driveway | pure CSS masonry layouts | Grid library

The way SVG mask works is a bit strange. The element it's applied to will only be rendered where the mask is white and where the mask is black (or just not white) it will be hidden. Another strange effect of SVG masks is that if you are animating an element and apply a mask to the element being animated then the mask will move with the element.

To account for the first part is simple, just add a white rect the size of the SVG itself inside the mask and use smaller black shapes to mask. The way to work around the moving mask is to apply the mask not to the element being animated but to a tag that wraps the element(s) being animated.

If you want the circles to go 'under' a section and then 'over' that same section however, then you'll need to do some animating inside the mask as well. In this example I'm using animateTransform inside the mask's rect child (the black part that does the masking) to shrink it after the circles pass 'under' the bridge, you could just as easily use CSS keyframes though.

I suggest strongly that you also cut down the viewbox since your visual elements are so small compared to the available space, in the example I just estimated but the best way is to re-render your graphics in Illustrator and crop the artboard better to your objects.

Also most of the additional markup produced by Illustrator is not needed if the SVG will be inline in HTML. You can lose pretty much everything except the viewBox as shown in my example since those other attributes are mostly only used when the SVG is rendered as an image, hope this helps.

I think the only problem is that you are not taking into account the resulting direction when displaying the results. Just apply the following changes:

Okay, the solution is a little bit ironic. The tool that stack overflow uses for images can be used for anything else, so I just had to right-click on the image in this question, click on 'copy image URL' then put it in my website and I got a working image!! LMAO

Ah, figured it out.

In shape_element() replace subdiv_dict = get_subdiv_dict() with just subdiv_dict = dict().

Then rewrite subdiv_key():

You can use javascript like this:

You can pass the cam data directly into your render_template and use a jinja for loop to output a container (or whatever you wish) per object in your camStreams, e.g:

Your view:

Note: If this should be entered as a comment, and not an answer, I apologise.

Vladlen445, I found a way to fix the mis-alignment issues. Like you, I created a helper column, AJ in your TEST tab. AJ is filled with the project name, onto each row. I add the project number on the end, but this is not necessary, just cosmetic.

Then I created a new FOLDERS-Demo tab to show the calculations in a different way. Instead of using Test! column E, I now use Test! column AJ, which is the project name.

The project names are listed alphabetically, then the values for SUM(N) and SUM(O). There is a sort option, in FOLDER-Demo!W1, to sort by project number (from Test! column E) but this then shows the same error as your yellow section, due to project 28 in C158.

For your original formulas, in the yellow section, I haven't changed anything. I assume turning C158 to TRUE will correct your mis-alignment there. Let me know if this has clearly answered your question.

Note that FOLDER-Demo!W1 has data validation (Proj# or ProjName) for the sort options, and FOLDER-Demo!T1 is a pointer cell to control the sort column for the queries. It can be hidden by making the ink color the same as the background, and locked/protected to prevent accidental changes.

You can create a list of axis instances with subplots and plot the columns one-by-one:

You can loop $arr1, and check its element that matched in keys of $arr2, and set the matched value to $arr1, set empty string to the element that cannot found in keys of $arr2: