Persist

This library allows to preserve structural sharing of immer containers while serializing and deserializing them.

Motivation: serialization

Structural sharing allows immer containers to be efficient. At runtime, two distinct containers can be operated on independently but internally they share nodes and use memory efficiently in that way. However when such containers are serialized in a simple direct way, for example, as lists, this sharing is lost: they become truly independent, same data is stored multiple times on disk and later, when it is read from disk, in memory.

This library operates on the internal structure of immer containers: allowing it to be serialized, deserialized and transformed. This enables more efficient storage, particularly when many nodes are reused, and, even more importantly, preserving structural sharing after deserializing the containers.

Motivation: transformation

Consider this scenario: an application has a document type that internally uses an immer container in multiple places, for example, an immer::vector<std::string>. Some of these vectors would be completely identical, while others would have just a few elements different (stored in an undo history, for example). The goal is to apply a transformation function to these vectors.

A direct approach would be to take each vector and create a new vector by applying the transformation function for each element. However, after this process, all the structural sharing of the original containers would be lost: the result would be multiple independent vectors without any structural sharing.

This library enables the application of the transformation function directly on the nodes, preserving structural sharing. Additionally, regardless of how many times a node is reused, the transformation needs to be performed only once.

Dependencies

In addition to the dependencies of immer, this library makes use of C++20, Boost.Hana, fmt and cereal.

First example

For this example, we’ll use a document type that contains two immer vectors.

// Set the BL constant to 1, so that only 2 elements are stored in leaves.
// This allows to demonstrate structural sharing even in vectors with just a few
// elements.
using vector_one =
    immer::vector<int, immer::default_memory_policy, immer::default_bits, 1>;

struct document
{
    vector_one ints;
    vector_one ints2;

    friend bool operator==(const document&, const document&) = default;

    // Make the struct serializable with cereal as usual, nothing special
    // related to immer-persist.
    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(ints), CEREAL_NVP(ints2));
    }
};

using json_t = nlohmann::json;

Let’s make the document struct compatible with boost::hana. This way, the persist library can determine what pool types are needed and to name the pools.

BOOST_HANA_ADAPT_STRUCT(document, ints, ints2);

Let’s say we have two vectors v1 and v2, where v2 is derived from v1 so that it shares data with it:

    const auto v1    = vector_one{1, 2, 3};
    const auto v2    = v1.push_back(4).push_back(5).push_back(6);
    const auto value = document{v1, v2};

We can serialize the document using cereal with this:

            auto os = std::ostringstream{};
            {
                auto ar = cereal::JSONOutputArchive{os};
                ar(value);
            }
            return os.str();

Generating a JSON like this one:

{"value0": {"ints": [1, 2, 3], "ints2": [1, 2, 3, 4, 5, 6]}}

As you can see, ints and ints2 contain the full linearization of each vector. The structural sharing between these two data structures is not represented in its serialized form. However, with immer-persist we can serialize it with:

        const auto policy =
            immer::persist::hana_struct_auto_member_name_policy(document{});
        const auto str = immer::persist::cereal_save_with_pools(value, policy);

Which generates some JSON like this:

        const auto expected_json = json_t::parse(R"(
{
  "value0": {"ints": 0, "ints2": 1},
  "pools": {
    "ints": {
      "B": 5,
      "BL": 1,
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "leaves": [[1, [3]], [2, [1, 2]], [4, [5, 6]], [5, [3, 4]]],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    }
  }
}
        )");

As you can see, the value is serialized with every immer container replaced by an identifier. This identifier is a key into a pool, which is serialized just after.

A pool represents a set of immer containers of a specific type. For example, we may have a pool that contains all immer::vector<int> of our document. You can think of it as a small database of immer containers. When serializing the pool, the internal structure of all those immer containers is written, preserving the structural sharing between those containers. The nodes of the trees that make up the immer containers are directly represented in the JSON and, because we are representing all the containers as a whole, those nodes that are referenced in multiple trees can be stored only once. That same structure is preserved when reading the pool back from disk and reconstructing the vectors (and other containers) from it, thus allowing us to preserve the structural sharing across sessions.

Note

Currently, immer-persist makes a distiction between pools used for saving containers (output pools) and for loading containers (input pools), similar to cereal with its InputArchive and OutputArchive distiction.

Currently, immer-persist focuses on JSON as the serialization format and uses the cereal library internally. In principle, other formats and serialization libraries could be supported in the future.

Custom policy

We can use policy to control the names of the pools for each container.

For this example, let’s define a new document type doc_2. It will also contain another type extra_data with a vector of strings in it. To demonstrate the responsibilities of the policy, the doc_2 type will not be a boost::hana::Struct and will not allow for compile-time reflection.

using vector_str = immer::
    vector<std::string, immer::default_memory_policy, immer::default_bits, 1>;

struct extra_data
{
    vector_str comments;

    friend bool operator==(const extra_data&, const extra_data&) = default;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(comments));
    }
};

struct doc_2
{
    vector_one ints;
    vector_one ints2;
    vector_str strings;
    extra_data extra;

    friend bool operator==(const doc_2&, const doc_2&) = default;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(ints),
           CEREAL_NVP(ints2),
           CEREAL_NVP(strings),
           CEREAL_NVP(extra));
    }
};

We define the doc_2_policy as following:

struct doc_2_policy
{
    auto get_pool_types(const auto&) const
    {
        return boost::hana::tuple_t<vector_one, vector_str>;
    }

    template <class Archive>
    void save(Archive& ar, const doc_2& doc2_value) const
    {
        ar(CEREAL_NVP(doc2_value));
    }

    template <class Archive>
    void load(Archive& ar, doc_2& doc2_value) const
    {
        ar(CEREAL_NVP(doc2_value));
    }

    auto get_pool_name(const vector_one&) const { return "vector_of_ints"; }
    auto get_pool_name(const vector_str&) const { return "vector_of_strings"; }
};

The get_pool_types function returns the types of containers that should be serialized with pools, in this case it’s both vector of ints and strings. The save and load functions control the name of the document node, in this case it is doc2_value. And the get_pool_name overloaded functions supply the name of the pool for each corresponding immer container. To create and serialize a value of doc_2, you can use the following approach:

    const auto v1   = vector_one{1, 2, 3};
    const auto v2   = v1.push_back(4).push_back(5).push_back(6);
    const auto str1 = vector_str{"one", "two"};
    const auto str2 =
        str1.push_back("three").push_back("four").push_back("five");
    const auto value = doc_2{v1, v2, str1, extra_data{str2}};

    const auto str =
        immer::persist::cereal_save_with_pools(value, doc_2_policy{});

The serialized JSON looks like this:

    const auto expected_json = json_t::parse(R"(
{
  "doc2_value": {"ints": 0, "ints2": 1, "strings": 0, "extra": {"comments": 1}},
  "pools": {
    "vector_of_ints": {
      "B": 5,
      "BL": 1,
      "leaves": [[1, [3]], [2, [1, 2]], [4, [5, 6]], [5, [3, 4]]],
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    },
    "vector_of_strings": {
      "B": 5,
      "BL": 1,
      "leaves": [[1, ["one", "two"]], [3, ["five"]], [4, ["three", "four"]]],
      "inners": [
        [0, {"children": [], "relaxed": false}],
        [2, {"children": [1, 4], "relaxed": false}]
      ],
      "vectors": [{"root": 0, "tail": 1}, {"root": 2, "tail": 3}]
    }
  }
}
    )");

And it can also be loaded from JSON like this:

    const auto loaded_value =
        immer::persist::cereal_load_with_pools<doc_2>(str, doc_2_policy{});

This example also demonstrates a scenario in which the main document type doc_2 contains another type extra_data with a vector. As you can see in the resulting JSON, nested types are also serialized with pools: "extra": {"comments": 1}. Only the ID of the comments vector is serialized instead of its content.

Transformations with pools

Suppose, we want to apply certain transforming functions to the immer containers inside a large document type. The most straightforward way would be to simply create new containers with the new data, running the transforming function over each element. However, this approach has some disadvantages:

• All new containers will be independent, no structural sharing will be preserved, leading to the same data being stored multiple times.
• The transformation would be applied more times than necessary when some of the data is shared. Example: one vector is built by appending elements to the other vector. Transforming shared elements multiple times could be unnecessary.

Let’s consider a simple case using the document from the First example. The desired transformation would be to multiply each element of the immer::vector<int> by 10.

First, the document value would be created in the same way:

    const auto v1    = vector_one{1, 2, 3};
    const auto v2    = v1.push_back(4).push_back(5).push_back(6);
    const auto value = document{v1, v2};

The next component we need is the pools of all the containers from the value:

    const auto pools = immer::persist::get_output_pools(value);

The get_output_pools function returns the output pools of all immer containers that would be serialized using pools, as controlled by the policy. Here we use the default policy hana_struct_auto_policy which will use pools for all immer containers inside the document type which must be a hana::Struct.

The other required component is the conversion_map:

        const auto conversion_map = hana::make_map(hana::make_pair(
            hana::type_c<vector_one>, [](int val) { return val * 10; }));

This is a hana::map that describes the desired transformations to be applied. The key of the map is an immer container and the value is the function to be applied to each element of the corresponding container type. In this case, it will apply [](int val) { return val * 10; } to each int of the vector_one type, we have two of those in the document.

Having these two parts, we can create new pools with the transformations:

        auto transformed_pools =
            immer::persist::transform_output_pool(pools, conversion_map);

At this point, we can start converting the immer containers and create the transformed document value with them, new_value:

        const auto new_v1 =
            immer::persist::convert_container(pools, transformed_pools, v1);
        const auto expected_new_v1 = vector_one{10, 20, 30};
        REQUIRE(new_v1 == expected_new_v1);

        const auto new_v2 =
            immer::persist::convert_container(pools, transformed_pools, v2);
        const auto expected_new_v2 = vector_one{10, 20, 30, 40, 50, 60};
        REQUIRE(new_v2 == expected_new_v2);

        const auto new_value = document{new_v1, new_v2};

In order to confirm that the structural sharing has been preserved after applying the transformations, let’s serialize the new_value and inspect the JSON:

        const auto policy =
            immer::persist::hana_struct_auto_member_name_policy(document{});
        const auto str =
            immer::persist::cereal_save_with_pools(new_value, policy);
        const auto expected_json = json_t::parse(R"(
{
  "pools": {
    "ints": {
      "B": 5,
      "BL": 1,
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "leaves": [[1, [30]], [2, [10, 20]], [4, [50, 60]], [5, [30, 40]]],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    }
  },
  "value0": {"ints": 0, "ints2": 1}
}
        )");
        REQUIRE(json_t::parse(str) == expected_json);

And indeed, we can see in the JSON that the node [2, [10, 20]] is reused in both vectors.

Transformation into a different type

The transforming function can even return a different type. In the following example, vector<int> is transformed into vector<std::string>. The first two steps are the same as in the previous example:

    const auto v1    = vector_one{1, 2, 3};
    const auto v2    = v1.push_back(4).push_back(5).push_back(6);
    const auto value = document{v1, v2};

    const auto pools = immer::persist::get_output_pools(value);

Only this time the transforming function will convert an integer into a string:

        const auto conversion_map = hana::make_map(hana::make_pair(
            hana::type_c<vector_one>,
            [](int val) -> std::string { return fmt::format("_{}_", val); }));

Then we convert the two vectors the same way as before:

        auto transformed_pools =
            immer::persist::transform_output_pool(pools, conversion_map);

        const auto new_v1 =
            immer::persist::convert_container(pools, transformed_pools, v1);
        const auto expected_new_v1 = vector_str{"_1_", "_2_", "_3_"};
        REQUIRE(new_v1 == expected_new_v1);

        const auto new_v2 =
            immer::persist::convert_container(pools, transformed_pools, v2);
        const auto expected_new_v2 =
            vector_str{"_1_", "_2_", "_3_", "_4_", "_5_", "_6_"};
        REQUIRE(new_v2 == expected_new_v2);

And in order to confirm that the structural sharing has been preserved, we can introduce a new document type with the two vectors being vector<std::string>.

namespace {
struct document_str
{
    vector_str str;
    vector_str str2;

    friend bool operator==(const document_str&, const document_str&) = default;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(str), CEREAL_NVP(str2));
    }
};
} // namespace
BOOST_HANA_ADAPT_STRUCT(document_str, str, str2);

And serialize it with pools:

        const auto new_value = document_str{new_v1, new_v2};
        const auto policy =
            immer::persist::hana_struct_auto_member_name_policy(document_str{});
        const auto str =
            immer::persist::cereal_save_with_pools(new_value, policy);
        const auto expected_json = json_t::parse(R"(
{
  "pools": {
    "str": {
      "B": 5,
      "BL": 1,
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "leaves": [
        [1, ["_3_"]],
        [2, ["_1_", "_2_"]],
        [4, ["_5_", "_6_"]],
        [5, ["_3_", "_4_"]]
      ],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    }
  },
  "value0": {"str": 0, "str2": 1}
}
        )");
        REQUIRE(json_t::parse(str) == expected_json);

In the resulting JSON we can confirm that the node [2, ["_1_", "_2_"]] is reused for both vectors.

