Question

我们使用iText + XHTMLRenderer将大型HTML文件转换为PDF。今天，它成功地占用了我们开发环境中的所有资源并使其无法使用：

This is a jstack dump:
02aaabc585000 nid=0x3af7 runnable [0x00002aaaf0269000]
   java.lang.Thread.State: RUNNABLE
    at java.awt.geom.Path2D$Double.rectCrossings(Path2D.java:1288)
    at java.awt.geom.Path2D.intersects(Path2D.java:2290)
    at java.awt.geom.Path2D.intersects(Path2D.java:2314)
    at org.xhtmlrenderer.layout.BoxCollector.intersectsAggregateBounds(BoxCollector.java:90)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:121)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:139)
    at org.xhtmlrenderer.layout.BoxCollector.collect(BoxCollector.java:46)
    at org.xhtmlrenderer.layout.Layer.paint(Layer.java:314)
    at org.xhtmlrenderer.pdf.ITextRenderer.paintPage(ITextRenderer.java:384)
    at org.xhtmlrenderer.pdf.ITextRenderer.writePDF(ITextRenderer.java:348)
    at org.xhtmlrenderer.pdf.ITextRenderer.createPDF(ITextRenderer.java:315)
    at org.xhtmlrenderer.pdf.ITextRenderer.createPDF(ITextRenderer.java:246)

这是一个组织转储：

num     instances         bytes  class name
1:       1344539      776639912  [B
2:       1798853      301253344  [C
3:        535059       72768024  org.xhtmlrenderer.render.InlineLayoutBox
4:        762761       52412032  [Ljava.lang.Object;
5:       1519522       48624704  java.lang.String
6:       1149491       45979640  com.someco.p.d
7:        203533       38674984  [I
8:        216490       31313568  &lt;constMethodKlass>
9:        216490       29455216  &lt;methodKlass>
10:        387065       24772160  org.xhtmlrenderer.render.InlineBox
11:         23732       23915216  &lt;constantPoolKlass>
12:        727350       23275200  java.awt.Rectangle
13:        243878       23095936  [Ljava.util.HashMap$Entry;
14:        147045       22350840  org.xhtmlrenderer.render.LineBox
15:        667914       21373248  java.util.HashMap$Entry
16:        855194       20524656  java.util.concurrent.LinkedBlockingQueue$Node
17:         23732       18543256  &lt;instanceKlassKlass>
18:        537890       17212480  org.xhtmlrenderer.css.style.derived.RectPropertySet
19:        688836       16532064  org.xhtmlrenderer.layout.PaintingInfo
20:        688836       16532064  java.awt.Dimension
21:        264061       15254448  &lt;symbolKlass>
22:        268028       15009568  org.xhtmlrenderer.render.InlineText

显然，这是因为转换了（非常非常）大的HTML文件。它让我们思考 - 更好的方法 - 阻止大型HTML文件被转换，或者找到一种更有效的方法将HTML转换为PDF 而不将HTML首先呈现为“屏幕”（这是XHTMLRednerer基本上做了什么。

在谷歌上搜索和阅读并没有带来好的选择。他们都是一些我们不相信的粗略公司。任何人都有另一种选择吗？

Answer 1

http://sourceforge.net/projects/xmlworker/

XMLWorker过去常常将iText作为HTMLWorker提供。它可能不适用于过于复杂的html，但可以尝试一下。

当前的解决方案当然是将pdfHTML（iText7附加组件）与iText7结合使用。

比XHTMLRenderer + iText更有效的将HTML转换为PDF的方法

1 个答案: