This drops fomantic-ui as css toolkit and introduces tailwindcss. With
tailwind there are no predefined components, but it's very easy to
create those. So customizing the look&feel is much simpler, most of
the time no additional css is needed.
This requires a complete rewrite of the markup + styles. Luckily all
logic can be kept as is. The now old ui is not removed, it is still
available by using a request header `Docspell-Ui` with a value of `1`
for the old ui and `2` for the new ui.
Another addition is "dev mode", where docspell serves assets with a
no-cache header, to disable browser caching. This makes developing a
lot easier.
Mails are filtered once by using an imap search and then by some globs
to filter files and subjects. Imap can search by subject via a
string-contains, but not via globs or patterns (afaik). The subject
filter is applied to all downloaded mail headers. Now for post
processing (moving to some target folder or deleting), it can be
chosen to post-process all "seen" mails or only those that matched all
filters.
Classifier don't work on each attachment, but on all. So the results
must not be stored at an attachment. This reverts some previous
changes to put the classifier results for item entities into its own
table.
When re-indexing everything, skip intermediate populating the index
and do this as the very last step.
Parameterize adding new fields by their language.
Improves and reorganizes how nlp pipelines are setup. Now users can
choose from many options, depending on their hardware and usage
scenario.
This is the base to use more languages without depending on what
stanford-nlp supports. Support then is involves to text extraction and
simple regex-ner processing.
The scaling factor can be given in the config file. When this changes,
images can be regenerated via POSTing to certain endpoints. It is
possible to regenerate just one attachment preview or all within a
collective.
There is a task to generate preview images per attachment. It can
either add them (if not present yet) or overwrite them (e.g. some
config has changed).
There is a task that selects all attachments without previews and
submits a task to create it. This is submitted on start automatically
to generate previews for all existing attachments.
For images the conversion also returns the extracted text. If this
would have failed to be saved, it is extracted in the following
text-extraction step.
This value is used to decide whether to try OCR or not. If text is
below this value, OCR is run and both results are compared. It was set
to 10, which is just one or two words. Since the context for docspell
are documents, this value is too low.
- Use another external tool to convert pdf to pdf which also adds the
extracted text as another layer into the pdf
- Although not used, the external conversion routine will now check
for an existing text file that is named as the pdf file with extension
`.txt`. If present it is included in the conversion result and will be
used as the extracted text.
- text extraction for pdf files happens now on the converted file,
because it may already contain the text from the conversion step and
thus avoids running OCR twice.
- All errors during conversion are not fatal; processing continues
without a converted file.
A request to cancel a job was not processed correctly. The cancelling
routine of a task must run, regardless of the (non-final) state. Now
it works like this: if a job is currently running, it is interrupted
and its cancel routine is invoked. It then enters "cancelled" state.
If it is stuck, it is loaded and only its cancel routine is run. If it
is in a final state or waiting, it is removed from the queue.
Many errors cannot be recovered from by retrying. There is currently
no way to distinguish these states so it is now set to a lower value
to have not long wait times until an item arrives.