Transforming hash-based containers

As it was shown, converting vectors is conceptually simple: the transforming function is applied to each element of each node, producing a new node with the transformed elements. When it comes to the hash-based containers, that is set, map and table, their structure is defined by the used hash function, so defining the transformation may become a bit more verbose.

In the following example, we’ll start with a simple case of transforming a map. For a map, only the hash of the key matters and we will not modify the key yet. We will focus on transformations here and not on the structural sharing within the document, so we will use the immer container itself as the document. Let’s define the following policy to indicate that we want to use pools only for our container:

template <class Container>
struct direct_container_policy : immer::persist::value0_serialize_t
{
    auto get_pool_types(const auto&) const
    {
        return boost::hana::tuple_t<Container>;
    }
};

By default, immer uses std::hash for the hash-based containers. While this hash is sufficient for runtime use, it can’t be used for persistence, as noted in the C++ reference:

Note

Hash functions are only required to produce the same result for the same input within a single execution of a program

We will use xxHash as the hash for this example. Let’s create a small map like this:

    using int_map_t =
        immer::map<std::string, int, immer::persist::xx_hash<std::string>>;

    const auto value = int_map_t{{"one", 1}, {"two", 2}};
    const auto pools = immer::persist::get_output_pools(
        value, direct_container_policy<int_map_t>{});

Our goal is to convert the value from int to std::string. Let’s create the conversion_map like this:

    namespace hana     = boost::hana;
    using string_map_t = immer::
        map<std::string, std::string, immer::persist::xx_hash<std::string>>;

    const auto conversion_map = hana::make_map(hana::make_pair(
        hana::type_c<int_map_t>,
        hana::overload(
            [](const std::pair<std::string, int>& item) {
                return std::make_pair(item.first,
                                      fmt::format("_{}_", item.second));
            },
            [](immer::persist::target_container_type_request) {
                return string_map_t{};
            })));

A few important details to note:

• For maps, the transforming function accepts a pair of key and value, std::pair<std::string, int>.
• The transforming function must also be able to handle an argument of type immer::persist::target_container_type_request. This is achieved by using hana::overload to combine 2 lambdas into one callable value. When called with that argument, it should return an empty container of the type we’re transforming to. This explicit approach is necessary because there is no reliable way to automatically determine the hash algorithm for the new container. Even though in this case the type of the key doesn’t change (and so the hash remains the same), in other scenarios it might.

Once the conversion_map is defined, the actual conversion is done as before:

    auto transformed_pools =
        immer::persist::transform_output_pool(pools, conversion_map);
    const auto new_value =
        immer::persist::convert_container(pools, transformed_pools, value);
    const auto expected_new = string_map_t{{"one", "_1_"}, {"two", "_2_"}};
    REQUIRE(new_value == expected_new);

And we can see that the original map’s values have been transformed into strings.

Transforming table’s ID

For this example, we’ll transform the type of the ID of the table element while keeping the hash of it the same. This can occur, for instance, if the member that serves as the ID gets wrapped in a wrapper type.

To begin, let’s define an item type for a table:

struct old_item
{
    std::string id;
    int data;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(id), CEREAL_NVP(data));
    }
};

We can create a table value with some data and get the pools for it like this:

    using table_t    = immer::table<old_item,
                                 immer::table_key_fn,
                                 immer::persist::xx_hash<std::string>>;
    const auto value = table_t{old_item{"one", 1}, old_item{"two", 2}};
    const auto pools = immer::persist::get_output_pools(
        value, direct_container_policy<table_t>{});

In this example, we want to change the type of the old_item's ID, which is std::string, while keeping its hash the same. Let’s define a wrapper for std::string and a new_item type like this:

struct new_id_t
{
    std::string id;

    friend bool operator==(const new_id_t&, const new_id_t&) = default;

    friend std::size_t xx_hash_value(const new_id_t& value)
    {
        return immer::persist::xx_hash<std::string>{}(value.id);
    }
};

struct new_item
{
    new_id_t id;
    std::string data;

    friend bool operator==(const new_item&, const new_item&) = default;
};

