Biblical Hebrew Forum

["Peter Kirk" <Peter_Kirk AT sil.org>]

The SIL program does not "critically rely on" these ?'s. It simply outputs
long lines where the input text has long lines as indicated by a long gap
between ?s. Your sample text shows why it would be wrong to output a new
line where the input text has a new line - the result would be very short
lines of Hebrew not matching BHS. The aim of the program is to reproduce the
line breaks of BHS. But of course garbage in means garbage out. WTS could
have verified their data e.g. with an AWK script like yours, but had not
done so in 1989. Maybe they had done for the later text for which the SIL
program was designed.

You are of course welcome to amend the SIL program if you have other
requirements. But the need to reverse each line means that the line breaks
must be explicitly inserted somewhere. Your program would have to be quite
intelligent to work out where to put them if it does not follow the ?s.

Peter Kirk

-----Original Message-----
From: Henry Churchyard [mailto:churchh AT usa.net]
Sent: 17 November 2000 20:11
To: Biblical Hebrew
Subject: Re: Conversion of OTA BHS text -- closing in

<snip>

>> From: Henry Churchyard

>> (it never occurred to me that anybody would write a software
>> program in such a way that the presence of the "?"  characters is
>> critically necessary to successfully process the text, since these
>> characters do not mark any feature of the manuscript L/B19a, and do
>> not represent any meaningful or ancient characteristic of the
>> Biblical text either).  So if the SIL software vitally depends on
>> an absolutely correct placement of "?" markers (which in fact does
>> not seem to have been present in the majority of the versions of
>> the BHS e-text over the years), then I would say that it is
>> somewhat finicky (the opposite of "robust").

> From: "Peter Kirk" <Peter_Kirk AT sil.org>

> Software programmers are used to parsing texts, such as documents
> encoded in standards such as XML, according to the published
> specifications, which in this case are that a line in the original
> is terminated by "?".  The way in which my SIL colleague programmed
> this feature was precisely correct.  The problem was only with the
> text, which did not match the published specification.

Yes, but "?" is just about the most inessential and non-meaningful
feature of the BHS e-text transcription, so it's a little odd to have
your program critically rely on it; and here's what the "specification"
documentation says about it (section 2.7.3 holds true both for the
ca. 1989 CCAT BHS e-text and the OTA BHS, so I don't know in which
version of the e-text the "?" marks are "systematically verified" --
apparently not a version of the e-text which has been widely
distributed non-commercially):

      SUPPLEMENT TO THE CODE MANUAL FOR THE MICHIGAN OLD TESTAMENT
  by Alan Groves  Westminster Theological Seminary Philadelphia, PA 19118
                           Last Revised 6/7/89

  2.7.3 The '?' has not been verified systematically at this point.

--
Henry Churchyard   churchh AT usa.net   http://www.crossmyt.com/hc/

---
You are currently subscribed to b-hebrew as: [Peter_Kirk AT sil.org]
To unsubscribe, forward this message to
$subst('Email.Unsub')
To subscribe, send an email to join-b-hebrew AT franklin.oit.unc.edu.