bitbake: toastergui: improvements in layer selection logic

This patch clearers and bring fixes for the layer selection
logic in order to enable information collected during build to be used
in configuring projects, specifically targeting the recipes
learned through the building process.

The patch also adds tests to verify the layer selection logic.

[YOCTO #7189]

(Bitbake rev: f0faba8ef0f08c98ac4bddf5b3954d540820d215)

Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

1 files changed, 24 insertions, 11 deletions

diff --git a/bitbake/lib/toaster/toastergui/views.py b/bitbake/lib/toaster/toastergui/views.py
index 6ccbf5452d..7353844bf1 100755
--- a/bitbake/lib/toaster/toastergui/views.py
+++ b/bitbake/lib/toaster/toastergui/views.py
@@ -22,7 +22,7 @@
 import operator,re
 import HTMLParser
 
-from django.db.models import Q, Sum, Count
+from django.db.models import Q, Sum, Count, Max
 from django.db import IntegrityError
 from django.shortcuts import render, redirect
 from orm.models import Build, Target, Task, Layer, Layer_Version, Recipe, LogMessage, Variable
@@ -236,7 +236,7 @@ def _get_queryset(model, queryset, filter_string, search_term, ordering_string,
     if search_term:
         queryset = _get_search_results(search_term, queryset, model)
 
-    if ordering_string and queryset:
+    if ordering_string:
         column, order = ordering_string.split(':')
         if column == re.sub('-','',ordering_secondary):
             ordering_secondary=''
@@ -2046,7 +2046,7 @@ if toastermain.settings.MANAGED:
                         "url": x.layercommit.layer.layer_index_url,
                         "layerdetailurl": reverse("layerdetails", args=(x.layercommit.pk,)),
                 # This branch name is actually the release
-                        "branch" : { "name" : x.layercommit.commit, "layersource" : x.layercommit.up_branch.layer_source.name}},
+                        "branch" : { "name" : x.layercommit.commit, "layersource" : x.layercommit.up_branch.layer_source.name if x.layercommit.up_branch != None else None}},
                     prj.projectlayer_set.all().order_by("id")),
             "targets" : map(lambda x: {"target" : x.target, "task" : x.task, "pk": x.pk}, prj.projecttarget_set.all()),
             "freqtargets": freqtargets,
@@ -2243,11 +2243,11 @@ if toastermain.settings.MANAGED:
 
             # returns layer versions that provide the named targets
             if request.GET['type'] == "layers4target":
-                # we returnd ata only if the recipe can't be provided by the current project layer set
-                if reduce(lambda x, y: x + y, [x.recipe_layer_version.filter(name="anki").count() for x in prj.projectlayer_equivalent_set()], 0):
+                # we return data only if the recipe can't be provided by the current project layer set
+                if reduce(lambda x, y: x + y, [x.recipe_layer_version.filter(name=request.GET['value']).count() for x in prj.projectlayer_equivalent_set()], 0):
                     final_list = []
                 else:
-                    queryset_all = prj.compatible_layerversions().filter(recipe_layer_version__name = request.GET.get('value', '__none__'))
+                    queryset_all = prj.compatible_layerversions().filter(recipe_layer_version__name = request.GET['value'])
 
                     # exclude layers in the project
                     queryset_all = queryset_all.exclude(pk__in = [x.id for x in prj.projectlayer_equivalent_set()])
@@ -2259,14 +2259,20 @@ if toastermain.settings.MANAGED:
 
             # returns targets provided by current project layers
             if request.GET['type'] == "targets":
-                queryset_all = Recipe.objects.all()
+                queryset_all = Recipe.objects.filter(name__icontains=request.GET.get('value',''))
                 layer_equivalent_set = []
                 for i in prj.projectlayer_set.all():
                     layer_equivalent_set += i.layercommit.get_equivalents_wpriority(prj)
                 queryset_all = queryset_all.filter(layer_version__in =  layer_equivalent_set)
+
+                # if we have more than one hit here (for distinct name and version), max the id it out
+                queryset_all_maxids = queryset_all.values('name').distinct().annotate(max_id=Max('id')).values_list('max_id')
+                queryset_all = queryset_all.filter(id__in = queryset_all_maxids)
+
+
                 return HttpResponse(jsonfilter({ "error":"ok",
-                    "list" : map ( lambda x: {"id": x.pk, "name": x.name, "detail":"[" + x.layer_version.layer.name+ (" | " + x.layer_version.up_branch.name + "]" if x.layer_version.up_branch is not None else "]")},
-                        queryset_all.filter(name__icontains=request.GET.get('value',''))[:8]),
+                    "list" : map ( lambda x: {"id": x.pk, "name": x.name, "detail":"[" + x.layer_version.layer.name + (" | " + x.layer_version.up_branch.name + "]" if x.layer_version.up_branch is not None else "]")},
+                        queryset_all[:8]),
 
                     }), content_type = "application/json")
 
@@ -2663,10 +2669,17 @@ if toastermain.settings.MANAGED:
 
         queryset_with_search = _get_queryset(Recipe, queryset_all, None, search_term, ordering_string, '-name')
 
-        queryset_with_search.prefetch_related("layer_source")
+        # get unique values for 'name' and 'version', and select the maximum ID for each entry (the max id is the newest one)
+        queryset_with_search_maxids = queryset_with_search.values('name').distinct().annotate(max_id=Max('id')).values_list('max_id')
+
+        queryset_with_search = queryset_with_search.filter(id__in=queryset_with_search_maxids).select_related('layer_version', 'layer_version__layer')
+
+        objects = list(queryset_with_search)
+        for e in objects:
+            e.preffered_layerversion = e.layer_version.get_equivalents_wpriority(prj)[0]
 
         # retrieve the objects that will be displayed in the table; targets a paginator and gets a page range to display
-        target_info = _build_page_range(Paginator(queryset_with_search, request.GET.get('count', 10)),request.GET.get('page', 1))
+        target_info = _build_page_range(Paginator(objects, request.GET.get('count', 10)),request.GET.get('page', 1))
 
 
         context = {