Implemented new metrics behaviour

README.md

@@ -35,19 +35,22 @@ Usage: dinglehopper [OPTIONS] GT OCR [REPORT_PREFIX]
   their text and falls back to plain text if no ALTO or PAGE is detected.
 
    The files GT and OCR are usually a ground truth document and the result of
-  an OCR software, but you may use dinglehopper to compare two OCR results.
-  In that case, use --no-metrics to disable the then meaningless metrics and
-  also change the color scheme from green/red to blue.
+  an OCR software, but you may use dinglehopper to compare two OCR results. In
+  that case, use --metrics='' to disable the then meaningless metrics and also
+  change the color scheme from green/red to blue.
 
   The comparison report will be written to $REPORT_PREFIX.{html,json}, where
-  $REPORT_PREFIX defaults to "report". The reports include the character
-  error rate (CER) and the word error rate (WER).
+  $REPORT_PREFIX defaults to "report". Depending on your configuration the
+  reports include the character error rate (CA|CER), the word error rate (WA|WER),
+  the bag of chars accuracy (BoC), and the bag of words accuracy (BoW).
+  The metrics can be chosen via a comma separated combination of their acronyms
+  like "--metrics=ca,wer,boc,bow".
 
   By default, the text of PAGE files is extracted on 'region' level. You may
   use "--textequiv-level line" to extract from the level of TextLine tags.
 
 Options:
-  --metrics / --no-metrics  Enable/disable metrics and green/red
+  --metrics                 Enable different metrics like ca|cer, wa|wer, boc and bow.
   --textequiv-level LEVEL   PAGE TextEquiv level to extract text from
   --progress                Show progress bar
   --help                    Show this message and exit.
@@ -80,12 +83,12 @@ The OCR-D processor has these parameters:
 
 | Parameter                 | Meaning                                                             |
 | ------------------------- | ------------------------------------------------------------------- |
-| `-P metrics false`        | Disable metrics and the green-red color scheme (default: enabled)   |
+| `-P metrics cer,wer`      | Enable character error rate and word error rate (default)           |
 | `-P textequiv_level line` | (PAGE) Extract text from TextLine level (default: TextRegion level) |
 
 For example:
 ~~~
-ocrd-dinglehopper -I ABBYY-FULLTEXT,OCR-D-OCR-CALAMARI -O OCR-D-OCR-COMPARE-ABBYY-CALAMARI -P metrics false
+ocrd-dinglehopper -I ABBYY-FULLTEXT,OCR-D-OCR-CALAMARI -O OCR-D-OCR-COMPARE-ABBYY-CALAMARI -P metrics cer,wer
 ~~~
 
 Developer information

qurator/dinglehopper/cli.py

@@ -10,9 +10,10 @@ from uniseg.graphemecluster import grapheme_clusters
 from .align import seq_align
 from .config import Config
 from .extracted_text import ExtractedText
-from .metrics.character_error_rate import character_error_rate_n
-from .metrics.word_error_rate import word_error_rate_n, words_normalized
-from .ocr_files import extract
+from .metrics import bag_of_chars_accuracy, bag_of_words_accuracy, character_accuracy, \
+    word_accuracy
+from .normalize import words_normalized
+from .ocr_files import text
 
 
 def gen_diff_report(gt_in, ocr_in, css_prefix, joiner, none):
@@ -85,9 +86,7 @@ def gen_diff_report(gt_in, ocr_in, css_prefix, joiner, none):
     )
 
 
-def generate_html_report(
-    gt, ocr, gt_text, ocr_text, report_prefix, metrics, cer, n_characters, wer, n_words
-):
+def generate_html_report(gt, ocr, gt_text, ocr_text, report_prefix, metrics_results):
     char_diff_report = gen_diff_report(
         gt_text, ocr_text, css_prefix="c", joiner="", none="·"
     )
@@ -112,57 +111,50 @@ def generate_html_report(
     template.stream(
         gt=gt,
         ocr=ocr,
-        cer=cer,
-        n_characters=n_characters,
-        wer=wer,
-        n_words=n_words,
         char_diff_report=char_diff_report,
         word_diff_report=word_diff_report,
-        metrics=metrics,
+        metrics_results=metrics_results,
     ).dump(out_fn)
 
 
-def generate_json_report(
-    gt, ocr, report_prefix, metrics, cer, n_characters, wer, n_words
-):
-    json_dict = {"gt": gt, "ocr": ocr, "n_characters": n_characters, "n_words": n_words}
-    if metrics:
-        json_dict = {**json_dict, "cer": cer, "wer": wer}
-    with open(f"{report_prefix}.json", 'w') as fp:
+def generate_json_report(gt, ocr, report_prefix, metrics_results):
+    json_dict = {"gt": gt, "ocr": ocr}
+    for result in metrics_results:
+        json_dict[result.metric] = {
+            key: value for key, value in result.get_dict().items() if key != "metric"
+        }
+    print(json_dict)
+    with open(f"{report_prefix}.json", "w") as fp:
         json.dump(json_dict, fp)
 
 
-def process(gt, ocr, report_prefix, *, metrics=True, textequiv_level="region"):
+def process(gt, ocr, report_prefix, *, metrics="cer,wer", textequiv_level="region"):
     """Check OCR result against GT.
 
     The @click decorators change the signature of the decorated functions,
     so we keep this undecorated version and use Click on a wrapper.
     """
 
-    gt_text = extract(gt, textequiv_level=textequiv_level)
-    ocr_text = extract(ocr, textequiv_level=textequiv_level)
+    gt_text = text(gt, textequiv_level=textequiv_level)
+    ocr_text = text(ocr, textequiv_level=textequiv_level)
 
-    cer, n_characters = character_error_rate_n(gt_text, ocr_text)
-    wer, n_words = word_error_rate_n(gt_text, ocr_text)
-
-    generate_json_report(
-        gt, ocr, report_prefix, metrics, cer, n_characters, wer, n_words
-    )
+    metrics_results = set()
+    if metrics:
+        metric_dict = {
+            "ca": character_accuracy,
+            "cer": character_accuracy,
+            "wa": word_accuracy,
+            "wer": word_accuracy,
+            "boc": bag_of_chars_accuracy,
+            "bow": bag_of_words_accuracy,
+        }
+        for metric in metrics.split(","):
+            metrics_results.add(metric_dict[metric.strip()](gt_text, ocr_text))
+    generate_json_report(gt, ocr, report_prefix, metrics_results)
 
     html_report = True
     if html_report:
-        generate_html_report(
-            gt,
-            ocr,
-            gt_text,
-            ocr_text,
-            report_prefix,
-            metrics,
-            cer,
-            n_characters,
-            wer,
-            n_words,
-        )
+        generate_html_report(gt, ocr, gt_text, ocr_text, report_prefix, metrics_results)
 
 
 @click.command()
@@ -170,7 +162,9 @@ def process(gt, ocr, report_prefix, *, metrics=True, textequiv_level="region"):
 @click.argument("ocr", type=click.Path(exists=True))
 @click.argument("report_prefix", type=click.Path(), default="report")
 @click.option(
-    "--metrics/--no-metrics", default=True, help="Enable/disable metrics and green/red"
+    "--metrics",
+    default="cer,wer",
+    help="Enable different metrics like cer, wer, boc and bow.",
 )
 @click.option(
     "--textequiv-level",
@@ -188,12 +182,15 @@ def main(gt, ocr, report_prefix, metrics, textequiv_level, progress):
 
     The files GT and OCR are usually a ground truth document and the result of
     an OCR software, but you may use dinglehopper to compare two OCR results. In
-    that case, use --no-metrics to disable the then meaningless metrics and also
+    that case, use --metrics='' to disable the then meaningless metrics and also
     change the color scheme from green/red to blue.
 
     The comparison report will be written to $REPORT_PREFIX.{html,json}, where
-    $REPORT_PREFIX defaults to "report". The reports include the character error
-    rate (CER) and the word error rate (WER).
+    $REPORT_PREFIX defaults to "report". Depending on your configuration the
+    reports include the character error rate (CA|CER), the word error rate (WA|WER),
+    the bag of chars accuracy (BoC), and the bag of words accuracy (BoW).
+    The metrics can be chosen via a comma separated combination of their acronyms
+    like "--metrics=ca,wer,boc,bow".
 
     By default, the text of PAGE files is extracted on 'region' level. You may
     use "--textequiv-level line" to extract from the level of TextLine tags.

qurator/dinglehopper/edit_distance.py

@@ -4,11 +4,9 @@ from functools import lru_cache, partial
 from typing import Sequence, Tuple
 
 import numpy as np
-from multimethod import multimethod
 from tqdm import tqdm
 
 from .config import Config
-from .extracted_text import ExtractedText
 from .normalize import chars_normalized
 
 
@@ -74,7 +72,6 @@ def levenshtein_matrix_cache_clear():
     _levenshtein_matrix.cache_clear()
 
 
-@multimethod
 def distance(s1: str, s2: str):
     """Compute the Levenshtein edit distance between two Unicode strings
 
@@ -86,11 +83,6 @@ def distance(s1: str, s2: str):
     return levenshtein(seq1, seq2)
 
 
-@multimethod
-def distance(s1: ExtractedText, s2: ExtractedText):
-    return distance(s1.text, s2.text)
-
-
 def seq_editops(seq1, seq2):
     """
     Return sequence of edit operations transforming one sequence to another.

qurator/dinglehopper/metrics/__init__.py

@@ -1,5 +1,5 @@
 from .bag_of_chars_accuracy import *
 from .bag_of_words_accuracy import *
-from .character_error_rate import *
-from .utils import Weights
-from .word_error_rate import *
+from .character_accuracy import *
+from .utils import MetricResult, Weights
+from .word_accuracy import *

qurator/dinglehopper/metrics/bag_of_chars_accuracy.py

@@ -7,7 +7,7 @@ from .utils import bag_accuracy, MetricResult, Weights
 
 
 def bag_of_chars_accuracy(
-    reference: str, compared: str, weights: Weights
+    reference: str, compared: str, weights: Weights = Weights(1, 0, 1)
 ) -> MetricResult:
     reference_chars = Counter(grapheme_clusters(normalize("NFC", reference)))
     compared_chars = Counter(grapheme_clusters(normalize("NFC", compared)))

qurator/dinglehopper/metrics/bag_of_words_accuracy.py

@@ -5,7 +5,7 @@ from ..normalize import words_normalized
 
 
 def bag_of_words_accuracy(
-    reference: str, compared: str, weights: Weights
+    reference: str, compared: str, weights: Weights = Weights(1, 0, 1)
 ) -> MetricResult:
     reference_words = Counter(words_normalized(reference))
     compared_words = Counter(words_normalized(compared))

qurator/dinglehopper/metrics/character_accuracy.py

@@ -0,0 +1,28 @@
+from __future__ import division
+
+from .utils import MetricResult, Weights
+from .. import distance
+from ..normalize import chars_normalized
+
+
+def character_accuracy(
+    reference: str, compared: str, weights: Weights = Weights(1, 1, 1)
+) -> MetricResult:
+    """Compute character accuracy and error rate.
+
+    :return: NamedTuple representing the results of this metric.
+    """
+
+    weighted_errors = distance(reference, compared)
+    n_ref = len(chars_normalized(reference))
+    n_cmp = len(chars_normalized(compared))
+
+    return MetricResult(
+        metric=character_accuracy.__name__,
+        weights=weights,
+        weighted_errors=int(weighted_errors),
+        reference_elements=n_ref,
+        compared_elements=n_cmp,
+    )
+
+    # XXX Should we really count newlines here?

qurator/dinglehopper/metrics/character_error_rate.py

@@ -1,46 +0,0 @@
-from __future__ import division
-
-from typing import Tuple
-
-from multimethod import multimethod
-
-from .. import distance
-from ..extracted_text import ExtractedText
-from ..normalize import chars_normalized
-
-
-@multimethod
-def character_error_rate_n(reference: str, compared: str) -> Tuple[float, int]:
-    """
-    Compute character error rate.
-
-    :return: character error rate and length of the reference
-    """
-
-    d = distance(reference, compared)
-    n = len(chars_normalized(reference))
-
-    if d == 0:
-        return 0, n
-    if n == 0:
-        return float("inf"), n
-    return d / n, n
-
-    # XXX Should we really count newlines here?
-
-
-@multimethod
-def character_error_rate_n(
-    reference: ExtractedText, compared: ExtractedText
-) -> Tuple[float, int]:
-    return character_error_rate_n(reference.text, compared.text)
-
-
-def character_error_rate(reference, compared) -> float:
-    """
-    Compute character error rate.
-
-    :return: character error rate
-    """
-    cer, _ = character_error_rate_n(reference, compared)
-    return cer

qurator/dinglehopper/metrics/utils.py

@@ -1,5 +1,5 @@
 from collections import Counter
-from typing import NamedTuple
+from typing import Dict, NamedTuple
 
 
 class Weights(NamedTuple):
@@ -25,10 +25,27 @@ class MetricResult(NamedTuple):
 
     @property
     def error_rate(self) -> float:
-        if self.reference_elements <= 0:
+        if self.reference_elements <= 0 and self.compared_elements <= 0:
+            return 0
+        elif self.reference_elements <= 0:
             return float("inf")
         return self.weighted_errors / self.reference_elements
 
+    def get_dict(self) -> Dict:
+        """Combines the properties to a dictionary.
+
+        We deviate from the builtin _asdict() function by including our properties.
+        """
+        return {
+            **{
+                key: value
+                for key, value in self._asdict().items()
+            },
+            "accuracy": self.accuracy,
+            "error_rate": self.error_rate,
+            "weights": self.weights._asdict(),
+        }
+
 
 def bag_accuracy(
     reference: Counter, compared: Counter, weights: Weights
@@ -44,7 +61,7 @@ def bag_accuracy(
     :param reference: Bag used as reference (ground truth).
     :param compared: Bag used to compare (ocr).
     :param weights: Weights/costs for editing operations.
-    :return: Tuple representing the results of this metric.
+    :return: NamedTuple representing the results of this metric.
     """
     n_ref = sum(reference.values())
     n_cmp = sum(compared.values())

qurator/dinglehopper/metrics/word_accuracy.py

@@ -0,0 +1,22 @@
+from .utils import MetricResult, Weights
+from ..edit_distance import levenshtein
+from ..normalize import words_normalized
+
+
+def word_accuracy(
+    reference: str, compared: str, weights: Weights = Weights(1, 1, 1)
+) -> MetricResult:
+    reference_seq = list(words_normalized(reference))
+    compared_seq = list(words_normalized(compared))
+
+    weighted_errors = levenshtein(reference_seq, compared_seq)
+    n_ref = len(reference_seq)
+    n_cmp = len(compared_seq)
+
+    return MetricResult(
+        metric=word_accuracy.__name__,
+        weights=weights,
+        weighted_errors=int(weighted_errors),
+        reference_elements=n_ref,
+        compared_elements=n_cmp,
+    )

qurator/dinglehopper/metrics/word_error_rate.py

@@ -1,43 +0,0 @@
-from __future__ import division
-
-from typing import Iterable, Tuple
-
-from multimethod import multimethod
-
-from ..edit_distance import levenshtein
-from ..extracted_text import ExtractedText
-from ..normalize import words_normalized
-
-
-@multimethod
-def word_error_rate_n(reference: str, compared: str) -> Tuple[float, int]:
-    reference_seq = list(words_normalized(reference))
-    compared_seq = list(words_normalized(compared))
-    return word_error_rate_n(reference_seq, compared_seq)
-
-
-@multimethod
-def word_error_rate_n(
-    reference: ExtractedText, compared: ExtractedText
-) -> Tuple[float, int]:
-    return word_error_rate_n(reference.text, compared.text)
-
-
-@multimethod
-def word_error_rate_n(reference: Iterable, compared: Iterable) -> Tuple[float, int]:
-    reference_seq = list(reference)
-    compared_seq = list(compared)
-
-    d = levenshtein(reference_seq, compared_seq)
-    n = len(reference_seq)
-
-    if d == 0:
-        return 0, n
-    if n == 0:
-        return float("inf"), n
-    return d / n, n
-
-
-def word_error_rate(reference, compared) -> float:
-    wer, _ = word_error_rate_n(reference, compared)
-    return wer

qurator/dinglehopper/ocr_files.py

@@ -163,8 +163,8 @@ def extract(filename, *, textequiv_level="region"):
         return alto_extract(tree)
 
