fix: honor --time-format rfc3339 for JSON output#127
Open
digizeph wants to merge 2 commits into
Open
Conversation
Previously, --time-format rfc3339 was silently ignored when using --format json, json-line, or json-pretty — the timestamp field always remained a numeric Unix value. build_json_object() now accepts a TimestampFormat parameter and emits a formatted RFC 3339 string when Rfc3339 is selected, while Unix (default) preserves the numeric f64 output byte-for-byte via the existing native serialization fallback. Fixes #123
There was a problem hiding this comment.
Pull request overview
Fixes #123 by making JSON outputs (json, json-line, json-pretty) honor --time-format rfc3339, so the timestamp field can be emitted as an RFC3339 string instead of always being a Unix numeric value.
Changes:
- Threaded
time_formatthrough JSON formatting sotimestampbecomes number (unix) or string (rfc3339) consistently for parse/search. - Updated CLI help text and README to state
--time-formatapplies to JSON formats too. - Added unit tests around JSON timestamp formatting.
Reviewed changes
Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| src/utils.rs | Updates TimestampFormat docs to reflect behavior across JSON and non-JSON formats. |
| src/bin/commands/search.rs | Updates clap help text for --time-format to include JSON formats. |
| src/bin/commands/parse.rs | Updates clap help text for --time-format to include JSON formats. |
| src/bin/commands/elem_format.rs | Threads time_format into JSON object construction and adds JSON timestamp tests. |
| README.md | Updates documentation to reflect JSON now honors --time-format. |
| CHANGELOG.md | Adds an Unreleased bug fix entry describing the behavior change. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
297
to
305
| // If all default parse fields are selected, no collector field, and the default | ||
| // Unix timestamp format is in use, fall back to the original element serialization | ||
| // to preserve backward-compatible output byte-for-byte. | ||
| if fields == DEFAULT_FIELDS_PARSE | ||
| && !fields.contains(&"collector") | ||
| && time_format == TimestampFormat::Unix | ||
| { | ||
| return json!(elem); | ||
| } |
Comment on lines
+542
to
+546
| let obj = build_json_object(&elem, DEFAULT_FIELDS_PARSE, None, TimestampFormat::Unix); | ||
| // With default fields + Unix, build_json_object should use native elem serialization | ||
| // which contains the full element (e.g. a `peer_ip` field). | ||
| assert!(obj.get("peer_ip").is_some()); | ||
| } |
build_json_object now always uses native BgpElem serialization for the default-fields case, overriding only the timestamp key when rfc3339 is selected. Previously, rfc3339 + default fields took the field-by-field path, dropping fields like origin_asns/only_to_customer that exist in native serialization but not in DEFAULT_FIELDS_PARSE. Also fix test to assert origin_asns (not in DEFAULT_FIELDS_PARSE) instead of peer_ip (which is), and add test for rfc3339 shape preservation. Addresses PR review feedback.
Member
Author
|
Addressed both review comments in 992979b: Line 305 (native serialization shape): Line 546 (test assertion): Replaced the |
Comment on lines
+464
to
+487
| fn test_elem() -> BgpElem { | ||
| BgpElem { | ||
| timestamp: 1234567890.0, | ||
| elem_type: ElemType::ANNOUNCE, | ||
| peer_ip: IpAddr::V4(Ipv4Addr::new(192, 0, 2, 1)), | ||
| peer_asn: 65000.into(), | ||
| prefix: NetworkPrefix::from_str("10.0.0.0/8").unwrap(), | ||
| next_hop: Some(IpAddr::V4(Ipv4Addr::new(192, 0, 2, 1))), | ||
| as_path: Some(AsPath { | ||
| segments: vec![AsPathSegment::AsSequence(vec![65000.into(), 65001.into()])], | ||
| }), | ||
| origin_asns: Some(vec![65001.into()]), | ||
| origin: Some(Origin::IGP), | ||
| local_pref: Some(100), | ||
| med: Some(0), | ||
| communities: None, | ||
| atomic: false, | ||
| aggr_asn: None, | ||
| aggr_ip: None, | ||
| only_to_customer: None, | ||
| unknown: None, | ||
| deprecated: None, | ||
| } | ||
| } |
Comment on lines
+493
to
+496
| let obj = build_json_object(&elem, &fields, None, TimestampFormat::Unix); | ||
| let ts = obj.get("timestamp").unwrap(); | ||
| assert_eq!(ts.as_f64().unwrap(), 1234567890.0); | ||
| } |
Comment on lines
+502
to
+506
| let obj = build_json_object(&elem, &fields, None, TimestampFormat::Rfc3339); | ||
| let ts = obj.get("timestamp").unwrap(); | ||
| assert!(ts.is_string(), "expected string for rfc3339 timestamp"); | ||
| let s = ts.as_str().unwrap(); | ||
| assert!(s.contains("2009")); |
Comment on lines
+515
to
+523
| let unix_out = format_elem( | ||
| &elem, | ||
| OutputFormat::JsonLine, | ||
| &fields, | ||
| None, | ||
| TimestampFormat::Unix, | ||
| ) | ||
| .unwrap(); | ||
| let rfc_out = format_elem( |
Comment on lines
+523
to
+531
| let rfc_out = format_elem( | ||
| &elem, | ||
| OutputFormat::JsonLine, | ||
| &fields, | ||
| None, | ||
| TimestampFormat::Rfc3339, | ||
| ) | ||
| .unwrap(); | ||
|
|
Comment on lines
+532
to
+537
| let unix_json: serde_json::Value = serde_json::from_str(&unix_out).unwrap(); | ||
| let rfc_json: serde_json::Value = serde_json::from_str(&rfc_out).unwrap(); | ||
|
|
||
| assert!(unix_json.get("timestamp").unwrap().is_number()); | ||
| assert!(rfc_json.get("timestamp").unwrap().is_string()); | ||
| } |
Comment on lines
+558
to
+560
| // Timestamp stays numeric for Unix | ||
| assert!(obj.get("timestamp").unwrap().is_number()); | ||
| } |
Comment on lines
+571
to
+575
| // But the timestamp should be overridden to a string | ||
| let ts = obj.get("timestamp").unwrap(); | ||
| assert!(ts.is_string(), "rfc3339 timestamp should be a string"); | ||
| assert!(ts.as_str().unwrap().contains('T')); | ||
| } |
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 join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Fixes #123
--time-format rfc3339was being ignored forjson,json-line, andjson-prettyoutput — thetimestampfield always stayed a numeric Unix value even when you asked for RFC 3339.Root cause was in
build_json_object(), which had a hardcoded "JSON always uses Unix timestamp" comment and didn't thread thetime_formatparameter through. Now it does:--time-format unix(default) → numerictimestampfield, unchanged for backward compat--time-format rfc3339→ stringtimestampfield like"2024-08-23T14:15:00+00:00"Applies to
parseandsearchconsistently. Updated the clap help text and README which previously said "non-JSON output" — that's no longer accurate.Added tests covering both formats for json-line output.