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.
(fd: TextIO)
| 273 | |
| 274 | |
| 275 | def text_fd_to_metric_families(fd: TextIO) -> Iterable[Metric]: |
| 276 | """Parse Prometheus text format from a file descriptor. |
| 277 | |
| 278 | This is a laxer parser than the main Go parser, |
| 279 | so successful parsing does not imply that the parsed |
| 280 | text meets the specification. |
| 281 | |
| 282 | Yields Metric's. |
| 283 | """ |
| 284 | name = '' |
| 285 | documentation = '' |
| 286 | typ = 'untyped' |
| 287 | samples: List[Sample] = [] |
| 288 | allowed_names = [] |
| 289 | |
| 290 | def build_metric(name: str, documentation: str, typ: str, samples: List[Sample]) -> Metric: |
| 291 | # Munge counters into OpenMetrics representation |
| 292 | # used internally. |
| 293 | if typ == 'counter': |
| 294 | if name.endswith('_total'): |
| 295 | name = name[:-6] |
| 296 | else: |
| 297 | new_samples = [] |
| 298 | for s in samples: |
| 299 | new_samples.append(Sample(s[0] + '_total', *s[1:])) |
| 300 | samples = new_samples |
| 301 | metric = Metric(name, documentation, typ) |
| 302 | metric.samples = samples |
| 303 | return metric |
| 304 | |
| 305 | for line in fd: |
| 306 | line = line.strip() |
| 307 | |
| 308 | if line.startswith('#'): |
| 309 | parts = _split_quoted(line, None, 3) |
| 310 | if len(parts) < 2: |
| 311 | continue |
| 312 | candidate_name, quoted = '', False |
| 313 | if len(parts) > 2: |
| 314 | # Ignore comment tokens |
| 315 | if parts[1] != 'TYPE' and parts[1] != 'HELP': |
| 316 | continue |
| 317 | candidate_name, quoted = _unquote_unescape(parts[2]) |
| 318 | if not quoted and not _is_valid_legacy_metric_name(candidate_name): |
| 319 | raise ValueError |
| 320 | if parts[1] == 'HELP': |
| 321 | if candidate_name != name: |
| 322 | if name != '': |
| 323 | yield build_metric(name, documentation, typ, samples) |
| 324 | # New metric |
| 325 | name = candidate_name |
| 326 | typ = 'untyped' |
| 327 | samples = [] |
| 328 | allowed_names = [candidate_name] |
| 329 | if len(parts) == 4: |
| 330 | documentation = _replace_help_escaping(parts[3]) |
| 331 | else: |
| 332 | documentation = '' |
no test coverage detected