We’re also changing the type for data from int to std::string but this change doesn’t affect the structure of the table. We define the xx_hash_value function for the new_id_t type to make it compatible with the immer::persist::xx_hash<new_id_t> hash. Then, we can define the target new_table_t type and the conversion_map that describes how to convert old_item into a new_item.

    using new_table_t = immer::
        table<new_item, immer::table_key_fn, immer::persist::xx_hash<new_id_t>>;

        const auto conversion_map = hana::make_map(hana::make_pair(
            hana::type_c<table_t>,
            hana::overload(
                [](const old_item& item) {
                    return new_item{
                        .id   = new_id_t{item.id},
                        .data = fmt::format("_{}_", item.data),
                    };
                },
                [](immer::persist::target_container_type_request) {
                    return new_table_t{};
                })));

Finally, to convert the value using the defined conversion_map we prepare the converted pools with transform_output_pool and use convert_container to convert the value table.

        auto transformed_pools =
            immer::persist::transform_output_pool(pools, conversion_map);
        const auto new_value =
            immer::persist::convert_container(pools, transformed_pools, value);
        const auto expected_new =
            new_table_t{new_item{{"one"}, "_1_"}, new_item{{"two"}, "_2_"}};
        REQUIRE(new_value == expected_new);

We can see that the new_value table contains the transformed data from the original value table.

Modifying the hash of the ID

If the key of a map, the ID of a table item or an element of a set changes its hash due to a transformation, the transformed hash-based container can no longer keep its shape and it can’t be efficiently transformed by simply applying transformations to its nodes.

immer::persist validates every container it creates from a pool. If such a hash modification occurs, a runtime exception will be thrown because it is not possible to detect this issue during compile-time. Let’s modify the previous example to also change the data of the ID:

        const auto conversion_map = hana::make_map(hana::make_pair(
            hana::type_c<table_t>,
            hana::overload(
                [](const old_item& item) {
                    return new_item{
                        // the ID's data is changed and its hash won't be the
                        // same
                        .id   = new_id_t{item.id + "_key"},
                        .data = fmt::format("_{}_", item.data),
                    };
                },
                [](immer::persist::target_container_type_request) {
                    return new_table_t{};
                })));

Now, if we attempt to convert the original table, a immer::persist::champ::hash_validation_failed_exception will be thrown:

        auto transformed_pools =
            immer::persist::transform_output_pool(pools, conversion_map);
        REQUIRE_THROWS_AS(
            immer::persist::convert_container(pools, transformed_pools, value),
            immer::persist::champ::hash_validation_failed_exception);

Even though such transformation can’t be performed efficiently, on a node level, we can still request these transformations to be applied. This will run for each value of the original container, creating a new independent container that doesn’t use structural sharing:

        const auto conversion_map = hana::make_map(hana::make_pair(
            hana::type_c<table_t>,
            hana::overload(
                [](const old_item& item) {
                    return new_item{
                        // the ID's data is changed and its hash won't be the
                        // same
                        .id   = new_id_t{item.id + "_key"},
                        .data = fmt::format("_{}_", item.data),
                    };
                },
                [](immer::persist::target_container_type_request) {
                    // We know that the hash is changing and requesting to
                    // transform in a less efficient manner
                    return immer::persist::incompatible_hash_wrapper<
                        new_table_t>{};
                })));

We can request for such container-level (as opposed to per-node level) transformation to be performed by wrapping the desired new container type new_table_t in a immer::persist::incompatible_hash_wrapper as the result of the immer::persist::target_container_type_request call.

        auto transformed_pools =
            immer::persist::transform_output_pool(pools, conversion_map);
        const auto new_value =
            immer::persist::convert_container(pools, transformed_pools, value);
        const auto expected_new = new_table_t{new_item{{"one_key"}, "_1_"},
                                              new_item{{"two_key"}, "_2_"}};
        REQUIRE(new_value == expected_new);

We can see that the transformation has been applied, the keys have the _key suffix.

