Fuzzy date — uncertain chronological dating

The fuzzy-date plugin adds a structured chronology section to any table. It stores uncertain, approximate, or range-based dates in five dedicated columns directly in the core table — making them searchable, filterable, and usable by other plugins such as the GeoFace temporal filter and the Harris Matrix absolute timeline.

When to use it

Use this plugin when your records have a dating that is:

• Approximate — "probably 4th century BC"
• A range — "between 350 and 300 BCE"
• One-sided — "not later than 300 BCE" (ante quem) or "not earlier than the late 4th century BCE" (post quem)
• Uncertain — "possibly Hellenistic"

Enabling the plugin for a table

Go to Config → Tables, select the table, scroll to the System plugins section and toggle Chronology (fuzzy date) on. The system immediately adds five columns to the table and confirms with a toast message. No other configuration is required.

To disable, toggle the switch off. The columns — and any data already entered — are preserved; the panel simply disappears from RecordView. Toggle it back on at any time to restore the panel without data loss.

The chronology panel

Once enabled, a Chronology section appears in RecordView below the regular fields.

Read mode

Displays the stored chronology as a formatted label (derived from the chronology string, or from chrono_from / chrono_to), a certainty badge, and an optional period name. Examples:

Late 4th cent. BCE Probable · Hellenistic

Ante quem: 4th cent. BCE Uncertain

350 BCE – 300 BCE Certain

Edit mode

The panel shows three inputs:

Input	Purpose
Chronology string	Primary input — see grammar below
Certainty	Certain / Probable / Uncertain
Period	Free text or vocabulary term (qualitative label only)

There is no separate label field — the chronology string is the label. Typing in it triggers live parsing, and a preview shows the generated label and the numeric range (from – to) so you can verify the result before saving.

Chronology string grammar

The chronology string is the primary input method. It supports two token types: century and year. The general structure is:

input = token             → single date or century range
      | token / token     → explicit range
      | ? / token         → ante quem
      | token / ?         → post quem
      | ?                 → undated

Century tokens

A century token starts with c (mandatory), followed by the century number, an optional qualifier, and the era:

c{N}{qualifier} {BCE|CE}

Qualifier	Meaning	Example	Range
(none)	Full century	c4 BCE	400–301 BCE
e	Early (~first 25%)	c4e BCE	400–376 BCE
m	Mid (~middle 50%)	c4m BCE	375–326 BCE
l	Late (~last 25%)	c4l BCE	325–301 BCE
h1	First half	c4h1 BCE	400–351 BCE
h2	Second half	c4h2 BCE	350–301 BCE
q1 – q4	Quarter 1–4	c4q3 BCE	350–326 BCE

Year tokens

Format	Example	Meaning
-N	-350	Year 350 BCE
N BCE	350 BCE	Year 350 BCE
N or N CE	300 or 300 CE	Year 300 CE

Disambiguation

4 BCE means year 4 BCE (not the 4th century). To express the 4th century BCE, write c4 BCE. The c prefix is what identifies a century token.

Range and one-sided dates

Input	Meaning	Stored as
c4l BCE / c3m CE	Late 4th BCE to mid 3rd CE	from=–325, to=275
-350 / -300	Years 350–300 BCE	from=–350, to=–300
? / c4q3 BCE	Ante quem: before 3rd quarter 4th BCE	from=null, to=–326
c4l BCE / ?	Post quem: after late 4th BCE	from=–325, to=null
?	Undated	from=null, to=null

Data model

Five columns are added to the core table when the plugin is activated:

Column	Type	Meaning
chrono_from	INTEGER	Start year (negative = BCE). NULL for ante quem or undated.
chrono_to	INTEGER	End year. NULL for post quem or undated.
chrono_label	VARCHAR(200)	The chronology string as typed (e.g. c4l BCE / c3m CE, -50/200, ?). The displayed label is computed from it — this is not a free-text field.
chrono_certainty	INTEGER	1 = certain · 2 = probable · 3 = uncertain. Legacy string values (certain / probable / possible) from older databases are still read.
chrono_period	VARCHAR(200)	Qualitative period name (e.g. "Hellenistic"). No numeric mapping.

The dating type is implicit — no separate type column is needed:

chrono_from	chrono_to	Type
null	null	Undated
X	X	Point date (exact year)
X	Y (X < Y)	Range / circa
null	Y	Ante quem
X	null	Post quem

Searching by chronology

_chrono_overlap — the primary operator

Use _chrono_overlap on chrono_from to find all records whose dating intersects a given range. It correctly handles the NULL semantics of ante quem and post quem records:

filter[chrono_from][_chrono_overlap][]=-400&filter[chrono_from][_chrono_overlap][]=-300

Or as a JSON body:

{ "chrono_from": { "_chrono_overlap": [-400, -300] } }

The operator generates a three-branch condition:

Record type	Included if
Normal range (from and to set)	range overlaps [low, high]
Ante quem (from is null)	to >= low
Post quem (to is null)	from <= high
Undated (both null)	never

WARNING

Use _chrono_overlap on chrono_from only. Applying it to any other field returns an error.

Filtering by other chrono fields

The remaining chrono columns support all standard operators:

# Records with probable certainty (1 = certain, 2 = probable, 3 = uncertain)
filter[chrono_certainty][_eq]=2

# Records in the Hellenistic period
filter[chrono_period][_icontains]=hellenistic

# Records with a label containing "BCE"
filter[chrono_label][_icontains]=BCE

# Records that have any chrono data (not undated)
filter[chrono_to][_nnull]=true

Integration with other plugins

Plugin	Integration
GeoFace	When the fuzzy-date plugin is active on a table, GeofaceView shows a Temporal filter bar with a dual-handle year slider (range −3000 to 2000, step 25 years). Dragging the handles filters map markers in real time using _chrono_overlap, correctly including ante quem and post quem records that intersect the selected window.
Harris Matrix	A Chronological toggle in the Matrix toolbar positions stratigraphic units on a vertical year axis using their chrono_from / chrono_to, keeping the stratigraphic columns so topology and calendar time are visible together. See Absolute chronological timeline.