.ietf.org/html/rfc4634
+        auto s = "abc";
+        auto hash = hashString(HashType::htSHA512, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),
+                "sha512:ddaf35a193617abacc417349ae20413112e6fa4e89a9"
+                "7ea20a9eeee64b55d39a2192992a274fc1a836ba3c23a3feebbd"
+                "454d4423643ce80e2a9ac94fa54ca49f");
+    }
+
+    TEST(hashString, testKnownSHA512Hashes2) {
+        // values taken from: https://tools.ietf.org/html/rfc4634
+        auto s = "abcdefghbcdefghicdefghijdefghijkefghijklfghijklmghijklmnhijklmnoijklmnopjklmnopqklmnopqrlmnopqrsmnopqrstnopqrstu";
+
+        auto hash = hashString(HashType::htSHA512, s);
+        ASSERT_EQ(hash.to_string(Base::Base16),
+                "sha512:8e959b75dae313da8cf4f72814fc143f8f7779c6eb9f7fa1"
+                "7299aeadb6889018501d289e4900f7e4331b99dec4b5433a"
+                "c7d329eeb6dd26545e96e55b874be909");
+    }
+
+    TEST(hashString, hashingWithUnknownAlgoExits) {
+        auto s = "unknown";
+        ASSERT_DEATH(hashString(HashType::htUnknown, s), "");
+    }
+}
diff --git a/src/libutil/tests/json.cc b/src/libutil/tests/json.cc
new file mode 100644
index 000000000..dea73f53a
--- /dev/null
+++ b/src/libutil/tests/json.cc
@@ -0,0 +1,193 @@
+#include "json.hh"
+#include <gtest/gtest.h>
+#include <sstream>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * toJSON
+     * --------------------------------------------------------------------------*/
+
+    TEST(toJSON, quotesCharPtr) {
+        const char* input = "test";
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "\"test\"");
+    }
+
+    TEST(toJSON, quotesStdString) {
+        std::string input = "test";
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "\"test\"");
+    }
+
+    TEST(toJSON, convertsNullptrtoNull) {
+        auto input = nullptr;
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "null");
+    }
+
+    TEST(toJSON, convertsNullToNull) {
+        const char* input = 0;
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "null");
+    }
+
+
+    TEST(toJSON, convertsFloat) {
+        auto input = 1.024f;
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "1.024");
+    }
+
+    TEST(toJSON, convertsDouble) {
+        const double input = 1.024;
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "1.024");
+    }
+
+    TEST(toJSON, convertsBool) {
+        auto input = false;
+        std::stringstream out;
+        toJSON(out, input);
+
+        ASSERT_EQ(out.str(), "false");
+    }
+
+    TEST(toJSON, quotesTab) {
+        std::stringstream out;
+        toJSON(out, "\t");
+
+        ASSERT_EQ(out.str(), "\"\\t\"");
+    }
+
+    TEST(toJSON, quotesNewline) {
+        std::stringstream out;
+        toJSON(out, "\n");
+
+        ASSERT_EQ(out.str(), "\"\\n\"");
+    }
+
+    TEST(toJSON, quotesCreturn) {
+        std::stringstream out;
+        toJSON(out, "\r");
+
+        ASSERT_EQ(out.str(), "\"\\r\"");
+    }
+
+    TEST(toJSON, quotesCreturnNewLine) {
+        std::stringstream out;
+        toJSON(out, "\r\n");
+
+        ASSERT_EQ(out.str(), "\"\\r\\n\"");
+    }
+
+    TEST(toJSON, quotesDoublequotes) {
+        std::stringstream out;
+        toJSON(out, "\"");
+
+        ASSERT_EQ(out.str(), "\"\\\"\"");
+    }
+
+    TEST(toJSON, substringEscape) {
+        std::stringstream out;
+        const char *s = "foo\t";
+        toJSON(out, s+3, s + strlen(s));
+
+        ASSERT_EQ(out.str(), "\"\\t\"");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * JSONObject
+     * --------------------------------------------------------------------------*/
+
+    TEST(JSONObject, emptyObject) {
+        std::stringstream out;
+        {
+            JSONObject t(out);
