The database is organized by SUBJECT and SOURCE.
SUBJECT listing: ABC/SUBJECTS
Subject file formatted:
{SUBJECT} {DESCRIPTION}
Issues
- Could be difficult to parse because of spacing
SOURCE listing (full bibliography): ABC/books
SOURCE listing (author, title): ABC/CODES
Only full biblography is needed
Source (full bibliography) formatted as:
{SOURCE_CODE} {BIBLIOGRAPHY}
Note: Bibliographies contain linebreaks
Issues
- SOURCE (full bibliography) may be hard to parse, contain numerous newlines
- Some SOURCE_CODES are invalid - they don't match to anything in the full SOURCE bibliography.
- Some SOURCES don't have SOURCE_CODES. These for the most part have the source appended to the end of the entry, after a
.. If no source code is available, we set the source to whatever is after the last.in the entry.
Fixes
- Find replace newlines
The entries themselves are formatted in the following manner:
Single subject:
{SUBJECT}, @{DATE/DATES}@ {ENTRY} <{SOURCE CODE} {PERIODICAL_DATE PAGE/PAGES}
Multi subject:
{SUBJECT,SUBJECT}~ @{DATE/DATES}@ {ENTRY} <{SOURCE CODE} {PERIODICAL_DATE PAGE/PAGES}
DATE, PERIODICAL_DATE are sometimes excluded
Within each ENTRY, there are some character replacements:
^should become a double quote~should become a comma
Issues
- SUBJECTS look to be duplicated in some ENTRIES
- Not all SUBJECTS are clearly delimited from ENTRIES
Will test with lunr.js, though this may be too large a dataset.
{
"SUBJECTS": "{csv}",
"SOURCE": "{full bibliography with periodical date, pages}",
"DATES": "{expand years to 19XX}"
"ENTRY": "{entry only}"
}