wkhtmltopdf with full page background - wkhtmltopdf

I am using wkhtmltopdf to generate a PDF file that is going to a printer and have some troubles with making the content fill up an entire page in the resulting PDF.
In the CSS I've set the width and height to 2480 X 3508 pixels (a4 300 dpi) and when creating the PDF I use 0 for margins but still end up with a small white border to the right and bottom. Also tried to use mm and percentage but with the same result.
I'd need someone to please provide an example on how to style the HTML and what options to use at command line so that the resulting PDF pages fill out the entire background. One way might be to include bleeding (this might be necessary anyway) but any tips are welcome. At the moment I am creating one big HTML page (without CSS page breaks - might help?) but if needed it would be fine to generate each page separately and then feed them all to wkhtmltopdf.

wkhtmltopdf v 0.11.0 rc2
What ended up working:
wkhtmltopdf --margin-top 0 --margin-bottom 0 --margin-left 0 --margin-right 0 <url> <output>
shortens to
wkhtmltopdf -T 0 -B 0 -L 0 -R 0 <url> <output>
Using html from stdin (Note dash)
echo "<h1>Testing Some Html</h2>" | wkhtmltopdf -T 0 -B 0 -L 0 -R 0 - <output>
Using html from stdin to stdout
echo "Testing Some Html" | wkhtmltopdf -T 0 -B 0 -L 0 -R 0 - test.pdf
echo "Testing Some Html" | wkhtmltopdf -T 0 -B 0 -L 0 -R 0 - - > test.pdf
What did not work:
Using --dpi
Using --page-width and --page-height
Using --zoom

We just solved the same problem by using the --disable-smart-shrinking option.

I realize this is old and cold, but just in case someone finds this and has the same/similar problem, here's a workaround that worked for me after some trial&error.
I created a simple filler.html as:
<!DOCTYPE html>
<html>
<head>
</head>
<body style="margin: 0; padding: 0;">
<div style="height: 30mm; background-color: #F7EBD4;">
</div>
</body>
</html>
Use valid HTML (!DOCTYPE is important) and only inline styles. Match the background color to that of the main document and use height equal or bigger than your margins.
I run version 0.12.0 with the following arguments:
wkhtmltopdf --print-media-type --orientation portrait --page-size A4
--encoding UTF-8 --T 10mm --B 10mm --L 0mm --R 0mm
--header-html filler.html --footer-html filler.html - - <file.html >file.pdf
Hoping this helps someone...

I'm using version 0.12.2.1 and setting:
body { padding: 0; margin 0; }
div.page-layout { height: 295.5mm; width: 209mm;}
worked for me.
Of course need to add 0 margins by:
wkhtmltopdf -T 0 -B 0 -L 0 -R 0

At http://code.google.com/p/wkhtmltopdf/issues/detail?id=359 I found out more people 'suffer' from this bug. The --dpi 300 workaround did not work for me, I had to set --zoom 1.045 to zoom in a bit which made the extra right and bottom border disappear...

Works fine for me with -B 0 -L 0 -R 0 -T 0 options and using your trick of setting up an A4 sized div.
Did you remember to use body {margin:0; padding:0;} in the top of your CSS?
I cannot help you with CSS page breaks as I have not trialled an errored those yet, however, you can run scripts on the page to do clever things. Here is a jQuery example of how to split content down into page size chunks based on the length of the content. If you can get that adapted to work with wkhtmltopdf then please post here!
http://www.script-tutorials.com/demos/79/index.html

What you are experiencing is a bug.
You'll need to set the --dpi option when converting the file. In you case you will probably want --dpi 300, but that can be set lower.

Solved it by increasing the DPI
I'm working with an A4 size in portrait mode. Had white space to the right.
I noticed that as the dpi is increased, the white space got thinner.
at 300 dpi the white space is not visible in chrome pdf view at (max) zoomed at 500%
In Adobe reader it's still visible. It got better at 600 DPI and at 1200 DPI it's become invisible even at 6500% zoom.
There's no disadvantage to this so far as I observed, all dpi generate the same file size and run at the same speed (tested on 1 page).
effectively my settings are as follows:
echo "<html style='padding=0;margin=0'><body style='background-color:black;padding=0;margin=0'></html>" | wkhtmltopdf -T 0 -B 0 -L 0 -R 0 --disable-smart-shrinking --orientation portrait --page-size A4 --dpi 1200 - happy.pdf

If using an unscaled PNG image (thus will be pixel perfect) the default ratio, for an A4 needs to be 120ppi thus # 210mm = 993 pixels wide x 1404 pixels high, if the source is 72 or 300 dpi it makes no difference for a default placement, its the 993 that's counted as 210 mm
No heights, no width, no stretch, nor shrink just default place image as background un-scaled.
wkhtmltopdf --enable-local-file-access -T "0mm" -L "0mm" -R "0mm" -B "0mm" test.html test.pdf
here is such an image reduced into A 4 pdf page 2 different densities same number of pixels
If you use scaling you can use different density values, but this is all that is needed by default's, since PDF works on overall pixel values not DPI as such. Note the PNG is actually smaller by insertion in a PDF than the source JPG which was over 372 KB

Related

Firefox does not render .svg properly

I manage a blog where I use .svg files as illustrations, you can see a live example here: https://salarship.com/article/dress-fast-food-job-interview/
The problem is that the .svg files do not render properly on Firefox. Here is how the image looks on Firefox and how it looks on other browsers. Here is the raw file of the image: https://salarship.com/wp-content/uploads/2022/01/wear-job-interview-fast-food.svg
Is it a problem with this particular file, how can I fix it? I have hundreds of articles with this problem, is there a way to fix each image relatively quickly?
There is a path in there with the following command sequence (shortened for clarity):
... v 0 a 0.25,0.25 0 0 0 0,0.07 0.19,0.19 0 0 1 0,0.07 8.510071e11,8.510071e11 0 0 1 0,0.18 ...
The arc command with radius 8.510071e11 seems to throw Firefox. That is a bug.
since an arc with such a large radius is straight anyway, the sequence could be changed such that a v line command is used instead:
... v 0 a 0.25,0.25 0 0 0 0,0.07 0.19,0.19 0 0 1 0,0.07 v 0.18 ...

Image Conversion - RAW to png/raw for game (Pac The Man X)

So I have raw image and I am just curious If I can edit such image to save as RGB-32 Packed transparent interlaced raw and what program I could use, there is specification:
Format of RAW image
I have tried using photoshop but then game crashes. Is it even possible? I should get file without thumbnail. I also tried using gimp, free converters and Raw viewer but no luck. Any suggestions?
Edit:
Used photoshop (interleaved with transparency format), game starts but images are just bunch of pixels.
file that i try to prepare (221bits)
We are still not getting a handle on what output format you are really trying to achieve. Let's try generating a file from scratch, to see if we can get there.
So, let's just use simple commands that are available on a Mac and generate some test images from first principles. Start with exactly the same ghost.raw image you shared in your question. We will take the first 12 bytes as the header, and then generate a file full of red pixels and see if that works:
# Grab first 12 bytes from "ghost.raw" and start a new file "red.raw"
head -c 12 ghost.raw > red.raw
# Now generate 512x108 pixels, where red=ff, green=00, blue=01, alpha=fe and append to "red.raw"
perl -E 'say "ff0001fe" x (512*108)' | xxd -r -p >> red.raw
So you can try using red.raw in place of ghost.raw and tell me what happens.
Now try generating a blue file just the same:
# Grab first 12 bytes from "ghost.raw" and start a new file "blue.raw"
head -c 12 ghost.raw > blue.raw
# Now generate 512x108 pixels, where red=00, green=01, blue=ff, alpha=fe and append to "blue.raw"
perl -E 'say "0001fffe" x (512*108)' | xxd -r -p >> blue.raw
And then try blue.raw.
Original Answer
AFAIK, your image is actually 512 pixels wide by 108 pixels tall in RGBA8888 format with a 12-byte header at the start - making 12 + 4*(512 * 108) bytes.
You can convert it to PNG or JPEG with ImageMagick like this:
magick -size 512x108+12 -depth 8 RGBA:ghost.raw result.png
I still don't understand from your question or comments what format you actually want - so if you clarify that, I am hopeful we can get you answered.
Try using online converters. They help most of the time.\
A Website like these can possibly help:
https://www.freeconvert.com/raw-to-png
https://cloudconvert.com/raw-to-png
https://www.zamzar.com/convert/raw-to-png/
Some are specific websites which ask you for detail and some are straight forward conversions.

