Your IP : 3.129.73.6


Current Path : /opt/hc_python/lib64/python3.8/site-packages/prometheus_client/
Upload File :
Current File : //opt/hc_python/lib64/python3.8/site-packages/prometheus_client/parser.py

import io as StringIO
import re
from typing import Dict, Iterable, List, Match, Optional, TextIO, Tuple

from .metrics_core import Metric
from .samples import Sample


def text_string_to_metric_families(text: str) -> Iterable[Metric]:
    """Parse Prometheus text format from a unicode string.

    See text_fd_to_metric_families.
    """
    yield from text_fd_to_metric_families(StringIO.StringIO(text))


ESCAPE_SEQUENCES = {
    '\\\\': '\\',
    '\\n': '\n',
    '\\"': '"',
}


def replace_escape_sequence(match: Match[str]) -> str:
    return ESCAPE_SEQUENCES[match.group(0)]


HELP_ESCAPING_RE = re.compile(r'\\[\\n]')
ESCAPING_RE = re.compile(r'\\[\\n"]')


def _replace_help_escaping(s: str) -> str:
    return HELP_ESCAPING_RE.sub(replace_escape_sequence, s)


def _replace_escaping(s: str) -> str:
    return ESCAPING_RE.sub(replace_escape_sequence, s)


def _is_character_escaped(s: str, charpos: int) -> bool:
    num_bslashes = 0
    while (charpos > num_bslashes
           and s[charpos - 1 - num_bslashes] == '\\'):
        num_bslashes += 1
    return num_bslashes % 2 == 1


def _parse_labels(labels_string: str) -> Dict[str, str]:
    labels: Dict[str, str] = {}
    # Return if we don't have valid labels
    if "=" not in labels_string:
        return labels

    escaping = False
    if "\\" in labels_string:
        escaping = True

    # Copy original labels
    sub_labels = labels_string
    try:
        # Process one label at a time
        while sub_labels:
            # The label name is before the equal
            value_start = sub_labels.index("=")
            label_name = sub_labels[:value_start]
            sub_labels = sub_labels[value_start + 1:].lstrip()
            # Find the first quote after the equal
            quote_start = sub_labels.index('"') + 1
            value_substr = sub_labels[quote_start:]

            # Find the last unescaped quote
            i = 0
            while i < len(value_substr):
                i = value_substr.index('"', i)
                if not _is_character_escaped(value_substr, i):
                    break
                i += 1

            # The label value is between the first and last quote
            quote_end = i + 1
            label_value = sub_labels[quote_start:quote_end]
            # Replace escaping if needed
            if escaping:
                label_value = _replace_escaping(label_value)
            labels[label_name.strip()] = label_value

            # Remove the processed label from the sub-slice for next iteration
            sub_labels = sub_labels[quote_end + 1:]
            next_comma = sub_labels.find(",") + 1
            sub_labels = sub_labels[next_comma:].lstrip()

        return labels

    except ValueError:
        raise ValueError("Invalid labels: %s" % labels_string)


# If we have multiple values only consider the first
def _parse_value_and_timestamp(s: str) -> Tuple[float, Optional[float]]:
    s = s.lstrip()
    separator = " "
    if separator not in s:
        separator = "\t"
    values = [value.strip() for value in s.split(separator) if value.strip()]
    if not values:
        return float(s), None
    value = float(values[0])
    timestamp = (float(values[-1]) / 1000) if len(values) > 1 else None
    return value, timestamp


def _parse_sample(text: str) -> Sample:
    # Detect the labels in the text
    try:
        label_start, label_end = text.index("{"), text.rindex("}")
        # The name is before the labels
        name = text[:label_start].strip()
        # We ignore the starting curly brace
        label = text[label_start + 1:label_end]
        # The value is after the label end (ignoring curly brace)
        value, timestamp = _parse_value_and_timestamp(text[label_end + 1:])
        return Sample(name, _parse_labels(label), value, timestamp)

    # We don't have labels
    except ValueError:
        # Detect what separator is used
        separator = " "
        if separator not in text:
            separator = "\t"
        name_end = text.index(separator)
        name = text[:name_end]
        # The value is after the name
        value, timestamp = _parse_value_and_timestamp(text[name_end:])
        return Sample(name, {}, value, timestamp)


def text_fd_to_metric_families(fd: TextIO) -> Iterable[Metric]:
    """Parse Prometheus text format from a file descriptor.

    This is a laxer parser than the main Go parser,
    so successful parsing does not imply that the parsed
    text meets the specification.

    Yields Metric's.
    """
    name = ''
    documentation = ''
    typ = 'untyped'
    samples: List[Sample] = []
    allowed_names = []

    def build_metric(name: str, documentation: str, typ: str, samples: List[Sample]) -> Metric:
        # Munge counters into OpenMetrics representation
        # used internally.
        if typ == 'counter':
            if name.endswith('_total'):
                name = name[:-6]
            else:
                new_samples = []
                for s in samples:
                    new_samples.append(Sample(s[0] + '_total', *s[1:]))
                    samples = new_samples
        metric = Metric(name, documentation, typ)
        metric.samples = samples
        return metric

    for line in fd:
        line = line.strip()

        if line.startswith('#'):
            parts = line.split(None, 3)
            if len(parts) < 2:
                continue
            if parts[1] == 'HELP':
                if parts[2] != name:
                    if name != '':
                        yield build_metric(name, documentation, typ, samples)
                    # New metric
                    name = parts[2]
                    typ = 'untyped'
                    samples = []
                    allowed_names = [parts[2]]
                if len(parts) == 4:
                    documentation = _replace_help_escaping(parts[3])
                else:
                    documentation = ''
            elif parts[1] == 'TYPE':
                if parts[2] != name:
                    if name != '':
                        yield build_metric(name, documentation, typ, samples)
                    # New metric
                    name = parts[2]
                    documentation = ''
                    samples = []
                typ = parts[3]
                allowed_names = {
                    'counter': [''],
                    'gauge': [''],
                    'summary': ['_count', '_sum', ''],
                    'histogram': ['_count', '_sum', '_bucket'],
                }.get(typ, [''])
                allowed_names = [name + n for n in allowed_names]
            else:
                # Ignore other comment tokens
                pass
        elif line == '':
            # Ignore blank lines
            pass
        else:
            sample = _parse_sample(line)
            if sample.name not in allowed_names:
                if name != '':
                    yield build_metric(name, documentation, typ, samples)
                # New metric, yield immediately as untyped singleton
                name = ''
                documentation = ''
                typ = 'untyped'
                samples = []
                allowed_names = []
                yield build_metric(sample[0], documentation, typ, [sample])
            else:
                samples.append(sample)

    if name != '':
        yield build_metric(name, documentation, typ, samples)

?>