• Defined pyparsing-specific warning classes so that they can be selectively enabled or disabled without affecting warnings raised by other libraries in the same Python app:

  • PyparsingWarning - base warning for all pyparsing-specific warnings (inherits from UserWarning)
  • PyparsingDeprecationWarning - warning for using deprecated features (inherits from PyparsingWarning and DeprecationWarning)
  • PyparsingDiagnosticWarning - warning raised when pyparsing diagnostics are enabled and a diagnostic feature is used (inherits from PyparsingWarning)

• Added as_datetime parse action to pyparsing.common - a more generalized version of the convert_to_datetime parse action (supports any expression that extracts date/time fields into "year", "month", "day", etc. results names), and validates that the parsed fields represent a valid date and time.

• Added iso8601_date_validated and iso8601_datetime_validated expressions to pyparsing.common, which return a Python datetime.datetime

• Various performance improvements in ParseResults class and core functions, with 10-20% performance overall.

• Added regex_inverter web page (using PyScript) to demonstrate using the inv_regex.py example.

• Expanded regex forms handled by the examples/inv_regex.py example:

  • named capturing groups (?P<name>)
  • partial repetition ({m,} and {,n})
  • negated character classes ([^...])

• Added SPy (Simplified Python) parser to examples.

• Added license info to metadata, following PEP-639. Thanks to Gedalia Pasternak and Marc Mueller for submitted issue and PR. Fixes #626.

=======================================================================================
The version 3.3.0 release will begin emitting DeprecationWarnings for pyparsing methods that have been renamed to PEP8-compliant names (introduced in pyparsing 3.0.0, in August, 2021, with legacy names retained as aliases). In preparation, I added in pyparsing 3.2.2 a utility for finding and replacing the legacy method names with the new names. This utility is located at pyparsing/tools/cvt_pep8_names.py. This script will scan all Python files specified on the command line, and if the -u option is selected, will replace all occurrences of the old method names with the new PEP8-compliant names, updating the files in place.

Here is an example that converts all the files in the pyparsing /examples directory:

  python -m pyparsing.tools.cvt_pyparsing_pep8_names -u examples/*.py

The new names are compatible with pyparsing versions 3.0.0 and later.

=======================================================================================

• Deprecated indentedBlock, when converted using the cvt_pyparsing_pep8_names utility, will emit UserWarnings that additional code changes will be required. This is because the new IndentedBlock class no longer requires the calling code to supply an indent stack, while adding support for nested indentation levels and grouping.

• Deprecated locatedExpr, when converted using the cvt_pyparsing_pep8_names utility, will emit UserWarnings that additional code changes may be required. The new Located class removes the extra grouping level of the parsed values. (If the original locatedExpr parser was defined with a results name, then the extra grouping is retained, so that the results name nesting works properly; in this case, no code changes would be required.)

• Updated all examples and test cases to use PEP8 names (unless the test case is specifically designed to test behavior of a legacy method). Added railroad diagrams for some examples.

• Added exception handling when calling formatted_message(), so that str(exception) always returns at least something.

• All unit tests pass with Python 3.14, including 3.14t. This does not necessarily mean that pyparsing is now thread-safe, just that when run in the free-threaded interpreter, there were no errors. None of the unit tests try to do any parsing with multiple threads - they test the basic functionality of the library, under various versions of packrat and left-recursive parsing.

• Added AI instructions so that AI agents can be prompted with best practices for generating parsers using pyparsing code. These instructions are in the ai/best_practices.md file, and can be accessed programmatically by calling pyparsing.show_best_practices() or running python -m pyparsing.ai.show_best_practices from the command line, after installing the pyparsing package.

• Implemented a TINY language parser/interpreter using pyparsing, in the examples/tiny directory. This is a little tutorial language that I used to demonstrate how to use pyparsing to build a simple interpreter, following a recommended parser+AST+engine+run structure. The docs sub-directory also includes transcripts of the AI session used to create the parser and the interpreter. The samples sub-directory includes a few sample TINY programs.

• Fixed minor formatting bugs in pyparsing.testing.with_line_numbers, found during development of the TINY language example.

• Added test in DelimitedList and nested_expr which auto-suppress delimiting commas to avoid wrapping in a Suppress if delimiter is already a Suppress.

• Added performance benchmarking tools and documentation:

  • tests/perf_pyparsing.py runs a series of benchmark parsing tests, exercising different aspects of the pyparsing package. For cross-version analysis, this script can export results as CSV and append to a consolidated data file.
  • Runner scripts run_perf_all_tags.bat (Windows) and run_perf_all_tags.sh (Ubuntu/bash) execute the benchmark across multiple Python versions (3.9–3.14) and pyparsing versions (3.1.1 through 3.3.0), aggregating results into perf_pyparsing.csv at the repo root.
  • See tests/README.md for usage instructions.

• Used performance benchmarking to identify and revert an inefficient utility method used in transform_string (introduced in pyparsing 3.2.0b2).

(added in 3.3.0b1)

• Implemented a TINY language parser/interpreter using pyparsing, in the examples/tiny directory. This is a little tutorial language that I used to demonstrate how to use pyparsing to build a simple interpreter, following a recommended parser+AST+engine+run structure. The docs sub-directory also includes transcripts of the AI session used to create the parser and the interpreter.

• Fixed minor formatting bugs in pyparsing.testing.with_line_numbers, found during development of the TINY language example.

• Added test in DelimitedList and nested_expr which auto-suppress delimiting commas to avoid wrapping in a Suppress if delimiter is already a Suppress.

• Added performance benchmarking tools and documentation:

  • tests/perf_pyparsing.py runs a series of benchmark parsing tests, exercising different aspects of the pyparsing package. For cross-version analysis, this script can export results as CSV and append to a consolidated data file.
  • Runner scripts run_perf_all_tags.bat (Windows) and run_perf_all_tags.sh (Ubuntu/bash) execute the benchmark across multiple Python versions (3.9–3.14) and pyparsing versions (3.1.1 through 3.3.0), aggregating results into perf_pyparsing.csv at the repo root.
  • See tests/README.md for usage instructions.

• Supports Python 3.14 and free-threaded 3.14t. (Note: this does not mean that pyparsing is thread-safe, simply that pyparsing can safely run using the free-threaded build of Python.)
• The version 3.3.0 release will begin emitting DeprecationWarnings for pyparsing methods that have been renamed to PEP8-compliant names (introduced in pyparsing 3.0.0, in August, 2021, with legacy names retained as aliases).
• AI instructions have been included for AI guidance when generating parsers using pyparsing. They can be accessed using python -m pyparsing.ai.show_best_practices.
• The cvt_pyparsing_pep8_names utility (released in pyparsing 3.2.2) will raise UserWarnings when converting locatedExpr and indentedBlock calls, since the new methods have slightly different call signatures and return types.
• Source code examples have all been upgraded to use new PEP8-compliant names.

Barring any catastrophic bugs in this release, this will be the last release in the 3.2.x line. The next release, 3.3.0, will begin emitting DeprecationWarnings when the pre-PEP8 methods are used (see header notes in CHANGES file for more information, including available automation for converting any existing code using pyparsing using the old names).

• JINX! Well, 3.2.4 had a bug for Word expressions that include a space character, if that expression was then copied, either directly with .copy() or by adding a results name, or including in another construct (like DelimitedList) that makes a copy internally. Issue #618, reported by mstinberg, among others - thanks, and sorry for the inconvenience.

Full Changelog: 3.2.4...3.2.5

Barring any catastrophic bugs in this release, this will be the last release in the 3.2.x line. The next release, 3.3.0, will begin emitting DeprecationWarnings when the pre-PEP8 methods are used (see header notes above for more information, including available automation for converting any existing code using pyparsing with the old names).

• Fixed bug when using a copy of a Word expression (either by using the explicit copy() method, or attaching a results name), and setting a new expression name, a raised ParseException still used the original expression name. Also affected Regex expressions with as_match or as_group_list = True. Reported by Waqas Ilyas, in Issue #612 - good catch!

• Fixed type annotation for replace_with, to accept Any type. Fixes Issue #602, reported by esquonk.

• Added locking around potential race condition in ParserElement.reset_cache, as well as other cache-related methods. Fixes Issue #604, reported by CarlosDescalziIM.

• Substantial update to docstrings and doc generation in preparation for 3.3.0, great effort by FeRD, thanks!

• Notable addition by FeRD to convert docstring examples to work with doctest! This was long overdue, thanks so much!

• Fixed bug released in 3.2.2 in which nested_expr could overwrite parse actions for defined content, and could truncate list of items within a nested list. Fixes Issue #600, reported by hoxbro and luisglft, with helpful diag logs and repro code.

The upcoming version 3.3.0 release will begin emitting DeprecationWarnings for pyparsing methods that have been renamed to PEP8-compliant names (introduced in pyparsing 3.0.0, in August, 2021, with legacy names retained as aliases). In preparation, I have added in pyparsing 3.2.2 a utility for finding and replacing the legacy method names with the new names. This utility is located at pyparsing/tools/cvt_pep8_names.py. This script will scan all Python files specified on the command line, and if the -u option is selected, will replace all occurrences of the old method names with the new PEP8-compliant names, updating the files in place.

Here is an example that converts all the files in the pyparsing /examples directory:

  python -m pyparsing.tools.cvt_pyparsing_pep8_names -u examples/*.py

The new names are compatible with pyparsing versions 3.0.0 and later.

• Released cvt_pyparsing_pep8_names.py conversion utility to upgrade pyparsing-based programs and libraries that use legacy camelCase names to use the new PEP8-compliant snake_case method names. The converter can also be imported into other scripts as

    from pyparsing.tools.cvt_pyparsing_pep8_names import pep8_converter
  

• Fixed bug in nested_expr where nested contents were stripped of whitespace when the default whitespace characters were cleared (raised in this StackOverflow question https://stackoverflow.com/questions/79327649 by Ben Alan). Also addressed bug in resolving PEP8 compliant argument name and legacy argument name.

• Fixed bug in rest_of_line and the underlying Regex class, in which matching a pattern that could match an empty string (such as ".*" or "[A-Z]*" would not raise a ParseException at or beyond the end of the input string. This could cause an infinite parsing loop when parsing rest_of_line at the end of the input string. Reported by user Kylotan, thanks! (Issue #593)

• Enhancements and extra input validation for pyparsing.util.make_compressed_re - see usage in examples/complex_chemical_formulas.py and result in the generated railroad diagram examples/complex_chemical_formulas_diagram.html. Properly escapes characters like "." and "*" that have special meaning in regular expressions.

• Fixed bug in one_of() to properly escape characters that are regular expression markers (such as '*', '+', '?', etc.) before building the internal regex.

• Better exception message for MatchFirst and Or expressions, showing all alternatives rather than just the first one. Fixes Issue #592, reported by Focke, thanks!

• Added return type annotation of "-> None" for all __init__() methods, to satisfy mypy --strict type checking. PR submitted by FeRD, thank you!

• Added optional argument show_hidden to create_diagram to show elements that are used internally by pyparsing, but are not part of the actual parser grammar. For instance, the Tag class can insert values into the parsed results but it does not actually parse any input, so by default it is not included in a railroad diagram. By calling create_diagram with show_hidden = True, these internal elements will be included. (You can see this in the tag_metadata.py script in the examples directory.)

• Fixed bug in number_words.py example. Also added ebnf_number_words.py to demonstrate using the ebnf.py EBNF parser generator to build a similar parser directly from EBNF.

• Fixed syntax warning raised in bigquery_view_parser.py, invalid escape sequence "\s". Reported by sameer-google, nice catch! (Issue #598)

• Added support for Python 3.14.

• Updated generated railroad diagrams to make non-terminal elements links to their related sub-diagrams. This greatly improves navigation of the diagram, especially for large, complex parsers.

• Simplified railroad diagrams emitted for parsers using infix_notation, by hiding lookahead terms. Renamed internally generated expressions for clarity, and improved diagramming.

• Improved performance of cpp_style_comment, c_style_comment, common.fnumber and common.ieee_float Regex expressions. PRs submitted by Gabriel Gerlero,
  nice work, thanks!

• Add missing type annotations to match_only_at_col, replace_with, remove_quotes, with_attribute, and with_class. Issue #585 reported by rafrafrek.

• Added generated diagrams for many of the examples.

• Replaced old examples/0README.html file with examples/README.md file.