This is a tool to convert map data from OpenStreetMap xml format, to Navmo Local Map format.
Best described by example, this is an xml file that looks like this:
TODO
The purpose of this format is to store map data in such a way that it is possible to quickly draw the map (in a local coordinate system), and to run a routing algorithim.
There are three parts to the Navmo Local map format:
- The metadata properties file. This is a standard java properties file.
- The online search data files. These are a collection of bzip2-compressed text files (one per data table); an ant script on the Navmo server is used to import these into the PostgreSQL database. These are used for online searching of roads and POIs.
- The rendering/routing data files. These are a collection of gzip-compressed binary files (one per data table), each containing a header and an array of serialized Java objects. These are used for routing and generating offline maps, and are loaded into the various Java processors that perform these tasks.
Standard java properties file. Contains the following keys (obviously the values can change)
CountryCode=GB # Tested with 'GB' (Great Britain), 'AT' (Austria), 'DE' (Germany)
MapName=London # Name of local map. Referenced in lots of other places.
CoordinateMapping=GeoTools # Records the method of coordinate transform. Not used anywhere in code.
CoordinateSystemId=27700 # the EPSG ID for the Ordnance Survey British Grid
BuildVersion=123 # ${build.version}
DataVersion=234 # ${perforce.build.number}
The following are bzip2-compressed text files containing map data suitable for searching for start and endpoints of routes. The text format is tab-separated-values, and the character set is CP-1252. Because these are imported directly into a Postgres database, the datatypes of the fields are specified as PostgreSQL data types.
The following files are required for online search:
- metadata.txt.bz2
- logicalroadjunction.txt.bz2
- poi.txt.bz2
- roadjunction.txt.bz2
- roadnamelookup.txt.bz2
This file contains the same keys (CountryCode, MapName, etc) as the metadata.properties file.
Field Data type Nullable
-------------- ----------- --------
keyname text not null
value text not null
Field Data type Nullable
-------------- ----------- --------
junctionid int not null
roadname1 text not null
roadname2 text not null
postalarea text not null
logicalroad1 int null
logicalroad2 int null
Field Data type Nullable
-------------- ----------- --------
logicalroadid int not null
junctionid int not null
Field Data type Nullable
-------------- ----------------- --------
poiid int not null
latitude numeric(11, 0) null
longitude numeric(11, 0) null
x real null
y real null
poiname text not null
poifullname text null
size smallint not null
poitypeid smallint not null
Field Data type Nullable
-------------- ----------- --------
logicalroadid int not null
roadname text not null
postalareas text not null
The following are gzip-compressed binary files containing map data suitable for rendering and routing (note that spatial indexes are generated on the server at the point when the files are read in). The actual physical file is written with a DataOutputStream, using the writeInt(), writeFloat() methods, etc. Strings are written using writeUTF().
The Metadata file contains the following sections:
header
record 1
record 2
...
record n
The header of the Metadata file contains the following fields:
Name Java data type Notes
----------------------------- --------------
File format version int Must be '2'
Number of records int
Each record in the Metadata file contains the following fields:
Name Java data type
----------------------------- --------------
key string
value string
The Place file contains the following sections:
header
record 1
record 2
...
record n
The header of the Place file contains the following fields:
Name Java data type Notes
----------------------------- --------------
File format version int
Minimum place ID int
Maximum place ID int
Number of records int
Each record in the Place file contains the following fields:
Name Java data type Notes
----------------------------- --------------
Place ID int
Name string
Size byte Arbitrary scale of place size/importance, where 0 = large capital city, 9 = tiny village
X coordinate float
Y coordinate float
The Junction file has the following structure:
header
record 1
record 2
...
record n
The Junction file header contains the following fields
Name Java data type Notes
--------------------- --------------
File format version int Must be '1'
Minimum junction ID int
Maximum junction ID int
Number of records int
Fields available int Bitfield of fields available in the records
Each record in the Junction file contains the following fields:
Name Java data type Notes
----------------------------- --------------
Junction ID int
X coordinate float
Y coordinate float
Attributes short (bitfield)
Number of attached sections byte
The Attached Section file has the following structure:
header
record 1
record 2
...
record n
The Attached Section file header structure contains the following fields
Name Java data type Notes
----------------------------- --------------
File format version int Must be '1'
Number of attached sections int
The Attached Section file record structure contains the following fields:
Name Java data type Notes
----------------------------- --------------
Junction ID int Foreign key into Junctions file
Sequence ID byte Incrementing from 0,
Section ID int ID of attached section. Foreign key into Sections file
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent