Skip to content

Latest commit

 

History

History
237 lines (168 loc) · 6.34 KB

NEWS.md

File metadata and controls

237 lines (168 loc) · 6.34 KB

1.9.0 2024-10-8 Has it been that long?

Revamp for a more modern Python style in the 3.12 era:

  • start adding type annotations to code.
  • use pyproject.toml
  • remove illegal escape sequences

BuildTree now builds trees iteratively rather than recursively. With this we can handle larger trees which are needed in the decompilers.

Revise README.rst for the above and go over spelling and grammar

Internally, more git branches have been created to support older Python.

1.8.9 2018-07-28 S

  • 3.8 change needed in Python 3.7.4

1.8.8 2018-07-28 Survived another year

  • Changes for Python 3.8

1.8.7 2018-05-18 Paper Tiger

  • Show LHS in error stack dump

1.8.6 2018-04-06

  • spark-parser-coverage: Canonicalize to reduce false-positive unused rules. Specifically we remove epsilon transition marks

1.8.5 2017-12-10 Dr Gecko

  • Create set of list-like and optional nonterminals Those are nonterminals used in *, +, and ? rules
  • example/python2/py2_parser.py: Pull constant out of nonterminal function
  • test/test_misc.py: Python 2.6 compatibility
  • example/gdb-loc/README.md, example/python2/README.md, example/python2/format-python.py, spark_parser/spark.py, test/test_misc.py: Create set of list-like nonterminals automatically

1.8.2-1.8.4 2017-11-27 johnnybamazing2

  • Fixes for removing a rules

  • remove_rules() improvements...

    • Allow comments in rules
    • Strip blank lines in rules
    • Remove rule from profiling

  • Document postoder/preorder better

1.8.0 2017-11-18 Cairole

  • Improve grammar checking
    • Add another grammar check in check_sets(): same non-null RHS
    • check_grammar() now returns the number of warnings issued
    • Improve output formatting in check_grammar()
  • Profiling improvements
    • Secondary key in bin/spark-parser-coverage is the rule
    • Do not profile epsilon productions, they don't go through reductions
  • Remove some Python lint warnings. (Even though Python formatting is imbecile, and for imbeciles)

spark 1.7.2 2017-11-14

  • Fix bug when used with profiling introduced in version 1.7.0
  • In python2 grammar example add check-nonterminal to template engine

spark 1.7.1 2017-10-24

  • Handle errorstack option when not set properly
  • more complete gdb-loc demo including better tests

1.7.0 2017-10-10

A number of incompatible changes, but necessary in code this old...

  • show the final token reduction tracing. This helps tracing a LOT
  • Improve ability to do reduction checks with AST (reduce_is_invalid), we now store tokens in parse object
  • xxx.type -> xxx.kind to reduce confusion with built-in type
  • some camelCase methods made "pythonic", e.g. check_grammar(), check_reduce()
  • Allow check_grammar to take a list of start symbols.
  • Add gdb-loc demo: gdb breakpoint and list expressions
  • Update docs

1.6.1 2017-05-17

  • parser.remove_rules is the same as remove_rule
  • fix parser.remove_rules so that it works even if rule has been removed

spark 1.6.0 2017-01-30

  • Add ability to track grammar coverage
  • Add ability to remove grammar rules

spark 1.5.2 2017-01-27

  • Fix bug in dumpGrammar()
  • print routines take an optional I/O parameter
  • Correct Jay Earley's name

1.5.1 2016-11-27 johnnybamazing

  • Add add limited Kleene closure symbols *, +, and optional suffix ? in grammar rules.
  • Use these to simplify Python2 grammar
  • Improve coverage in tests of Python2 grammar
  • Start a NEW-FEATURES file
  • Error if start symbol is not on LHS of grammar
  • Check for totally recursive rules

1.5.0 2016-11-26

  • User-definable reduction check. This allows one to specify an additional check before a reduction is made. For example in uncompyle6 we check whether a SETUP_LOOP target jumps to the last token of the rule. Or that an augmented assignment statement doesn't have inplace-binary operators

  • User defined reduce printing rule. For example, the uncompyle6 token class has line numbers and offsets associated with them. We can show these as part of the reduction rule.

  • Go over python 2 grammar and improve, but some indentation things are slightly worse. Sigh.

  • spark.py Fix small python2 print bug

1.4.3 2016-11-22

  • handle Python 2.3 via the 2.4 branch
  • change grammar dup debug switch from 'rules' (already used) to 'dup'

1.4.2 2016-11-22

  • Fix botched release
  • handle Python 2.5 via the 2.4 branch

1.4.1 2016-11-21

  • Eliminate duplicate grammar rules
  • Allow hook for additional reduction testing
  • Fix up setup.py: tests_require not test_requires
  • Python 2.5 tolerance
  • update example doc

1.4.0 2016-06-22

  • Add a grammar checker
  • Improve AST printing
  • More complete examples
    • More complete Python 2 example
    • expr2 example does evaluation

1.3.0 2016-06-08

  • Incompatible change: error now returns token stack and index rather than just the error token
  • add debug options 'context' to show tokens around the error token
  • More useful GenericToken class that works better with the table-driven AST routines often used.
  • Add GenericParser().dumpgrammar to show grammar rules
  • Add Python2 parser example

1.2.1 2016-05-14

  • Allow full 'errorstack': 'full' and abbreviated 'errorstack': True to show error stack with full dotted rules or just the part up to the dot
  • fix packaging

1.2.0 2016-05-10

  • Add routine to find/show error stack
  • Allow setting paring debug option on parse method
  • Add an additional and longer example parsing expressions.

1.1.2 2016-05-05

  • Add LICENSE to setup.py - this time for sure!

1.1.1 2016-05-03

  • Add LICENSE and other administrivia

1.1.0 2016-04-28

  • change module name to from spark to spark_parser to avoid conflict other python spark modules in existence

1.0.2 2016-04-27

  • Administrivia

1.0.1 2016-04-27

  • Fix metadata for Python3 upload
  • Fix module exports

1.0.0 2016-04-26

  • First Python package release: 8 years in the making