- rules is already used by logseq-query and will be useful for
nbb-logseq
- Updated db jobs to run in deps/db by default. Less typing but
more importantly discourages independent libraries from depending on
external scripts.
- Moved lint-rules task into db since it is db specific now
- Also remove nbb-logseq from top-level as all nbb compatibility is in
deps.
- Parser now parses all graph files like the app does, not just pages and journals.
This required extracting another fn from repo-handler
- Add and tweak CI steps that are specific to graph-parser. All
namespaces in this library are checked for nbb compatibility
- Cleaned up parser cli API so only one fn is needed for scripts
- Tests were updated to match new parsing behavior
- large_vars.clj can run with a smaller max-line-count after only refactoring two fns
- Add docs
- Also reuse docs graph test assertions to ensure main app
and graph-parser are in sync.
- Removed promesa from extract tests
since they are no longer async
- Added cljs and npm deps that were missing and caught by cljs tests
Had to tweak rewrite-clj to read vars with metadata.
Added fail fast behavior to fns to prevent silent failure
which was impacting tasks. Started looking at a few vars. More to look
at
- Updated docs with more accurate picture of translation state
- invalid task found keys that were no longer used from 2 commands
- Cleaned up rewrite-clj
* Remove expensive parsing when saving files
* Add limit to page blocks query
* Don't collapse block's body to make it compatible with other tools
* Alert if there're unsaved changes when switching graphs
* DB schema migration for :block/collapsed? from it's property
Co-authored-by: Andelf <andelf@gmail.com>
Start using :clj-kondo/ignore as lint violations that need to be
revisited. They may side effect and should be removed with better
understanding of their context