Merge pull request #118 from procesor2017/master

- better log to image

README.md

@@ -4,14 +4,6 @@
 
 ## [Documentation](https://tesena-smart-testing.github.io/WatchUI/) | [Tesena](https://www.tesena.com/) | [Pypi](https://pypi.org/project/WatchUI/)
 
-## Important notice for users
-
-WatchUI 2.0 brings breaking changes. Dev team decided to streamline the library and focus it solely on the image and text comparison. This allows us to remove the implicit dependency on browser automation libraries - namely SeleniumLibrary, which was implicit part of the library via RF `BuiltIn()` import of the SeleniumLibrary instance.
-
-This is no longer the case - user of the WatchUI therefore **can and have to choose, what UI automation library will use** and provide screenshots to the WatchUI keywords to be compared. It could be now used with SeleniumLibrary, Browser library, Sikuli, Appium or any other UI library where visual validation is required.
-
-Version 1.x.x is no longer supported, but it is still available on [Pypi](pip install WatchUI==1.0.11).
-
 ### Basic Info
 
 Custom library for works with image, pdf and tesseract with RF.
@@ -52,3 +44,25 @@ _Image where the differences are stored + You can see two black box in left corn
 _The red rectangles outlining missing elements on compared screens_
 
 <img src="assets/example-difference.jpg">
+
+
+## Important notice for users 
+
+## For users with Windows
+
+The scikit image library is used for image matching. Unfortunately, this is a scientific library. 
+You need to have these libraries for installed to use them properly on windows:
+- Windows 10 SDK
+- C++ x64/x86 build tools
+
+These libraries can be downloaded by using [VS studio](https://visualstudio.microsoft.com/cs/) (not vscode) and installing the msbuild tool from VS studio.
+Alternatively, [this library](https://visualstudio.microsoft.com/cs/visual-cpp-build-tools/) should also work.
+
+
+## WatchUI 1.0 vs 2.x.x
+
+WatchUI 2.0 brings breaking changes. Dev team decided to streamline the library and focus it solely on the image and text comparison. This allows us to remove the implicit dependency on browser automation libraries - namely SeleniumLibrary, which was implicit part of the library via RF `BuiltIn()` import of the SeleniumLibrary instance.
+
+This is no longer the case - user of the WatchUI therefore **can and have to choose, what UI automation library will use** and provide screenshots to the WatchUI keywords to be compared. It could be now used with SeleniumLibrary, Browser library, Sikuli, Appium or any other UI library where visual validation is required.
+
+Version 1.x.x is no longer supported, but it is still available on [Pypi](pip install WatchUI==1.0.11).

WatchUI/keywords/Image.py

@@ -98,7 +98,8 @@ def _write_and_log(
         url: str = f"{checked_save_folder}/Img{time_}{img_format}"
         # Show image
         if score < ssim:
-            cv.imwrite(url, target_image)
+            img_diff: Any = cv.hconcat([base_image, target_image])
+            cv.imwrite(url, img_diff)
             self.set_log_message(
                 work_object="Image", type_of_messages="Error", path_to_image=url
             )
@@ -133,7 +134,7 @@ def compare_images(
             base_image_path (str): path to base image
             compared_image_path (str): path to compared image
             save_folder (str): path to the save folder
-            ssim (float): SSIM threshold value
+            ssim (float): SSIM threshold value 0 to 1 e.g. "0.5"
             image_format (str): image format, e.g. "png", "jpg"
         Raises:
             TypeError:
@@ -182,7 +183,7 @@ def compare_screen_without_areas(
             base_image_path (str): path to the base image
             compared_image_path (str): path to the compared image
             save_folder (str): path to the save folder
-            ssim (float): SSIM threshold value
+            ssim (float): SSIM threshold value 0 to 1 e.g. "0.5"
             image_format (str): format of the image, e.g. "png", "jpg"
         """
         checked_save_folder, my_ssim, img_format = self._do_checks(

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "WatchUI"
-version = "2.0.11"
+version = "2.0.12"
 description = "RobotFramework library package for automated visual testing."
 readme = "README.md"
 repository = "https://github.com/Tesena-smart-testing/WatchUI"

setup.py

@@ -2,4 +2,4 @@
 
 # setup for development purposes
 # run with python setup.py develop
-setup(name="WatchUI", version="2.0.11", packages=find_packages())
+setup(name="WatchUI", version="2.0.12", packages=find_packages())

tests/test.robot

@@ -1,6 +1,8 @@
 *** Variables ***
 ${TESERACT_PATH}     /usr/bin/tesseract
 ${BASELINE_IMAGE}    assets/img.jpg
+${IMAGE_1}           resource/data/forpres.png
+${IMAGE_2}           resource/data/forpres1.png
 ${PDF_FILE}          assets/ok.pdf
 
 
@@ -10,7 +12,11 @@ Library          ../WatchUI/          outputs_folder=tests/outputs  tesseract_pa
 
 *** Test Cases ***
 Compare Same Images
-    Compare images    ${BASELINE_IMAGE}    ${BASELINE_IMAGE}
+    Compare images    ${BASELINE_IMAGE}    ${BASELINE_IMAGE}    
+
+Compare different Images
+    [Tags]              1
+    Compare images    ${IMAGE_1}    ${IMAGE_2}      ssim=0.5
 
 Check Compare Screen Without Areas
     Compare screen without areas    ${BASELINE_IMAGE}    ${BASELINE_IMAGE}    0    0    100    100    125    125    250    250