Note

While different transformed containers will not have structural sharing, transforming the same container multiple times will reuse previously transformed data. In other words, transformation will be cached on the container level but not on the nodes level.

        const auto new_value_2 =
            immer::persist::convert_container(pools, transformed_pools, value);
        REQUIRE(new_value_2.impl().root == new_value.impl().root);

Transforming nested containers

Let’s consider a scenario where a transforming function works on an item within an immer container and also needs to transform another immer container. We define the types as follows:

struct nested_t
{
    vector_one ints;

    friend bool operator==(const nested_t&, const nested_t&) = default;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(ints));
    }
};

struct with_nested_t
{
    immer::vector<nested_t> nested;

    friend bool operator==(const with_nested_t&,
                           const with_nested_t&) = default;

    template <class Archive>
    void serialize(Archive& ar)
    {
        ar(CEREAL_NVP(nested));
    }
};

The important property here is that we have a vector<nested_t> where nested_t contains vector<int>, so we can say a vector is nested inside another vector. We can prepare a value with some structural sharing and then serialize it:

    const auto v1    = vector_one{1, 2, 3};
    const auto v2    = v1.push_back(4).push_back(5).push_back(6);
    const auto value = with_nested_t{
        .nested =
            {
                nested_t{.ints = v1},
                nested_t{.ints = v2},
            },
    };

    const auto policy =
        immer::persist::hana_struct_auto_member_name_policy(with_nested_t{});
    const auto str = immer::persist::cereal_save_with_pools(value, policy);

The resulting JSON looks like:

    const auto expected_json = json_t::parse(R"(
{
  "pools": {
    "ints": {
      "B": 5,
      "BL": 1,
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "leaves": [[1, [3]], [2, [1, 2]], [4, [5, 6]], [5, [3, 4]]],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    },
    "nested": {
      "B": 5,
      "BL": 3,
      "inners": [[0, {"children": [], "relaxed": false}]],
      "leaves": [[1, [{"ints": 0}, {"ints": 1}]]],
      "vectors": [{"root": 0, "tail": 1}]
    }
  },
  "value0": {"nested": 0}
}
    )");

Looking at the JSON we can confirm that the node [2, [1, 2]] is reused. Let’s define a conversion_map like this:

    const auto conversion_map = hana::make_map(
        hana::make_pair(
            hana::type_c<vector_one>,
            [](int val) -> std::string { return fmt::format("_{}_", val); }),
        hana::make_pair(
            hana::type_c<immer::vector<nested_t>>,
            [](const nested_t& item, const auto& convert_container) {
                return new_nested_t{
                    .str =
                        convert_container(hana::type_c<vector_str>, item.ints),
                };
            }));

The transforming function for vector_one is simple as it transforms an int into a std::string. However, the function for the vector<nested_t> is more involved. When we attempt to transform one item of that vector, nested_t, we realize that inside that function we have a vector<int> to deal with. This brings us back to the problems described in the beginning of the Transformations with pools section. To solve this issue, immer::persist provides an optional second argument to the transforming function, a function called convert_container. This function can be called with two arguments: the desired container type and the immer container to convert. This allows us to access the conversion_map we’re defining. This transformation will be performed using pools and will preserve structural sharing as expected.

Having defined the conversion_map, we apply it in the usual way and get the new_value:

    const auto pools = immer::persist::get_output_pools(value, policy);
    auto transformed_pools =
        immer::persist::transform_output_pool(pools, conversion_map);

    const auto new_value = with_new_nested_t{
        .nested = immer::persist::convert_container(
            pools, transformed_pools, value.nested),
    };

We can verify that the new_value has the expected content:

    const auto expected_new = with_new_nested_t{
        .nested =
            {
                new_nested_t{.str = {"_1_", "_2_", "_3_"}},
                new_nested_t{.str = {"_1_", "_2_", "_3_", "_4_", "_5_", "_6_"}},
            },
    };
    REQUIRE(new_value == expected_new);

