# HG changeset patch
# User František Kučera <franta-hg@frantovo.cz>
# Date 1318797406 -7200
# Node ID 21f41354135721e7aa001fcaddb9fec14eae2908
# Parent b51ab80c7a9d276cf40bac071ef0c59981c9d4f1
Drupal: tidy, odstavce, XSLT
diff -r b51ab80c7a9d -r 21f413541357 src/org/sonews/storage/DrupalMessage.java
--- a/src/org/sonews/storage/DrupalMessage.java Sun Oct 16 20:55:46 2011 +0200
+++ b/src/org/sonews/storage/DrupalMessage.java Sun Oct 16 22:36:46 2011 +0200
@@ -91,57 +91,67 @@
Multipart multipart = new MimeMultipart("alternative");
setContent(multipart);
+ /** XHTML part */
+ MimeBodyPart htmlPart = new MimeBodyPart();
+ multipart.addBodyPart(htmlPart);
+ String xhtmlText = readXhtmlText(rs);
+ htmlPart.setContent(xhtmlText, XHTML_CONTENT_TYPE);
+
/** Plain text part */
MimeBodyPart textPart = new MimeBodyPart();
multipart.addBodyPart(textPart);
- textPart.setText(readPlainText(rs));
-
- /** XHTML part */
- MimeBodyPart htmlPart = new MimeBodyPart();
- multipart.addBodyPart(htmlPart);
- htmlPart.setContent(readXhtmlText(rs), XHTML_CONTENT_TYPE);
+ textPart.setText(readPlainText(rs, xhtmlText));
} else {
+ /** empty body, just headers */
setText("");
}
}
- private String readPlainText(ResultSet rs) {
+ private String readPlainText(ResultSet rs, String xhtmlText) {
/**
* TODO: převést na prostý text
*/
- return "TODO: obyčejný text";
+ return "TODO: obyčejný text\n\n\n" + xhtmlText;
}
private String readXhtmlText(ResultSet rs) {
/**
- * TODO: znovupoužívat XSL transformér
+ * TODO:
+ * - znovupoužívat XSL transformér
+ * - používat cache, ukládat si vygenerované články
*/
try {
- String originalText = rs.getString("text");
+ String inputText = "<html><body>" + rs.getString("text") + "</body></html>";
- /**
- * TODO: používat cache, ukládat si vygenerované články
- *
- *
- * Místo markdownu jen ošetřit:
- * - odstavce
- * - nesmyslné entity v odkazech
- * - neuzavřené značky: br, hr, img
- */
- String tidyTexy = tidyXhtml("<html><body>" + originalText + "</body></html>");
+ TransformerFactory tf = TransformerFactory.newInstance();
+ Transformer paragraphTransformer = tf.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart-make-paragraphs.xsl")));
+ String paragraphedText;
+ boolean tidyWasUsed = false;
+ try {
+ StringReader input = new StringReader(inputText);
+ StringWriter output = new StringWriter(2 * inputText.length());
+ paragraphTransformer.transform(new StreamSource(input), new StreamResult(output));
+ paragraphedText = output.toString();
+ } catch (Exception e) {
+ log.log(Level.FINER, "HTML input was shitty – Tidy had to be called.", e);
+ StringReader input = new StringReader(tidyXhtml(inputText));
+ StringWriter output = new StringWriter(2 * inputText.length());
+ paragraphTransformer.transform(new StreamSource(input), new StreamResult(output));
+ paragraphedText = output.toString();
+ tidyWasUsed = true;
+ }
-
- StringReader input = new StringReader(tidyTexy);
- StringWriter output = new StringWriter(2 * tidyTexy.length());
- TransformerFactory tf = TransformerFactory.newInstance();
- Transformer t = tf.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart.xsl")));
- t.setParameter("isRoot", (rs.getInt("parent_id") == 0));
- t.setParameter("title", rs.getString("subject"));
- t.setParameter("urlBase", rs.getString("urlBase"));
- t.setParameter("wwwRead", rs.getString("wwwRead"));
- t.setParameter("wwwPost", rs.getString("wwwPost"));
- t.transform(new StreamSource(input), new StreamResult(output));
+ Transformer xhtmlTransformer = tf.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart.xsl")));
+ xhtmlTransformer.setParameter("isRoot", (rs.getInt("parent_id") == 0));
+ xhtmlTransformer.setParameter("title", rs.getString("subject"));
+ xhtmlTransformer.setParameter("urlBase", rs.getString("urlBase"));
+ xhtmlTransformer.setParameter("wwwRead", rs.getString("wwwRead"));
+ xhtmlTransformer.setParameter("wwwPost", rs.getString("wwwPost"));
+ xhtmlTransformer.setParameter("headComment", String.format("Drupal-NNTP bridge. Transformed: %1$tc. Tidy had to be used: %2$b", new Date(), tidyWasUsed));
+ StringReader input = new StringReader(paragraphedText);
+ StringWriter output = new StringWriter(2 * paragraphedText.length());
+ xhtmlTransformer.transform(new StreamSource(input), new StreamResult(output));
return output.toString();
} catch (Exception e) {
@@ -157,15 +167,21 @@
* TODO: refaktorovat, přesunout
*/
private static String tidyXhtml(String inputText) throws IOException {
+ // https://sourceforge.net/tracker/index.php?func=detail&aid=3424437&group_id=27659&atid=390966
+ inputText = inputText.replaceAll("\\n", "◆\n");
+
Runtime r = Runtime.getRuntime();
- Process p = r.exec(new String[]{"tidy",
- "-asxml",
- "-numeric",
- "-utf8",
- "-quiet",
- "--doctype", "omit",
- "--logical-emphasis", "true",
- "--show-errors", "0"});
+ Process p = r.exec(new String[]{"tidy", // http://tidy.sourceforge.net
+ "-asxml", // well formed XHTML
+ "-numeric", // číselné entity
+ "-utf8", // kódování
+ "--show-warnings", "false", // žádná varování nás nezajímají
+ "--show-errors", "0", // ani chyby
+ "--doctype", "omit", // doctype nepotřebujeme (doplníme si případně vlastní v XSLT)
+ "--logical-emphasis", "true", // em a strong místo i a b
+ "--literal-attributes", "true", // zachovat mezery a konce řádků v atributech
+ "--force-output", "true" // neznámé značky zahodíme, vložíme jen jejich obsah
+ });
PrintStream vstupProcesu = new PrintStream(p.getOutputStream());
vstupProcesu.print(inputText);
@@ -173,6 +189,9 @@
String outputText = streamToString(p.getInputStream());
+ outputText = outputText.replaceAll("◆\\n", "\n");
+ outputText = outputText.replaceAll("◆", "\n");
+
return outputText;
}