From 3da557d82c75e6775be825c3b9197ed80f351d45 Mon Sep 17 00:00:00 2001 From: Georgy Litvinov Date: Fri, 28 Feb 2020 18:09:37 +0100 Subject: [PATCH] Progress on applying csv metadata to rdf --- build.gradle | 1 + .../java/writer2latex/rdf/DocumentPart.java | 7 +++++ .../writer2latex/rdf/DocumentStructure.java | 26 ++++++++++++++----- .../writer2latex/rdf/RDFDocumentResult.java | 8 +++--- .../java/writer2latex/xhtml/XhtmlConfig.java | 7 ++++- 5 files changed, 37 insertions(+), 12 deletions(-) diff --git a/build.gradle b/build.gradle index e0e5921..8b9d199 100644 --- a/build.gradle +++ b/build.gradle @@ -22,6 +22,7 @@ dependencies{ bundledLibs 'org.libreoffice:unoil:5.3.2' compile 'junit:junit:4.12' bundledLibs group: 'org.apache.jena', name: 'jena-core', version: '3.14.0' + bundledLibs group: 'com.opencsv', name: 'opencsv', version: '5.1' bundledLibs files('idl') configurations.compile.extendsFrom(configurations.bundledLibs) } diff --git a/src/main/java/writer2latex/rdf/DocumentPart.java b/src/main/java/writer2latex/rdf/DocumentPart.java index d6e73fe..3a871de 100644 --- a/src/main/java/writer2latex/rdf/DocumentPart.java +++ b/src/main/java/writer2latex/rdf/DocumentPart.java @@ -28,6 +28,7 @@ public class DocumentPart { private String body; private String parentPath; private String name; + private String order; public DocumentPart(XhtmlDocument document) { this.excerptDoc = document; @@ -63,6 +64,12 @@ public class DocumentPart { public String getBody() { return body; } + public void setOrder(String order) { + this.order = order; + } + public String getOrder() { + return order; + } public String getParentPath() { return parentPath; } diff --git a/src/main/java/writer2latex/rdf/DocumentStructure.java b/src/main/java/writer2latex/rdf/DocumentStructure.java index eea9d50..83fd25b 100644 --- a/src/main/java/writer2latex/rdf/DocumentStructure.java +++ b/src/main/java/writer2latex/rdf/DocumentStructure.java @@ -1,18 +1,15 @@ package writer2latex.rdf; import java.io.File; -import java.io.FileOutputStream; import java.io.FileWriter; import java.io.IOException; import java.io.OutputStream; import java.util.HashMap; -import java.util.Iterator; import java.util.Set; import java.util.Vector; import org.apache.jena.ontology.OntClass; import org.apache.jena.ontology.OntModel; -import org.apache.jena.rdf.model.Model; import org.apache.jena.rdf.model.ModelFactory; import org.apache.jena.rdf.model.Resource; import org.apache.jena.vocabulary.*; @@ -20,7 +17,6 @@ import org.apache.jena.vocabulary.*; import writer2latex.xhtml.XhtmlDocument; import org.apache.jena.rdf.model.Property; -import org.apache.jena.rdf.model.RDFWriter; public class DocumentStructure { @@ -56,12 +52,17 @@ public class DocumentStructure { public DocumentStructure(Vector files,String fileName) { this(); this.documentID = fileName; - Iterator filesIterator = files.iterator(); - while (filesIterator.hasNext()) { - XhtmlDocument inputDoc = filesIterator.next(); + //Iterator filesIterator = files.iterator(); + //while (filesIterator.hasNext()) { + //XhtmlDocument inputDoc = filesIterator.next(); + for(int i = 0 ; i< files.size();i++) { + XhtmlDocument inputDoc = files.get(i); DocumentPart part = new DocumentPart(inputDoc); + part.setOrder(Integer.toString(i)); addPart(part); } + + //} addEmptyParts(); } @@ -205,5 +206,16 @@ public class DocumentStructure { } } } + public void applyMetadata(Metadata metadata) { + for (DocumentPart part: inputParts.values()) { + String partOrder = part.getOrder(); + //Set blank main doc number as it is in CSV + if (partOrder.equals("0")) { + partOrder = ""; + } + + + } + } } diff --git a/src/main/java/writer2latex/rdf/RDFDocumentResult.java b/src/main/java/writer2latex/rdf/RDFDocumentResult.java index 88aa66f..cb80ad6 100644 --- a/src/main/java/writer2latex/rdf/RDFDocumentResult.java +++ b/src/main/java/writer2latex/rdf/RDFDocumentResult.java @@ -1,12 +1,9 @@ package writer2latex.rdf; -import java.io.ByteArrayOutputStream; import java.io.IOException; import java.io.OutputStream; -import java.util.Iterator; import java.util.Vector; -import java.util.zip.ZipEntry; -import java.util.zip.ZipOutputStream; + import writer2latex.api.ConverterResult; import writer2latex.api.OutputFile; @@ -28,6 +25,9 @@ public class RDFDocumentResult implements OutputFile { this.sFileName = Misc.removeExtension(fileName); this.config = config; rdfStructure = new DocumentStructure(outFiles,fileName); + Metadata metadata = new Metadata(); + metadata.read(config.getCSVMetadataFile()); + rdfStructure.applyMetadata(metadata); rdfStructure.createTree(); } diff --git a/src/main/java/writer2latex/xhtml/XhtmlConfig.java b/src/main/java/writer2latex/xhtml/XhtmlConfig.java index 3c82936..0fc7cc4 100644 --- a/src/main/java/writer2latex/xhtml/XhtmlConfig.java +++ b/src/main/java/writer2latex/xhtml/XhtmlConfig.java @@ -40,7 +40,7 @@ import writer2latex.util.Misc; public class XhtmlConfig extends writer2latex.base.ConfigBase { // Implement configuration methods - protected int getOptionCount() { return 64; } + protected int getOptionCount() { return 65; } protected String getDefaultConfigPath() { return "/writer2latex/xhtml/config/"; } // Override setOption: To be backwards compatible, we must accept options @@ -163,6 +163,7 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase { private static final int MIN_LETTER_SPACING = 61; private static final int PAGE_BREAK_STYLE = 62; private static final int ALIGN_SPLITS_TO_PAGES = 63; + private static final int CSV_METADATA = 64; protected ComplexOption xheading = addComplexOption("heading-map"); protected ComplexOption xpar = addComplexOption("paragraph-map"); @@ -292,6 +293,8 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase { options[PAGINATION] = new BooleanOption("pagination","true"); options[MIN_LETTER_SPACING] = new Option("min_letter_spacing","0.15"); options[PAGE_BREAK_STYLE] = new Option("page_break_style",""); + options[CSV_METADATA] = new Option("csv_metadata",""); + options[CSS_INLINE] = new BooleanOption("css_inline","true"); options[ALIGN_SPLITS_TO_PAGES] = new BooleanOption("align_splits_to_pages","false"); @@ -439,6 +442,8 @@ public class XhtmlConfig extends writer2latex.base.ConfigBase { public String getXhtmlUplink() { return options[UPLINK].getString(); } public String getXhtmlDirectoryIcon() { return options[DIRECTORY_ICON].getString(); } public String getXhtmlDocumentIcon() { return options[DOCUMENT_ICON].getString(); } + public String getCSVMetadataFile() { return options[CSV_METADATA].getString(); } + public boolean getGreenstoneSeparation() { if ( ((IntegerOption) options[SPLIT_LEVEL]).getValue() != 0) { return false;