[processing] fix description files for SVM SAGA algs (fix #20796)

(cherry picked from commit c6540a6)

Author: alexbruy

python/plugins/processing/algs/saga/description/SVMClassification.txt

@@ -5,10 +5,10 @@ QgsProcessingParameterRasterDestination|CLASSES|Classification
 QgsProcessingParameterEnum|SCALING|Scaling|[0] none;[1] normalize (0-1);[2] standardize|False|2
 QgsProcessingParameterBoolean|MESSAGE|Verbose Messages|False
 QgsProcessingParameterEnum|MODEL_SRC|Model Source|[0] create from training areas;[1] restore from file|False|0
-QgsProcessingParameterFile|MODEL_LOAD|Restore Model from File|QgsProcessingParameterFile.File|None|False
+QgsProcessingParameterFile|MODEL_LOAD|Restore Model from File|QgsProcessingParameterFile.File|None|None|True
 QgsProcessingParameterFeatureSource|ROI|Training Areas|-1|None|False
-QgsProcessingParameterFeatureSource|ROI_ID|Class Identifier|5|None|False
-QgsProcessingParameterFile|MODEL_SAVE|Store Model to File|QgsProcessingParameterFile.File|None|False
+QgsProcessingParameterField|ROI_ID|Class Identifier|None|ROI|-1|False|False
+QgsProcessingParameterFile|MODEL_SAVE|Store Model to File|QgsProcessingParameterFile.File|None|None|True
 QgsProcessingParameterEnum|SVM_TYPE|SVM Type|[0] C-SVC;[1] nu-SVC;[2] one-class SVM;[3] epsilon-SVR;[4] nu-SVR|False|0
 QgsProcessingParameterEnum|KERNEL_TYPE|Kernel Type|[0] linear;[1] polynomial;[2] radial basis function;[3] sigmoid|False|2
 QgsProcessingParameterNumber|DEGREE|Degree|QgsProcessingParameterNumber.Integer|3|False|None|None

python/plugins/processing/algs/saga/description/SupportVectorMachineClassification(OpenCV).txt

@@ -3,7 +3,7 @@ imagery_opencv
 QgsProcessingParameterMultipleLayers|FEATURES|Features|3|None|False
 QgsProcessingParameterBoolean|NORMALIZE|Normalize|False
 QgsProcessingParameterFeatureSource|TRAIN_AREAS|Training Areas|-1|None|False
-QgsProcessingParameterFeatureSource|TRAIN_CLASS|Class Identifier|5|None|False
+QgsProcessingParameterField|TRAIN_CLASS|Class Identifier|None|TRAIN_AREAS|-1|False|False
 QgsProcessingParameterRasterDestination|CLASSES|Classification
 QgsProcessingParameterEnum|SVM_TYPE|SVM Type|[0] c-support vector classification;[1] nu support vector classification;[2] distribution estimation (one class);[3] epsilon support vector regression;[4] nu support vector regression|False|0
 QgsProcessingParameterNumber|C|C|QgsProcessingParameterNumber.Double|1.000000|False|0.000000|None

python/plugins/processing/tests/testdata/custom/saga/training_areas.cpg

@@ -0,0 +1 @@
+UTF-8

python/plugins/processing/tests/testdata/custom/saga/training_areas.dbf

@@ -0,0 +1 @@
+GEOGCS["GCS_WGS_1984",DATUM["D_WGS_1984",SPHEROID["WGS_1984",6378137,298.257223563]],PRIMEM["Greenwich",0],UNIT["Degree",0.017453292519943295]]

python/plugins/processing/tests/testdata/custom/saga/training_areas.prj

@@ -0,0 +1 @@
+GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AUTHORITY["EPSG","4326"]]

python/plugins/processing/tests/testdata/custom/saga/training_areas.qpj

@@ -563,7 +563,7 @@ tests:
       OUTTAB:
         name: expected/saga/zonal_grid_stats.dbf
         type: vector
-  
+
   - algorithm: saga:resampling
     name: Resample down
     params:
@@ -668,3 +668,41 @@ tests:
         hash: aa197cdf1b0bac1595bd337bab08ae3cddb82ddb4d06c17b528a1d43
         type: rasterhash
 
+  - algorithm: saga:svmclassification
+    name: SVM Classification
+    params:
+      CACHE_SIZE: 100.0
+      COEF0: 0.0
+      COST: 1.0
+      CROSSVAL: 1
+      DEGREE: 3
+      EPS: 0.001
+      EPS_SVR: 0.1
+      GAMMA: 0.0
+      GRIDS:
+        params:
+        - name: dem.tif
+          type: raster
+        type: multi
+      KERNEL_TYPE: 2
+      MESSAGE: false
+      MODEL_LOAD:
+        name: ''
+        type: file
+      MODEL_SAVE:
+        name: ''
+        type: file
+      MODEL_SRC: 0
+      NU: 0.5
+      PROBABILITY: false
+      ROI:
+        name: custom/saga/training_areas.shp
+        type: vector
+      ROI_ID: class
+      SCALING: 2
+      SHRINKING: false
+      SVM_TYPE: 0
+    results:
+      CLASSES:
+        hash: deeb8769979916ab544a824cd5400d4a4207e7df37fbeaaa6e4830af
+        type: rasterhash