import FJ .wsp files with 'CurlyQuad' gates

[phbradley]

Hi there, thank you for this great package, to which I am confess I am very new. Feel free to close this issue if it is not a prioriry-- I see that the README clearly states "limited support for FJ workspace files"!

Anyhow, I have been able to read almost all of the FlowJo workspace files I've tried, except for two, which both failed with this error:

File ~/envs/flowkit/lib/python3.11/site-packages/flowkit/_utils/wsp_utils.py:259, in _parse_population_node(pop_el, parent_id, gating_ns, data_type_ns)
256 # determine gate type
257 # TODO: this string parsing seems fragile, may need to be shored up
258 gate_type = gate_child_el.tag.partition('}')[-1]
--> 259 gate_class = wsp_gate_constructor_lut[gate_type]
261 # Lookup 'eventsInside' attribute value, determines whether gate results
262 # should within the gate area or outside (i.e. the gate's complement).
263 # The value is a text string of an integer: '1' for events inside,
264 # '0' for events outside.
265 try:

KeyError: 'CurlyQuad'

It looks like this is some particular kind of gate type which is not recognized, hence this issue. Again, thought, I totally understand if this is not a priority. Thanks again!

[whitews]

Hi Philip,

Glad you find the library useful. I'm willing to give it a shot, though no promises as it really depends on whether I can reverse engineer it properly. I'm currently working on spectral compensation support, which is taking longer than expected. And there's always other projects...

The best way to get this moving is if you can provide an example (.wsp file plus at least 1 FCS file in it). Even better if you can make a copy of the 8_color_ICS.wsp file with a CurlyQuad gate added (saves space in the repo). If you could do this it would be a great help, then I can parse through the XML and get a better sense of how difficult it will be to implement.

Kind regards,
Scott

[phbradley]

Hi Scott,
Thanks so much for the quick reply! I just left for a trip but I will see what I can do when I get back. Unfortunately in addition to being new to Flowkit I am also new to flow cytometry, so I may not be the best person, but I will consult with the collaborator who generated the data I am trying to parse.
Take care,
Phil

[whitews]

No worries Phil, the example doesn't have to make any biological sense. It would just be good to have a distribution of events in each quadrant. If your colleague can provide an example that would be great, but if not I may play around to generate something, though it may take some time for me to get to it.