mrc_metrics.py

#   Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import division
from __future__ import absolute_import
from __future__ import print_function
from __future__ import unicode_literals


import sys
import re
import six
import logging
import math
import collections
import nltk

import unicodedata
from collections import namedtuple

RawResult = namedtuple("RawResult", ["unique_id", "start_logits", "end_logits"])

log = logging.getLogger(__name__)


def _is_whitespace(char):
    """Checks whether `chars` is a whitespace character."""
    # \t, \n, and \r are technically contorl characters but we treat them
    # as whitespace since they are generally considered as such.
    if char == " " or char == "\t" or char == "\n" or char == "\r":
        return True
    cat = unicodedata.category(char)
    if cat == "Zs":
        return True
    return False


def _is_control(char):
    """Checks whether `chars` is a control character."""
    # These are technically control characters but we count them as whitespace
    # characters.
    if char == "\t" or char == "\n" or char == "\r":
        return False
    cat = unicodedata.category(char)
    if cat.startswith("C"):
        return True
    return False


def _is_punctuation(char):
    """Checks whether `chars` is a punctuation character."""
    cp = ord(char)
    # We treat all non-letter/number ASCII as punctuation.
    # Characters such as "^", "$", and "`" are not in the Unicode
    # Punctuation class but we treat them as punctuation anyways, for
    # consistency.
    if ((cp >= 33 and cp <= 47) or (cp >= 58 and cp <= 64) or
        (cp >= 91 and cp <= 96) or (cp >= 123 and cp <= 126)):
        return True
    cat = unicodedata.category(char)
    if cat.startswith("P"):
        return True
    return False


def whitespace_tokenize(text):
    """Runs basic whitespace cleaning and splitting on a peice of text."""
    text = text.strip()
    if not text:
        return []
    tokens = text.split()
    return tokens


def convert_to_unicode(text):
    """Converts `text` to Unicode (if it's not already), assuming utf-8 input."""
    if six.PY3:
        if isinstance(text, str):
            return text
        elif isinstance(text, bytes):
            return text.decode("utf-8", "ignore")
        else:
            raise ValueError("Unsupported string type: %s" % (type(text)))
    elif six.PY2:
        if isinstance(text, str):
            return text.decode("utf-8", "ignore")
        elif isinstance(text, unicode):
            return text
        else:
            raise ValueError("Unsupported string type: %s" % (type(text)))
    else:
        raise ValueError("Not running on Python2 or Python 3?")


class _BasicTokenizer(object):
    """Runs basic tokenization (punctuation splitting, lower casing, etc.)."""

    def __init__(self, do_lower_case=True):
        """Constructs a BasicTokenizer.

        Args:
            do_lower_case: Whether to lower case the input.
        """
        self.do_lower_case = do_lower_case

    def tokenize(self, text):
        """Tokenizes a piece of text."""
        text = convert_to_unicode(text)
        text = self._clean_text(text)

        # This was added on November 1st, 2018 for the multilingual and Chinese
        # models. This is also applied to the English models now, but it doesn't
        # matter since the English models were not trained on any Chinese data
        # and generally don't have any Chinese data in them (there are Chinese
        # characters in the vocabulary because Wikipedia does have some Chinese
        # words in the English Wikipedia.).
        text = self._tokenize_chinese_chars(text)

        orig_tokens = whitespace_tokenize(text)
        split_tokens = []
        for token in orig_tokens:
            if self.do_lower_case:
                token = token.lower()
                token = self._run_strip_accents(token)
            split_tokens.extend(self._run_split_on_punc(token))

        output_tokens = whitespace_tokenize(" ".join(split_tokens))
        return output_tokens

    def _run_strip_accents(self, text):
        """Strips accents from a piece of text."""
        text = unicodedata.normalize("NFD", text)
        output = []
        for char in text:
            cat = unicodedata.category(char)
            if cat == "Mn":
                continue
            output.append(char)
        return "".join(output)

    def _run_split_on_punc(self, text):
        """Splits punctuation on a piece of text."""
        chars = list(text)
        i = 0
        start_new_word = True
        output = []
        while i < len(chars):
            char = chars[i]
            if _is_punctuation(char):
                output.append([char])
                start_new_word = True
            else:
                if start_new_word:
                    output.append([])
                start_new_word = False
                output[-1].append(char)
            i += 1

        return ["".join(x) for x in output]

    def _tokenize_chinese_chars(self, text):
        """Adds whitespace around any CJK character."""
        output = []
        for char in text:
            cp = ord(char)
            if self._is_chinese_char(cp):
                output.append(" ")
                output.append(char)
                output.append(" ")
            else:
                output.append(char)
        return "".join(output)

    def _is_chinese_char(self, cp):
        """Checks whether CP is the codepoint of a CJK character."""
        # This defines a "chinese character" as anything in the CJK Unicode block:
        #     https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_(Unicode_block)
        #
        # Note that the CJK Unicode block is NOT all Japanese and Korean characters,
        # despite its name. The modern Korean Hangul alphabet is a different block,
        # as is Japanese Hiragana and Katakana. Those alphabets are used to write
        # space-separated words, so they are not treated specially and handled
        # like the all of the other languages.
        if ((cp >= 0x4E00 and cp <= 0x9FFF) or  #
            (cp >= 0x3400 and cp <= 0x4DBF) or  #
            (cp >= 0x20000 and cp <= 0x2A6DF) or  #
            (cp >= 0x2A700 and cp <= 0x2B73F) or  #
            (cp >= 0x2B740 and cp <= 0x2B81F) or  #
            (cp >= 0x2B820 and cp <= 0x2CEAF) or
            (cp >= 0xF900 and cp <= 0xFAFF) or  #
            (cp >= 0x2F800 and cp <= 0x2FA1F)):  #
            return True

        return False

    def _clean_text(self, text):
        """Performs invalid character removal and whitespace cleanup on text."""
        output = []
        for char in text:
            cp = ord(char)
            if cp == 0 or cp == 0xfffd or _is_control(char):
                continue
            if _is_whitespace(char):
                output.append(" ")
            else:
                output.append(char)
        return "".join(output)


def _get_best_indexes(logits, n_best_size):
    """Get the n-best logits from a list."""
    index_and_score = sorted(
        enumerate(logits), key=lambda x: x[1], reverse=True)

    best_indexes = []
    for i in range(len(index_and_score)):
        if i >= n_best_size:
            break
        best_indexes.append(index_and_score[i][0])
    return best_indexes


def _compute_softmax(scores):
    """Compute softmax probability over raw logits."""
    if not scores:
        return []

    max_score = None
    for score in scores:
        if max_score is None or score > max_score:
            max_score = score

    exp_scores = []
    total_sum = 0.0
    for score in scores:
        x = math.exp(score - max_score)
        exp_scores.append(x)
        total_sum += x

    probs = []
    for score in exp_scores:
        probs.append(score / total_sum)
    return probs


def _get_final_text(pred_text, orig_text, tokenizer):
    """Project the tokenized prediction back to the original text."""

    # When we created the data, we kept track of the alignment between original
    # (whitespace tokenized) tokens and our WordPiece tokenized tokens. So
    # now `orig_text` contains the span of our original text corresponding to the
    # span that we predicted.
    #
    # However, `orig_text` may contain extra characters that we don't want in
    # our prediction.
    #
    # For example, let's say:
    #   pred_text = steve smith
    #   orig_text = Steve Smith's
    #
    # We don't want to return `orig_text` because it contains the extra "'s".
    #
    # We don't want to return `pred_text` because it's already been normalized
    # (the SQuAD eval script also does punctuation stripping/lower casing but
    # our tokenizer does additional normalization like stripping accent
    # characters).
    #
    # What we really want to return is "Steve Smith".
    #
    # Therefore, we have to apply a semi-complicated alignment heruistic between
    # `pred_text` and `orig_text` to get a character-to-charcter alignment. This
    # can fail in certain cases in which case we just return `orig_text`.

    def _strip_spaces(text):
        ns_chars = []
        ns_to_s_map = collections.OrderedDict()
        for (i, c) in enumerate(text):
            if c == " ":
                continue
            ns_to_s_map[len(ns_chars)] = i
            ns_chars.append(c)
        ns_text = "".join(ns_chars)
        return (ns_text, ns_to_s_map)

    # We first tokenize `orig_text`, strip whitespace from the result
    # and `pred_text`, and check if they are the same length. If they are
    # NOT the same length, the heuristic has failed. If they are the same
    # length, we assume the characters are one-to-one aligned.

    tok_text = " ".join(tokenizer.tokenize(orig_text))

    start_position = tok_text.find(pred_text)
    if start_position == -1:
        return orig_text
    end_position = start_position + len(pred_text) - 1

    (orig_ns_text, orig_ns_to_s_map) = _strip_spaces(orig_text)
    (tok_ns_text, tok_ns_to_s_map) = _strip_spaces(tok_text)

    if len(orig_ns_text) != len(tok_ns_text):
        return orig_text

    # We then project the characters in `pred_text` back to `orig_text` using
    # the character-to-character alignment.
    tok_s_to_ns_map = {}
    for (i, tok_index) in six.iteritems(tok_ns_to_s_map):
        tok_s_to_ns_map[tok_index] = i

    orig_start_position = None
    if start_position in tok_s_to_ns_map:
        ns_start_position = tok_s_to_ns_map[start_position]
        if ns_start_position in orig_ns_to_s_map:
            orig_start_position = orig_ns_to_s_map[ns_start_position]

    if orig_start_position is None:
        return orig_text

    orig_end_position = None
    if end_position in tok_s_to_ns_map:
        ns_end_position = tok_s_to_ns_map[end_position]
        if ns_end_position in orig_ns_to_s_map:
            orig_end_position = orig_ns_to_s_map[ns_end_position]

    if orig_end_position is None:
        return orig_text

    output_text = orig_text[orig_start_position:(orig_end_position + 1)]
    return output_text


def make_results(vocab, all_examples, all_features, all_results, n_best_size,
            max_answer_length, do_lower_case):
    """Write final predictions to the json file and log-odds of null if needed."""
    tokenizer = _BasicTokenizer(do_lower_case)
    example_index_to_features = collections.defaultdict(list)
    for feature in all_features:
        example_index_to_features[feature.example_index].append(feature)

    unique_id_to_result = {}
    for result in all_results:
        unique_id_to_result[result.unique_id] = result

    _PrelimPrediction = collections.namedtuple(  # pylint: disable=invalid-name
        "PrelimPrediction", [
            "feature_index", "start_index", "end_index", "start_logit",
            "end_logit"
        ])

    all_predictions = collections.OrderedDict()
    all_nbest_json = collections.OrderedDict()

    for (example_index, example) in enumerate(all_examples):
        features = example_index_to_features[example_index]

        prelim_predictions = []
        # keep track of the minimum score of null start+end of position 0
        for (feature_index, feature) in enumerate(features):
            result = unique_id_to_result[feature.unique_id]
            start_indexes = _get_best_indexes(result.start_logits, n_best_size)
            end_indexes = _get_best_indexes(result.end_logits, n_best_size)
            #log.debug(start_indexes)
            #log.debug(end_indexes)
            for start_index in start_indexes:
                for end_index in end_indexes:
                    # We could hypothetically create invalid predictions, e.g., predict
                    # that the start of the span is in the question. We throw out all
                    # invalid predictions.
                    if start_index >= len(feature.tokens):
                        continue
                    if end_index >= len(feature.tokens):
                        continue
                    if start_index not in feature.token_to_orig_map:
                        continue
                    if end_index not in feature.token_to_orig_map:
                        continue
                    if not feature.token_is_max_context.get(start_index, False):
                        continue
                    if end_index < start_index:
                        continue
                    length = end_index - start_index + 1
                    if length > max_answer_length:
                        continue
                    prelim_predictions.append(
                        _PrelimPrediction(
                            feature_index=feature_index,
                            start_index=start_index,
                            end_index=end_index,
                            start_logit=result.start_logits[start_index],
                            end_logit=result.end_logits[end_index]))

        prelim_predictions = sorted(
            prelim_predictions,
            key=lambda x: (x.start_logit + x.end_logit),
            reverse=True)

        _NbestPrediction = collections.namedtuple(  # pylint: disable=invalid-name
            "NbestPrediction", ["text", "start_logit", "end_logit"])

        seen_predictions = {}
        nbest = []
        for pred in prelim_predictions:
            if len(nbest) >= n_best_size:
                break
            feature = features[pred.feature_index]
            if pred.start_index > 0:  # this is a non-null prediction
                tok_tokens = feature.tokens[pred.start_index:(pred.end_index + 1
                                                              )]
                orig_doc_start = feature.token_to_orig_map[pred.start_index]
                orig_doc_end = feature.token_to_orig_map[pred.end_index]
                orig_tokens = example.doc_tokens[orig_doc_start:(orig_doc_end +
                                                                 1)]
                tok_text = " ".join(tok_tokens)

                # De-tokenize WordPieces that have been split off.
                tok_text = tok_text.replace(" ##", "")
                tok_text = tok_text.replace("##", "")

                # Clean whitespace
                tok_text = tok_text.strip()
                tok_text = " ".join(tok_text.split())
                orig_text = "".join(orig_tokens)
                final_text = _get_final_text(tok_text, orig_text, tokenizer)
                if final_text in seen_predictions:
                    continue

                seen_predictions[final_text] = True
            else:
                final_text = ""
                seen_predictions[final_text] = True

            nbest.append(
                _NbestPrediction(
                    text=final_text,
                    start_logit=pred.start_logit,
                    end_logit=pred.end_logit))

        # In very rare edge cases we could have no valid predictions. So we
        # just create a nonce prediction in this case to avoid failure.
        if not nbest:
            nbest.append(
                _NbestPrediction(
                    text="empty", start_logit=0.0, end_logit=0.0))

        total_scores = []
        best_non_null_entry = None
        for entry in nbest:
            total_scores.append(entry.start_logit + entry.end_logit)

        probs = _compute_softmax(total_scores)

        nbest_json = []
        for (i, entry) in enumerate(nbest):
            output = collections.OrderedDict()
            output["text"] = entry.text
            output["probability"] = probs[i]
            output["start_logit"] = entry.start_logit
            output["end_logit"] = entry.end_logit
            nbest_json.append(output)
        #log.debug(nbest_json[0])
        #log.debug(example.qas_id)

        assert len(nbest_json) >= 1

        all_predictions[example.qas_id] = nbest_json[0]["text"]
        all_nbest_json[example.qas_id] = nbest_json
    return all_predictions, all_nbest_json


# split Chinese with English
def mixed_segmentation(in_str, rm_punc=False):
    """mix segmentation"""
    in_str = in_str.lower().strip()
    segs_out = []
    temp_str = ""
    sp_char = ['-', ':', '_', '*', '^', '/', '\\', '~', '`', '+', '=', 
               '，', '。', '：', '？', '！', '“', '”', '；', '’', '《', '》', '……', '·', '、', 
               '「', '」', '（', '）', '－', '～', '『', '』']
    for char in in_str:
        if rm_punc and char in sp_char:
            continue
        if re.search(r'[\u4e00-\u9fa5]', char) or char in sp_char:
            if temp_str != "":
                ss = nltk.word_tokenize(temp_str)
                segs_out.extend(ss)
                temp_str = ""
            segs_out.append(char)
        else:
            temp_str += char

    #handling last part
    if temp_str != "":
        ss = nltk.word_tokenize(temp_str)
        segs_out.extend(ss)

    return segs_out


# remove punctuation
def remove_punctuation(in_str):
    """remove punctuation"""
    in_str = in_str.lower().strip()
    sp_char = ['-', ':', '_', '*', '^', '/', '\\', '~', '`', '+', '=', 
               '，', '。', '：', '？', '！', '“', '”', '；', '’', '《', '》', '……', '·', '、', 
               '「', '」', '（', '）', '－', '～', '『', '』']
    out_segs = []
    for char in in_str:
        if char in sp_char:
            continue
        else:
            out_segs.append(char)
    return ''.join(out_segs)


# find longest common string
def find_lcs(s1, s2):
    """find_lcs"""
    m = [[0 for i in range(len(s2)+1)] for j in range(len(s1)+1)]
    mmax = 0
    p = 0
    for i in range(len(s1)):
        for j in range(len(s2)):
            if s1[i] == s2[j]:
                m[i + 1][j + 1] = m[i][j] + 1
                if m[i + 1][j + 1] > mmax:
                    mmax = m[i + 1][j + 1]
                    p = i + 1
    return s1[p - mmax: p], mmax


def calc_f1_score(answers, prediction):
    """calc_f1_score"""
    f1_scores = []
    for ans in answers:
        ans_segs = mixed_segmentation(ans, rm_punc=True)
        prediction_segs = mixed_segmentation(prediction, rm_punc=True)
        lcs, lcs_len = find_lcs(ans_segs, prediction_segs)
        if lcs_len == 0:
            f1_scores.append(0)
            continue
        precision     = 1.0 * lcs_len / len(prediction_segs)
        recall         = 1.0 * lcs_len / len(ans_segs)
        f1             = (2 * precision * recall) / (precision + recall)
        f1_scores.append(f1)
    return max(f1_scores)


def calc_em_score(answers, prediction):
    """calc_f1_score"""
    em = 0
    for ans in answers:
        ans_ = remove_punctuation(ans)
        prediction_ = remove_punctuation(prediction)
        if ans_ == prediction_:
            em = 1
            break
    return em


def evaluate(ground_truth_file, prediction_file):
    """evaluate"""
    f1 = 0
    em = 0
    total_count = 0
    skip_count = 0
    for instances in ground_truth_file["data"]:
        for instance in instances["paragraphs"]:
            context_text = instance['context'].strip()
            for qas in instance['qas']:
                total_count += 1
                query_id    = qas['id'].strip()
                query_text  = qas['question'].strip()
                answers     = [ans["text"] for ans in qas["answers"]]

                if query_id not in prediction_file:
                    sys.stderr.write('Unanswered question: {}\n'.format(query_id))
                    skip_count += 1
                    continue

                prediction     = prediction_file[query_id]
                f1 += calc_f1_score(answers, prediction)
                em += calc_em_score(answers, prediction)

    f1_score = f1 / total_count
    em_score = em / total_count
    return [f1_score, em_score, total_count, skip_count]