Changes in Version 1.3.0 (2019-10-30)

• Package bibtex moved to Suggests in DESCRIPTION due to it currently being orphaned on CRAN. Work is underway by the ROpenSci team to rectify this. The package can still be installed from GitHub. In the event that bibtex is not installed, the functions ReadBib(), GetBibEntryWithDOI(), ReadCrossRef(), and ReadZotero() throw an appropriate message asking the user to install bibtex and invisibly return NULL.
• The old CrossRef API can no longer be used. If use.old.api is set to in , it will be ignored with a warning.
• GetDOIs() had to be removed due to changes to the CrossRef API. It will hopefully return in the next release.

Changes in Version 1.2.13 (2019-04-03)

BUG FIXES

• When working in single-byte locales, the print method for BibEntry objects is more robust against accented characters being converted to incorrect ones when bib.style = "authoryear". Additionally, for this style, a period could be removed from the last initial in the first author’s given name when first.inits = TRUE. This has been corrected.

Changes in Version 1.2.12 (2019-04-02)

• The serial comma is now used when formatting name lists

BUG FIXES

• Fix issue that could lead to Unicode characters being converted to latin1 when printing (h/t joaochenriques #62)
• Fix issue with extracting DOI from CrossRef results in GetDOIs
• Fixes for ReadCrossRef when use.old.api is TRUE. Scores are sometimes not returned by the API call and when this occurs the entries will now be added to the output BibEntry object with a message indicating that no score was available.
• A comma no longer appears before “et al.” when the max.names options is set to 1 (h/t davidaknowles #56)

Changes in Version 1.2.8 (2018-12-10)

BUG FIXES

• Fix extraction of citation counts in ReadGS that was occasionally causing errors from some scholar.ids (h/t Miao Sun, #59)
• Fix printing of thesis and report entries types in authoryear style when “type” field missing (h/t Hugo Grunson, #58)
• Fix for PrintBibliography for the case of BibEntry objects with a single entry, an NA value could appear next to the year in the output (#60)
• Fix issue that could lead to Unicode characters being converted to latin1 when printing (h/t joaochenriques #62)

Changes in Version 1.2.2 (2018-05-31)

BUG FIXES

• GetPubMedByID is better at extracting years and months from the results returned by NCBI Entrez (h/t Dale Steele #52)
• The as.data.framemethod for BibEntry objects now correctly handles the case of a single entry with name list fields containing multiple names (h/t Damon Bayer #51)

Changes in Version 1.2.0 (2018-04-24)

NEW FEATURES

• +.BibEntry and merge.BibEntry gain an argument ignore.case, which defaults to BibOptions()$ignore.case (TRUE) so that case is ignore when checking for duplicate fields (h/t Justin Calabrese #47)
• Improved warning message when printing entries with unknown LaTeX macros (the entry key is now included). (h/t Justin Calabrese #49)
• The entry key is now included in warning messages when entries are missing fields and BibOptions()$check.entries == "warn" (h/t Justin Calabrese #48)

BUG FIXES

• Entries are now only checked once to ensure all required fields are present in ReadBib

Changes in Version 1.1.0 (2018-04-02)

• PrintBibliography gains parameters “start” and “end” to allow for printing only a subset of all cited entries from a BibEntry object (h/t Joseph Casillas #45, #46)

Changes in Version 1.0.0 (2018-02-19)

• Use https for all links (h/t Katrin Leinweber)
• Use preferred DOI resovler (h/t Katrin Leinweber)
• Add support for latex macro ast for asterisks (h/t Melinda Higgins)

Changes in Version 0.14.25 (2017-12-25)

BUG FIXES

• Fix GetDOIs to use https
• Fix download of bibliographic info from DOI in ReadPDFs

Changes in Version 0.14.23 (2017-11-12)

BUG FIXES

• Fix writing of BibEntry object to stdout in WriteBib (h/t Stephane Plaisance)
• ReadBib won’t add an attribute “strings” if there are none present in read bib file (h/t Stephane Plaisance)

Changes in Version 0.14.21 (2017-09-05)

BUG FIXES

• Fix deletion of temporary file if user supplies a DOI to ReadCrossRef (h/t Ben Raymond)

Changes in Version 0.14.20 (2017-07-27)

NEW FEATURES

• Documentation example improvements
• Improve error handling for API query functions
• Package peer reviewed and accepted by rOpenSci (h/t Noam Ross, Carl Boettiger, and Amelia McNamara)

BUG FIXES

• Remove missing plot from Rhtml vignette
• URL field returned by GetBibEntryWithDOI is now decoded properly
• Fix hyperlinks from bibliography to citations in vignettes
• Remove a incorrect message occasionally output from the addition operator for BibEntry objects

Changes in Version 0.14.12 (2017-06-30)

NEW FEATURES

• Package now uses httr, xml2, jsonlite packages instead of RCurl, XML, RJSONIO for scaffolding
• No more R CMD check NOTE regarding foreign function call to bibtex (h/t Romain Francois)

BUG FIXES

• Fix printing when BibOptions(style = 'yaml)
• Remove invalid character in inst/Bib/RJC.bib
• Correct parsing of interval dates when creating unique labels for authoryear style citations
• c.BibEntry throws an error if not all objects are bibentry objects
• Fix typos in documentation
• Literal ampersands are now printed correctly (not as ‘&’) (h/t Yue Hu)
• Ensure BibTeX month macros are processed properly by lubridate in non-English locales (h/t Sergio Oller)

Changes in Version 0.13.4 (2017-04-25)

BUG FIXES

• Unescape special characters in URL fields returned by CrossRef (h/t Michael Schubert)
• Remove square brackets from custom entry type names (h/t Hugh Parsonage)

Changes in Version 0.13.1 (2016-11-14)

BUG FIXES

• Feature involving LaTeX macros added in package version 0.12.0 can only be used for R 3.3.z and higher; this corrects cause of failed checks on R 3.2.z

Changes in Version 0.13.0 (2016-11-09)

BUG FIXES

• Updated calls to NCBI Entrez for functions ReadPubMed, GetPubMedByID, etc. to use https as now required by NCBI (h/t Dale Steele and Anthony Crane)
• Change reference to www.omegahat.org to www.omegahat.net (h/t Kurt Hornik)
• Documentation for ReadPubMed is updated to reflect that the default number of entries returned (controlled by the argument retmax) is 20 (h/t Dale Steele)

Changes in Version 0.12.0 (2016-09-30)

NEW FEATURES

• Some LaTeX macros unknown to R are now defined as macros in the package, and will be parsed using macros arg in tools::parse_Rd (assuming getRversion() >= "3.2.0" Note: corrected in 0.13.1 to be getRversion() >= "3.3.0")

BUG FIXES

• Parse LaTeX macro \textquotesingle in author names (h/t Bill Denney)
• Avoid “Request-URI too large” errors in GetPubMedByID if requesting a large number of IDs (h/t Maurits Evers)

Changes in Version 0.11.0 (2016-09-10)

NEW FEATURES

• ReadCrossRef now uses the newer CrossRef API and gains arguments filter and offset to use with the new API; an additional argument use.old.api is added if the user wishes to use the old API (h/t Carl Boettiger)
• ReadCrossRef now parses the results returned by CrossRef to create the BibEntry object when using the new API; for the old API (and hence, older versions of the package) the query only returns DOIs and ReadCrossRef would then use the DOIs to request the corresponding BibTeX entries from CrossRef (i.e. less HTTP requests when using the new API)

BUG FIXES

• Fix generation of entry keys when the word used from the title for key generation contains a non-ascii character (h/t Mark Johnson)
• RefManageR will no longer hang due to a bug in tools::latexToUtf8 (PR#17138) that is occasionally encountered when that function processes an unknown macro (h/t Eric Bryant)
• Entries with no title field can now be printed without error when BibOptions()$check.entries is not set to “error” (default is “error”)

Changes in Version 0.10.15 (2016-06-06)

BUG FIXES

• Removed unnecessary use of local/parent.frame; fixes execution with bytecode compiler (h/t Tomas Kalibera)

Changes in Version 0.10.12 (2016-03-25)

BUG FIXES

• Fixed broken test involving ReadPDFs due to changed URL (h/t Kurt Hornik)
• as.data.frame.BibEntry works for length one BibEntry with multiple authors (h/t Dale Steele)
• Use httr::GET to fix ReadGS
• Fixed broken tests in test-authors.R owing to changes to person class
• Fixed ReadCrossRef tests and error message
• Fixed printing for authoryear style (h/t Joseph Casillas)
• Name list fields (author, editor, etc.) provided to the function BibEntry are now properly parsed when specified as they would be in BibTeX/BibLaTeX; e.g. author = "Smith, Bob A. and Doe, Jane".

Changes in Version 0.10.5 (2016-01-03)

• The ‘key’ field in BibEntry objects is now always enforced to be unique
• as.data.frame.BibEntry is faster and now works if duplicate keys are present; keys in (h/t Dale Steele)
• Fix for ReadCrossRef if downloaded BibTeX had leading whitespace (h/t Carl Boettiger)
• useBytes = TRUE used for all calls to grep, sub, etc. (h/t HI&RH Lord Ripley of England)
• remove use of deprecated function lubridate::new_interval
• updated URL for the BibLaTeX manual
• Fix test in test-search.R that broke because of new year (h/t HI&RH Lord Ripley of England)
• add additional functions from utils and stats to NAMESPACE

Changes in Version 0.9.0 (2015-06-10)

• Use bibtex >= 0.4.0.9000 function do_read_bib to avoid .External call and R check note (request of HI&RH Lord Ripley of England)

Changes in Version 0.8.63 (2015-06-08)

NEW FEATURES

• Improve parsing of dates in ReadPDFs
• Citations using Cite family of functions can now be pandoc style, e.g. [@abibkey] by setting BibOptions(cite.style = "pandoc") (h/t Dale Steele)
• Added note about locales when parsing string ‘month’ fields to ReadBib help page (h/t Dieter Menne)

BUG FIXES

• Fixed merging BibEntry objects by multiple fields when no duplicates
• open.BibEntry fixed to not use partial matching of field names; e.g. an error would occur if the specified entry had a ‘urldate’ field, but no ‘url’ field
• open.BibEntry will message and not throw error if entry cannot be opened
• Fixes for ReadPDFs when argument use.metadata is FALSE
• Fix for ReadPDFs when when reading one file which is a JSTOR pdf
• Fix for sorting by volume (BibOptions(sorting = "anyvt") and BibOptions(sorting = "nyvt")
• Fix for sorting by label (BibOptions()sorting equal to “anyvt” or “anyt”)
• GetBibEntryWithDOI.R will not stop if an error occurs downloading any of the DOIs (e.g., if one entry in the doi vector has a typo and the rest are valid)

Changes in Version 0.8.52 (2015-01-26)

NEW FEATURES

• GetPubMedByID: Now returns some additional fields including ‘month’ and ‘issn’ for articles; will print a warning if PubMed does not return the complete list of authors; will use the name of a collective if one is available and the individual authors are missing (h/t Dale Steele)

BUG FIXES

• ReadBib: If a name list field in an entry cannot be parsed in the bib file, the entry will be ignored, but the rest of the file will still be processed and returned. In the past, this caused an error and no output would be returned.
• ‘Book’ entries will now be parsed correctly by GetPubMedByID (h/t Dale Steele)
• Fix error/warning messages when entry is missing required fields (bug introduced in Version 0.8.45)
• Name lists containing a comma in braces will now be parsed correctly, e.g. “Buchalter, Louis and {Murder, Inc.} and Anastasia, Albert”

Changes in Version 0.8.45 (2014-12-29)

BUG FIXES

• ReadCrossRef now correctly handles the small number of cases where BibTeX information cannot be obtained for a particular DOI, which resulted in ‘stack imbalance’ warnings and no results being returned (h/t Norman L Guinasso Jr).
• ReadGS fixed to account for changes to Google “API” (h/t Norman L Guinasso Jr).
• Improved parsing for BibTeX format names ending with a ‘}’ (h/t Henrik Bengtsson).
• Printing references with style = "html" would not always add an opening tag when bib.style = "numeric" or bib.style = "alphabetic" (h/t Henrik Bengtsson).
• format.BibEntry would ignore the .style argument if called directly by the user. Note, this function should normally not need to be called directly. (h/t Henrik Bengtsson)

