[
https://issues.apache.org/jira/browse/BATIK-1112?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16928900#comment-16928900 ]
Samuel Penn commented on BATIK-1112:
------------------------------------
Not sure there's an easy workaround for this, since it involves exporting the image as a bitmap and importing a bitmap rather than the SVG. This results in loss of quality, and a much larger PDF. To me, this implies it's more important than a 'minor' bug. I'm using the FOP command line application in scripts, so 'simply' adding extra code isn't trivial.
> [PATCH] CSS style Parser does not handle style names with leading dash
> ----------------------------------------------------------------------
>
> Key: BATIK-1112
> URL:
https://issues.apache.org/jira/browse/BATIK-1112> Project: Batik
> Issue Type: Bug
> Components: CSS
> Affects Versions: 1.7
> Reporter: Tony BenBrahim
> Priority: Minor
> Attachments: BATIK-1112.patch
>
>
> Consider the following fragment:
> {code}
> <text
> id="text8072"
> y="15.697294"
> x="1.0898001"
> style="font-style:normal;font-variant:normal;font-weight:normal;font-stretch:normal;font-size:15.67250013px;line-height:125%;font-family:'Arial Unicode MS';-inkscape-font-specification:'Arial Unicode MS, Normal';fill:#ff0000>
> {code}
> the style Parser stops parsing with an exception when it encounters -inkscape-font-specification, which Inkscape adds to every font declaration and which cannot be removed.The rendered text is then missing any style attribute specified afterwards, such as color, anchor, alignment, etc...
> possible patch that parses styles correctly:
> {code}
> protected void parseStyleDeclaration(final boolean inSheet) throws CSSException {
> boolean leadingDash = false;
> for (;;) {
> switch (current) {
> case LexicalUnits.EOF:
> if (inSheet) {
> throw createCSSParseException("eof");
> }
> return;
> case LexicalUnits.RIGHT_CURLY_BRACE:
> if (!inSheet) {
> throw createCSSParseException("eof.expected");
> }
> nextIgnoreSpaces();
> return;
> case LexicalUnits.SEMI_COLON:
> nextIgnoreSpaces();
> continue;
> case LexicalUnits.MINUS:
> leadingDash = true;
> next();
> break;
> default:
> throw createCSSParseException("identifier");
> case LexicalUnits.IDENTIFIER:
> }
> final String name = (leadingDash ? "-" : "") + scanner.getStringValue();
> leadingDash = false;
> if (nextIgnoreSpaces() != LexicalUnits.COLON) {
> throw createCSSParseException("colon");
> }
> nextIgnoreSpaces();
> LexicalUnit exp = null;
> try {
> exp = parseExpression(false);
> } catch (final CSSParseException e) {
> reportError(e);
> }
> if (exp != null) {
> boolean important = false;
> if (current == LexicalUnits.IMPORTANT_SYMBOL) {
> important = true;
> nextIgnoreSpaces();
> }
> documentHandler.property(name, exp, important);
> }
> }
> }
> {code}
> workaround:
> Create your own parser class that extends org.apache.batik.css.parser.Parser, overriding parseStyleDeclaration with the code above, and add the following line your code
> {code}
> XMLResourceDescriptor.setCSSParserClassName(MyParser.class.getName());
> {code}
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
---------------------------------------------------------------------
To unsubscribe, e-mail:
[hidden email]
For additional commands, e-mail:
[hidden email]