2024-02-19 15:06:44 +00:00
|
|
|
|
package eu.faircode.email;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
This file is part of FairEmail.
|
|
|
|
|
|
|
|
|
|
FairEmail is free software: you can redistribute it and/or modify
|
|
|
|
|
it under the terms of the GNU General Public License as published by
|
|
|
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
|
|
FairEmail is distributed in the hope that it will be useful,
|
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
|
GNU General Public License for more details.
|
|
|
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
|
|
|
along with FairEmail. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
|
|
|
|
|
|
Copyright 2018-2024 by Marcel Bokhorst (M66B)
|
|
|
|
|
*/
|
|
|
|
|
|
2024-02-20 09:59:17 +00:00
|
|
|
|
import com.vladsch.flexmark.html2md.converter.FlexmarkHtmlConverter;
|
2024-02-20 11:35:01 +00:00
|
|
|
|
import com.vladsch.flexmark.util.data.DataHolder;
|
|
|
|
|
import com.vladsch.flexmark.util.data.MutableDataSet;
|
2024-02-20 09:59:17 +00:00
|
|
|
|
|
2024-02-20 07:54:20 +00:00
|
|
|
|
import org.commonmark.Extension;
|
|
|
|
|
import org.commonmark.ext.gfm.strikethrough.StrikethroughExtension;
|
|
|
|
|
import org.commonmark.ext.gfm.tables.TablesExtension;
|
2024-02-20 11:48:50 +00:00
|
|
|
|
import org.commonmark.ext.ins.InsExtension;
|
|
|
|
|
import org.commonmark.ext.task.list.items.TaskListItemsExtension;
|
2024-02-19 15:06:44 +00:00
|
|
|
|
import org.commonmark.node.Node;
|
|
|
|
|
import org.commonmark.parser.Parser;
|
|
|
|
|
import org.commonmark.renderer.html.HtmlRenderer;
|
2024-02-24 09:18:13 +00:00
|
|
|
|
import org.jsoup.nodes.Document;
|
|
|
|
|
import org.jsoup.nodes.Element;
|
|
|
|
|
import org.jsoup.select.Elements;
|
2024-02-19 15:06:44 +00:00
|
|
|
|
|
2024-02-20 07:54:20 +00:00
|
|
|
|
import java.util.Arrays;
|
2024-02-24 09:18:13 +00:00
|
|
|
|
import java.util.HashMap;
|
2024-02-20 07:54:20 +00:00
|
|
|
|
import java.util.List;
|
2024-02-24 09:18:13 +00:00
|
|
|
|
import java.util.Map;
|
2024-02-20 07:54:20 +00:00
|
|
|
|
|
2024-02-19 15:06:44 +00:00
|
|
|
|
public class Markdown {
|
|
|
|
|
static String toHtml(String markdown) {
|
2024-02-21 06:55:48 +00:00
|
|
|
|
// https://github.com/commonmark/commonmark-java#usage
|
|
|
|
|
// https://github.com/commonmark/commonmark-java/issues/294
|
2024-02-20 09:56:33 +00:00
|
|
|
|
markdown = markdown.replace('\u00a0', ' ');
|
2024-02-24 09:18:13 +00:00
|
|
|
|
|
2024-02-20 07:54:20 +00:00
|
|
|
|
List<Extension> extensions = Arrays.asList(
|
2024-02-20 11:48:50 +00:00
|
|
|
|
InsExtension.create(),
|
|
|
|
|
TaskListItemsExtension.create(),
|
2024-02-20 07:54:20 +00:00
|
|
|
|
TablesExtension.create(),
|
|
|
|
|
StrikethroughExtension.create());
|
|
|
|
|
Parser p = Parser.builder()
|
|
|
|
|
.extensions(extensions)
|
|
|
|
|
.build();
|
2024-02-19 15:06:44 +00:00
|
|
|
|
Node d = p.parse(markdown);
|
2024-02-20 07:54:20 +00:00
|
|
|
|
HtmlRenderer r = HtmlRenderer.builder()
|
|
|
|
|
.extensions(extensions)
|
|
|
|
|
.build();
|
2024-02-23 11:16:29 +00:00
|
|
|
|
String html = r.render(d);
|
|
|
|
|
if (BuildConfig.DEBUG) {
|
|
|
|
|
Log.i("Markdown md=" + markdown.replace('\n', '|'));
|
2024-02-24 09:18:13 +00:00
|
|
|
|
Log.i("Markdown html=" + html.replace('\n', '|'));
|
2024-02-23 11:16:29 +00:00
|
|
|
|
}
|
|
|
|
|
return html;
|
2024-02-19 15:06:44 +00:00
|
|
|
|
}
|
|
|
|
|
|
2024-02-20 09:56:33 +00:00
|
|
|
|
static String fromHtml(String html) {
|
2024-02-20 09:59:17 +00:00
|
|
|
|
// https://github.com/vsch/flexmark-java/wiki/Extensions#html-to-markdown
|
2024-02-24 09:18:13 +00:00
|
|
|
|
Map<String, String> specialCharsMap = new HashMap<>();
|
|
|
|
|
//specialCharsMap.put("“", "\"");
|
|
|
|
|
//specialCharsMap.put("”", "\"");
|
|
|
|
|
specialCharsMap.put("“", "\"");
|
|
|
|
|
specialCharsMap.put("”", "\"");
|
|
|
|
|
//specialCharsMap.put("‘", "'");
|
|
|
|
|
//specialCharsMap.put("’", "'");
|
|
|
|
|
specialCharsMap.put("‘", "'");
|
|
|
|
|
specialCharsMap.put("’", "'");
|
|
|
|
|
specialCharsMap.put("'", "'");
|
|
|
|
|
//specialCharsMap.put("«", "<<");
|
|
|
|
|
specialCharsMap.put("«", "<<");
|
|
|
|
|
//specialCharsMap.put("»", ">>");
|
|
|
|
|
specialCharsMap.put("»", ">>");
|
|
|
|
|
//specialCharsMap.put("…", "...");
|
|
|
|
|
specialCharsMap.put("…", "...");
|
|
|
|
|
//specialCharsMap.put("–", "--");
|
|
|
|
|
specialCharsMap.put("&endash;", "--");
|
|
|
|
|
//specialCharsMap.put("—", "---");
|
|
|
|
|
specialCharsMap.put("&emdash;", "---");
|
|
|
|
|
|
2024-02-20 11:35:01 +00:00
|
|
|
|
DataHolder options = new MutableDataSet()
|
|
|
|
|
.set(FlexmarkHtmlConverter.SETEXT_HEADINGS, false)
|
2024-02-24 09:18:13 +00:00
|
|
|
|
.set(FlexmarkHtmlConverter.OUTPUT_ATTRIBUTES_ID, false)
|
|
|
|
|
.set(FlexmarkHtmlConverter.TYPOGRAPHIC_REPLACEMENT_MAP, specialCharsMap)
|
2024-02-20 11:35:01 +00:00
|
|
|
|
.toImmutable();
|
2024-02-24 09:18:13 +00:00
|
|
|
|
|
|
|
|
|
// Remove nested/empty tables
|
|
|
|
|
Document doc = JsoupEx.parse(html);
|
|
|
|
|
for (Element table : doc.select("table")) {
|
|
|
|
|
boolean empty = false;
|
|
|
|
|
Elements children = table.children().select("table");
|
|
|
|
|
if (children.size() == 0)
|
|
|
|
|
for (Element tr : table.children()) {
|
|
|
|
|
if (tr.children().size() == 1) {
|
|
|
|
|
empty = true;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
if (empty)
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
if (children.size() > 0 || empty) {
|
|
|
|
|
table.tagName("div");
|
|
|
|
|
for (Element child : table.children())
|
|
|
|
|
if ("tr".equals(child.tagName()))
|
|
|
|
|
child.tagName("div");
|
|
|
|
|
else if ("td".equals(child.tagName()))
|
|
|
|
|
child.tagName("span");
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2024-02-20 11:35:01 +00:00
|
|
|
|
String markdown = FlexmarkHtmlConverter.builder(options)
|
2024-02-20 09:59:17 +00:00
|
|
|
|
.build()
|
2024-02-24 09:18:13 +00:00
|
|
|
|
.convert(doc.html());
|
|
|
|
|
|
2024-02-23 11:16:29 +00:00
|
|
|
|
if (BuildConfig.DEBUG) {
|
2024-02-24 09:18:13 +00:00
|
|
|
|
Log.i("Markdown html=" + html.replace('\n', '|'));
|
2024-02-23 11:16:29 +00:00
|
|
|
|
Log.i("Markdown md=" + markdown.replace('\n', '|'));
|
|
|
|
|
}
|
2024-02-24 09:18:13 +00:00
|
|
|
|
|
2024-02-22 09:22:49 +00:00
|
|
|
|
return markdown
|
|
|
|
|
.replaceAll("(?m)^( *)(\\d+)\\.( +)", "$1$2\\\\.$3")
|
2024-02-24 09:02:46 +00:00
|
|
|
|
.replaceAll("<br />", "")
|
2024-02-22 09:22:49 +00:00
|
|
|
|
.replaceAll("\n\n\\s+<!-- -->\n", "");
|
2024-02-19 15:06:44 +00:00
|
|
|
|
}
|
|
|
|
|
}
|