Tags Usage

db - column name in db


package main

       type _ struct {
          ID int `db:"id"`
       }

pk - if set then column is primary key


package main

      type _ struct {
         ID int `db:"id" pk:"true"`
      }

nullable - if set then column is nullable(without NOT NULL)


package main

      type _ struct {
         Code *int `db:"code" nullable:"true"`
      }

length - column length(useful for varchar)


package main

      type _ struct {
         Name string `db:"name" length:"64"`
      }

uniq - if set then column is unique(can be placed to many properties - grouped, on single column can be many uniq’s, separated by ,)


package main

      // CREATE UNIQUE INDEX uniq_name ON example_table (name);
      type _ struct {
         Name string `db:"name" uniq:"uniq_name"`
      }

      // CREATE UNIQUE INDEX uniq_name_parted ON example_table (name_part1, name2);
      type _ struct {
         NamePart1 string `db:"name_part1" uniq:"uniq_name_parted"`
         NamePart2 string `db:"name2" uniq:"uniq_name_parted"`
      }

uniq_cond - work in pair with uniq, uniq_cond is a condition for uniq, can be placed on any column(of grouped uniq’s, do not set to every column of grouped uniq’s, can be multiple, separated by ,)


package main

      // CREATE UNIQUE INDEX uniq_name_parted ON example_table (name1, name2) WHERE is_selected = true;
      type _ struct {
         Checked bool `db:"is_checked"`
         NamePart1 string `db:"name1" uniq:"uniq_name_parted" uniq_cond:"uniq_name_parted:(is_checked = true)"`
         NamePart2 string `db:"name2" uniq:"uniq_name_parted"`
      }

index - if set then column is index(logic same as uniq)


package main

      // CREATE INDEX idx_name ON example_table (name);
      type _ struct {
         Name string `db:"name" index:"idx_name"`
      }

      // CREATE INDEX idx_name_parted ON example_table (name_part1, name2);
      type _ struct {
         NamePart1 string `db:"name_part1" index:"idx_name_parted"`
         NamePart2 string `db:"name2" index:"idx_name_parted"`
      }

index_cond - work in pair with index(logic same as uniq_cond)


package main

      // CREATE INDEX idx_name_parted ON example_table (name1, name2) WHERE is_selected = true;
      type _ struct {
         Checked bool `db:"is_checked"`
         NamePart1 string `db:"name1" index:"idx_name_parted" index_cond:"idx_name_parted:(is_checked = true)"`
         NamePart2 string `db:"name2" index:"idx_name_parted"`
      }

default - if set then column has default value


package main

      type _ struct {
         Name string `db:"name" default:"123"`
      }

type - if set then column has type(manually set, without struct property type checking)


package main

      type ExampleType int

      type _ struct {
         Type ExampleType `db:"type" type:"int"`
      }

Possible types: text, varchar, json, jsonb, integer