Graphviz not rendering node image

despite using the most simple boilerplate I can find, I can't seem to get graphviz to render a png image onto a node
Some may consider this a duplicate of How do I get DOT to display an image for a node? however this is the question I was following to get to this point and it STILL doesn't work
I've tried rendering already in several programs including vscode's preview extension, the dot -Tpng graph.png chickens2.dot -v command line method, and GVEdit and none of them correctly render the image into the node
here is my code
digraph graph1 {
node [shape=record];
white [image="dye_powder_white.png", label=""];
}
and the folder structure
Chickens/
|chickens2.dot
|dye_powder_white.png
I wanted it to simply show the icon inside of the node, but it just renders a blank node.
here is the output from dot.exe:
>dot -Tpng -o graph.png chickens2.dot -v
dot - graphviz version 2.38.0 (20140413.2041)
Using render: cairo:cairo
Using device: png:cairo:cairo
libdir = "C:\Program Files (x86)\Graphviz2.38\bin"
Activated plugin library: gvplugin_dot_layout.dll
Using layout: dot:dot_layout
The plugin configuration file:
C:\Program Files (x86)\Graphviz2.38\bin\config6
was successfully loaded.
render : cairo dot fig gd gdiplus map pic pov ps svg tk vml vrml xdot
layout : circo dot fdp neato nop nop1 nop2 osage patchwork sfdp twopi
textlayout : textlayout
device : bmp canon cmap cmapx cmapx_np dot emf emfplus eps fig gd gd2 gif gv imap imap_np ismap jpe jpeg jpg metafile pdf pic plain plain-ext png pov ps ps2 svg svgz tif tiff tk vml vmlz vrml wbmp xdot xdot1.2 xdot1.4
loadimage : (lib) bmp eps gd gd2 gif jpe jpeg jpg png ps svg
pack info:
mode undefined
size 0
flags 0
margin 8
pack info:
mode node
size 0
flags 0
fontname: "Times-Roman" resolved to: (ps:pango Times New Roman, ) (PangoCairoWin32Font) "Times New Roman, 13.9990234375"
network simplex: 1 nodes 0 edges maxiter=2147483647 balance=1
network simplex: 1 nodes 0 edges 0 iter 0.00 sec
Maxrank = 0, minrank = 0
mincross: pass 0 iter 0 trying 0 cur_cross 0 best_cross 0
mincross graph1: 0 crossings, 0.00 secs.
network simplex: 1 nodes 0 edges maxiter=2147483647 balance=2
network simplex: 1 nodes 0 edges 0 iter 0.00 sec
routesplines: 0 edges, 0 boxes 0.00 sec
Using render: cairo:cairo
Using device: png:cairo:cairo
dot: allocating a 19K cairo image surface (83 x 60 pixels)
gvRenderJobs graph1: 0.00 secs.
here is the exact png image I'm trying to embed just in case it's incompatible with graphviz and I don't realize:
https://mega.nz/#!M8EHEITJ!Mm489BA4sd5JLoeWlY7BM-YCIkHeAni96d-e1IY4UAQ
and finally, the resulting image which is the same in every method I've tried:
https://i.imgur.com/IcIJSMA.png
some things I'm considering:
is there some kind of configuration I'm not seeing which disables images by default?
do I have some kind of outdated version? (literally downloaded the stable this morning to try it out so I doubt it)
am I simply doing something wrong?
are graphviz images incompatible with windows 10?
am I missing a prerequisite library?
is my png image incompatible with graphviz somehow?
I've been trying this for an hour or so now and none of my attempts seem to be changing anything.
what is the correct way to get graphviz to render an image onto a node?
Old question, but I also couldn't figure this out. The attribute image doesn't seem to work as expected, and the manpage doesn't help, nor do the docs. Posting what worked for me for others who come across this question.
This answered it for me, from 2011. For example, in some image.gv file:
digraph {
ratio="fill";
size="10,10!";
margin="0,0";
node [shape=plain];
root [label=<<TABLE border="0"><TR><TD><IMG SRC="image.png"/></TD></TR>
<TR><TD>text under</TD></TR></TABLE>>];
}

How to crop AI (PDF embedded) to PNG using Ghostscript?

