226 railroad updates (#298)

* Add line separators to HowToUsePyparsing.rst to call attention to PEP-8 naming in this document

* Update railroad diagram generation code, to show results names as group annotations, and break out all expressions with a name set using setName.

* Revert dataclasses back to NamedTuples for 3.6-7 compat; add setName calls in simpleBool.py; add simpleBool to make_diagram.py

* Remove default setName calls on delimitedList

* Add setName calls to simpleSQL for better diagram

* Remove hard-coded debug mode

* Move setName on delimitedList into test code

* Restore default setName() calls for delimitedList; set default vertical=3; update jsonParser.py and simpleSQL.py with better setName() calls (and update test_diagram.py accordingly); update test_diagram.py to move asserts after tempfiles are written, moved tempfiles to local dir instead of hard-to-find temp dir

* Get proper railroad diags for infixNotation

* Undo forced railroad_debug

* Code cleanup from PR comments

* Remove hard-coded base_expr name from infix_notation

* Add special EachItem to compose DiagramItem for Group-OneOrMore-Choice; refactored tests to move duplicated code to function; added names to mozillaCalendarParser.py for better diagram

* Make sure root element gets in the diagram, even if it has no custom name

* Update tests to reflect diagram structure changes

* Add LOOKAHEAD and LOOKBEHIND annotations for FollowedBy and PrecededBy elements, and changed the annotation on Each to [ALL]; renamed _first to _element_diagram_states; add expr.streamline() in create_diagram() to collapse nested exprs; added railroad_diagram_demo.py example general blackening; update CHANGES with latest enhancements; bump version date

* Fix pip command

* Update CHANGES and whats_new_in_3_0_0.rst with some features and acknowledgements

* Updates from PR review: change user instructions to use pyparsing[diagrams]; consistent annotations for NotAny along with FollowedBy and PrecededBy; fixed up comments and type annotations

* Remove unneeded pip installs for tox (already handled in tox.ini)

* Refactor duplicate code into decorator; drop unused group_results_name argument

* Add diagram handling for SkipTo, and for And's constructed using `expr*N` notation (use a OneOrMore diagram with a repeat count instead of a sequence of N exprs)

* Fix parsing ambiguity in railroad_diagram_demo.py so that parser can actually parse a valid input string

1 files changed, 2 insertions, 2 deletions

diff --git a/examples/simpleBool.py b/examples/simpleBool.py
index 5ff1728..ac75187 100644
--- a/examples/simpleBool.py
+++ b/examples/simpleBool.py
@@ -73,7 +73,7 @@ class BoolNot:
 TRUE = Keyword("True")

 FALSE = Keyword("False")

 boolOperand = TRUE | FALSE | Word(alphas, max=1)

-boolOperand.setParseAction(BoolOperand)

+boolOperand.setParseAction(BoolOperand).setName("bool_operand")

 

 # define expression, based on expression operand and

 # list of operations in precedence order

@@ -84,7 +84,7 @@ boolExpr = infixNotation(
         ("and", 2, opAssoc.LEFT, BoolAnd),

         ("or", 2, opAssoc.LEFT, BoolOr),

     ],

-)

+).setName("boolean_expression")

 

 

 if __name__ == "__main__":