Re: cvs commit: xml-fop/src/java/org/apache/fop/layoutmgrKnuthElement.java KnuthBox.java KnuthGlue.java KnuthPenalty.java
Glen Mazza wrote: Luca, I think we should be using getWidth() instead of getW(), correct? Yes, it would be much clearer! I'm going to rename: getW() - getWidth() getY() - getStretch() getZ() - getShrink() getP() - getPenaltyValue() The last name is quite long: I first thought of getPenalty() or getValue(), but they seemed less clear to me. Regards Luca
Re: cvs commit: xml-fop/src/java/org/apache/fop/layoutmgrKnuthElement.java KnuthBox.java KnuthGlue.java KnuthPenalty.java
Sounds good. --- Luca Furini [EMAIL PROTECTED] wrote: Glen Mazza wrote: Luca, I think we should be using getWidth() instead of getW(), correct? Yes, it would be much clearer! I'm going to rename: getW() - getWidth() getY() - getStretch() getZ() - getShrink() getP() - getPenaltyValue() The last name is quite long: I first thought of getPenalty() or getValue(), but they seemed less clear to me. Regards Luca
Marker.rebind()
Hi, I have just committed a patch which fixes bug 32253. One problem remains: the text in the marker does not always have the right properties, e.g. the right size. This is probably due to the fact that Marker.rebind() does not rebind the whole subtree below a marker. Regards, Simon -- Simon Pepping home page: http://www.leverkruid.nl
Re: Marker.rebind()
[Simon] I have just committed a patch which fixes bug 32253. One problem remains: the text in the marker does not always have the right properties, e.g. the right size. This is probably due to the fact that Marker.rebind() does not rebind the whole subtree below a marker. Your right! It seems that FOTexts isn't bind()'ed by the Marker.rebind(). I'm certain that I had that working at some point before I submitted the patch. I guess it was the change in FOText's inheritance, which happened just before I submitted, that caused the problem. The solution is to move the FObj.createPropertyList() method to FONode and to call ft.createPropertList from FObjMixed.addCharacter(). That will register the the FOText with the Marker.children map. I'll do it later this week. regards, finn
Knuth maximum demerit value.
Hi, What is the logic behind the INFINITE_DEMERITS = 1E11 ? Having a threshold of 20 can result in a demerit value of (1 + 100 * 20 ** 3) ** 2 == 6.400016E11 for the first KnuthNode which is then ignored since it is worse than the initial demerit value. I would guess that when we use a ration of 20, the initial demerit value should somehow match. regards, finn