Time-Based Log Segmenter

Time is the organising principle of every log, and most troubleshooting questions are asked in time-shaped units: what happened in that hour, on that day, in that week. Time-Based Log Segmenter reads the timestamp of every line and buckets the file into hour, day, or week segments, giving you a clean chronological stack you can page through, copy, or download one piece at a time instead of scrolling through one undifferentiated wall of text.

Each line’s timestamp is extracted with a configurable regular expression (default ISO-8601, adjustable for any other format your logs use). Segments are named after the start of their window — the hour, the calendar date, or the Monday of the ISO week — and appear in time order, earliest first. Day and hour boundaries are read directly from the digits of the timestamp text itself rather than through any timezone-sensitive date parsing, so a line stamped just before midnight lands in the correct day segment regardless of what timezone your browser happens to be running in. The same care applies to week segmentation: the Monday that starts each ISO week is computed from the calendar date alone, in UTC, so late-night and early-morning lines are never nudged into the wrong week by a local time zone offset.

Lines that carry no matching timestamp — blank lines, stack-trace continuations, or anything in an unexpected format — are kept in their own “unknown” segment rather than silently discarded, so every single line from the input is accounted for somewhere in the output. You can verify this yourself: the total lines scanned and the sum of lines across every rendered segment always match.

• Segment by hour, day, or week — the unit you pick defines the windows the log is cut into.
• Configurable timestamp pattern — regex-based extraction that adapts to any timestamp format, not just ISO-8601.
• Auto-named segments — labeled by hour, calendar date, or the Monday of the ISO week.
• Chronological ordering — earliest segment first, unknown-timestamp lines listed last.
• Timezone-safe boundaries — day, hour, and week cutoffs are computed from the timestamp digits and UTC calendar math, not local-time date parsing.
• Unknown timestamps kept — lines that do not match the pattern land in their own segment instead of being dropped.
• Per-segment copy and download, plus a one-click download of the whole set.
• Summary stats — total lines, segment count, input size, and largest segment size.
• 100% local — nothing leaves the browser.

1. Paste your log into the text box.
2. Pick a segment unit — Hour, Day, or Week (Monday to Sunday).
3. Adjust the timestamp pattern if your format is not ISO-8601; the default expects YYYY-MM-DD followed by T or a space and HH:MM:SS.
4. Segment the log. Every line is scanned once and grouped into the appropriate time window.
5. Review the time-ordered segments as they appear with their labels, line counts, and byte sizes.
6. Export. Copy or download each segment individually, or download the entire set at once.

Example 1 — Pinpoint the bad hour. A slow-down starts mid-morning and clears up by lunch. Segment by hour and the two degraded hours stand out as their own self-contained blocks of diagnostics, ready to paste into an incident timeline.

Example 2 — Per-day incident splits. A three-day outage archive becomes three day-segments, one per calendar day, ready to be filed by date or attached separately to a postmortem.

Example 3 — Weekly reporting. Split a month of activity into four Monday-to-Sunday week-segments for a weekly operations report, with each segment labeled by the Monday that starts its window.

Example 4 — Midnight-boundary safety. A batch job logs lines at 23:59:58 and 00:00:02 spanning a calendar day boundary. Segmenting by day puts each line in the correct date regardless of the browser’s local timezone, because the boundary is read from the timestamp’s own date digits rather than reconstructed through local-time parsing.

• Time-shaped views — hour, day, or week segmentation on demand, matching how incidents are actually discussed.
• Clean chronological output — natural windows, auto-labeled, always in time order.
• Configurable patterns — works with any timestamp format via a regular expression, not just ISO-8601.
• Reliable boundaries — hour, day, and week cutoffs are computed consistently regardless of the viewer’s local timezone.
• Nothing dropped — unknown-timestamp and blank lines land in their own segment instead of vanishing.
• Private — all processing stays local, nothing is ever uploaded.