 
-def text(filename):
-    return extract(filename).text
+def text(filename, *args, **kwargs):
+    return extract(filename, *args, **kwargs).text
 
 
 if __name__ == "__main__":

qurator/dinglehopper/ocrd-tool.json

@@ -19,9 +19,10 @@
       ],
       "parameters": {
         "metrics": {
-          "type": "boolean",
-          "default": true,
-          "description": "Enable/disable metrics and green/red"
+          "type": "string",
+          "enum": ["", "boc", "boc,bow", "bow", "ca", "ca,boc", "ca,boc,bow", "ca,bow", "ca,wa", "ca,wa,boc", "ca,wa,boc,bow", "ca,wa,bow", "ca,wer", "ca,wer,boc", "ca,wer,boc,bow", "ca,wer,bow", "cer", "cer,boc", "cer,boc,bow", "cer,bow", "cer,wa", "cer,wa,boc", "cer,wa,boc,bow", "cer,wa,bow", "cer,wer", "cer,wer,boc", "cer,wer,boc,bow", "cer,wer,bow", "wa", "wa,boc", "wa,boc,bow", "wa,bow", "wer", "wer,boc", "wer,boc,bow", "wer,bow"],
+          "default": "cer,wer",
+          "description": "Enable different metrics like ca|cer, wa|wer, boc and bow."
         },
         "textequiv_level": {
           "type": "string",

qurator/dinglehopper/templates/report.html.j2

@@ -6,7 +6,7 @@
 
     <link rel="stylesheet" href="https://stackpath.bootstrapcdn.com/bootstrap/4.3.1/css/bootstrap.min.css" integrity="sha384-ggOyR0iXCbMQv3Xipma34MD+dH/1fQ784/j6cY/iJTQUOhcWr7x9JvoRxT2MZw1T" crossorigin="anonymous">
     <style type="text/css">
-    {% if metrics %}
+    {% if metrics_results %}
     .gt .diff {
         color: green;
     }
@@ -38,10 +38,44 @@
 {{ ocr }}
 
 
-{% if metrics %}
+{% if metrics_results %}
 <h2>Metrics</h2>
-<p>CER: {{ cer|round(4) }}</p>
-<p>WER: {{ wer|round(4) }}</p>
+<table class="table">
+  <thead>
+    <tr>
+      <th scope="col" rowspan="2">#</th>
+      <th scope="col" colspan="2">Results</th>
+      <th scope="col" colspan="2">Elements</th>
+      <th scope="col" colspan="2">Calculation</th>
+    </tr>
+    <tr>
+      <th scope="col">Acc</th>
+      <th scope="col">ER</th>
+      <th scope="col">Ref</th>
+      <th scope="col">Cmp</th>
+      <th scope="col">Errors</th>
+      <th scope="col">Weights</th>
+    </tr>
+  </thead>
+  <tbody>
+  {% for result in metrics_results %}
+    <tr>
+      <th scope="row">{{ result.metric.replace("_", " ") }}</th>
+      <td>{{ result.accuracy }}</td>
+      <td>{{ result.error_rate }}</td>
+      <td>{{ result.reference_elements }}</td>
+      <td>{{ result.compared_elements }}</td>
+      <td>{{ result.weighted_errors }}</td>
+      <td>d={{ result.weights.deletes }},i={{ result.weights.inserts }},r={{ result.weights.replacements }}</td>
+    </tr>
+  {% endfor %}
+  </tbody>
+  <tfoot>
+    <tr>
+        <td colspan="7">Legend: Acc. = Accuracy, ER = Error Rate, Ref. = Reference, Cmp. = Compared</td>
+    </tr>
+  </tfoot>
+</table>
 {% endif %}
 