And we can serialize it again to confirm that the structural sharing of the nested vectors has been preserved:

    const auto transformed_str = immer::persist::cereal_save_with_pools(
        new_value,
        immer::persist::hana_struct_auto_member_name_policy(
            with_new_nested_t{}));
    const auto expected_transformed_json = json_t::parse(R"(
{
  "pools": {
    "nested": {
      "B": 5,
      "BL": 3,
      "inners": [[0, {"children": [], "relaxed": false}]],
      "leaves": [[1, [{"str": 0}, {"str": 1}]]],
      "vectors": [{"root": 0, "tail": 1}]
    },
    "str": {
      "B": 5,
      "BL": 1,
      "inners": [
        [0, {"children": [2], "relaxed": false}],
        [3, {"children": [2, 5], "relaxed": false}]
      ],
      "leaves": [
        [1, ["_3_"]],
        [2, ["_1_", "_2_"]],
        [4, ["_5_", "_6_"]],
        [5, ["_3_", "_4_"]]
      ],
      "vectors": [{"root": 0, "tail": 1}, {"root": 3, "tail": 4}]
    }
  },
  "value0": {"nested": 0}
}
    )");

We can see that the [2, ["_1_", "_2_"]] node is still being reused in the two vectors.

Policy

concept immer::persist::Policy = requires(Value value, T policy) { policy.get_pool_types(value); }

Policy is a type that describes certain aspects of serialization for immer-persist.

• How to call into the cereal archive to save and load the user-provided value. Can be used to serealize the value inline (without the value0 node) by taking a dependency on cereal-inline, for example.
• Types of immer containers that will be serialized using pools. One pool contains nodes of only one immer container type.
• Names for each per-type pool.

auto immer::persist::via_get_pools_names_policy(const auto &value)

Create an immer-persist policy that uses the user-provided get_pools_names function (located in the same namespace as the value user serializes) to determine:

• the types of immer containers that should be serialized in a pool
• the names of those pools in JSON (and possibly other formats).

The get_pools_names function is expected to return a boost::hana::map where key is a container type and value is the name for this container’s pool as a BOOST_HANA_STRING.

Parameters

• value: Value that is going to be serialized, only type of the value matters.

auto immer::persist::hana_struct_auto_member_name_policy(const auto &value)

Create an immer-persist policy that recursively finds all immer containers in a serialized value and uses member names to name the pools.

The value must be a boost::hana::Struct.

Parameters

• value: Value that is going to be serialized, only type of the value matters.

struct immer::persist::value0_serialize_t

#include <policy.hpp>

This struct provides functions that immer-persist uses to serialize the user-provided value using cereal.

In this case, we use cereal’s default name, value0. It’s used in all policies provided by immer-persist.

Other possible way would be to use a third-party library to serialize the value inline (without the value0 node) by taking a dependency on cereal-inline, for example.

Subclassed by immer::persist::hana_struct_auto_member_name_policy_t< T >, immer::persist::via_get_pools_names_policy_t< T >, immer::persist::via_get_pools_types_policy

struct immer::persist::demangled_names_t

#include <policy.hpp>

This struct is used in some policies to provide names to each pool by using a demangled name of the immer container corresponding to the pool.

Subclassed by immer::persist::hana_struct_auto_policy, immer::persist::via_get_pools_types_policy

struct immer::persist::via_get_pools_types_policy

#include <policy.hpp>

An immer-persist policy that uses the user-provided get_pools_types function to determine the types of immer containers that should be serialized in a pool.

The get_pools_types function is expected to return a boost::hana::set of types of the desired containers.

The names for the pools are determined via demangled_names_t.

Inherits from immer::persist::demangled_names_t, immer::persist::value0_serialize_t

struct immer::persist::hana_struct_auto_policy

#include <policy.hpp>

An immer-persist policy that recursively finds all immer containers in a serialized value.

The value must be a boost::hana::Struct.

The names for the pools are determined via demangled_names_t.

Inherits from immer::persist::demangled_names_t

API Overview

template <class T, class Archive = cereal::JSONInputArchive, Policy< T > Policy = default_policy, class… Args>

T immer::persist::cereal_load_with_pools(std::istream &is, const Policy &policy = Policy{}, Args&&... args)

Load a value of the given type T from the provided stream using pools.

By default, cereal::JSONInputArchive is used but a different cereal input archive can be provided.

template <class T, class Archive = cereal::JSONInputArchive, Policy< T > Policy = default_policy>

T immer::persist::cereal_load_with_pools(const std::string &input, const Policy &policy = Policy{})

Load a value of the given type T from the provided string using pools.

By default, cereal::JSONInputArchive is used but a different cereal input archive can be provided.

template <class Archive = cereal::JSONOutputArchive, class T, Policy< T > Policy = default_policy, class… Args>

void immer::persist::cereal_save_with_pools(std::ostream &os, const T &value0, const Policy &policy = Policy{}, Args&&... args)

Serialize the provided value with pools using the provided policy outputting into the provided stream.