I've read a number of post and tried to follow but it's not working.
Using GS (gsdll32.dll) with the following arguments:
Info from bbox
%%BoundingBox: 33 244 577 546 %%HiResBoundingBox: 33.611976 244.201633
576.009896 545.351819
render and crop AI2PNG
-P-
-dNOPAUSE
-dBATCH
-dSAFER
-q
-IC:/Program Files (x86)/Gerber Scientific Products/OMEGA 6.50/Software/gs/fonts;C:/Program Files (x86)/Gerber Scientific Products/OMEGA 6.50/Software/gs/lib;C:/Program Files (x86)/Gerber
Scientific Products/OMEGA 6.50/Software/gs/resource
-sDEVICE=pngalpha
-g544x302
-c <> setpagedevice
-sOutputFile=E:/Images/AI from PLM/captain-america [Converted].png E:/Images/AI from PLM/captain-america [Converted].ai
Without any cropping logic I get the image on an 8.5 x 11, with cropping(above commands) the objects are translated mostly off the top of the page and do not seem to move to the left.
The size of the result image is correct.
Does anyone see anything wrong?
Thanks
You've put the /Install after the input file, that means it will be executed after the input file is complete. Which means it takes effect after the input is completely processed, which is too late to have nay effect.
Order of switches, and particularly order of input, is important in Ghostscript.
That's assuming that 'AI2PNG' is a synonym for Ghostscript.

want eps file and text into single compound path using ghostscript

what I want is to add eps file into temporary ps file which has text written, then I convert my ps file to eps file using ghostscript, but when I see my eps file in AI outline mode, I see extra square around my eps file which is size box, which should not be there, It should be part of single compound box
Ghostscript version is 9.05 and before I include eps into ps I need to resize it. So resized eps file shows page border into outline mode. Which is actually not there, but when it goes to machine it will cut out that path which should not be case.
Alright, I think I have some understanding of what you're doing and where the trouble may be creeping in. As I commented, you're running the file through ghostscript multiple times. Each time it has to interpret the postscript code and construct an internal display-list representation and then recreate appropriate postscript code on the output end. So it's the clone of a clone of a clone problem. Any little hiccup can cause cascade failures.
So, enough of being preachy. The alternative is to manipulate the eps file as text.
So, if we want the image to be scaled to fill a 500x500 square, we will be guided by the number in the BoundingBox comment. I'll quote this silly file from the linked question as an example:
%!PS-Adobe-2.0 EPSF-2.0
%%BoundingBox: 72 700 127 708 %<-- modify this
%%HiResBoundingBox: 72.000000 700.000000 127.000000 707.500000 %<-- delete this
%%EndComments
% EPSF created by ps2eps 1.68
%%BeginProlog
save
countdictstack
mark
newpath
/showpage {} def
/setpagedevice {pop} def
%%EndProlog
%%Page 1 1 %<-- insert translate and scale after this line
/Times-Roman findfont
11 scalefont setfont
72 700 moveto
(This is a test)show
%%Trailer
cleartomark
countdictstack
exch sub { end } repeat
restore
%%EOF
So, the BoundingBox was %%BoundingBox: 72 700 127 708 and it needs to be 0 0 500 500 or rather (to preserve the aspect ratio) 0 0 x 500 or 0 0 500 y where x or y (whichever it happens to be) is < 500. The existing size is 127-72 x 708 - 700 = 55 x 8. So our scaling factor is 500/55. But we also want to translate the lower-left corner to the origin, and its simplest to do that first, so the scaling doesn't affect the interpretation of the numbers.
So, to take 72 700 127 708 to 0 0 500 y, first we add -72 -700 translate to the file, and modify the bounding box to 0 0 55 8, and delete that silly HiRes line: we don't really need it.
Then, we add 500 55 div dup scale (let the interpreter do the math, hee hee). So the maximum x will now be 500, but, oh, what to put for the y? A quick calculation yields 72!
So, this awk program will modify an eps file to be 500 points wide, with y scaled appropriately.
/%%BoundingBox: ([^ ]*) ([^ ]*) ([^ ]*) ([^ ]*)/{x=$2;y=$3;w=$4-x;h=$5-y;print $1,0,0,500,(500/w)*h}
!/%%BoundingBox:/&&!/%%HiRes/{print}
/%%Page /{print -x,-y,"translate"; print 500,w,"div dup scale"}
Usage:
$ awk -f epsscale.awk etest.eps

Resources