update: use cheerio to parse local novel

This commit is contained in:
nyagami
2024-01-28 18:43:20 +07:00
parent bf4d761c05
commit 58f6475749
7 changed files with 188 additions and 228 deletions
@@ -1,174 +0,0 @@
package com.rajarsheechatterjee.EpubParser;
import androidx.annotation.NonNull;
import com.facebook.react.bridge.Promise;
import com.facebook.react.bridge.ReactApplicationContext;
import com.facebook.react.bridge.ReactContextBaseJavaModule;
import com.facebook.react.bridge.ReactMethod;
import com.facebook.react.bridge.ReadableArray;
import com.facebook.react.bridge.ReadableMap;
import com.facebook.react.bridge.WritableArray;
import com.facebook.react.bridge.WritableMap;
import com.facebook.react.bridge.WritableNativeArray;
import com.facebook.react.bridge.WritableNativeMap;
import org.w3c.dom.Document;
import org.w3c.dom.Element;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.util.regex.Pattern;
import java.util.zip.ZipEntry;
import java.util.zip.ZipInputStream;
import javax.xml.parsers.DocumentBuilder;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.xpath.XPath;
import javax.xml.xpath.XPathConstants;
import javax.xml.xpath.XPathFactory;
public class EpubParser extends ReactContextBaseJavaModule {
String rootPath;
String contentDirPath;
XPath xPath = XPathFactory.newInstance().newXPath();
DocumentBuilder db = DocumentBuilderFactory.newInstance().newDocumentBuilder();
EpubParser(ReactApplicationContext context) throws Exception {
super(context);
}
@NonNull
@Override
public String getName() {
return "EpubParser";
}
@ReactMethod
public void parse(String epubFilePath, String epubDirPath, Promise promise) {
try{
this.rootPath = epubDirPath;
String novelCover = this.unzip(epubFilePath, epubDirPath);
File contentFile = this.parseContainer(epubDirPath);
this.contentDirPath = contentFile.getParent();
WritableMap novel = (WritableMap) this.parseContent(contentFile);
if(novelCover != null){
novel.putString("cover", new File(epubDirPath, novelCover).getAbsolutePath());
}
promise.resolve(novel);
} catch (Exception e) {
promise.reject(e.getCause());
}
}
public File parseContainer(String epubDir) throws Exception {
Document doc = this.parserDocument(new File(epubDir, "META-INF/container.xml"));
return new File(this.rootPath,
((Element) doc.getElementsByTagName("rootfile").item(0)).getAttribute("full-path"));
}
public ReadableMap parseContent(File contentFile) throws Exception {
Document content = this.parserDocument(contentFile);
WritableMap novel = new WritableNativeMap();
novel.putString("name", this.getFirstElementByTag(content, "dc:title").getTextContent().trim());
novel.putString("url", contentFile.getParent());
novel.putString("author", this.getFirstElementByTag(content, "dc:creator").getTextContent().trim());
novel.putArray("chapters", this.getChapters(content));
return novel;
}
public ReadableArray getChapters(Document doc) throws Exception {
WritableArray chapters = new WritableNativeArray();
Document toc = null;
if (this.getFirstElementByTag(doc, "spine").getAttribute("toc") != null) {
toc = this.parserDocument(new File(this.contentDirPath, "toc.ncx"));
}
NodeList itemrefs = this.getNodeListByTag(doc, "itemref");
for (int i = 0; i < itemrefs.getLength(); i++) {
WritableMap chapter = new WritableNativeMap();
Element itemref = (Element) itemrefs.item(i);
chapter.putString("name", itemref.getAttribute("idref"));
Element item = this.getElementById(doc, chapter.getString("name"));
chapter.putString("url", new File(this.contentDirPath, item.getAttribute("href")).getAbsolutePath());
if(toc != null){
chapter.putString("name", this.getChapterName(toc, item.getAttribute(("href"))));
if(chapter.getString("name") == null) continue;
}
chapters.pushMap(chapter);
}
return chapters;
}
public String getChapterName(Document doc, String href) throws Exception {
String path = String.format("//*[starts-with(@src,'%1$s')]", href);
Node node = ((NodeList) this.xPath.compile(path).evaluate(doc, XPathConstants.NODESET)).item(0);
if(node != null){
String text = node.getPreviousSibling().getTextContent().trim();
if(text.isEmpty()) text = node.getPreviousSibling().getPreviousSibling().getTextContent().trim();
return !text.isEmpty() ? text : null;
}
return null;
}
public Element getElementById(Document doc, String id) throws Exception {
return (Element) ((NodeList) this.xPath
.compile(String.format("//*[@id='%1$s']", id))
.evaluate(doc, XPathConstants.NODESET))
.item(0);
}
public NodeList getNodeListByTag(Document doc, String tag) throws Exception {
return (NodeList)
(
this.xPath.compile(String.format("//*[contains(name(), '%1$s')]", tag))
.evaluate(doc, XPathConstants.NODESET)
);
}
public Element getFirstElementByTag(Document doc, String tag) throws Exception {
return (Element) this.getNodeListByTag(doc, tag).item(0);
}
public Document parserDocument(File docFile) throws Exception {
Document doc = this.db.parse(docFile);
doc.getDocumentElement().normalize();
return doc;
}
public String escapeFilePath(String filePath){
return filePath.replaceAll(":", "\uA789");
}
public String unzip(String epubFilePath, String epubDirPath) throws Exception {
ZipInputStream zis = new ZipInputStream(new FileInputStream(epubFilePath));
ZipEntry zipEntry;
int len;
String novelCover = null;
byte[] buffer = new byte[4096];
while ((zipEntry = zis.getNextEntry()) != null) {
String escapedFilePath = escapeFilePath(zipEntry.getName());
if(Pattern.compile(".*cover.*\\.(png|jpeg|jpg)", Pattern.CASE_INSENSITIVE).matcher(escapedFilePath).matches()){
novelCover = escapedFilePath;
}
File newFile = new File(epubDirPath, escapedFilePath);
if(newFile.exists()) continue;
newFile.getParentFile().mkdirs();
FileOutputStream fos = new FileOutputStream(newFile);
int writeCount = 0;
while ((len = zis.read(buffer)) > 0) {
writeCount += 1;
fos.write(buffer, 0, len);
}
fos.close();
if(writeCount == 0){
newFile.delete();
}
}
zis.closeEntry();
zis.close();
return novelCover;
}
}
@@ -0,0 +1,58 @@
package com.rajarsheechatterjee.EpubParser;
import androidx.annotation.NonNull;
import com.facebook.react.bridge.Promise;
import com.facebook.react.bridge.ReactApplicationContext;
import com.facebook.react.bridge.ReactContextBaseJavaModule;
import com.facebook.react.bridge.ReactMethod;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.util.zip.ZipEntry;
import java.util.zip.ZipInputStream;
public class ZipArchive extends ReactContextBaseJavaModule {
ZipArchive(ReactApplicationContext context) throws Exception {
super(context);
}
@NonNull
@Override
public String getName() {
return "ZipArchive";
}
public String escapeFilePath(String filePath){
return filePath.replaceAll(":", "\uA789");
}
@ReactMethod
public void unzip(String epubFilePath, String epubDirPath, Promise promise) {
try{
ZipInputStream zis = new ZipInputStream(new FileInputStream(epubFilePath));
ZipEntry zipEntry;
int len;
byte[] buffer = new byte[4096];
while ((zipEntry = zis.getNextEntry()) != null) {
String escapedFilePath = this.escapeFilePath(zipEntry.getName());
File newFile = new File(epubDirPath, escapedFilePath);
newFile.getParentFile().mkdirs();
FileOutputStream fos = new FileOutputStream(newFile);
boolean isWritten = false; // ignore folder entry;
while ((len = zis.read(buffer)) > 0) {
isWritten = true;
fos.write(buffer, 0, len);
}
fos.close();
if(!isWritten) newFile.delete();
}
zis.closeEntry();
zis.close();
promise.resolve(null);
} catch (Exception e) {
promise.reject(e.getCause());
}
}
}
@@ -11,13 +11,13 @@ import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
public class EpubParserPackage implements ReactPackage {
public class ZipArchivePackage implements ReactPackage {
@NonNull
@Override
public List<NativeModule> createNativeModules(@NonNull ReactApplicationContext reactApplicationContext) {
List<NativeModule> modules = new ArrayList<>();
try {
modules.add(new EpubParser(reactApplicationContext));
modules.add(new ZipArchive(reactApplicationContext));
} catch (Exception e) {
throw new RuntimeException(e);
}
@@ -14,7 +14,7 @@ import com.facebook.soloader.SoLoader;
import com.rajarsheechatterjee.NavigationBarColor.NavigationBarColorPackage;
import com.rajarsheechatterjee.VolumeButtonListener.VolumeButtonListenerPackage;
import com.rajarsheechatterjee.EpubParser.EpubParserPackage;
import com.rajarsheechatterjee.EpubParser.ZipArchivePackage;
import expo.modules.ApplicationLifecycleDispatcher;
import expo.modules.ReactNativeHostWrapper;
@@ -35,7 +35,7 @@ public class MainApplication extends Application implements ReactApplication {
List<ReactPackage> packages = new PackageList(this).getPackages();
packages.add(new NavigationBarColorPackage());
packages.add(new VolumeButtonListenerPackage());
packages.add(new EpubParserPackage());
packages.add(new ZipArchivePackage());
return packages;
}