前几天翻译了一篇关于EOS智能合约数据库的内容，今天来演示一下数据库的使用方法。
目录
增
查
改
删
新增
新增内容往往用到emplace构造函数，来进行数据库对象的新增。
.cpp
void test_da::create(account_name user, string title, string content)
{
require_auth( user ); //验证权限
das datable( _self, user); //定义数据库对象
datable.emplace(user, [&]( da & d){
d.title = title;
d.content = content;
d.post_id = datable.available_primary_key();
d.poster = user;
}); //数据库内容创建
}
这里需要注意的是：
定义数据库对象， 其中第一个参数是合约的拥有者_self，第二个变量就是数据库的payer，也就是数据库是谁的，数据库存储在谁的账户下。
emplace函数接收两个参数，一个payer,和一个lamada表达式。这个结构是固定的。
那么我们现在来看一下我们的test_da类是怎么定义的。
.hpp
class test_da : public contract {
public:
test_da( account_name self ):contract(self){}
// @abi action
void create(account_name user, string title, string content);
private:
// @abi table data i64
struct da {
uint64_t post_id;
account_name poster;
string title;
string content;
uint64_t primary_key()const { return post_id; }
account_name get_poster() const { return poster; }
EOSLIB_SERIALIZE(da, (post_id)(poster)(title)(content))
};
typedef eosio::multi_index> > das;
};
} /// namespace eosio
所有的智能合约都继承自contract合约。`test_da( account_name self ):contract(self){}
是test_da`合约的构造函数。
下面是对create函数的声明。
接下来是对数据字段的定义。这里我们定义了数据结构da.
primary_key函数是定义主键的函数。
接下来我们定义了辅助主键返回poster。
EOSLIB_SERIALIZE宏的第一个参数是数据结构，其他参数是数据结构中的数据成员。
typedef我们在这里定义了一个名字为das的类型，它用来定义数据库对象。这里我们定义的是一个具有主键及一个辅助键的数据库对象。
.abi
abi 非常重要，错误的abi会导致合约执行失败。
{
"types": [],
"structs": [{
"name": "da",
"base": "",
"fields": [{
"name": "post_id",
"type": "uint64"
},{
"name": "poster",
"type": "account_name"
},{
"name": "title",
"type": "string"
},{
"name": "content",
"type": "string"
}
]
},{
"name": "create",
"base": "",
"fields": [{
"name": "user",
"type": "account_name"
},{
"name": "title",
"type": "string"
},{
"name": "content",
"type": "string"
}
]
}}
],
"actions": [{
"name": "create",
"type": "create",
}
],
"tables": [{
"name": "data",
"index_type": "i64",
"key_names": [
"post_id"
],
"key_types": [
"uint64"
],
"type": "da"
}
],
"ricardian_clauses": []
}
接下来来演示一下:
发布合约
cleos set contract eosio test_da
Reading WAST/WASM from test_da/test_da.wasm...
Using already assembled WASM...
Publishing contract...
executed transaction: 3d6f04278617d3807fe876a33057f1155acf9c9e5a392ac6ed8ad51e79506009 6752 bytes 24679 us
# eosio <= eosio::setcode {"account":"eosio","vmtype":0,"vmversion":0,"code":"0061736d0100000001ad011a60037f7e7e0060057f7e7e7f...
# eosio <= eosio::setabi {"account":"eosio","abi":{"types":[],"structs":[{"name":"da","base":"","fields":[{"name":"post_id","...
创建数据
cleos push action eosio create '{"user":"eosio","title":"first","content":"create a first one"}' -p eosio
executed transaction: 830057f270fa499b1d61b82e80ad8cda1774cdc1786c1e786f558a3e0a48974c 216 bytes 17229 us
# eosio <= eosio::create {"user":"eosio","title":"first","content":"create a first one"}
下面我们来查一下数据表：
cleos get table eosio eosio data
{
"rows": [{
"post_id": 0,
"poster": "eosio",
"title": "first",
"content": "create a first one"
}
],
"more": false
}
创建信息成功，那么我们用别的账号创建会怎样呢？
cleos push action eosio create '{"user":"eostea","title":"eostea first","content":"eostea create a first one"}' -p eostea
executed transaction: 8542a87e563a9c62b7dbe46ae09ccf829c7821f8879167066b658096718de148 232 bytes 2243 us
# eosio <= eosio::create {"user":"eostea","title":"eostea first","content":"eostea create a first one"}
查看数据表：
cleos get table eosio eostea data
{
"rows": [{
"post_id": 0,
"poster": "eostea",
"title": "eostea first",
"content": "eostea create a first one"
}
],
"more": false
}
到这里，相信大家对创建数据已经没有什么疑惑了。
查询
对于数据库，最重要的功能就是查询，如果没有查询功能，数据库里的数据就不能呈现，也就没有意义。查询数据库主要分为两方面，一方面是主键查询，一方面是通过二级索引查询。
这里为了使表中数据多样化我会做一些修改，将所有数据都合到一张表中。
我将以上.cpp中的das datable( _self, user);改为das datable( _self, _self);.这样数据都存在合约账户的表中。
主键查询
这里我添加了一个方法来查询数据并打印：
void test_da::getd(uint64_t post_id){
das datable(_self, _self);
auto post_da = datable.find( post_id);
eosio::print("Post_id: ", post_da->post_id, " Post_Tile: ", post_da->title.c_str(), " Content: ", post_da->content.c_str());
}
abi文件也做了相应的调整：
执行：
cleos push action eosio getd '{"post_id":1}' -p eosio
executed transaction: ac8663235462d947c74542af848cca54a059c3991d193237025da7d4767d6725 192 bytes 1724 us
# eosio <= eosio::getd {"post_id":1}
>> Post_id: 1 Post_Tile: first Content: eosio create a first one
二级索引查询
添加二级索引查询代码如下：
auto poster_index = datable.template get_index();
auto pos = poster_index.find( user );
for (; pos != poster_index.end(); pos++)
{
eosio::print("content:", pos->content.c_str(), " post_id:", pos->post_id, " title:", pos->title.c_str());
}
获取二级索引并获取数据，这里我只用了find查询，其他查询就不在一一介绍。
执行getd操作：
cleos push action eosio getd '{"post_id":2,"user": "eostea"}' -p eosio
executed transaction: 2370e1fb1ee8a581f7321f02fb40645e51269e579d183c33ef470dba0b3afdbc 200 bytes 5403 us
# eosio <= eosio::getd {"post_id":2,"user":"eostea"}
>> Post_id: 2 Post_Tile: eostea first Content: eostea create a first onecontent:eostea create a first one post_id:2 title:eostea first
数据库中数据如下：
cleos get table eosio eosio data
{
"rows": [{
"post_id": 0,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 1,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 2,
"poster": "eostea",
"title": "eostea first",
"content": "eostea create a first one"
}
],
"more": false
}
更改
更改数据库内容。
这里我们先看一下目前的数据库内容。
cleos get table eosio eosio data
{
"rows": [{
"post_id": 0,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 1,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 2,
"poster": "eostea",
"title": "eostea first",
"content": "eostea create a first one"
}
],
"more": false
}
写一个更改数据的action代码如下：
void test_da::change(account_name user, uint64_t post_id, string title, string content)
{
require_auth(user);
das datable( _self, user);
auto post = datable.find(post_id);
eosio_assert(post->poster == user, "yonghucuowu");
datable.modify(post, user, [&](auto& p){
if (title != "")
p.title = title;
if (content != "")
p.content = content;
});
}
1. 前几行代码已经之前讲解过，现在直接说modify方法，他的第一个参数是你查询出的要更改的对象，第二个参数是payer,其他的不用多说。
下面我们执行一下命令：
cleos push action eosio change '{"user":"eosio","post_id":1,"title":"change","content":"change action"}' -p eosio
executed transaction: 8cb561a712f2741560118651aefd49efd161e3d73c56f6d24cf1d699c265e2dc 224 bytes 2130 us
# eosio <= eosio::change {"user":"eosio","post_id":1,"title":"change","content":"change action"}
下面我们看一下数据库：
cleos get table eosio eosio data
{
"rows": [{
"post_id": 0,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 1,
"poster": "eosio",
"title": "change",
"content": "change action"
},{
"post_id": 2,
"poster": "eostea",
"title": "eostea first",
"content": "eostea create a first one"
}
],
"more": false
}
post_id=1的记录已经被改变，说明我们成功了。
删除数据
删除数据我又加了一个action,如下所示:
void test_da::dele(account_name user, uint64_t post_id)
{
require_auth(user);
das datable( _self, user);
auto post = datable.find(post_id);
eosio::print(post->title.c_str());
eosio_assert(post->poster == user, "yonghucuowu");
datable.erase(post);
}
这里调用了erase方法删除数据，参数为一个数据对象。下面我们来看一下执行结果：
cleos push action eosio dele '{"user":"eosio","post_id":1}' -p eosioexecuted transaction: 3affbbbbd1da328ddcf37753f1f2f6c5ecc36cd81a0e12fea0c789e75b59714e 200 bytes 2383 us
# eosio <= eosio::dele {"user":"eosio","post_id":1}
现在再来看一下我们的数据库：
cleos get table eosio eosio data
{
"rows": [{
"post_id": 0,
"poster": "eosio",
"title": "first",
"content": "eostea create a first one"
},{
"post_id": 2,
"poster": "eostea",
"title": "eostea first",
"content": "eostea create a first one"
}
],
"more": false
}
post_id=1的数据已经被我们删除。
到这里，数据库的增删改查已经讲解完毕。
本文内容为原创内容，如需转载请联系作者。谢谢！！！
原文链接: https://eosfans.io/topics/484深度强化学习在指代消解中的一种尝试-博客-云栖社区-阿里云
