udtools 0.2.7__tar.gz → 0.2.9__tar.gz

{udtools-0.2.7/src/udtools.egg-info → udtools-0.2.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: udtools
-Version: 0.2.7
+Version: 0.2.9
 Summary: Python tools for Universal Dependencies
 Author: Filip Ginter, Sampo Pyysalo, Daniel Zeman, Milan Straka, Martin Popel, Gosse Bouma, John Bauer, Arianna Masciolini, Ludovica Pannitto
 Maintainer-email: Daniel Zeman <zeman@ufal.mff.cuni.cz>
@@ -237,6 +237,20 @@ print(state)
 
 ## The official UD parsing scorer
 
+Evaluates the accuracy of a UD tokenizer / lemmatizer / tagger / parser against gold-standard data.
+The script was originally developed for the [CoNLL 2017](http://universaldependencies.org/conll17/)
+and [2018 shared tasks](http://universaldependencies.org/conll18/) in UD parsing, and later extended
+to handle the enhanced dependency representation in the [IWPT 2020](https://universaldependencies.org/iwpt20/)
+and [2021 shared tasks](https://universaldependencies.org/iwpt21/).
+
+```
+python eval.py -v goldstandard.conllu systemoutput.conllu
+```
+
+For more details on the metrics reported, see the overview papers of the shared tasks linked above.
+
+### The scorer module
+
 Reads two CoNLL-U files: gold standard (annotated manually) and system output (predicted by a parsing model). Both
 files must be valid at least at level 2, and their underlying text must be compatible, i.e., it can differ in
 whitespace but not in other characters. The scorer evaluates similarity of the system output to the gold standard

{udtools-0.2.7 → udtools-0.2.9}/README.md

@@ -217,6 +217,20 @@ print(state)
 
 ## The official UD parsing scorer
 
+Evaluates the accuracy of a UD tokenizer / lemmatizer / tagger / parser against gold-standard data.
+The script was originally developed for the [CoNLL 2017](http://universaldependencies.org/conll17/)
+and [2018 shared tasks](http://universaldependencies.org/conll18/) in UD parsing, and later extended
+to handle the enhanced dependency representation in the [IWPT 2020](https://universaldependencies.org/iwpt20/)
+and [2021 shared tasks](https://universaldependencies.org/iwpt21/).
+
+```
+python eval.py -v goldstandard.conllu systemoutput.conllu
+```
+
+For more details on the metrics reported, see the overview papers of the shared tasks linked above.
+
+### The scorer module
+
 Reads two CoNLL-U files: gold standard (annotated manually) and system output (predicted by a parsing model). Both
 files must be valid at least at level 2, and their underlying text must be compatible, i.e., it can differ in
 whitespace but not in other characters. The scorer evaluates similarity of the system output to the gold standard

{udtools-0.2.7 → udtools-0.2.9}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "udtools"
-version = "0.2.7"
+version = "0.2.9"
 authors = [
   { name="Filip Ginter" },
   { name="Sampo Pyysalo" },
@@ -22,7 +22,10 @@ maintainers = [
 description = "Python tools for Universal Dependencies"
 readme = "README.md"
 requires-python = ">=3.10"
-dependencies = ["udapi>=0.5.0", "regex>=2020.09.27"]
+dependencies = [
+  "udapi>=0.5.0",
+  "regex>=2020.09.27"
+]
 classifiers = [
   "Development Status :: 2 - Pre-Alpha",
   "Intended Audience :: Science/Research",
@@ -45,4 +48,3 @@ where = ["src"]
 
 [tool.setuptools.package-data]
 udtools = ["data/*.json"]
-

{udtools-0.2.7 → udtools-0.2.9}/src/udtools/__init__.py

@@ -4,7 +4,7 @@ udtools.
 Python tools for Universal Dependencies.
 """
 
-__version__ = "0.2.7"
+__version__ = "0.2.9"
 __author__ = 'Daniel Zeman'
 __credits__ = 'Universal Dependencies community'
 

{udtools-0.2.7 → udtools-0.2.9}/src/udtools/argparser.py

@@ -20,19 +20,16 @@ def build_argparse_validator():
                           Exit with 0 on pass, non-zero on fail.""")
     io_group.add_argument('--no-warnings',
                           dest='no_warnings', action='store_true', default=False,
-                          help="""Print only errors but no warnings.
-                          The final summary will still include the number of warnings, although they were not printed.""")
+                          help="""Print and count only errors but no warnings.""")
     io_group.add_argument('-e', '--exclude',
                           nargs='+',
-                          help="""One or more ids of tests whose incidents should not be printed.
-                          For example: --exclude missing-sent-id invalid-sent-id.
-                          The tests will still be performed and resulting incidents counted in the final summary.""")
+                          help="""One or more ids of tests whose incidents should not be printed and counted.
+                          For example: --exclude missing-sent-id invalid-sent-id.""")
     io_group.add_argument('-i', '--include-only',
                           nargs='+',
-                          help="""One or more ids of tests whose incidents should be printed.
+                          help="""One or more ids of tests whose incidents should be printed and counted.
                           For example: --include-only missing-sent-id invalid-sent-id.
-                          Any other incidents will not be printed.
-                          The tests will still be performed and resulting incidents counted in the final summary.""")
+                          Any other incidents will not be printed.""")
     io_group.add_argument('--max-err',
                           action='store', type=int, default=20,
                           help="""How many incidents to print per test class? 0 for all.

{udtools-0.2.7 → udtools-0.2.9}/src/udtools/cli.py

@@ -8,6 +8,11 @@ import sys
 from udtools.argparser import parse_args_validator, parse_args_scorer
 from udtools.validator import Validator
 from udtools.udeval import evaluate_wrapper, build_evaluation_table
+###!!!import logging
+###!!!import udtools.logging_utils as logging_utils
+
+###!!!logger = logging.getLogger(__name__)
+###!!!logging_utils.setup_logging(logger)
 
 
 
@@ -19,6 +24,7 @@ from udtools.udeval import evaluate_wrapper, build_evaluation_table
 
 def main():
     args = parse_args_validator()
+    ###!!!logger.info("Arguments: \n%s", logging_utils.pprint(vars(args)))
     validator = Validator(lang=args.lang, level=args.level, max_store=10, args=args)
     state = validator.validate_files(args.input)
     # Summarize the warnings and errors.