This is our internal run-once utility, which purpose is to import data from relational database to MongoDB using JSON mapping file.
As it was written as internal run-once tool, it contains a lot of dirty hacks and poorly-written code, sorry. After all, we decided that it seem to be rather universal and useful to be opensourced.
- Should work with virtually any SQL database (using JDBC). Tested with H2 and MySQL.
- Collections are mapped to SQL SELECTs, making possible to use JOINs and SQL functions e.g. LOWER().
- Nested structures creation using dot notation.
- Embedded arrays, counters, maps, using child SELECTs for each entity.
- Placeholders for entity fields in child selects (e.g. in "where" clause)
- Explicit conversion to int/long/string
- Cached column-to-objectId bindings, which allows to refer not-yet-inserted documents.
- Batch select/insert, may significally improve performance
jmm_2.8.1-1.0.min.jar contains all necessary dependencies so after creating config and mapping files, simply run it like java -jar jmm_2.8.1-1.0.min.jar [config]
The migration tool jar contains MySQL and H2 JDBC drivers, but it should work for every DB you'll provide JDBC driver for. Simply add necessary jars to the classpath and set proper driver class name and JDBC URI in config file.
Example of mapping file:
{ "collections": {
//MongoDB collection name. There could be several mappings for single collection,
//which result in consecutive inserts
"users": {
//FROM part, it is possible to use JOINs here
"from": "users AS u"
//columns-fields mapping
"mapping": {
// Generate mongodb objectId.
// All subsequent usages of {"$oid" : { "collection" : "users", "key" : ... }} (or $oidString) with the same key
// will generate the same objectId
"_id" : { "$oid" : "u.id" }
// Same as above, but converts objectId's to string
"stringId" : {"$oidString" : "u.id"}
// Simple column-to-field value binding
"oldId" : "u.id"
"username" : "u.username"
// Explicitly convert rating column to integer type
// (to prevent automatic conversion to long by JDBC driver)
// Dot notation in field name will create nested "counters" structure
"counters.rating" : { "$int" : "u.rating"}
// Integer field, countaining count(*) of user's posts.
// Note the reference to oldId field
"counters.posts" : {
"$count" : {
"from" : "posts AS p"
"where" : "p.user_id = ${oldId}"
}
}
// generate short url for each user using link_id column
// like "http://exmpl.me/u/fQsd"
"shortUrl" : { "$surl" : "http://exmpl.me/u/${u.link_id}"}
// Array of strings field, containing values of 'slug' columns from referenced groups
"groups" : {
"$array" : {
"from" : "groups AS g LEFT JOIN members AS m ON g.id = m.group_id"
"mapping" : "LOWER(g.slug)"
"where" : "banned = 0 AND m.user_id = ${oldId}"
}
}
// Map of counters key = group slug, value = posts count in group
"groupPostCount" : {
"$countMap" : {
"from" : "posts AS p LEFT JOIN groups AS g ON p.group_id = g.id"
"key" : "LOWER(g.slug)"
"where" : "g.slug IS NOT NULL AND p.user_id = ${oldId}"
}
}
// Array of structures
"invites" : {
"$array" : {
"from" : "group_invites AS i LEFT JOIN groups AS g ON i.group_id = g.id"
"mapping" : {
"group" : "LOWER(g.slug)",
"invitedBy" : {"$oidString" : "i.invited_by_id"}
}
"where" : "i.user_id = ${oldId}"
}
}
// Column values converted to array. Resulting array will contain only non-null values.
"roles": {
"$colArray" : [
"CASE u.is_admin WHEN 1 THEN 'admin' ELSE NULL END",
"CASE u.is_staff WHEN 1 THEN 'staff' ELSE NULL END"
]
}
}
}
}
}
Migration tool takes config file as the first (and only) argument, ./jmm.conf
by default.
The tool uses configgy for configuration and logging. Here is a config file example:
mapping {
# Mapping file (see above)
file = "mapping.json"
# Short url generation parameters
short_url {
alphabet = "fBMhDcvQd27l4ePtH0RspY3NFbXVzna6iG8oygkqULwWAmC5ZJEKjISuOTx91r"
blockSize = 12
minLength = 4
}
}
# JDBC parameters
jdbc {
driver = "com.mysql.jdbc.Driver"
uri = "jdbc:mysql://192.168.1.1:3306/prod?user=root&password=mysecret&characterEncoding=UTF8"
# Batch size for selects
limit = 5000
}
mongo {
# Set to true for dry run (no actual inserts will be performed)
dry-run = false
host = "127.0.0.1"
port= 27017
database = "migrated"
#drop database before import
clean = true
# user=
# password=
}
#logging parameters, see configgy readme for details
log {
filename = "migration.log"
level = "info"
utc = false
console = false
}
tl;dr: MIT
Copyright (C) 2011 by e-Legion
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.