* removed first-header-extractor
get first heading from markdown rende
* don't remove editor or renderer just hide them
this way both are always up to date and can be shown very fast
* extracted image alt attribute into first title. too
* added tests as suggested by @mrdrogdrog
* Add caching of user-data for 600 seconds
* Make cache-entry interface commonly usable
* Extract revision types
* Remove revision-cache rule
* Use seconds as cache-time interval (Date.now uses milliseconds)
* Fix import error
* Extract cache logic into common cache-class
* Add cache class that was forgotten to commit in last commit
* Start adding unit tests
* Fix bug detected during unit-testing
* Add unit tests for cache
* Made entry-limit test more explicit
* Renamed files to lower-case starting letter
* Change element.io links to matrix.to links
Following to the discussion in codimd/server#496, this commit changes element.io links to matrix.to in the new react-client as well.
* Updated cypress test for new link
which puts the markdown text directly to a variable in the route function.
Content-type of the http request is checked to be text/markdown because we dealing with markdown. Technically by now there can be any content which can be encoded. There could be features in the software which do not work properly if the text can't be parsed as markdown.
Signed-off-by: Yannick Bungers <git@innay.de>
ConsoleLoggerService is based on the default Nest LoggerService, but adds the ability to give context about the function that is logging something. It also removes the `[Nest]` string and the PID at the beginning of each log line.
NestConsoleLoggerService is a wrapper around ConsoleLoggerService and makes it possible to use our implementation as a default Nest LoggerService
Signed-off-by: David Mehren <git@herrmehren.de>
this was just a simple index problem. The 0. entry of a regex result is the whole string and not the first capture group
added e2e tests to fix this in the future
cypress does currently not support copy to clipboard in firefox so this part of the test is commented out (for the time being)
It might be handy to have access to the original `Note` after creating one, so the creation and conversion to a `NoteDto` is now split.
Signed-off-by: David Mehren <git@herrmehren.de>
NoteRevisionsMetadata is an array containing revision data and not an object with a single property containing an array.
Revision timestamps are ISO strings, not UNIX timestamps.
Signed-off-by: David Mehren <git@herrmehren.de>
It was helpful to inspect database contents while the code was stopped by the debugger. Therefore the E2E test database is now persisted on disk and cleared before every test-run.
Signed-off-by: David Mehren <git@herrmehren.de>
The precision of sqlites datetime() timestamp is only one second (see https://www.sqlite.org/lang_datefunc.html). Therefore we could not order revisions of one note that were created in the same second. To remedy this, the primary key was changed to a monotonically increasing number, which solves the ordering problem.
Signed-off-by: David Mehren <git@herrmehren.de>
