chore: autoformat sparql-files with pre-commit #327
Conversation
danhje
force-pushed
the
autoformat
branch
3 times, most recently
from
e67f5ba to
b71b0ea
Compare
| insert {graph ${g} {?s cim:IdentifiedObject.mRID ?mrid}} | ||
| where { | ||
| graph ${g} {?s a ${rdf_type}} | ||
| filter (not exists {?s cim:IdentifiedObject.mRID ?mrid}) | ||
| bind(str(?s) as ?mrid) | ||
|
|
||
| INSERT { | ||
| GRAPH ?g { | ||
| ?s cim:IdentifiedObject.mRID ?mrid . | ||
| } | ||
| } | ||
| WHERE { | ||
| GRAPH ?g { | ||
| ?s a ?rdf_type . | ||
| } | ||
| FILTER NOT EXISTS { | ||
| ?s cim:IdentifiedObject.mRID ?mrid . | ||
| } | ||
| BIND (str(?s) AS ?mrid) |
There was a problem hiding this comment.
I struggled a bit with the substitutions ${g} and ${rdf_type} here. They trip up the formatter. Temporarily replacing them with placeholder strings, and putting the substitutions back after formatting, caused other issues. So I've removed them. But I haven't made the corresponding changes in the code yet. Are the substitutions necessary?
|
I unfortunately couldn't find a pip-installable CLI that actually works, this npm package is the only relatively bug-free formatter I could find. The substitutions are causing some issues, hence the script that temporarily replaces them. If we remove the substitutions we don't need the script. |
| filter(?analysis_enabled) | ||
| filter(?connected) | ||
| FILTER (?analysis_enabled = true) | ||
| FILTER (?connected = true) |
There was a problem hiding this comment.
When the formatter finds a FILTER call with just a bool variable as argument, it removes the parenthesis. Run it again, and it complains about the missing parenthesis ... I think that's a bug, and will file an issue. In the meantime, I'm explicitly comparing against true to avoid the issue.
No description provided.