I want to have a rectangle that takes all the place in a SVG file. It should also have a border (3px stroke width). The size of the graphic should be easy changeable (by changing attributes "width" and "height" of the "svg" node). I came up with following construction:
<svg width="150" height="35" >
<g>
<rect
id="rect6648"
style="fill:#ffffff; fill-opacity:1; stroke:#000000; stroke-width:3;"
x="0"
y="0"
width="100%"
height="100%" />
</g>
</svg>
But it produces following image with dirty border:
I need something like this:
Is it possible at all? As mentioned before it must work for any size of the graphic.
Thanks in advance!
Alas, no, at least not with purely declarative SVG. The stroke on a shape is painted on both sides of the geometric line that defines that shape (in your case, there's 1.5 on either side). Because of that, it will get clipped for a shape that fills the whole viewbox.
In which context are you using this? You should be able to script it: get the size of the viewbow, on rect set x and y to stroke-width/2, width to width - stroke-width and height to height - stroke-width. If in a dynamic context you will need to detect resizes, but that's often possible.
You need to place the ractangle at half pixel coordinates like x="0.5" y="0.5", then the borders won't be blurry. Also add vector-effect:non-scaling-stroke to the rectangle's CSS to be sure that the border is always 3px wide regardless of zoom level.
Related
According to the standards I'm reading -- for example https://www.w3.org/TR/css-values-4/#viewport-relative-lengths -- a vmin unit should be 1% of the smallest dimension of the containing viewport.
Going for a minimal example illustrating my dilemma, this is square in my current instance of chrome:
<svg><rect height="30vmin" width="30vmin" fill="red">
But this is not:
<svg><rect height="50vmin" width="50vmin" fill="red">
Playing with variations on this theme (closing tags, adding width and height to the svg element, etc.) suggests that the rect is not using the svg viewport as its reference, but instead is using some containing browser context as its reference viewport.
So, my question is: how do I specify to the browser that I want vmin units to refer to the innermost containing svg viewport? (Specifically when working with svg elements embedded in html documents.)
Browser support for those units that were added in CSS3 may still be spotty. I haven't checked recently.
But the rule is that these units are resolved relative to the whole document. So in a browser, that will be the whole browser window.
This SVGWG issue may help clarify things.
https://github.com/w3c/svgwg/issues/207
how do I specify to the browser that I want vmin units to refer to the innermost containing svg viewport?
You can use percentage values for coordinates,
<rect height="50%" width="50%" fill="red">
However in SVG, percentage values are always relative to their associated axis. So percentage width values are relative to the X axis, and percentage height values are relative to the Y axis.
Alternatively you could use a suitable viewBox and appropriate coordinate values relative to that viewBox. For example, if your viewBox has a width and height of 100:
viewBox="0 0 100 100"
All coordinates values in the SVG would effectively be percentage values. However the same axis rule applies as described above.
svg {
width: 200px;
background-color: linen;
}
<svg viewBox="0 0 100 100">
<!-- rectnagle 50% x 33.3% -->
<rect width="50" height="33.3"/>
</svg>
I noticed the the root coordinates for a text element are not at the top left corner like a rect element:
Is there a way to set it such that when a text element is at (0,0), it fits inside the parent element?
If I understood you well, you can use this:
https://developer.mozilla.org/en-US/docs/Web/SVG/Attribute/dominant-baseline
A) Chromium browsers
svg {
dominant-baseline: hanging;
}
https://jsfiddle.net/e7vc4bqj/
B) Chromium and Firefox
.text {
dominant-baseline: hanging;
}
https://jsfiddle.net/3zskd148/
SVG text coordinates are used to define its left bottom corner by default:
https://developer.mozilla.org/en-US/docs/Web/SVG/Attribute/text-anchor
Hope this help :)
Why are the SVG Text elements too high?
The x and y coordinates of a <text> element specify the start of the baseline of the text. This makes complete sense. You wouldn't want it to be the top left of the first character - because it would then be a difficult job to get text of different sizes and styles to line up.
There is no global option in SVG that changes that behaviour. However see below for alternatives)
Is there a way to set it such that when a text element is at (0,0), it fits inside the parent element?
Normally you would just adjust the y coordinate based on the font size.
However there are a couple of alternatives you can use:
One is the xxx-baseline properties (as #gengns has pointed out), that can alter how the character glyphs are positioned relative to the baseline. Note however, that those attributes are not entirely reliable, due to mixed browser support. Plus they depend on the font containing the correct data tables. Not all fonts have those tables.
A better option IMO is to use the dy attribute. This adds a relative offset to the text position. Meaning the text is actually positioned at (x, y + dy). And it is supported by all browsers.
<svg width="200" height="150">
<rect x="0" y="0" width="200" height="150" fill="skyblue"/>
<text x="0" y="0"
font-size="25px" dy="1em">asd</text>
</svg>
Q: How can I use the viewBox coordinate system whilst still filling the viewPort completely and preserving aspect ratio?
I'm new to svg programming, so hopefully I'm just mis-understanding a basic concept.
I want to create an interactive & responsive map with , based on a background image that the user uploads.