+        }
+        ASSERT_EQ(out.str(), "{}");
+    }
+
+    TEST(JSONObject, objectWithList) {
+        std::stringstream out;
+        {
+            JSONObject t(out);
+            auto l = t.list("list");
+            l.elem("element");
+        }
+        ASSERT_EQ(out.str(), R"#({"list":["element"]})#");
+    }
+
+    TEST(JSONObject, objectWithListIndent) {
+        std::stringstream out;
+        {
+            JSONObject t(out, true);
+            auto l = t.list("list");
+            l.elem("element");
+        }
+        ASSERT_EQ(out.str(),
+R"#({
+  "list": [
+    "element"
+  ]
+})#");
+    }
+
+    TEST(JSONObject, objectWithPlaceholderAndList) {
+        std::stringstream out;
+        {
+            JSONObject t(out);
+            auto l = t.placeholder("list");
+            l.list().elem("element");
+        }
+
+        ASSERT_EQ(out.str(), R"#({"list":["element"]})#");
+    }
+
+    TEST(JSONObject, objectWithPlaceholderAndObject) {
+        std::stringstream out;
+        {
+            JSONObject t(out);
+            auto l = t.placeholder("object");
+            l.object().attr("key", "value");
+        }
+
+        ASSERT_EQ(out.str(), R"#({"object":{"key":"value"}})#");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * JSONList
+     * --------------------------------------------------------------------------*/
+
+    TEST(JSONList, empty) {
+        std::stringstream out;
+        {
+            JSONList l(out);
+        }
+        ASSERT_EQ(out.str(), R"#([])#");
+    }
+
+    TEST(JSONList, withElements) {
+        std::stringstream out;
+        {
+            JSONList l(out);
+            l.elem("one");
+            l.object();
+            l.placeholder().write("three");
+        }
+        ASSERT_EQ(out.str(), R"#(["one",{},"three"])#");
+    }
+}
+
diff --git a/src/libutil/tests/local.mk b/src/libutil/tests/local.mk
index a297edb64..815e18560 100644
--- a/src/libutil/tests/local.mk
+++ b/src/libutil/tests/local.mk
@@ -8,7 +8,7 @@ libutil-tests_INSTALL_DIR :=
 
 libutil-tests_SOURCES := $(wildcard $(d)/*.cc)
 
-libutil-tests_CXXFLAGS += -I src/libutil
+libutil-tests_CXXFLAGS += -I src/libutil -I src/libexpr
 
 libutil-tests_LIBS = libutil
 
diff --git a/src/libutil/tests/logging.cc b/src/libutil/tests/logging.cc
new file mode 100644
index 000000000..fbdc91253
--- /dev/null
+++ b/src/libutil/tests/logging.cc
@@ -0,0 +1,255 @@
+#include "logging.hh"
+#include "nixexpr.hh"
+#include "util.hh"
+
+#include <gtest/gtest.h>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * logEI
+     * --------------------------------------------------------------------------*/
+
+    TEST(logEI, catpuresBasicProperties) {
+
+        MakeError(TestError, Error);
+        ErrorInfo::programName = std::optional("error-unit-test");
+
+        try {
+            throw TestError("an error for testing purposes");
+        } catch (Error &e) {
+            testing::internal::CaptureStderr();
+            logger->logEI(e.info());
+            auto str = testing::internal::GetCapturedStderr();
+
+            ASSERT_STREQ(str.c_str(),"\x1B[31;1merror:\x1B[0m\x1B[34;1m --- TestError ------------------------------------ error-unit-test\x1B[0m\nan error for testing purposes\n");
+        }
+    }
+
+    TEST(logEI, appendingHintsToPreviousError) {
+
+        MakeError(TestError, Error);
+        ErrorInfo::programName = std::optional("error-unit-test");
+
+        try {
+            auto e = Error("initial error");
+            throw TestError(e.info());
+        } catch (Error &e) {
+            ErrorInfo ei = e.info();
+            ei.hint = hintfmt("%s; subsequent error message.", normaltxt(e.info().hint ? e.info().hint->str() : ""));
+
+            testing::internal::CaptureStderr();
+            logger->logEI(ei);
+            auto str = testing::internal::GetCapturedStderr();
+
+            ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- TestError ------------------------------------ error-unit-test\x1B[0m\n\x1B[33;1m\x1B[0minitial error\x1B[0m; subsequent error message.\n");
+        }
+
+    }
+
+    TEST(logEI, picksUpSysErrorExitCode) {
+
+        MakeError(TestError, Error);
+        ErrorInfo::programName = std::optional("error-unit-test");
+
+        try {
+            auto x = readFile(-1);
+        }
+        catch (SysError &e) {
+            testing::internal::CaptureStderr();
+            logError(e.info());
+            auto str = testing::internal::GetCapturedStderr();
+
+            ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- SysError ------------------------------------- error-unit-test\x1B[0m\n\x1B[33;1m\x1B[0mstatting file\x1B[0m: \x1B[33;1mBad file descriptor\x1B[0m\n");
+
+        }
+    }
+
+    TEST(logEI, loggingErrorOnInfoLevel) {
+        testing::internal::CaptureStderr();
+
+        logger->logEI({ .level = lvlInfo,
+                .name = "Info name",
+                .description = "Info description",
+                });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[32;1minfo:\x1B[0m\x1B[34;1m --- Info name ------------------------------------- error-unit-test\x1B[0m\nInfo description\n");
+    }
+
+    TEST(logEI, loggingErrorOnTalkativeLevel) {
+        verbosity = lvlTalkative;
+
+        testing::internal::CaptureStderr();
+
+        logger->logEI({ .level = lvlTalkative,
+                        .name = "Talkative name",
+                        .description = "Talkative description",
+                        });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[32;1mtalk:\x1B[0m\x1B[34;1m --- Talkative name -------------------------------- error-unit-test\x1B[0m\nTalkative description\n");
+    }
+
+    TEST(logEI, loggingErrorOnChattyLevel) {
+        verbosity = lvlChatty;
+
+        testing::internal::CaptureStderr();
+
+        logger->logEI({ .level = lvlChatty,
+                        .name = "Chatty name",
+                        .description = "Talkative description",
+                        });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[32;1mchat:\x1B[0m\x1B[34;1m --- Chatty name ----------------------------------- error-unit-test\x1B[0m\nTalkative description\n");
+    }
+
+    TEST(logEI, loggingErrorOnDebugLevel) {
+        verbosity = lvlDebug;
+
+        testing::internal::CaptureStderr();
+
+        logger->logEI({ .level = lvlDebug,
+                        .name = "Debug name",
+                        .description = "Debug description",
+                        });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[33;1mdebug:\x1B[0m\x1B[34;1m --- Debug name ----------------------------------- error-unit-test\x1B[0m\nDebug description\n");
+    }
+
+    TEST(logEI, loggingErrorOnVomitLevel) {
+        verbosity = lvlVomit;
+
+        testing::internal::CaptureStderr();
+
+        logger->logEI({ .level = lvlVomit,
+                        .name = "Vomit name",
+                        .description = "Vomit description",
+                        });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[32;1mvomit:\x1B[0m\x1B[34;1m --- Vomit name ----------------------------------- error-unit-test\x1B[0m\nVomit description\n");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * logError
+     * --------------------------------------------------------------------------*/
+
+
+    TEST(logError, logErrorWithoutHintOrCode) {
+        testing::internal::CaptureStderr();
