Standards

Document Encoding Standards

These coding standards apply to new files scanned at sacred-texts.
There is a bibliography of all files scanned at this site here.
Because these conventions have evolved over time, not all texts will
have all of the markup specified here, particularly the page numbering
and footnote conventions.
We are constantly refining our methods of document preparation and
production.
However, there are still quite a few files at this site which were scanned
prior to these features being added, and many texts were developed elsewhere.
These files are incrementally being upgraded to reflect the newer standards;
this process is likely to take quite a while.

General Standards

The text is, as far as practically possible,
transcribed letter for letter from the original book.
The etext includes all of the original illustrations and graphics,
where possible.

The etext includes a transcription of the title page,
all prefatory material, a linked table of contents,
and all footnotes and other apparatus, except (in most cases) for the index.
We omit the index because it is sometimes impossible or difficult
to convert to etext, and once the text is on the Internet, every
word is indexed by multiple search engines.
We reproduce italics and bold text using standard HTML markup.

Etexts are posted in HTML format (and sometimes plain ASCII text)
due to open source concerns.
PDF and other ebook formats are vendor-specific,
and it is often difficult to migrate text out of them, particularly
with regard to formatting.

The etexts are uncensored and uncut.
In no case is any text omitted from the transcription,
even in case of egregious factual errors or language
which might be considered offensive.

Non-English passages are reproduced verbatim, and no
translation is supplied (unless the translation is part of the original text).

Each text is carefully checked against a specific printed copy of the book.
We attempt to find an early edition of the book,
(if possible the first edition).
If that is not practical, a photographic reproduction of an early edition
is used.
In some cases a later printing may be used, if we have a high degree of
confidence that it accurately reflects the complete, uncut original text,
with no major editing.
If a later edition is used, that will be noted.
We do not normally use multiple editions of the book to prepare the text.

Page numbers
are included in the transcription.
We feel that is important to indicate page numbers so the
etexts can be cited in academic and other publications.

Use of Unicode

Characters not within the standard ISO-8859-1 HTML
coding are represented using the closest
or exact equivalent in the Unicode character set.
In older files, these were transcribed systematically
using close 8 bit equivalents.
Such substitutions and omissions are noted on a case by case basis.

Unicode is inserted using extended character entities,
rather than UTF-8, since this is documents the character more clearly.
In some cases UTF-8 may be employed for longer files for reasons
of space.

In some cases vowel diacritics may be omitted from the transcription,
particularly in the case of well-known words such as ‘Sufi’ or ‘Nirvana’.
This will be noted.

Macron Vowels

Where vowels with a ‘macron’ (a straight line above the vowel,
usually indicating a ‘long’ vowel) in a text
are used consistently, these
may be transcribed using the ‘circumflex’ (the ‘hat’ diacritic, such
as â, ê, etc.).

In-text commentary

It is the policy of this site to not add any commentary
in the body of the etext.
Editorial matter written at sacred-texts about
the text is reserved for the index page, and will have a byline.
Where it is necessary to add brief in-text commentary,
it is printed in a green font.
Such editorial annotations may also be italicized and the
initials of the transcriber (normally, ‘jbh’) noted.
In-text commentary is usually limited to technical notes
about the source book, such as notes of illegible or ambiguous type,
missing pages, or the redactors’ attribution paragraph on the title page.

Errata

Each text is spell-checked during the proof-reading stage using the
standard MS Word spell-check dictionary.
Specialized vocabulary is added to a custom dictionary on a per-text basis.
The text is also vetted for known ‘OCR bums’–words that are OCRed incorrectly,
but spell-check as valid words,
such as ‘burn’ and ‘bum’, ‘bad’ and ‘had’, ‘arid’ and ‘and’.

British, archaic and dialect spelling has been retained where it occurs.
If an idiosyncratic spelling occurs more than twice in a given text,
it is normally not marked as errata.

Typos are corrected and those corrections are noted.
For instance, if ‘Greenland’ is spelled ‘Greeenland’,
that is considered a typo.
In such cases, the word (correctly spelled) is linked
to an errata file which documents the original spelling.
This file is named ‘errata.htm’ in the same directory as the text.

In some cases minor punctuation errors have been silently corrected.
However, care has been taken to preserve as closely as possible the
original punctuation, particularly in the case of older books.
Some books (for instance the Ganguli translation of the Mahabharata)
have major problems with punctuation, particularly nested dialog,
and this has been corrected where it was deemed necessary for comprehension.

Poetry

Poetry often requires complex indentation to transcribe its
appearance on the printed page.
This is simulated using nested DIR tags and non-breaking spaces at the start
(and in some cases the middle) of the line.
Where possible, each block of poetry has line breaks (BR) between
each line, and paragraph breaks at the end of each stanza.
Page numbers are placed within the body of a block of poetry to avoid
breaking up the layout.

Quotation Marks

Where a space appears on both sides of a double or single quote, the
inner space is deleted:

For the most part, only ‘straight’ quotes are used, except where it
may be required to resolve ambiguity such as

Page Breaks, Numbering and Continuations

Page breaks are indicated as p. NNN,
where NNN is the page number.
The HTML page markup is as follows:

Thus you could link to this specific page using the HTML anchor ‘page_91’.
For instance, if the name of the file was ‘foo42.htm’ in the directory ‘https://www.sacred-texts.com/bar’
you could link to it using the
following HTML markup: