3.5.2 hotfix: Imagery Search multi-image selection#289
Open
AlexeyTrekin wants to merge 6 commits into
Open
Conversation
Member
AlexeyTrekin
commented
May 26, 2026
AlexeyTrekin
commented
- Send every selected image (not just row 0) in cost and create-processing requests via get_search_images_ids
- Allow combining orbview_* providers (e.g. orbview_msi + orbview_pan) in a single processing when product type is Image; backend OrbviewService dispatches any orbview-prefixed name through the same client
- Mirror the orbview-family exception in validate_provider_params so the zoom/product-type gate doesn't re-block the same selection
- duplicate_imagery_search now expands the metadata table and pseudo metadata layer to one row per imageId and selects all rows so cost recalculation fires on the full payload
- Send every selected image (not just row 0) in cost and create-processing requests via get_search_images_ids - Allow combining orbview_* providers (e.g. orbview_msi + orbview_pan) in a single processing when product type is Image; backend OrbviewService dispatches any orbview-prefixed name through the same client - Mirror the orbview-family exception in validate_provider_params so the zoom/product-type gate doesn't re-block the same selection - duplicate_imagery_search now expands the metadata table and pseudo metadata layer to one row per imageId and selects all rows so cost recalculation fires on the full payload Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Findings from testing: - Backend rejects mixed orbview_* providers (orbview_msi + orbview_pan) in a single order, so the client-side allowance was wrong. Drop _is_orbview_family and the multi-provider exception from both get_search_images_ids and validate_provider_params. Multi-image processing remains supported when all selected rows share the same provider name (or are all Mosaic). - Provider minimum-area check needs to operate on the actually-processable area (user AOI ∩ union of selected image footprints), not the raw AOI. Introduce app_context.processing_aoi, populated by AreaCalculatorService with the cropped geometry, and use it as the processing request geometry in get_processing_schema (falling back to the raw AOI when no cropping was performed). Tests adjusted: orbview-family allow-cases removed; mixed-provider Image case now asserts the selection error is raised. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Previously the mismatch was caught only inside get_search_images_ids, which called dlg.disable_processing_start() as a side effect. That flashes the error in the UI but does not propagate to validate_all_processing_params, so update_processing_cost still sent /cost/v2 with the first provider name and all selected image IDs from different providers. Lift the same check into validate_provider_params so the error reaches the validation chain and both the cost and create-processing requests are short-circuited. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…request, duplicate fix
- Confirmation dialog now shows "(N images selected)" when more than one row
is queued in setup_provider_info.
- update_processing_cost short-circuits before any work when billing_type
is not CREDITS, so /cost calls are skipped under AREA / NONE billing.
- get_processing_schema now guards processing_aoi with explicit
isNull/isEmpty checks instead of relying on Python truthiness of
QgsGeometry, which varies across PyQGIS builds. AreaCalculator also
clears processing_aoi alongside aoi when the layer is reset, so stale
cropped state can't leak into a later request.
- duplicate_imagery_search rebuilt search_footprints with the raw
feature.attribute("local_index") value, which the in-memory layer
stores as a string. The rest of the codebase uses int keys
(get_local_image_indices casts via int(...)), so every lookup was
KeyError-ing — that left provider_names empty (no dataProvider in the
request → backend 400) and made the AOI cropping fall through to
QgsGeometry() (area displayed as 0). Coerce to int when building the
dict.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
When the user deselected every row, get_provider_params skipped its ImagerySearchProvider branch entirely, leaving imagery_search_provider_instance.image_ids at the value from the prior selection. validate_provider_params then checked `image_ids == None`, which is false for the stale list, so validate_all_processing_params returned no error and update_processing_cost dispatched /cost with dataProvider=None (no provider_name could be resolved) and the stale imageIds — backend 400. Fix: - Reset image_ids and requires_id on the instance when selectedItems() is empty. - Loosen validate_provider_params to treat the empty list the same as None so any path that ends up with no IDs short-circuits the request. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…l params validation; bump to 30s start processing request timeout
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.