Add rasterization upon export

Author	DaveJarvis <email>
Date	2021-03-07 23:15:25 GMT-0800
Commit	bfef24b3221246f4b618546281edfb1d508aa90e
Parent	b368ed8

src/main/java/com/keenwrite/MainPane.java

		.stream()
		.collect(
		- groupingBy( path -> bin.apply( MediaType.valueFrom( path ) ) )
		+ groupingBy( path -> bin.apply( MediaType.fromFilename( path ) ) )
		);

src/main/java/com/keenwrite/io/HttpMediaType.java

		map.forEach( ( key, values ) -> {
		if( "content-type".equalsIgnoreCase( key ) ) {
		- var header = values.get( 0 );
		- // Trim off the character encoding.
		- var i = header.indexOf( ';' );
		- header = header.substring( 0, i == -1 ? header.length() : i );
		-
		- // Split the type and subtype.
		- i = header.indexOf( '/' );
		- i = i == -1 ? header.length() : i;
		- final var type = header.substring( 0, i );
		- final var subtype = header.substring( i + 1 );
		-
		- mediaType[ 0 ] = MediaType.valueFrom( type, subtype );
		+ mediaType[ 0 ] = MediaType.valueFrom( values.get( 0 ) );
		clue( "Main.status.image.request.success", mediaType[ 0 ] );
		}

		return mediaType[ 0 ];
		}
		+
		}

src/main/java/com/keenwrite/io/MediaType.java

		*/
		public static MediaType valueFrom( final File file ) {
		- return valueFrom( file.getName() );
		+ return fromFilename( file.getName() );
		+ }
		+
		+ /**
		+ * Returns the {@link MediaType} associated with the given file name.
		+ *
		+ * @param filename The file name that may contain an extension associated
		+ * with a known {@link MediaType}.
		+ * @return {@link MediaType#UNDEFINED} if the extension has not been
		+ * assigned, otherwise the {@link MediaType} associated with this
		+ * URL's file name extension.
		+ */
		+ public static MediaType fromFilename( final String filename ) {
		+ return getMediaType( getExtension( filename ) );
		}



		/**
		- * Returns the {@link MediaType} associated with the given file name.
		+ * Determines the media type an IANA-defined, semi-colon-separated string.
		+ * This is often used after making an HTTP request to extract the type
		+ * and subtype from the content-type.
		*
		- * @param filename The file name that may contain an extension associated
		- * with a known {@link MediaType}.
		- * @return {@link MediaType#UNDEFINED} if the extension has not been
		- * assigned, otherwise the {@link MediaType} associated with this
		- * URL's file name extension.
		+ * @param header The content-type header value.
		+ * @return The data type for the resource or {@link MediaType#UNDEFINED} if
		+ * unmapped.
		*/
		- public static MediaType valueFrom( final String filename ) {
		- return getMediaType( getExtension( filename ) );
		+ public static MediaType valueFrom( String header ) {
		+ // Trim off the character encoding.
		+ var i = header.indexOf( ';' );
		+ header = header.substring( 0, i == -1 ? header.length() : i );
		+
		+ // Split the type and subtype.
		+ i = header.indexOf( '/' );
		+ i = i == -1 ? header.length() : i;
		+ final var type = header.substring( 0, i );
		+ final var subtype = header.substring( i + 1 );
		+
		+ return valueFrom( type, subtype );
		}



		/**
		- * Used by {@link MediaTypeExtension} to initialize associations where the
		- * subtype name and the file name extension have a 1:1 mapping.
		+ * Returns the IANA-defined subtype classification. Primarily used by
		+ * {@link MediaTypeExtension} to initialize associations where the subtype
		+ * name and the file name extension have a 1:1 mapping.
		*
		* @return The IANA subtype value.
		*/
		- String getSubtype() {
		+ public String getSubtype() {
		return mSubtype;
		}

src/main/java/com/keenwrite/preview/SvgRasterizer.java

		package com.keenwrite.preview;

		-import javafx.scene.image.ImageView;
		import org.apache.batik.anim.dom.SAXSVGDocumentFactory;
		import org.apache.batik.gvt.renderer.ImageRenderer;

		import static javax.xml.transform.OutputKeys.*;
		import static org.apache.batik.transcoder.SVGAbstractTranscoder.KEY_WIDTH;
		+import static org.apache.batik.transcoder.image.ImageTranscoder.KEY_PIXEL_UNIT_TO_MILLIMETER;
		import static org.apache.batik.util.XMLResourceDescriptor.getXMLParserClassName;



		/**
		- * Rasterizes the resource specified by the path into an image.
		+ * Rasterizes the given SVG input stream into an image at 96 DPI.
		*
		- * @param svg The SVG data to rasterize.
		- * @return The resource at the given path as an {@link ImageView}.
		+ * @param svg The SVG data to rasterize, must be closed by caller.
		+ * @return The given input stream converted to a rasterized image.
		*/
		public static BufferedImage rasterize( final InputStream svg )
		throws TranscoderException {
		- final var in = new TranscoderInput( svg );
		+ return rasterize( svg, 96 );
		+ }
		+
		+ /**
		+ * Rasterizes the given SVG input stream into an image.
		+ *
		+ * @param svg The SVG data to rasterize, must be closed by caller.
		+ * @param dpi Resolution to use when rasterizing (default is 96 DPI).
		+ * @return The given input stream converted to a rasterized image at the
		+ * given resolution.
		+ */
		+ public static BufferedImage rasterize(
		+ final InputStream svg, final float dpi )
		+ throws TranscoderException {
		final var transcoder = new BufferedImageTranscoder();
		- transcoder.transcode( in, null );
		+ transcoder.addTranscodingHint(
		+ KEY_PIXEL_UNIT_TO_MILLIMETER, 1f / dpi * 25.4f );
		+ transcoder.transcode( new TranscoderInput( svg ), null );
		return transcoder.getImage();
		}

		}

		+ /**
		+ * Rasterizes the given vector graphic file using the width dimension
		+ * specified by the document's width attribute.
		+ *
		+ * @param document The {@link Document} containing a vector graphic.
		+ * @return A rasterized image as an instance of {@link BufferedImage}, or
		+ * {@link #BROKEN_IMAGE_PLACEHOLDER} if the graphic could not be rasterized.
		+ */
		public static BufferedImage rasterize( final Document document ) {
		try {

src/main/java/com/keenwrite/preview/SvgReplacedElementFactory.java


		if( getProtocol( source ).isHttp() ) {
		- var mediaType = MediaType.valueFrom( source );
		+ var mediaType = MediaType.fromFilename( source );

		if( isSvg( mediaType ) \|\| mediaType == UNDEFINED ) {

		}
		}
		- else if( isSvg( MediaType.valueFrom( source ) ) ) {
		+ else if( isSvg( MediaType.fromFilename( source ) ) ) {
		// Attempt to rasterize based on file name.
		final var path = Path.of( new URI( source ).getPath() );

src/main/java/com/keenwrite/processors/ProcessorFactory.java


		import com.keenwrite.AbstractFileFactory;
		+import com.keenwrite.preferences.Workspace;
		import com.keenwrite.preview.HtmlPreview;
		import com.keenwrite.processors.markdown.MarkdownProcessor;

		? createHtmlPreviewProcessor()
		: context.isExportFormat( XHTML_TEX )
		- ? createXhtmlProcessor()
		+ ? createXhtmlProcessor( context.getWorkspace() )
		: createIdentityProcessor();


		* @return An instance of {@link Processor} that completes an HTML document.
		*/
		- private Processor<String> createXhtmlProcessor() {
		- return new XhtmlProcessor();
		+ private Processor<String> createXhtmlProcessor( final Workspace workspace ) {
		+ return new XhtmlProcessor( workspace );
		}

src/main/java/com/keenwrite/processors/XhtmlProcessor.java

		package com.keenwrite.processors;

