usage rule

Thrift

Supported data types

The set in mongodb corresponds to the struct in thrift.

The struct field supports all basic data types, container types (list, map, set), binary, enumeration types (both in file and out of file references are supported), struct types (both in file and out of file references are supported), and legal combinations of the above types in thrift.

Supported naming conventions

Struct and various fields can be named using either the Big Camel Naming Method or the Snake Naming Method.

Add Mongodb tag

Add tags using annotations.

format: (go.tag="bson:\"{field_name}{,omitempty}(Choose whether to add as needed)\"")

example: (go.tag="bson:\"id,omitempty\"") or (go.tag="bson:\"id\"")

Add Mongodb curd function

Add functions using annotations.

format: mongo.{Parse the function name used} = "{The actual function name used}({input params...}) (returns...)"

example: mongo.InsertOne = "InsertUser(ctx context.Context, user *user.User) (interface{}, error)"

For specific rules, please refer to usage rule.

Protobuf

Supported data types

The set in mongodb corresponds to the message in protobuf.

The message field supports all basic data types in proto, keywords such as repeated, bytes, enumeration types (both in file and out of file references are supported), message types (both in file and out of file references are supported), and legal combinations of the above types.

Supported naming conventions

Message and various fields can be named using either the Big Camel Naming Method or the Snake Naming Method.

Add Mongodb tag

Add tags using comments.

format: // go.tag=|bson:"{field_name}{,omitempty}(Choose whether to add as needed)"|

example: // go.tag=|bson:"id,omitempty"| or // go.tag=|bson:"id"|

Add Mongodb curd function

Add functions using comments.

format: // mongo.{Parse the function name used} = |{The actual function name used}({input params...}) (returns...)|

example: // mongo.InsertOne = |InsertUser(ctx context.Context, user *user.User) (interface{}, error)|

For specific rules, please refer to usage rule.

Usage Rule

The function name adopts a camel shaped naming, with the first letter capitalized, and the first word must be one of the following:

• Insert
• Find
• Update
• Delete
• Count
• Bulk
• Transaction

Divide all operations into two categories: acting on a single document and acting on multiple documents. The specific judgment is determined by the combination of passed in parameters and return values.

Insert

The Insert operation only has two options: insert a single document and insert multiple documents, which are determined by both the incoming parameters and the return value.

Function Naming: The first word is Insert, and the rest can be defined arbitrarily.

Input Params: Fix two,(ctx(fix), structure pointer or structure pointer slice)

Returns: Fix two, (interface{} or []interface{}, error(fix))

Find

The Find operation includes two types: single entity and multiple entities, determined by the return value. The return value is a pointer slice, which is a single entity operation. Otherwise, it is a multiple entity operation.

Function Naming: The first word is Find, followed by the following rules: Find{The field names you want to obtain, if not, obtain all of them}{Sort and page operations, if not, do not add}By{Query condition}/All(No Query Condition)

Input Params: The first parameter is fixed as ctx, and the following parameters are passed in order based on the comparison conditions.

Returns: Fix two, (structure pointer or structure pointer slice, error(fix))

Example: Find Name OrderbyNameSkipLimit By GenderEqualOrAgeGreaterThan

• Field Name：Corresponding structural field names in go

  If there is a nested structure, the field name format is: {Field name of the child structure in the parent structure}{Specific field name in the child structure}. For example, if User nested Video and wants to find the Id field of the Video, it is: VideoId

• Sorting and pagination operations

• Query Condition

  Start with By or All, followed By specific query conditions, and All does not need to be followed by conditions

  By must be followed by specific fields and comparison conditions. If necessary, use And or Or to identify the And Or condition and connect each field and comparison condition. When using the And or Or connector, if it is necessary to indicate query priority, use Lb or Rb to identify sub connection conditions. For example: ByLbLbUsernameEqualOrUsernameEqualRbAndAgeGreaterThanRb == By((UsernameEqualOrUsernameEqual)AndAgeGreaterThan)

• Grammar Format：By{Field Name}{Query Condition}{And/Or}{Field Name}{Query Condition}

• Compare Condition：

Update

The Update operation includes two types: single entity and multiple entities, which are determined by the return value. The return value is bool, indicating whether the single entity operation is successful; The return value is int, which represents a multi entity operation and the number of successful updates.

The Update operation is divided into two types based on the perspective of updating objects:

• Update the entire document(No field added after Update, structure pointer needs to be passed in);

• Update some fields(Add a field name after Update and pass in the updated field value).

Function Naming: The first word is Update, followed by the following rules:

Update(Upsert){The name of the field to be updated, if not, it will be updated as the entire structure, and a structure pointer needs to be passed in}By{Query Condition}/All(No Query Condition) The query conditions are the same as Find.

The update operation currently only supports the use of $set.

Input Params: The first parameter is fixed as ctx, and the following parameters are passed in order based on the updated field name and comparison criteria.

Returns: Fix two, (bool or int, error(fix))

Delete

The Delete operation includes two types: single entity and multiple entities, and the judgment method is the same as Update.

Function Naming:

DeleteBy{Query Condition}/All(No Query Condition) The query conditions are the same as Find

Input Params: The first parameter is fixed as ctx, and the following parameters can be passed in order based on the comparison conditions.

Returns: Fix two, (bool or int, error(fix))

Count

Only multi entity operations.

Function Naming:

CountBy{Query Condition}/All(No Query Condition) The query conditions are the same as Find.

Input Params: The first parameter is fixed as ctx, and the following parameters can be passed in order based on the comparison conditions.

Returns: Fix (int, error)

Bulk

Bulk supports Insert, Update, Delete Operation

Among them, Insert only supports InsertOne, while the other two operations, One and Many are all supported.

Insert fixed write InsertOne, Update, Delete needs to specify whether it is a single entity operation or a multi entity operation, with One or Many added after it, Example: UpdateOne UpdateMany

Function Naming:

Bulk{Operation1, example InsertOne}{Operation2, example DeleteOneAll}

Input Params: The first one is fixed as ctx, and the following parameters are passed in the order of the parameters passed in each operation.

Returns: Fix (*mongo.BulkWriteResult, error)

Transaction

Transaction supports Insert, Update, Delete, Bulk operation, and support transaction operations for multiple collections.

Insert, Update, and Delete all support One Many, and it is necessary to specify whether a single entity operation or a multi entity operation is required, such as InsertOne and InsertMany; The bulk operation needs to be marked with Lb and Rb to start and end, for example: BulkLbInsertOneUpdateManyByIdEqualRb.

If using operations from other sets, you need to add the Collection keyword and the parameter’s camel hump naming capitalization before the Insert, Update, Delete, and Bulk operations, and pass in the corresponding collection parameters after the client in sequence (parameter naming requirements: camel hump naming, lowercase first letter)

For example: CollectionUserCollectionInsertMany, input param: userCollection *mongo.Collection; If the operation of this collection is used, there is no need to add Collection and related parameters.

Function Naming:

Transaction{Operation1, example CollectionUserCollectionInsertMany}{Operation2, example UpdateOneByAgeEqual}

Input Params: The first block is fixed as ctx, the second block is fixed as client, and the third block is fixed as other sets used. The following parameters are passed in the order in which they are passed in for each operation.

Returns: Fix error

Keywords

When naming structure fields, do not include the following keywords:

• Insert
• Find
• Update
• Delete
• Count
• Bulk
• Transaction
• Collection
• Lb
• Rb
• One
• Many
• By
• All
• Orderby
• Skip
• Limit
• Equal
• NotEqual
• LessThan
• LessThanEqual
• GreaterThan
• GreaterThanEqual
• Between
• NotBetween
• In
• NotIn
• True
• False
• Exists
• NotExists

Update

When there is an update to a field or function, the behavior of the tool is to compare the function name read in the annotation with the function name generated in the code file. If there is an annotation in the code file that is not present, append it after the file, and leave the rest unchanged.