[pull] master from DataDog:master#570
Merged
Merged
Conversation
The backslashes here were not escaping anything, instead being treated as literal backslashes that would show up in the logs.
…se (#23741) * Add CronExpression and CronScheduler utilities to datadog_checks_base CronExpression parses a standard 5-field UTC cron string and computes the next/previous scheduled tick relative to any timestamp. CronScheduler wraps an expression with a non-blocking 'has my tick elapsed?' state machine designed to be polled once per AgentCheck.check() invocation, with an optional startup_lookback for recovering a missed tick across check restarts. Supports *, n, n-m, n,m,k, */k, and n-m/k in each field, plus 7 as a Sunday alias for day-of-week. Vixie OR semantics for day-of-month vs. day-of-week when both are restricted. * Add changelog entry for cron utility * Fix N/step cron semantics, tighten error types, derive equality from parsed fields - N/step in a single-integer base now extends to the field's high (Vixie semantics): "5/15 * * * *" fires {5, 20, 35, 50} instead of {5}. - CronExpression raises TypeError for non-string input and ValueError for all malformed expressions, including parse-time rejection of DOM/month combos that can never fire ("0 0 31 2 *"). - Equality and hashing derive from the parsed fields, so "0 0 * * 7" and "0 0 * * 0" compare equal. - _jump_month iterates self._months directly to match the other _jump_* helpers; the _roll_month helper is removed. - CronScheduler.due_ticks returns list[float] matching the materialised behaviour; the typing.Iterator alias drops with it. - Rename test helpers to their fully spelled form (utc_timestamp, iso_format, BASE_TIMESTAMP); drop em-dashes and motivational framing from comments and docstrings. * Promote MAX_DAYS_PER_MONTH to module scope and tighten DOM/month guard - Move the per-month max-day table from inside _reject_unsatisfiable_dom_month to a module-level MAX_DAYS_PER_MONTH constant, alongside the other range constants. - Short-circuit _reject_unsatisfiable_dom_month when DOM is unrestricted as well: only the DOM-restricted + DOW-unrestricted case can be impossible at parse time. - Add a NEXT_PREV_CASES row that crosses a month boundary backward while Vixie OR semantics are active ("0 0 31 1 3" anchored in Feb 2026), pinning the interaction between _jump_month and _day_matches. - Drop the now-redundant list(...) wrappers around CronScheduler.due_ticks calls in the scheduler tests; due_ticks already returns list[float]. * Split _parse_field into single-purpose helpers and align backward jumps - _parse_field now returns only the sorted values tuple; the is_restricted flag was discarded by three of five call sites. CronExpression.__init__ computes _dom_restricted and _dow_restricted directly from the raw fields where the Vixie OR rule actually needs them. - _parse_field delegates to four helpers, each with a single responsibility: _expand_chunk: one comma piece, with optional /step _split_step: extracts and validates /step _parse_base: dispatches between '*', 'n-m', and 'n' _parse_range: parses an 'n-m' inclusive range Each helper raises with its own context so error messages still pinpoint the specific malformation. - _jump_hour and _jump_minute backward paths now use self._minutes[-1] (and self._hours[-1]) instead of hardcoded 59/23 when crossing a boundary, so a sparse-minute schedule does not need an extra walk iteration to recover. * Add edge-case tests adapted from croniter's test suite Covers behavior surface our parser claims to support but had no dedicated test: - Duplicate and overlapping values in comma lists normalize to the same parsed form (e.g. "5,5,1-6 * * * *" equals "1-6 * * * *"). - Vixie OR with stepped DOM and a weekday ("0 16 */2 * 6"), pinning the interaction between */k expansion in DOM and DOW match in _day_matches. - next_tick and previous_tick around stepped DOM ("5 0 */2 * *") from an even-numbered day where the anchor itself does not match. - previous_tick crossing into the previous month with a simple DOM ("0 0 22 * *" anchored mid-March -> 22 February). - Weekend DOW list ("0 0 * * 0,6") matching the most recent Sat/Sun in both directions. - Comma list in the month field ("0 0 1 1,3,6,9,12 *") rolling forward and backward across allowed months. * Reject cron expressions with whitespace inside fields A field like "1, 1, 3,3" looks at a glance like one comma list, but the spaces split it into multiple whitespace-separated fields, so the overall expression no longer has five fields. The parser already rejects this correctly via the field-count check; pinning the behaviour here covers the common ways a user might accidentally introduce intra-field whitespace (after a comma, before a comma, around a dash). * Accept whitespace around commas in cron fields A field like "1, 2, 3" is visually unambiguous: the writer means a comma list. The previous parser was strict and rejected it as multiple whitespace-separated fields. Normalise whitespace adjacent to commas before splitting into fields, so "0 0 1 1, 2, 3 *" parses the same as "0 0 1 1,2,3 *". Dashes stay strict ("1- 3" is more likely a typo than a style choice, so it is still rejected). * Reject any whitespace inside a cron field The previous commit accepted whitespace around commas but not around dashes, which is inconsistent: every kind of intra-field whitespace has only one plausible reading (the punctuation it sits next to), so "unambiguous to a human" applies the same way to "1, 2" and "1- 3". A single strict rule (no whitespace inside fields) gives the user the same clear failure for every variant and matches standard cron, croniter, and systemd timers. * Slim MAX_DAYS_PER_MONTH to non-31 months and merge tick-table tests - Drop the always-31 entries from MAX_DAYS_PER_MONTH and look them up via .get(month, 31) so the table communicates "31 is the rule, these are the exceptions." Add a short comment explaining why February uses 29. - Collapse test_next_tick and test_previous_tick into a single test_bracketing_ticks with the cases inlined into the parametrize block. Removes the top-level NEXT_PREV_CASES constant and the underscore-prefixed unused arguments, while preserving the joint (prev, next) specification at each anchor. Failure messages now name the broken direction. * Share a single parser between construction and is_valid - Extract _parse() as the one grammar both __init__ and is_valid use, so is_valid validates without instantiating a CronExpression. - Explain the scheduling sense of "due" in the due_ticks docstring. * Drop is_valid; document walk budget and CronExpression exceptions - Remove the is_valid predicate and update tests to validate through the constructor. - Add a comment explaining WALK_ITERATION_BUDGET. - Document the ValueError and TypeError CronExpression raises on bad input.
* [Release] Bumped datadog_checks_base version to 37.38.0 * [Release] Update metadata
* [Release] Bumped datadog_checks_base version to 37.39.0 * [Release] Update metadata
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )