restructured to use docker compose #8
Conversation
Co-authored-by: Jakob Miksch <[email protected]>
Co-authored-by: Jakob Miksch <[email protected]>
Co-authored-by: Jakob Miksch <[email protected]>
|
@JakobMiksch I also included an update to the query function to use SQL instead of plpgsql in this PR |
There was a problem hiding this comment.
nice, this is much cleaner.
IMHO we can also drop the prefix OSM_PG_FEATURESERV_
or is there a specific need to have it?
There was a problem hiding this comment.
Just namespacing. I try to do it on all my project. If you prefer, I can remove it.
There was a problem hiding this comment.
yes, currently I do not see an advantage of the prefix. For me it makes the config files more verbose
|
|
||
| Then, create the necessary function for pg_featureserv to return features via API request: | ||
|
|
||
| ```sh | ||
| docker compose exec db psql -f /data/query_function.sql | ||
| ``` |
There was a problem hiding this comment.
I tried the README and apparently the function must be created after the OSM tables, because the function references the table
There was a problem hiding this comment.
I suggest we either move the creation of the function after the osm2pgsql or we revert the change of the function and handle the issue #4 in a separate PR
| * Specify slim mode to keep the "middle tables" so we can enable automatic updates [(relevant osm2pgsql docs)](https://osm2pgsql.org/doc/manual.html#middle) | ||
| * Specify output mode as "flex" so that we can use the flex output [(relevant osm2pgsql docs)](https://osm2pgsql.org/doc/manual.html#output-options) | ||
| * Use the flex style config file (`data/flex_syle.lua`) [(relevant osm2pgsql docs)](https://osm2pgsql.org/doc/manual.html#lua-library-for-flex-output) | ||
| * Specify "raw" as the prefix so that the tables being with `raw` [(relevant osm2pgsql docs)](https://osm2pgsql.org/doc/manual.html#database-layout) |
There was a problem hiding this comment.
IMHO we do not need this explanation. People can just read the offical docs or ask a LLM.
| db: | ||
| condition: service_healthy | ||
| db: | ||
| image: postgis/postgis:16-3.4 |
There was a problem hiding this comment.
the official postgis docker image automatically installs the tiger extension (see postgis/docker-postgis#187), which we do not need at all.
that's why I originally used the kartoza postgis image.
if we want to stick to the officical postgis image, we must mount a sql file with the content create extension postgis to the start directory
There was a problem hiding this comment.
I created a PR on this PR that fixes this: #11 just merge or approve it if you agree
|
@iboates thanks for the updates, I tried it and it worked mostly as expected. See my comments inline |
Co-authored-by: Jakob Miksch <[email protected]>
Closes #1
Closes #4
Everything is shoved into a single .env file because I have had a weird amount of problems when trying to specify non-default .env file locations.
Also changed the query to use SQL instead of plpgsql
In any case, everything should work as expected, just dockerized