 <h2>Character differences</h2>

qurator/dinglehopper/tests/metrics/test_character_accuracy.py

@@ -0,0 +1,40 @@
+import math
+import unicodedata
+
+from ...metrics import character_accuracy
+
+
+def test_character_accuracy():
+    assert character_accuracy("a", "a").error_rate == 0
+    assert character_accuracy("a", "b").error_rate == 1 / 1
+    assert character_accuracy("Foo", "Bar").error_rate == 3 / 3
+
+    assert character_accuracy("Foo", "").error_rate == 3 / 3
+
+    assert character_accuracy("", "").error_rate == 0
+    assert math.isinf(character_accuracy("", "Foo").error_rate)
+
+    assert character_accuracy("Foo", "Food").error_rate == 1 / 3
+    assert character_accuracy("Fnord", "Food").error_rate == 2 / 5
+    assert character_accuracy("Müll", "Mull").error_rate == 1 / 4
+    assert character_accuracy("Abstand", "Sand").error_rate == 4 / 7
+
+
+def test_character_accuracy_hard():
+    s1 = unicodedata.normalize("NFC", "Schlyñ lorem ipsum.")
+    s2 = unicodedata.normalize("NFD", "Schlyñ lorem ipsum!")  # Different, decomposed!
+    assert character_accuracy(s1, s2).error_rate == 1 / 19
+
+    s1 = "Schlyñ"
+    assert (
+        len(s1) == 6
+    )  # This ends with LATIN SMALL LETTER N WITH TILDE, so 6 code points
+    s2 = "Schlym̃"
+    assert (
+        len(s2) == 7
+    )  # This, OTOH, ends with LATIN SMALL LETTER M + COMBINING TILDE, 7 code points
+
+    # Both strings have the same length in terms of grapheme clusters.
+    # So the CER should be symmetrical.
+    assert character_accuracy(s2, s1).error_rate == 1 / 6
+    assert character_accuracy(s1, s2).error_rate == 1 / 6

qurator/dinglehopper/tests/metrics/test_character_error_rate.py

@@ -1,41 +0,0 @@
-from __future__ import division, print_function
-
-import math
-import unicodedata
-
-from ...metrics import character_error_rate
-
-
-def test_character_error_rate():
-    assert character_error_rate("a", "a") == 0
-    assert character_error_rate("a", "b") == 1 / 1
-    assert character_error_rate("Foo", "Bar") == 3 / 3
-
-    assert character_error_rate("Foo", "") == 3 / 3
-
-    assert character_error_rate("", "") == 0
-    assert math.isinf(character_error_rate("", "Foo"))
-
-    assert character_error_rate("Foo", "Food") == 1 / 3
-    assert character_error_rate("Fnord", "Food") == 2 / 5
-    assert character_error_rate("Müll", "Mull") == 1 / 4
-    assert character_error_rate("Abstand", "Sand") == 4 / 7
-
-
-def test_character_error_rate_hard():
-    s1 = unicodedata.normalize("NFC", "Schlyñ lorem ipsum.")
-    s2 = unicodedata.normalize("NFD", "Schlyñ lorem ipsum!")  # Different, decomposed!
-    assert character_error_rate(s1, s2) == 1 / 19
-
-    s1 = "Schlyñ"
-    assert (
-        len(s1) == 6
-    )  # This ends with LATIN SMALL LETTER N WITH TILDE, so 6 code points
-    s2 = "Schlym̃"
-    assert (
-        len(s2) == 7
-    )  # This, OTOH, ends with LATIN SMALL LETTER M + COMBINING TILDE, 7 code points
-
-    # Both strings have the same length in terms of grapheme clusters. So the CER should be symmetrical.
-    assert character_error_rate(s2, s1) == 1 / 6
-    assert character_error_rate(s1, s2) == 1 / 6

qurator/dinglehopper/tests/metrics/test_integ_character_accuracy_ocr.py

@@ -7,7 +7,7 @@ from lxml import etree as ET
 from uniseg.graphemecluster import grapheme_clusters
 
 from ... import page_text, alto_text
-from ...metrics import character_error_rate
+from ...metrics import character_accuracy
 
 data_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), "../", "data")
 
@@ -22,7 +22,7 @@ def test_character_error_rate_between_page_files():
     gt_len = len(list(grapheme_clusters(gt)))
     expected_cer = 2 / gt_len
 
-    assert character_error_rate(gt, ocr) == expected_cer
+    assert character_accuracy(gt, ocr).error_rate == expected_cer
 
 
 @pytest.mark.integration
@@ -39,7 +39,7 @@ def test_character_error_rate_between_page_alto():
     )
 
     assert gt == ocr
-    assert character_error_rate(gt, ocr) == 0
+    assert character_accuracy(gt, ocr).error_rate == 0
 