By default, cereal::JSONOutputArchive is used but a different cereal output archive can be provided.

See

Policy

template <class Archive = cereal::JSONOutputArchive, class T, Policy< T > Policy = default_policy, class… Args>

std::string immer::persist::cereal_save_with_pools(const T &value0, const Policy &policy = Policy{}, Args&&... args)

Serialize the provided value with pools using the provided policy.

By default, cereal::JSONOutputArchive is used but a different cereal output archive can be provided.

Return

std::string The resulting JSON.

template <class T>

struct immer::persist::xx_hash

#include <xxhash.hpp>

xxHash is a good option to be used with immer::persist as it produces hashes identical across all platforms.

See

https://xxhash.com/

template <class Previous, class Pools, class WrapFn, class PoolNameFn>

class immer::persist::output_pools_cereal_archive_wrapper

#include <archives.hpp>

A wrapper type that wraps a cereal::OutputArchive (for example, JSONOutputArchive), provides access to the Pools object stored inside, and serializes the pools object alongside the user document.

Normally, the function cereal_save_with_pools should be used instead of using this wrapper directly.

See

cereal_save_with_pools

template <class Previous, class Pools, class WrapFn, class PoolNameFn>

class immer::persist::input_pools_cereal_archive_wrapper

#include <archives.hpp>

A wrapper type that wraps a cereal::InputArchive (for example, JSONInputArchive) and provides access to the pools object.

Normally, the function cereal_load_with_pools should be used instead of using this wrapper directly.

See

cereal_load_with_pools

Transform API

template <typename T, Policy< T > Policy = hana_struct_auto_policy>

auto immer::persist::get_output_pools(const T &value0, const Policy &policy = Policy{})

Return just the pools of all the containers of the provided value serialized using the provided policy.

See

convert_container

template <class Storage, class ConversionMap>

auto immer::persist::transform_output_pool(const detail::output_pools<Storage> &old_pools, const ConversionMap &conversion_map)

Given output_pools and a map of transformations, produce a new type of input pools with those transformations applied.

conversion_map is a boost::hana::map where keys are types of immer containers and values are the transforming functions.

See

get_output_pools

Transformations with pools Transforming nested containers

template <class SaveStorage, class LoadStorage, class Container>

auto immer::persist::convert_container(const detail::output_pools<SaveStorage> &output_pools, detail::input_pools<LoadStorage> &new_input_pools, const Container &container)

Given output_pools and new (transformed) input_pools, effectively convert the given container.

See

get_output_pools

Transformations with pools Transforming nested containers

template <class Container>

struct immer::persist::incompatible_hash_wrapper

#include <hash_container_conversion.hpp>

The wrapper is used to enable the incompatible hash mode which is required when the key of a hash-based container transformed in a way that changes its hash.

A value of this type should be returned from a transforming function accepting target_container_type_request.

See

Modifying the hash of the ID

struct immer::persist::target_container_type_request

#include <hash_container_conversion.hpp>

This type is used as an argument for a transforming function.

The return type of the function is used to specify the desired container type to contain the transformed values.

See

Transforming hash-based containers

Exceptions

class immer::persist::pool_exception

#include <errors.hpp>

Base class from which all the exceptions in immer::persist are derived.

Subclassed by immer::persist::duplicate_name_pool_detected, immer::persist::invalid_children_count, immer::persist::invalid_container_id, immer::persist::invalid_node_id, immer::persist::pool_has_cycles

class immer::persist::pool_has_cycles

#include <errors.hpp>

Thrown when a cycle is detected in the pool of vectors.

Inherits from immer::persist::pool_exception

class immer::persist::invalid_node_id

#include <errors.hpp>

Thrown when a non-existent node is mentioned.

Inherits from immer::persist::pool_exception

class immer::persist::invalid_container_id

#include <errors.hpp>

Thrown when a non-existent container is mentioned.

Inherits from immer::persist::pool_exception

class immer::persist::invalid_children_count

#include <errors.hpp>

Thrown when a node has more children than expected.

Inherits from immer::persist::pool_exception

class immer::persist::duplicate_name_pool_detected

#include <errors.hpp>

Thrown when duplicate pool name is detected.

Inherits from immer::persist::pool_exception