> 1) First of all, make fails right of the bat, due to the control > characters in the Makefile, so after unjarring the file > Xerces-J-src.1.0.1.jar, I wrote a perl script to recursively > go through all the files and remove the control characters ie ^M.
I looked into that and thought I was going to loose my mind for a while, but I think I finally understand what this is about. The Makefiles actually are fine within CVS, and the carriage returns only get there when extracted on a Windows box. When extracted directly to a unix box, everything is fine. I guess CVS is doing some voodoo here. So, things aren't as bad as they appear. However, the zip distribution file is currently generated on a Windows machine, so Makefiles are screwed when got from there. Hopefully, this will be fixed for the next release. In the meantime, don't delete your perl script, or check the source out from anonymous CVS. > 2) Then tried run make, after setting the environment variable > JDK12BIN and setting the classpath as per the 'Readme' file. > Files were getting compiled, till it came to this file: > xerces-1_0_1/src/org/apache/xml/serialize/OutputFormat.java > where it failed with the message: > OutputFormat.java:67: Class org.w3c.dom.html.HTMLDocument not found in > import. > >> import org.w3c.dom.html.HTMLDocument > > I searched for the HTMLDocument.java file and could not find it any > where and > also xerces-1_0_1/src/org/w3c/dom/html directory is missing. > > Please let me know where I can get the HTMLDocument.java file and what > > are the contents of the directory "xerces-1_0_1/src/org/w3c/dom/html". The release that was just posted (1.0.2) fixes that problem. -- Arnaud Le Hors - IBM Cupertino, XML Technology Group