 
 @pytest.mark.integration
@@ -57,4 +57,4 @@ def test_character_error_rate_between_page_alto_2():
         )
     )
 
-    assert character_error_rate(gt, ocr) == 8 / 591  # Manually verified
+    assert character_accuracy(gt, ocr).error_rate == 8 / 591  # Manually verified

qurator/dinglehopper/tests/metrics/test_integ_word_accuracy.py

@@ -6,7 +6,7 @@ import pytest
 from lxml import etree as ET
 
 from ... import alto_text, page_text
-from ...metrics import word_error_rate
+from ...metrics import word_accuracy
 from ...normalize import words
 
 data_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), "../", "data")
@@ -24,7 +24,7 @@ def test_word_error_rate_between_page_files():
     assert len(list(words(gt))) == gt_word_count
 
     ocr = page_text(ET.parse(os.path.join(data_dir, "test-fake-ocr.page2018.xml")))
-    assert word_error_rate(gt, ocr) == 2 / gt_word_count
+    assert word_accuracy(gt, ocr).error_rate == 2 / gt_word_count
 
 
 @pytest.mark.integration
@@ -41,7 +41,7 @@ def test_word_error_rate_between_page_alto():
     )
 
     assert gt == ocr
-    assert word_error_rate(gt, ocr) == 0
+    assert word_accuracy(gt, ocr).error_rate == 0
 
 
 @pytest.mark.integration
@@ -66,5 +66,5 @@ def test_word_error_rate_between_page_alto_2():
     )
 
     assert (
-        word_error_rate(gt, ocr) == 7 / gt_word_count
+        word_accuracy(gt, ocr).error_rate == 7 / gt_word_count
     )  # Manually verified, 6 words are wrong, 1 got split (=2 errors)

qurator/dinglehopper/tests/metrics/test_word_accuracy.py

@@ -1,8 +1,6 @@
-from __future__ import division, print_function
-
 import math
 
-from ...metrics import word_error_rate
+from ...metrics import word_accuracy
 from ...normalize import words
 
 
@@ -55,33 +53,44 @@ def test_words_private_use_area():
 
 def test_word_error_rate():
     assert (
-        word_error_rate("Dies ist ein Beispielsatz!", "Dies ist ein Beispielsatz!") == 0
-    )
-    assert (
-        word_error_rate("Dies. ist ein Beispielsatz!", "Dies ist ein Beispielsatz!")
+        word_accuracy(
+            "Dies ist ein Beispielsatz!", "Dies ist ein Beispielsatz!"
+        ).error_rate
         == 0
     )
     assert (
-        word_error_rate("Dies. ist ein Beispielsatz!", "Dies ist ein Beispielsatz.")
+        word_accuracy(
+            "Dies. ist ein Beispielsatz!", "Dies ist ein Beispielsatz!"
+        ).error_rate
+        == 0
+    )
+    assert (
+        word_accuracy(
+            "Dies. ist ein Beispielsatz!", "Dies ist ein Beispielsatz."
+        ).error_rate
         == 0
     )
 
     assert (
-        word_error_rate("Dies ist ein Beispielsatz!", "Dies ist ein Beispielsarz:")
+        word_accuracy(
+            "Dies ist ein Beispielsatz!", "Dies ist ein Beispielsarz:"
+        ).error_rate
         == 1 / 4
     )
     assert (
-        word_error_rate("Dies ist ein Beispielsatz!", "Dies ein ist Beispielsatz!")
+        word_accuracy(
+            "Dies ist ein Beispielsatz!", "Dies ein ist Beispielsatz!"
+        ).error_rate
         == 2 / 4
     )
 
-    assert word_error_rate("Dies ist ein Beispielsatz!", "") == 4 / 4
-    assert math.isinf(word_error_rate("", "Dies ist ein Beispielsatz!"))
-    assert word_error_rate("", "") == 0
+    assert word_accuracy("Dies ist ein Beispielsatz!", "").error_rate == 4 / 4
+    assert math.isinf(word_accuracy("", "Dies ist ein Beispielsatz!").error_rate)
+    assert word_accuracy("", "").error_rate == 0
 
     assert (
-        word_error_rate(
+        word_accuracy(
             "Schlyñ lorem ipsum dolor sit amet,", "Schlym̃ lorem ipsum dolor sit amet."
-        )
+        ).error_rate
         == 1 / 6
     )

qurator/dinglehopper/tests/test_integ_cli_valid_json.py

@@ -1,14 +1,43 @@
 import json
+from itertools import combinations
 
 import pytest
-from .util import working_directory
 
+from .util import working_directory
 from ..cli import process
 
+METRIC_DICT = {
+    "": "",
+    "ca": "character_accuracy",
+    "cer": "character_accuracy",
+    "wa": "word_accuracy",
+    "wer": "word_accuracy",
+    "boc": "bag_of_chars_accuracy",
+    "bow": "bag_of_words_accuracy",
+}
+
 
 @pytest.mark.integration
-def test_cli_json(tmp_path):
-    """Test that the cli/process() yields a loadable JSON report"""
+@pytest.mark.parametrize(
+    "metrics",
+    [
+        *(("",), ("cer",), ("wer",), ("ca",), ("wa",), ("boc",), ("bow",)),
+        *combinations(("ca", "wa", "boc", "bow"), 2),
+        *combinations(("cer", "wer", "boc", "bow"), 2),
+        *combinations(("ca", "wa", "boc", "bow"), 3),
+        *combinations(("cer", "wer", "boc", "bow"), 3),
+        ("ca", "wa", "boc", "bow"),
+        ("cer", "wer", "boc", "bow"),
+    ],
+)
+def test_cli_json(metrics, tmp_path):
+    """Test that the cli/process() yields a loadable JSON report."""
+    expected_values = {
+        "character_accuracy": 0.2,
+        "word_accuracy": 1.0,
+        "bag_of_chars_accuracy": 0.2,
+        "bag_of_words_accuracy": 1.0,
+    }
 
     with working_directory(str(tmp_path)):
         with open("gt.txt", "w") as gtf:
@@ -18,25 +47,38 @@ def test_cli_json(tmp_path):
 
         with open("gt.txt", "r") as gtf:
             print(gtf.read())
-        process("gt.txt", "ocr.txt", "report")
+        process("gt.txt", "ocr.txt", "report", metrics=",".join(metrics))
         with open("report.json", "r") as jsonf:
             print(jsonf.read())
+
         with open("report.json", "r") as jsonf:
             j = json.load(jsonf)
-            assert j["cer"] == pytest.approx(0.2)
+            metrics_translated = {METRIC_DICT[metric] for metric in metrics}
+            for metric, expected_value in expected_values.items():
+                if metric in metrics_translated:
+                    assert j[metric]["error_rate"] == pytest.approx(expected_value)
+                else:
+                    assert metric not in j.keys()
 
 
 @pytest.mark.integration
-def test_cli_json_cer_is_infinity(tmp_path):
-    """Test that the cli/process() yields a loadable JSON report when CER == inf"""
+@pytest.mark.parametrize(
+    "gt,ocr,err",
+    [("", "Not important", float("inf")), ("Lorem Ipsum", "Lorem Ipsum", 0)],
+)
+def test_cli_json_extremes(gt, ocr, err, tmp_path):
+    """Test that the cli/process() yields a loadable JSON reports."""
 
     with working_directory(str(tmp_path)):
         with open("gt.txt", "w") as gtf:
-            gtf.write("")  # Empty to yield CER == inf
+            gtf.write(gt)
         with open("ocr.txt", "w") as ocrf:
-            ocrf.write("Not important")
+            ocrf.write(ocr)
 
-        process("gt.txt", "ocr.txt", "report")
+        process("gt.txt", "ocr.txt", "report", metrics="ca,wa,boc,bow")
         with open("report.json", "r") as jsonf:
             j = json.load(jsonf)
-            assert j["cer"] == pytest.approx(float("inf"))
+            for metric in set(METRIC_DICT.values()):
+                if not metric:
+                    continue
+                assert j[metric]["error_rate"] == pytest.approx(err)

requirements.txt

@@ -7,5 +7,4 @@ colorama
 MarkupSafe
 ocrd >= 2.20.1
 attrs
-multimethod == 1.3  # latest version to officially support Python 3.5
 tqdm