+
+        logError({
+                .name = "name",
+                .description = "error description",
+                });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- name ----------------------------------------- error-unit-test\x1B[0m\nerror description\n");
+    }
+
+    TEST(logError, logErrorWithPreviousAndNextLinesOfCode) {
+        SymbolTable testTable;
+        auto problem_file = testTable.create("myfile.nix");
+
+        testing::internal::CaptureStderr();
+
+        logError({
+                .name = "error name",
+                .description = "error with code lines",
+                .hint = hintfmt("this hint has %1% templated %2%!!",
+                        "yellow",
+                        "values"),
+                .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13),
+                .prevLineOfCode = "previous line of code",
+                .errLineOfCode = "this is the problem line of code",
+                .nextLineOfCode = "next line of code",
+                }});
+
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- error name ----------------------------------- error-unit-test\x1B[0m\nin file: \x1B[34;1mmyfile.nix (40:13)\x1B[0m\n\nerror with code lines\n\n    39| previous line of code\n    40| this is the problem line of code\n      |             \x1B[31;1m^\x1B[0m\n    41| next line of code\n\nthis hint has \x1B[33;1myellow\x1B[0m templated \x1B[33;1mvalues\x1B[0m!!\n");
+    }
+
+    TEST(logError, logErrorWithoutLinesOfCode) {
+        SymbolTable testTable;
+        auto problem_file = testTable.create("myfile.nix");
+        testing::internal::CaptureStderr();
+
+        logError({
+                .name = "error name",
+                .description = "error without any code lines.",
+                .hint = hintfmt("this hint has %1% templated %2%!!",
+                        "yellow",
+                        "values"),
+                .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13)
+                }});
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- error name ----------------------------------- error-unit-test\x1B[0m\nin file: \x1B[34;1mmyfile.nix (40:13)\x1B[0m\n\nerror without any code lines.\n\nthis hint has \x1B[33;1myellow\x1B[0m templated \x1B[33;1mvalues\x1B[0m!!\n");
+    }
+
+    TEST(logError, logErrorWithOnlyHintAndName) {
+        SymbolTable testTable;
+        auto problem_file = testTable.create("myfile.nix");
+        testing::internal::CaptureStderr();
+
+        logError({
+                .name = "error name",
+                .hint = hintfmt("hint %1%", "only"),
+                .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13)
+                }});
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[31;1merror:\x1B[0m\x1B[34;1m --- error name ----------------------------------- error-unit-test\x1B[0m\nin file: \x1B[34;1mmyfile.nix (40:13)\x1B[0m\n\nhint \x1B[33;1monly\x1B[0m\n");
+
+    }
+
+    /* ----------------------------------------------------------------------------
+     * logWarning
+     * --------------------------------------------------------------------------*/
+
+    TEST(logWarning, logWarningWithNameDescriptionAndHint) {
+        testing::internal::CaptureStderr();
+
+        logWarning({
+            .name = "name",
+            .description = "error description",
+            .hint = hintfmt("there was a %1%", "warning"),
+        });
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[33;1mwarning:\x1B[0m\x1B[34;1m --- name --------------------------------------- error-unit-test\x1B[0m\nerror description\n\nthere was a \x1B[33;1mwarning\x1B[0m\n");
+    }
+
+    TEST(logWarning, logWarningWithFileLineNumAndCode) {
+
+        SymbolTable testTable;
+        auto problem_file = testTable.create("myfile.nix");
+
+        testing::internal::CaptureStderr();
+
+        logWarning({
+                .name = "warning name",
+                .description = "warning description",
+                .hint = hintfmt("this hint has %1% templated %2%!!",
+                        "yellow",
+                        "values"),
+                .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13),
+                .prevLineOfCode = std::nullopt,
+                .errLineOfCode = "this is the problem line of code",
+                .nextLineOfCode = std::nullopt
+                }});
+
+
+        auto str = testing::internal::GetCapturedStderr();
+        ASSERT_STREQ(str.c_str(), "\x1B[33;1mwarning:\x1B[0m\x1B[34;1m --- warning name ------------------------------- error-unit-test\x1B[0m\nin file: \x1B[34;1mmyfile.nix (40:13)\x1B[0m\n\nwarning description\n\n    40| this is the problem line of code\n      |             \x1B[31;1m^\x1B[0m\n\nthis hint has \x1B[33;1myellow\x1B[0m templated \x1B[33;1mvalues\x1B[0m!!\n");
+    }
+
+}
diff --git a/src/libutil/tests/lru-cache.cc b/src/libutil/tests/lru-cache.cc
new file mode 100644
index 000000000..091d3d5ed
--- /dev/null
+++ b/src/libutil/tests/lru-cache.cc
@@ -0,0 +1,130 @@
+#include "lru-cache.hh"
+#include <gtest/gtest.h>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * size
+     * --------------------------------------------------------------------------*/
+
+    TEST(LRUCache, sizeOfEmptyCacheIsZero) {
+        LRUCache<std::string, std::string> c(10);
+        ASSERT_EQ(c.size(), 0);
+    }
+
+    TEST(LRUCache, sizeOfSingleElementCacheIsOne) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("foo", "bar");
+        ASSERT_EQ(c.size(), 1);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * upsert / get
+     * --------------------------------------------------------------------------*/
+
+    TEST(LRUCache, getFromEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        auto val = c.get("x");
+        ASSERT_EQ(val.has_value(), false);
+    }
+
+    TEST(LRUCache, getExistingValue) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("foo", "bar");
+        auto val = c.get("foo");
+        ASSERT_EQ(val, "bar");
+    }
+
+    TEST(LRUCache, getNonExistingValueFromNonEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("foo", "bar");
+        auto val = c.get("another");
+        ASSERT_EQ(val.has_value(), false);
+    }
+
+    TEST(LRUCache, upsertOnZeroCapacityCache) {
+        LRUCache<std::string, std::string> c(0);
+        c.upsert("foo", "bar");
+        auto val = c.get("foo");
+        ASSERT_EQ(val.has_value(), false);
+    }
+
+    TEST(LRUCache, updateExistingValue) {
+        LRUCache<std::string, std::string> c(1);
+        c.upsert("foo", "bar");
+
+        auto val = c.get("foo");
+        ASSERT_EQ(val.value_or("error"), "bar");
+        ASSERT_EQ(c.size(), 1);
+
+        c.upsert("foo", "changed");
+        val = c.get("foo");
+        ASSERT_EQ(val.value_or("error"), "changed");
+        ASSERT_EQ(c.size(), 1);
+    }
+
+    TEST(LRUCache, overwriteOldestWhenCapacityIsReached) {
+        LRUCache<std::string, std::string> c(3);
+        c.upsert("one", "eins");
+        c.upsert("two", "zwei");
+        c.upsert("three", "drei");
+
+        ASSERT_EQ(c.size(), 3);
+        ASSERT_EQ(c.get("one").value_or("error"), "eins");
+
+        // exceed capacity
+        c.upsert("another", "whatever");
+
+        ASSERT_EQ(c.size(), 3);
+        // Retrieving "one" makes it the most recent element thus
+        // two will be the oldest one and thus replaced.