Changes in Version 0.8.40 (2014-10-28)

NEW FEATURES

• Improved formatting of citation given to CrossRef for increased chances of finding matches with GetDOIs function (h/t Erich Studerus)
• Additional parsing of ‘month’ field to accomodate days and ranges of days and months. Example bib entries that will be parsed correctly include month = jun # "/" # jul, month = "20~" # jan, month = "20==25~" # dec, month = "10~" # jan # "/" # feb (request of Stephen Eglen)
• Added argument ‘group’ to ReadZotero for specifying a groupID to query a group library instead of a user library (h/t Greg Blomquist).

BUG FIXES

• DOI’s hyperlinks in Markdown format are now correct (h/t Stephen Eglen)
• print.BibEntry with BibOptions(style = "Biblatex") fixed (h/t Artem Klevtsov)
• unlist.BibEntry and RelistBibEntry now retains @strings and mheader and mfooter attributes (see ?BibEntry) if they are present

Changes in Version 0.8.34 (2014-08-18)

NEW FEATURES

• Added function GetDOIs which searches CrossRef for DOIs for the citations stored in a BibEntry object

BUG FIXES

• ReadCrossRef fixed to account for change to CrossRef API endpoint. (h/t Carl Boettiger)
• Abstracts returned by NCBI Entrez can be multiple parts. This is now handled correctly and the complete abstract will be returned in the ‘abstract’ field. (h/t Erich Studerus)
• DOI’s were too naively extracted from NCBI Entrez results, resulting in some entries having ‘doi’ fields with length greater than one. Now fixed. (h/t Erich Studerus)

Changes in Version 0.8.32 (2014-08-14)

NEW FEATURES

• Functions for interacting with NCBI Entrez return abstract of each article (request of Erich Studerus)

BUG FIXES

• print.BibEntry with BibOptions(style = "citation") now works properly
• Examples calling web resources should no longer upset the check farm (h/t HI&RH Lord Ripley of England)

Changes in Version 0.8.3 (2013-07-30)

NEW FEATURES

• as.BibEntry will create entry key if given a bibentry object with no key. Useful when citing packages with citation.
• PrintBibliography and Cite functions (Cite, Citet, etc.) accept bibentry objects in addition to BibEntry objects.
• $<-.BibEntry will now accept a single person object, so that a single author in a multi-author entry may be updated. An example may be found at help("$<-.BibEntry"). (h/t Carl Boettiger)

BUG FIXES

• validated html
• changed example for WriteBib that occasionally failed check

Changes in Version 0.8.2 (2013-06-01)

BUG FIXES

• Cite functions work if a specified entry has no key. Note that keys should always be provided for all entries as they are required for all entries in a BibLaTeX bib file (h/t Carl Boettiger)
• Entries returned by Crossref that have entry type ‘Data’ which is not supported by default in BibLaTeX are converted to type ‘Online’ (h/t Carl Boettiger)
• Fix for ReadGS when argument check.entries is FALSE or “warn” (h/t Francisco Rodriguez Sanchez)
• Family names from Scholar in all caps are handled correctly in ReadGS

NEW FEATURES

• Functions for interacting with PubMed return language of each article (h/t Dale Steele)
• Added CITATION file
• Updated License to explicitly include GPL-2 and GPL-3

Changes in Version 0.8.1 (2013-03-09)

BUG FIXES

• Fix for names<-.BibEntry
• Fix for print.BibEntry when entry has urldate field but no url field
• Corrections for some documentation typos
• Fix pmidrelated field when batch.mode = FALSE in GetPubMedRelated
• Fix for LookupPubMedID when index argument specified
• open.BibEntry now works properly
• Fix for converting thesis entries in toBibtex.BibEntry
• Fix for WriteBib with biblatex argument

NEW FEATURES

• Added Vignettes including user manual and Rmd citation examples
• Added NEWS
• Added HTML output of Rmd and RHTML citation examples to doc/