Update Morphological Stemming and N-Gram Indexing

Synchronization of software shared with ActiveWatch.

v2.2

Let PyElly to call external Python code as a special subprocedure. Clean up morphological stemming. Define NGRAM example application to support n-gram indexing of text content. Generalize checking for space char at the start or end of an output buffer.

Extend Integration Testing, Add Undefined Subprocedure Checking

Fixed a bug in detection of undefined sub procedures when compiling grammar rules. Made dumpEllyGrammar output easier to read. Updated documentation.

Extend Integration Test, Add Error Checking at Rule Compilation

This is an extension of v2.1.2. It corrects errors in MINIMARKING grammar rules and adds more examples for PyElly integration testing. It allows for detection of missing semantic subprocedures at the time that grammar rules are compiled. Previously, this problem was caught only at run time for rules.

Cleanup for New Example Application

This fixes a bug in loading grammar table rules with a semantic subprocedure with an undefined command. It introduces a new example application that is a major simplification of the MARKING application.

Clean Up Example Application Rules

This is an intermediate release in preparation for major work.

Bug Fixes and Cleanup

This corrects and cleans up the code, testing, and commentary for the deinflectionMatching module. It fixes a bug in dumping the arcs of the patternTable FSA and problems in unit testing with utf-8 decoding of text input. Integration testing with the MARKING example application has been updated.

Conversion of PyElly to Python 3.8 Code

This replicates the functionality of PyElly v1.6.1 in Python 3.8 code, which is incompatible with Python 2.7.5. Documentation was updated.