Check if input is USFM itself before attempting to parse

[kavitharaju]

How usfm_grammar 3.x would behave if we gave it a random text file?
Can we do some checks like, if no \id found in the first 3 content lines of the file, then bail?

[cmahte]

Here's what I use:

\id GEN ENG-US (p.sfm) - [GTP] Galilee Translation Project 2021[CC0] Hackett [7]
\id AAA BBB-CC (DDDD)  - [EEE] Fffffff Fffffffffff Fffffff 2021[CC0] Kkkkkkk [L]

Where the ID line is (theoretically) parsed into variables

Var	Example	Definitiion [spec] (data form)
id (&)	(all)	Project ID-complete field
id0 (&)	AAA	Project ID-Book ID [USFM]
id1 (&)	BBB	Project ID-ISO639 Language [p.sfm]
id2 (&)	CC	Project ID-ISO3166 Country [p.sfm]
id3 (&)	DDD	Project ID-Tagging Language [p.sfm]
id4 (&)	EEE	Project ID-Acronym (3Letter) [p.sfm]
id5 (&)	FFF	Project ID Title [p.sfm]
id6 (&)	GGG	Project ID Text Freeze Date [p.sfm]
id7 (&)	HHH	Project ID Rights Code [p.sfm] (creative commons)
id8 (&)	KKK	Project ID Rights Owner [p.sfm] (of final work)
id9 (&)	LLL	Project ID Status Level [p.sfm] (1-7 p.sfm publishing status, not 1-3 USFM community acceptance.)

So, specifically to USFM conformance:

If exactly "(USFM)" is found before the first dash on the ID line, then the work should conform to the USFM standard listed with the \usfm tag, or USFM 2.5 if no \usfm tag is found.

This affects linking, tables and images.

• Links \jmp are only intra-document.
• Tables (\tr#) will have no preceeding definition \rem or closing \b tags, and will fail somewhat gracefully as blue text paragraphs instead.
• Images will have no print/display size information embedded into them.

[joelthe1]

Thank you @cmahte for sharing this! I could not, however, find any official documentation to support this syntax specification. Is this something Paratext (or some such software) does? If so, could you point to the documentation?