sodago
is an implementation of a subset of the soda4LCA
API. It is not intended to be used in
production but only for testing purposes. However, it comes with a very fast
server and is trivial to set up which makes it a fun tool when
developing/testing against the soda4LCA service API in
openLCA or the EPD
Editor.
sodago
is written in Go and compiles to a single binary:
cd sodago
go build # compile it
./sodago # run it
All data are stored as plain files in the data
folder. Within this folder
the content of each data stock is stored in a sub-folder. The name of the
data stock is the name of this sub-folder and the UUID of the data stock is
stored in a file .stock
in that folder. Additionally, the content of a
data stock is stored in the index.json
file. The data sets are stored under
their respective paths in the data folder (flows in flows
, processes in
processes
etc.). The file name of a data set is simply <uuid>_<version>.xml
.
External documents of sources are stored under external_docs/<source uuid>/<file>
.
Profiles are stored in the sub-folder profiles
of the data
directory. Thus,
you should not name a data stock profiles
. Each profile is stored in a
separate JSON file where the name has the following pattern: <profile ID>.json
The prefix /resource
is always added to all service routes (as in soda4LCA):
GET /datastocks
GET /profiles
GET /profiles/{id}
/authenticate/login
(returns always OK!)GET [/datastocks/{datastock}]/{path}
GET [/datastocks/{datastock}]/{path}/{id}[?version={version}]
GET [/datastocks/{datastock}]/sources/{id}/{file}
POST /{path}
POST /sources/withBinaries
TODO:
- implement: GET /sources/{uuid}/digitalfile
- put types to lower case (e.g. Version)
- normalize versions (1 == 1.00 == 1.00.000)
- simple search
- LCIA methods