TODO.txt

This file contains notes, todo items, and an incomplete record of features
implemented. It's just a scratch file and shouldn't be considered
authoritative. 


TODO:
=====
- add option to disable cleaning steps?
- merge adjacent {} nodes?
    mRNA, DNA are great -> {mRNA,} {DNA} are great
    should be {mRNA, DNA} are great
- Pages that only contain symbols ?
- Check for volumes, numbers that are not INTs
- implement biblint extract in.bib listofids.aux
    listofids.aux can be either:
    a list of \citation{a,b,c,d,e} commands
    or a list of a b c d entry keys 
- write additional unit tests and have them acutally check the output
- handle #
- check for lone "{" inside quoted string

DONE:
=====
x check outputs messages to stdout instead of stderr
x Fix preamble string bug so it outputs {" ... "} instead of { }
x only strange-case handle the title and booktitle?
x remove whitespace at start and end of fields
x correctly read preamble
x handle @comment on input (deleted on output)
x allow () to deliminate entries
x check for duplicate key=value pairs
x handle moo\"{e}bar correctly -> {moo\"{e}bar}
x deal with et al.

- Find " that are not in {}
- Find titles that are entirely enclosed in {}
- Find all caps words that are not inside {} and strange caps words
    (i.e. any word that isn't Xxxx or xxxx)
- remove empty fields
- recognize quoted ints
- parse author names
- handle @preamble
- handle integer IDENT
- write out the encountered errors
- include locations in errors

In Clean:

* when sorting, compare {} strings correctly
* when sorting, expand symbols
* add {} around " in fields
* Handle Word-Word-Word in IsStrangeCase
* Replace lone - in pages with --
* remove . from the end of titles
* back replace months
* Sorts entries by year
* Removes empty fields
* Removes non-blessed fields
* Removes {} around entire fields (trying to {} smaller sections}
* Converts {INT} -> INT
* Puts @preamble at the top
* Puts @string at the top (after @preamble)
* Remove all comments
* Normalize whitespace within fields
* Format all authors as von Last, First or von Last, Jr, First
* Use symbols whenever unambiguious
* Format the entries consistently, ordering fields consistently
* PAges that look like NUMBER - NUMBER -> NUMBER--NUMBER
* Pages A--B where A > B -> replace B with the expansion based on A
* lowercase words in Journal name
* Remove entries where onne is contained in another

LINT
* Check for malformed author names
* Check for unmatched $
* Lone - instead of --- in title?
* "et al" appearing in author lists
* Non-ascii characters anyplace
* Check for "years" that are not INTs
* Pages A--B where A > B 
* Check for undefined symbols
* Check for duplicate symbol definitions
* Check for duplicate keys
* Check for missing required fields
* check for two symbols defining same thing

DUPS
* look for duplicate entries that have different keys?
    * cluster by cleaned title

NOTES AND IDEAS FOR FUTURE:
===========================

IMPORT to.bib from.bib > combined.bib

if a field exists in from.key add it to to.key if it is not already in to.key

LIST
    biblint list author
* list all the values for the given field, along with the pubs that have
  that value
  sort the values by "similar" values
* for authors "have that value" means "has that author"
* list all the authors and the pubs with them

REMOVETAG
* remove all tags of the given type

SORT
* by key, year, first author, last author, journal (any tag)
    biblint sort -by [key|...]

GENKEYS
* generate keys from records using various patterns
    biblint genkeys "{author1}:{year}"