How Int./Ext. Is Guessed from Slugline Delimiters

When SlateX processes a slugline, it reads Int./Ext. and time of day from the text’s delimiters and word shapes rather than matching the line against a vocabulary of Int./Ext. terms. The result depends on how the heading is written: the same rule can accept one Chinese form and miss another English form, or the reverse.

Slugline reading follows the text’s shape, not a keyword whitelist; when it misses, check how the heading is written.

Time of day is read from the end, Int./Ext. from the beginning

Time of day is read from the end of the heading and Int./Ext. from the beginning; both follow the text’s shape. When no recognizable form is present, Int./Ext. or time of day may stay empty, or only an incomplete fragment may be read. There is no keyword whitelist here, and the project’s Int./Ext. presets or time-of-day presets are not the basis of this step.

Four examples show the actual dividing line

内景 控制室 - 日: the Int./Ext. value reads 内景. 内景：控制室 - 日: a full-width colon with no space after the prefix, so no Int./Ext. value appears. EXTERIOR. OFFICE - DAY: the opening segment is too long to be recognized. INT./EXT. 屋顶 - 夜: when the opening segment contains a slash it is read as one compound marker, and INT./EXT. is preserved in full.

Across these examples, the visible dividing line is whether the opening segment falls inside a recognizable shape; the judgement looks at the whole prefix and its length, not a word count, so CJK text also participates by character shape.

Two shapes pull the wrong text into a field

With 1. INT. 控制室 - DAY, the leading 1. is stripped before the interior/exterior marker is looked for, so it is not taken as that prefix. With the all-caps, no-space title ROOFTOP-PARTY, the trailing PARTY is read as time of day while the location field keeps the whole line as written and is not shortened; the value is not rejected at that point and continues into the later flow.

The em dash, the en dash and the full-width hyphen are all inside the recognized forms. A compound form written without a slash, such as INT.,EXT., leaves only the leading INT (unlike the case above), and that is a partial read, not a rejection. That should not be summarized as one language working better than another, because the same reading rules miss different inputs in different languages.

A wrong value stays instead of being rejected

A prefix read as Int./Ext. is added to the project’s Int./Ext. presets as-is and then appears in the editor dropdown. The 1. example can therefore become a selectable preset, and a trailing value cut from a no-space title also enters the corresponding preset list.

This boundary matters because “a value was produced” does not mean “the value matches the on-set meaning.” A receiver should verify the Int./Ext. and time-of-day presets that appear after import, especially for sluglines with numbering, compound prefixes, or nonstandard hyphen characters.

A miscut prefix goes into the preset list; the wrong value stays instead of being rejected on the spot.

What this rule does not cover

What a Slugline Is Made Of explains the overall parts of a slugline; this page is narrower and only explains how Int./Ext. and time of day are cut from one line of text. For the export-and-return behavior covered by What an Exported FDX Does Not Bring Back, a receiver should verify the actual round trip rather than assuming that one successful cut will return unchanged.

Five separator shapes are recognised here — the hyphen, the spaced hyphen, the em dash, the en dash and the full-width hyphen — and anything outside that set is not read as a time separator. A first segment that contains a slash is read as one compound marker. When an input falls outside the space, period, and current hyphen branches, it should be treated as something to verify manually rather than as evidence about the source format.

FAQ

Do the project’s Int./Ext. presets participate in slugline parsing?

No. The reading follows the heading text itself, not the project’s Int./Ext. or time-of-day preset lists.

Why does EXTERIOR. OFFICE - DAY fail to produce Int./Ext.?

The opening segment is too long to fall inside a recognizable shape. That judgement runs on the whole prefix and its length rather than a word count, so another spelling can fail to be recognised — check those headings by eye.

What happens to INT./EXT.?

It is read as one compound marker, so INT./EXT. is preserved in full rather than cut down to INT.

Is a miscut value cleared at this step?

It is not rejected at this step. A produced prefix can enter the preset list, so it should be checked after import.

Axiom One LLC — SlateX