public class PageParserTest
- extends Object
Regression testing since eyeballing parser output just ain't cutting it.
For developers only, useful when the molly parser is hacked/changed and
we want to ensure that the new parser does not break any existing behavior.
All regression tests are in the
test subdirectory (relative to the java
source of this class). These tests exist as *.mp files. This program will
run the parser on each file and a) either check actual output with expected
output and/or b) see if an expected error occurs.
The expected/canonical output (generated by the good working parser) also
always exists in the
test subdirectory. The actual
output generated by the latest hacked parser is output in a temporary
directory (the location of which is specified on the command line). The
output are then compared and should be identical.
The expected output can be regenerated via the
argument but this should be used by caution. Future/actual output of the
changed/hacked parser is compared with the last good generated expected
output so the expected output should only be updated when the parser is
in a known working state.
public static void main(String args)