		+import com.keenwrite.io.MediaType;
		+import com.keenwrite.preferences.Workspace;
		+import com.keenwrite.util.ProtocolScheme;
		+
		+import javax.imageio.ImageIO;
		+import java.io.File;
		+import java.io.IOException;
		+import java.net.URL;
		+import java.nio.file.Files;
		+
		+import static com.keenwrite.Bootstrap.APP_TITLE_LOWERCASE;
		+import static com.keenwrite.events.StatusEvent.clue;
		+import static com.keenwrite.io.MediaTypeExtension.valueFrom;
		+import static com.keenwrite.preview.SvgRasterizer.rasterize;
		+import static java.io.File.createTempFile;
		+import static java.nio.file.StandardCopyOption.REPLACE_EXISTING;
		+import static org.jsoup.Jsoup.parse;
		+import static org.jsoup.nodes.Document.OutputSettings.Syntax;
		+
		/**
		* Responsible for making the body of an HTML document complete by wrapping
		- * it with html and body elements.
		+ * it with html and body elements. This doesn't have to be super-efficient
		+ * because it's not run in real-time.
		*/
		public final class XhtmlProcessor extends ExecutorProcessor<String> {
		+ private final Workspace mWorkspace;
		+
		+ public XhtmlProcessor( final Workspace workspace ) {
		+ mWorkspace = workspace;
		+ }
		+
		@Override
		public String apply( final String html ) {
		- return "<html><body>" + html + "</body></html>";
		+ final var doc = parse( html );
		+ doc.outputSettings().syntax( Syntax.xml );
		+
		+ for( final var img : doc.getElementsByTag( "img" ) ) {
		+ final var src = img.absUrl( "src" );
		+
		+ try {
		+ final var url = new URL( src );
		+ final var protocol = ProtocolScheme.valueFrom( url );
		+
		+ if( protocol.isRemote() ) {
		+ final var conn = url.openConnection();
		+ conn.setUseCaches( false );
		+
		+ final var type = conn.getContentType();
		+ final var media = MediaType.valueFrom( type );
		+
		+ try( final var in = conn.getInputStream() ) {
		+ File imageFile;
		+
		+ if( media == MediaType.IMAGE_SVG_XML ) {
		+ // Rasterize.
		+ final var image = rasterize( in, 300f );
		+ final var mt = MediaType.IMAGE_PNG;
		+ imageFile = createTemporaryFile( mt );
		+ ImageIO.write( image, mt.getSubtype(), imageFile );
		+ }
		+ else {
		+ // Download into temporary directory.
		+ imageFile = createTemporaryFile( media );
		+ Files.copy( in, imageFile.toPath(), REPLACE_EXISTING );
		+ }
		+
		+ img.attr( "src", imageFile.getAbsolutePath() );
		+ }
		+ }
		+ } catch( final Exception ex ) {
		+ clue( ex );
		+ }
		+ }
		+
		+ return doc.html();
		+ }
		+
		+ private static File createTemporaryFile( final MediaType media )
		+ throws IOException {
		+ final var file = createTempFile(
		+ APP_TITLE_LOWERCASE, '.' + valueFrom( media ).getExtension() );
		+ file.deleteOnExit();
		+ return file;
		}
		}

src/main/java/com/keenwrite/util/BoundedCache.java


		@Override
		- protected boolean removeEldestEntry(
		- final Map.Entry<K, V> eldest ) {
		+ protected boolean removeEldestEntry( final Map.Entry<K, V> eldest ) {
		return size() > mCacheSize;
		}

src/test/java/com/keenwrite/io/MediaTypeTest.java


		/**
		- * Test that {@link MediaType#valueFrom(String)} can lookup by file name.
		+ * Test that {@link MediaType#fromFilename(String)} can lookup by file name.
		*/
		@Test

