Reftest analyzer: load raw structured log

Either paste your log into this textarea:
<br /> <input type="button" value="Process pasted log" onclick="log_pasted()" /> </p> <p> ... or load it from a file:<br /> <input type="file" id="fileentry" onchange="fileentry_changed()" /> </p> </div> <div id="loading" style="display: none">Loading log...</div> <div id="viewer" style="display: none"> <div id="pixelarea"> <div id="pixelinfo"> <table> <tbody> <tr> <th>Pixel at:</th> <td colspan="2" id="coords" /> </tr> <tr> <th>Image 1:</th> <td id="pix1rgb"></td> <td id="pix1hex"></td> </tr> <tr> <th>Image 2:</th> <td id="pix2rgb"></td> <td id="pix2hex"></td> </tr> </tbody> </table> <div> <div id="pixelhint"> ★ <div> <p> Move the mouse over the reftest image on the right to show magnified pixels on the left. The color information above is for the pixel centered in the magnified view. </p> <p> Image 1 is shown in the upper triangle of each pixel and Image 2 is shown in the lower triangle. </p> </div> </div> </div> </div> <div id="magnification"> <svg xmlns="http://www.w3.org/2000/svg" width="84" height="84" shape-rendering="optimizeSpeed" > <g id="mag" /> </svg> </div> </div> <div id="itemlist"></div> <div id="images" style="display: none"> <form id="imgcontrols"> <input id="radio1" type="radio" name="which" value="0" onchange="show_image(1)" checked="checked" /><label id="label1" title="1" for="radio1">Image 1</label> <input id="radio2" type="radio" name="which" value="1" onchange="show_image(2)" /><label id="label2" title="2" for="radio2">Image 2</label> <label ><input id="differences" type="checkbox" onchange="show_differences(this)" />Circle differences</label > </form> <svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" version="1.1" width="800" height="1000" id="svg" > <defs> <!-- use sRGB to avoid loss of data --> <filter id="showDifferences" x="0%" y="0%" width="100%" height="100%" style="color-interpolation-filters: sRGB" > <feImage id="feimage1" result="img1" xlink:href="#image1" /> <feImage id="feimage2" result="img2" xlink:href="#image2" /> <!-- inv1 and inv2 are the images with RGB inverted --> <feComponentTransfer result="inv1" in="img1"> <feFuncR type="linear" slope="-1" intercept="1" /> <feFuncG type="linear" slope="-1" intercept="1" /> <feFuncB type="linear" slope="-1" intercept="1" /> </feComponentTransfer> <feComponentTransfer result="inv2" in="img2"> <feFuncR type="linear" slope="-1" intercept="1" /> <feFuncG type="linear" slope="-1" intercept="1" /> <feFuncB type="linear" slope="-1" intercept="1" /> </feComponentTransfer> <!-- w1 will have non-white pixels anywhere that img2 is brighter than img1, and w2 for the reverse. It would be nice not to have to go through these intermediate states, but feComposite type="arithmetic" can't transform the RGB channels and leave the alpha channel untouched. --> <feComposite result="w1" in="img1" in2="inv2" operator="arithmetic" k2="1" k3="1" /> <feComposite result="w2" in="img2" in2="inv1" operator="arithmetic" k2="1" k3="1" /> <!-- c1 will have non-black pixels anywhere that img2 is brighter than img1, and c2 for the reverse --> <feComponentTransfer result="c1" in="w1"> <feFuncR type="linear" slope="-1" intercept="1" /> <feFuncG type="linear" slope="-1" intercept="1" /> <feFuncB type="linear" slope="-1" intercept="1" /> </feComponentTransfer> <feComponentTransfer result="c2" in="w2"> <feFuncR type="linear" slope="-1" intercept="1" /> <feFuncG type="linear" slope="-1" intercept="1" /> <feFuncB type="linear" slope="-1" intercept="1" /> </feComponentTransfer> <!-- c will be nonblack (and fully on) for every pixel+component where there are differences --> <feComposite result="c" in="c1" in2="c2" operator="arithmetic" k2="255" k3="255" /> <!-- a will be opaque for every pixel with differences and transparent for all others --> <feColorMatrix result="a" type="matrix" values="0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0" /> <!-- a, dilated by 1 pixel --> <feMorphology result="dila1" in="a" operator="dilate" radius="1" /> <!-- a, dilated by 2 pixels --> <feMorphology result="dila2" in="dila1" operator="dilate" radius="1" /> <!-- all the pixels in the 2-pixel dilation of a but not in the 1-pixel dilation, to highlight the diffs --> <feComposite result="highlight" in="dila2" in2="dila1" operator="out" /> <feFlood result="red" flood-color="red" /> <feComposite result="redhighlight" in="red" in2="highlight" operator="in" /> <feFlood result="black" flood-color="black" flood-opacity="0.5" /> <feMerge> <feMergeNode in="black" /> <feMergeNode in="redhighlight" /> </feMerge> </filter> </defs> <g onmousemove="magnify(evt)"> <image x="0" y="0" width="100%" height="100%" id="image1" /> <image x="0" y="0" width="100%" height="100%" id="image2" /> </g> <rect id="diffrect" filter="url(#showDifferences)" pointer-events="none" x="0" y="0" width="100%" height="100%" /> </svg> </div> </div> </body> </html>