Here's the basic example I'm trying to get to work (JSFiddle):
<svg version="1.1"
xmlns="http://www.w3.org/2000/svg"
xmlns:xlink="http://www.w3.org/1999/xlink"
width="200px"
height="400px"
preserveAspectRatio="xMidYMid meet"
style="border: 1px solid black;">
<image x="0" y="0" width="100%" height="100%"
preserveAspectRatio="xMidYMid meet"
xlink:href="http://www.bized.co.uk/sites/bized/modules/bized_cb_navigation/images/floorplan_info.gif">
</image>
</svg>
This works nicely. since however the viewPort changes, the image always fills it whilst maintaining its aspect ratio. (See wide-Screen example)
Next I add a coordinate system viewBox="0 0 100 100":
the wide-screen view still fills nicely
but the vertical-screen view now does not fill the viewPort anymore
If you take a different image that is wider than tall, then the wide-screen view breaks, and the vertical-screen view still works.
When I inspect the SVG in Chrome DOM Element inspector, for the first two examples without using viewBox="0 0 100 100" The svg element has the same size as the viewPort. Once the viewBox attribute is added, the element becomes a square with sides equal to the lesser of the viewPort's sides.
This behavior is explained in this Tutorial as:
"... the view box is scaled according to the smaller of the two aspect ratios..."
I need the viewBox attribute so that I can zoom and pan on the image within the viewPort.
This is because you effectively have two competing viewBox transformations.
Because of your square viewBox, you are fitting the image into a square, and then fitting the square into your SVG rectangle.
If you make your SVG viewBox the same dimensions as your image (or the same aspect ratio will do), then the problem will be resolved.
viewBox="0 0 155 210"
http://jsfiddle.net/2qexypLs/15/
http://jsfiddle.net/2qexypLs/16/
My current solution is to use JavaScript to dynamically set the viewBox width and height values to the same value as svg width and height values. That way the aspect ratio for x and y are the same and the fill returns to 100% of viewPort. (JsFiddle)
For the interactive elements layer of the map I have a separate coordinate system that is mapped to the background image scaling ratio when the svg viewPort is defined. That means all coordinates need to be recalculated on svg define/change width/height event.
After this first re-calculation, the map can be zoomed and panned by changing the viewBox parameters without any further calculations.
TL;DR: I need to mask out a portion of one rectangle in SVG, based on the size and position of another existing rectangle, which will be changing dynamically. A Chrome bug is blocking the mask + use approach I tried. How can I do a mask or inverted clip path based on an existing shape?
Full Overview:
I'm using D3.js, and I am using the brush control to add a brush to a rectangle in an embedded SVG. By default, this adds some extra elements to the SVG, including a rect with class extent that shows the size of the brushed area.
Rather than have the brush extent be rendered as a semi-transparent overlay on top of the rectangle, as in most D3 examples, I am trying to "cut out" the extent from a semi-transparent overlay, so that the brush area shows the true color below. Per this question, I am trying to do this with a mask element, with a child use element referencing the extent. With some D3 magic, I now have a structure like this:
<svg width="100" height="100">
<g class="brush-layer inverted">
<defs>
<mask id="mask835">
<rect fill="#fff" width="100%" height="100%"></rect>
<use fill="#000" xlink:href="#extent848"></use>
</mask>
</defs>
<g class="brush" style="pointer-events: none;">
<rect class="overlay" mask="url(#mask835)" width="100%" height="17"></rect>
<rect class="extent" x="30" width="52" height="17" id="extent848"></rect>
</g>
</g>
</svg>
This works great... sort of. It turns out that there appears to be a tricky Chrome bug, which I've filed here, which prevents the mask from being applied if there's a #media query in the CSS. You can see the working version here and the failing version here (fails in Chrome, works in FF).
I need this to work in Chrome, and can't drop the #media query. I also need to make the use element work, because D3 will automatically resize the extent rectangle, and that's the shape I need to mask out.
So, how can I mask out a portion of one rect, based on another rect, without using the mask + use strategy above?
One possible workaround might be to use a custom clip-path, but it's probably not going to be as elegant. Some examples of how to do do this with clip-path can be found in this question.
I am using multiple differently colored rectangles to build a SVG data visualization. This works great but sometimes background color bleeds through between the rectangles. I am browsing with Chrome but other browsers seem similarly affected.
http://jsfiddle.net/dVEPk/
<svg xmlns="http://www.w3.org/2000/svg" version="1.1">
<rect x="10.5" y="10" height="100" width="100"
style="stroke:none; fill: #00cc00"/>
<rect x="110.5" y="10" height="100" width="100"
style="stroke:none; fill: #00cc00"/>
</svg>
In Chrome, if the x offset is an integer, the line is not visible.
I'm sure I can avoid lines by making rectangles a little larger than the space they have to occupy. But this seems like a hack: is there an SVG idiom or best practice to achieve perfectly tiled shapes without "grout"?
I'm also concerned by rendering performance because my visualizations can be very large (say 100MB XML .svg). I'd like to be able to give the renderer hints like "none of the shapes in this <g> are overlapping"?
This is antialiasing at work between the shape and the background. If you want to turn it off set shape-rendering="crispEdges" on the shapes. You can either set that on the rect elements or on the <svg> in which case the rect elements will inherit it.
You may be able to adjust the line's positions by adding 0.5 to the co-ordinates. See the cairo FAQ for more details on this.