		);

		- map.forEach( ( k, v ) -> assertEquals( v, valueFrom( "f." + k ) ) );
		+ map.forEach( ( k, v ) -> assertEquals( v, fromFilename( "f." + k ) ) );
		}

tex/build.sh

		+#!/usr/bin/env bash
		+
		+#
		+# Example command-line to be invoked directly by the application.
		+#
		+
		+context --environment="setups,entities,style,classes" "$1.xml"
		+

tex/classes.tex

		+\defineframedtext[projection][
		+ style=tt,
		+ width=\textwidth,
		+]
		+

tex/entities.tex

		+% Map XHTML document entities to ConTeXt symbols and TeX macros.
		+
		+\xmltexentity{ldquo}{\symbol[leftquotation]{}}
		+\xmltexentity{rdquo}{\symbol[rightquotation]{}}
		+\xmltexentity{rsquo}{\quotesingle{}}
		+\xmltexentity{mdash}{\emdash{}}
		+\xmltexentity{ndash}{\endash{}}
		+\xmltexentity{hellip}{\dots{}}
		+

tex/setups.tex

		-\startxmlsetups xml:xhtmlsetups
		+% XML setups map ConTeXt commands to HTML elements.
		+
		+\startxmlsetups xml:xhtml
		% Do not typeset the HTML document's header title element.
		\xmlsetsetup{\xmldocument}{*}{-}

		- % Typeset the elements that follow.
		- \xmlsetsetup{\xmldocument}{html\|body\|h1\|h2\|h3\|p\|em\|strong\|q\|div}{xml:*}
		+ % Only elements included in this list are typeset.
		+ \xmlsetsetup{\xmldocument}{%
		+ html\|body\|div\|h1\|h2\|h3\|h4\|h5\|h6\|p\|blockquote\|span\|em\|q\|b\|strong\|%
		+ a\|ul\|ol\|li\|dl\|dt\|dd\|hr\|br\|sup\|sub\|code\|img%
		+ }{xml:*}
		\stopxmlsetups

		-\xmlregistersetup{xml:xhtmlsetups}
		+\xmlregistersetup{xml:xhtml}

		\startxmlsetups xml:html

		\stopxmlsetups

		+% Paragraphs are followed by a paragraph break.
		\startxmlsetups xml:p
		\xmlflush{#1}\par
		\stopxmlsetups

		+% Emphasized text is italicized, typically.
		\startxmlsetups xml:em
		\dontleavehmode{\em\xmlflush{#1}}
		\stopxmlsetups

		+% Strong text is bolded, typically.
		\startxmlsetups xml:strong
		\dontleavehmode{\bf\xmlflush{#1}}
		+\stopxmlsetups
		+
		+\startxmlsetups xml:img
		+ \placefigure[none]{}{%
		+ \externalfigure[\xmlatt{#1}{src}]
		+ }
		\stopxmlsetups

		\startxmlsetups xml:q
		\quotation{\xmlflush{#1}}
		\stopxmlsetups

		+% Map arbitrary div classes, defined by fenced divs.
		\startxmlsetups xml:div
		- GOT DIV?
		+ \start[\xmlatt{#1}{class}]
		+ \xmlflush{#1}
		+ \stop
		\stopxmlsetups

tex/style.tex

		\setupindenting[medium, yes]

		+\setupexternalfigures[
		+ %order={svg,pdf,png},
		+ %directory={images},
		+ %maxwidth=\makeupwidth,
		+ width=\makeupwidth,
		+]
		+
		+% Force images to flow exactly where they fall in the text.
		+\setupfloat[figure][default=force]
		+
		+% Indent the paragraph following each image.
		+\setupfloats[indentnext=yes]
		+

Delta	211 lines added, 47 lines removed, 164-line increase