+        ASSERT_EQ(c.get("two").has_value(), false);
+        ASSERT_EQ(c.get("another").value(), "whatever");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * clear
+     * --------------------------------------------------------------------------*/
+
+    TEST(LRUCache, clearEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        c.clear();
+        ASSERT_EQ(c.size(), 0);
+    }
+
+    TEST(LRUCache, clearNonEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("one", "eins");
+        c.upsert("two", "zwei");
+        c.upsert("three", "drei");
+        ASSERT_EQ(c.size(), 3);
+        c.clear();
+        ASSERT_EQ(c.size(), 0);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * erase
+     * --------------------------------------------------------------------------*/
+
+    TEST(LRUCache, eraseFromEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        ASSERT_EQ(c.erase("foo"), false);
+        ASSERT_EQ(c.size(), 0);
+    }
+
+    TEST(LRUCache, eraseMissingFromNonEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("one", "eins");
+        ASSERT_EQ(c.erase("foo"), false);
+        ASSERT_EQ(c.size(), 1);
+        ASSERT_EQ(c.get("one").value_or("error"), "eins");
+    }
+
+    TEST(LRUCache, eraseFromNonEmptyCache) {
+        LRUCache<std::string, std::string> c(10);
+        c.upsert("one", "eins");
+        ASSERT_EQ(c.erase("one"), true);
+        ASSERT_EQ(c.size(), 0);
+        ASSERT_EQ(c.get("one").value_or("empty"), "empty");
+    }
+}
diff --git a/src/libutil/tests/tests.cc b/src/libutil/tests/tests.cc
index 5abfe2c9c..8e77ccbe1 100644
--- a/src/libutil/tests/tests.cc
+++ b/src/libutil/tests/tests.cc
@@ -5,6 +5,8 @@
 
 namespace nix {
 
+/* ----------- tests for util.hh ------------------------------------------------*/
+
     /* ----------------------------------------------------------------------------
      * absPath
      * --------------------------------------------------------------------------*/
@@ -15,6 +17,9 @@ namespace nix {
         ASSERT_EQ(p, "/");
     }
 
+
+
+
     TEST(absPath, turnsEmptyPathIntoCWD) {
         char cwd[PATH_MAX+1];
         auto p = absPath("");
@@ -581,5 +586,4 @@ namespace nix {
 
         ASSERT_EQ(filterANSIEscapes(s, true), "foo     bar     baz" );
     }
-
 }
diff --git a/src/libutil/tests/url.cc b/src/libutil/tests/url.cc
new file mode 100644
index 000000000..80646ad3e
--- /dev/null
+++ b/src/libutil/tests/url.cc
@@ -0,0 +1,266 @@
+#include "url.hh"
+#include <gtest/gtest.h>
+
+namespace nix {
+
+/* ----------- tests for url.hh --------------------------------------------------*/
+
+    string print_map(std::map<string, string> m) {
+        std::map<string, string>::iterator it;
+        string s = "{ ";
+        for (it = m.begin(); it != m.end(); ++it) {
+            s += "{ ";
+            s += it->first;
+            s += " = ";
+            s += it->second;
+            s += " } ";
+        }
+        s += "}";
+        return s;
+    }
+
+
+    std::ostream& operator<<(std::ostream& os, const ParsedURL& p) {
+        return os << "\n"
+                  << "url:       " << p.url << "\n"
+                  << "base:      " << p.base << "\n"
+                  << "scheme:    " << p.scheme << "\n"
+                  << "authority: " << p.authority.value() << "\n"
+                  << "path:      " << p.path << "\n"
+                  << "query:     " << print_map(p.query) << "\n"
+                  << "fragment:  " << p.fragment << "\n";
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrl) {
+        auto s = "http://www.example.org/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://www.example.org/file.tar.gz",
+            .base = "http://www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpsUrl) {
+        auto s = "https://www.example.org/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "https://www.example.org/file.tar.gz",
+            .base = "https://www.example.org/file.tar.gz",
+            .scheme = "https",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrlWithQueryAndFragment) {
+        auto s = "https://www.example.org/file.tar.gz?download=fast&when=now#hello";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "https://www.example.org/file.tar.gz",
+            .base = "https://www.example.org/file.tar.gz",
+            .scheme = "https",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "download", "fast" }, { "when", "now" } },
+            .fragment = "hello",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrlWithComplexFragment) {
+        auto s = "http://www.example.org/file.tar.gz?field=value#?foo=bar%23";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://www.example.org/file.tar.gz",
+            .base = "http://www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "field", "value" } },
+            .fragment = "?foo=bar#",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+
+    TEST(parseURL, parseIPv4Address) {
+        auto s = "http://127.0.0.1:8080/file.tar.gz?download=fast&when=now#hello";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://127.0.0.1:8080/file.tar.gz",
+            .base = "https://127.0.0.1:8080/file.tar.gz",
+            .scheme = "http",
+            .authority = "127.0.0.1:8080",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "download", "fast" }, { "when", "now" } },
+            .fragment = "hello",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parseIPv6Address) {
+        auto s = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .base = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .scheme = "http",
+            .authority = "[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .path = "",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+
+    }
+
+    TEST(parseURL, parseEmptyQueryParams) {
+        auto s = "http://127.0.0.1:8080/file.tar.gz?&&&&&";
+        auto parsed = parseURL(s);
+        ASSERT_EQ(parsed.query, (StringMap) { });
+    }
+
+    TEST(parseURL, parseUserPassword) {
+        auto s = "http://user:pass@www.example.org:8080/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://user:pass@www.example.org/file.tar.gz",
+            .base = "http://user:pass@www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "user:pass@www.example.org:8080",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parseFileURLWithQueryAndFragment) {
+        auto s = "file:///none/of/your/business";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "",
+            .base = "",
+            .scheme = "file",
+            .authority = "",
+            .path = "/none/of/your/business",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+
+    }
+
+    TEST(parseURL, parsedUrlsIsEqualToItself) {
+        auto s = "http://www.example.org/file.tar.gz";
+        auto url = parseURL(s);
+
+        ASSERT_TRUE(url == url);
+    }
+
+    TEST(parseURL, parseFTPUrl) {
+        auto s = "ftp://ftp.nixos.org/downloads/nixos.iso";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "ftp://ftp.nixos.org/downloads/nixos.iso",
+            .base = "ftp://ftp.nixos.org/downloads/nixos.iso",
+            .scheme = "ftp",
+            .authority = "ftp.nixos.org",
+            .path = "/downloads/nixos.iso",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesAnythingInUriFormat) {
+        auto s = "whatever://github.com/NixOS/nixpkgs.git";
+        auto parsed = parseURL(s);
+    }
+
+    TEST(parseURL, parsesAnythingInUriFormatWithoutDoubleSlash) {
+        auto s = "whatever:github.com/NixOS/nixpkgs.git";
+        auto parsed = parseURL(s);
+    }
+
+    TEST(parseURL, emptyStringIsInvalidURL) {
+        ASSERT_THROW(parseURL(""), Error);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * decodeQuery
+     * --------------------------------------------------------------------------*/
+
+    TEST(decodeQuery, emptyStringYieldsEmptyMap) {
+        auto d = decodeQuery("");
+        ASSERT_EQ(d, (StringMap) { });
+    }
+
+    TEST(decodeQuery, simpleDecode) {
+        auto d = decodeQuery("yi=one&er=two");
+        ASSERT_EQ(d, ((StringMap) { { "yi", "one" }, { "er", "two" } }));
+    }
+
+    TEST(decodeQuery, decodeUrlEncodedArgs) {
+        auto d = decodeQuery("arg=%3D%3D%40%3D%3D");
+        ASSERT_EQ(d, ((StringMap) { { "arg", "==@==" } }));
+    }
+
+    TEST(decodeQuery, decodeArgWithEmptyValue) {
+        auto d = decodeQuery("arg=");
+        ASSERT_EQ(d, ((StringMap) { { "arg", ""} }));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * percentDecode
+     * --------------------------------------------------------------------------*/
+
+    TEST(percentDecode, decodesUrlEncodedString) {
+        string s = "==@==";
+        string d = percentDecode("%3D%3D%40%3D%3D");
+        ASSERT_EQ(d, s);
+    }
+
+    TEST(percentDecode, multipleDecodesAreIdempotent) {
+        string once = percentDecode("%3D%3D%40%3D%3D");
+        string twice = percentDecode(once);
+
+        ASSERT_EQ(once, twice);
+    }
+
+    TEST(percentDecode, trailingPercent) {
+        string s = "==@==%";
+        string d = percentDecode("%3D%3D%40%3D%3D%25");
+
+        ASSERT_EQ(d, s);
+    }
+
+}
diff --git a/src/libutil/tests/xml-writer.cc b/src/libutil/tests/xml-writer.cc
new file mode 100644
index 000000000..adcde25c9
--- /dev/null
+++ b/src/libutil/tests/xml-writer.cc
@@ -0,0 +1,105 @@
+#include "xml-writer.hh"
+#include <gtest/gtest.h>
+#include <sstream>
+
+namespace nix {
+
+    /* ----------------------------------------------------------------------------
+     * XMLWriter
+     * --------------------------------------------------------------------------*/
+
+    TEST(XMLWriter, emptyObject) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n");
+    }
+
+    TEST(XMLWriter, objectWithEmptyElement) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            t.openElement("foobar");
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar></foobar>");
+    }
+
+    TEST(XMLWriter, objectWithElementWithAttrs) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            XMLAttrs attrs = {
+                { "foo", "bar" }
+            };
+            t.openElement("foobar", attrs);
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar foo=\"bar\"></foobar>");
+    }
+
+    TEST(XMLWriter, objectWithElementWithEmptyAttrs) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            XMLAttrs attrs = {};
+            t.openElement("foobar", attrs);
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar></foobar>");
+    }
+
+    TEST(XMLWriter, objectWithElementWithAttrsEscaping) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            XMLAttrs attrs = {
+                { "<key>", "<value>" }
+            };
+            t.openElement("foobar", attrs);
+        }
+
+        // XXX: While "<value>" is escaped, "<key>" isn't which I think is a bug.
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar <key>=\"&lt;value&gt;\"></foobar>");
+    }
+
+    TEST(XMLWriter, objectWithElementWithAttrsIndented) {
+        std::stringstream out;
+        {
+            XMLWriter t(true, out);
+            XMLAttrs attrs = {
+                { "foo", "bar" }
+            };
+            t.openElement("foobar", attrs);
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar foo=\"bar\">\n</foobar>\n");
+    }
+
+    TEST(XMLWriter, writeEmptyElement) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            t.writeEmptyElement("foobar");
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar />");
+    }
+
+    TEST(XMLWriter, writeEmptyElementWithAttributes) {
+        std::stringstream out;
+        {
+            XMLWriter t(false, out);
+            XMLAttrs attrs = {
+                { "foo", "bar" }
+            };
+            t.writeEmptyElement("foobar", attrs);
+
+        }
+
+        ASSERT_EQ(out.str(), "<?xml version='1.0' encoding='utf-8'?>\n<foobar foo=\"bar\" />");
+    }
+
+}
diff --git a/src/libutil/url.cc b/src/libutil/url.cc
index 5d5328e5d..88c09eef9 100644
--- a/src/libutil/url.cc
+++ b/src/libutil/url.cc
@@ -4,6 +4,7 @@
 namespace nix {
 
 std::regex refRegex(refRegexS, std::regex::ECMAScript);
+std::regex badGitRefRegex(badGitRefRegexS, std::regex::ECMAScript);
 std::regex revRegex(revRegexS, std::regex::ECMAScript);
 std::regex flakeIdRegex(flakeIdRegexS, std::regex::ECMAScript);
 
diff --git a/src/libutil/url.hh b/src/libutil/url.hh
index a184eadce..2ef88ef2a 100644
--- a/src/libutil/url.hh
+++ b/src/libutil/url.hh
@@ -49,6 +49,12 @@ const static std::string pathRegex = "(?:" + segmentRegex + "(?:/" + segmentRege
 const static std::string refRegexS = "[a-zA-Z0-9][a-zA-Z0-9_.-]*"; // FIXME: check
 extern std::regex refRegex;
 
+// Instead of defining what a good Git Ref is, we define what a bad Git Ref is
+// This is because of the definition of a ref in refs.c in https://github.com/git/git
+// See tests/fetchGitRefs.sh for the full definition
+const static std::string badGitRefRegexS = "//|^[./]|/\\.|\\.\\.|[[:cntrl:][:space:]:?^~\[]|\\\\|\\*|\\.lock$|\\.lock/|@\\{|[/.]$|^@$|^$";
+extern std::regex badGitRefRegex;
+
 // A Git revision (a SHA-1 commit hash).
 const static std::string revRegexS = "[0-9a-fA-F]{40}";
 extern std::regex revRegex;