Skip to content

Commit

Permalink
Finish API
Browse files Browse the repository at this point in the history
  • Loading branch information
@dieghernan
dieghernan committed Mar 3, 2024
1 parent e924743 commit 8c751c9
Show file tree
Hide file tree
Showing 12 changed files with 296 additions and 192 deletions.
280 changes: 146 additions & 134 deletions R/as_bibentry.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -118,18 +118,10 @@ as_bibentry <- function(x,
}


# Three cases:
# A) Full cff reference object or
# B) Individual reference list
# C) List of references
# Guess case
cff_type <- guess_cff_part(obj)


# Detect case A
is_full_cff <- "cff-version" %in% names(obj)
# Detect case B
is_single_ref <- "type" %in% names(obj)

if (is_full_cff) {
if (cff_type == "cff_full") {
# Try to generate preferred if not present
if (!("preferred-citation" %in% names(obj))) {
prefcit <- obj
Expand All @@ -146,7 +138,7 @@ as_bibentry <- function(x,
"references" = obj$references,
c(list(obj$`preferred-citation`), obj$references)
)
} else if (is_single_ref) {
} else if (cff_type == "cff_ref") {
obj_extract <- list(obj)
} else {
obj_extract <- obj
Expand All @@ -158,20 +150,18 @@ as_bibentry <- function(x,
return(NULL)
}

ref <- lapply(obj_extract, cff_bibtex_parser)
ref <- lapply(obj_extract, make_bibentry)
ref <- do.call(c, ref)

return(ref)
}


cff_bibtex_parser <- function(x) {
make_bibentry <- function(x) {

Check warning on line 160 in R/as_bibentry.R

View workflow job for this annotation

GitHub Actions / Run lintr scanning 

file=R/as_bibentry.R,line=160,col=1,[cyclocomp_linter] Functions should have cyclomatic complexity of less than 15, this has 31.

Check notice

Code scanning / lintr

Functions should have cyclomatic complexity of less than 15, this has 31. Note

Functions should have cyclomatic complexity of less than 15, this has 31.
if (is.null(x)) {
return(NULL)
}

stopifnotcff(x)

# Partially based on ruby parser
# https://github.com/citation-file-format/ruby-cff/blob/main/lib/cff/ >>
# (cont) formatter/bibtex_formatter.rb
Expand All @@ -184,79 +174,13 @@ cff_bibtex_parser <- function(x) {
# edition institution journal month number pages publisher title volume year

# Guess type of entry----
tobibentry$bibtype <- switch(tolower(x$type),
"article" = "article",
"book" = "book",
"manual" = "manual",
"unpublished" = "unpublished",
"conference" = "inproceedings",
"conference-paper" = "inproceedings",
"proceedings" = "proceedings",
"magazine-article" = "article",
"newspaper-article" = "article",
"pamphlet" = "booklet",
"report" = "techreport",
"thesis" = "mastersthesis",
# We would need to guess
"misc"
)
# Try guess thesis
ttype <- clean_str(gsub("[[:punct:]]", "",
x$`thesis-type`,
perl = TRUE
))

if (!is.null(ttype) && x$type == "thesis") {
if (grepl("Phd", ttype, ignore.case = TRUE)) {
tobibentry$bibtype <- "phdthesis"
}
}

# Check if it may be an incollection
# Hint: is misc with collection-title and publisher

if (all(
tobibentry$bibtype == "misc", !is.null(x$`collection-title`),
!is.null(x$publisher), !is.null(x$year)
)) {
tobibentry$bibtype <- "incollection"
}


tobibentry$bibtype <- guess_bibtype(x)
# address----
# BibTeX 'address' is taken from the publisher (book, others) or the
# conference (inproceedings).
# Set logic: conference > institution > publisher
if (!is.null(x$conference)) {
addr_search <- x$conference
} else if (!is.null(x$institution)) {
addr_search <- x$institution
} else {
addr_search <- x$publisher
}


tobibentry$address <- clean_str(paste(
c(
addr_search$address,
addr_search$city,
addr_search$region,
addr_search$country
),
collapse = ", "
))

# As a fallback, use also location
if (is.null(tobibentry$address) && !is.null(x$location)) {
tobibentry$address <- x$location$name
}


tobibentry$address <- guess_address(x)

# author----
aut <- x$authors
author <- as.person(aut)
tobibentry$author <- author
tobibentry$author <- as.person(x$authors)


# booktitle ----
Expand All @@ -268,7 +192,10 @@ cff_bibtex_parser <- function(x) {

# Fallback to conference name

if (tobibentry$bibtype == "inproceedings" && is.null(tobibentry$booktitle)) {
if (all(
tobibentry$bibtype == "inproceedings",
is.null(tobibentry$booktitle)
)) {
tobibentry$booktitle <- x$conference$name
}

Expand All @@ -280,21 +207,10 @@ cff_bibtex_parser <- function(x) {

# editor----
# Same case than authors
editors <- as.person(x$editors)
tobibentry$editor <- editors
tobibentry$editor <- as.person(x$editors)

# howpublished----

howpublished <- x$medium

if (!is.null(howpublished)) {
# Capitalize first letter
letts <- unlist(strsplit(howpublished, "|"))
howpublished <-
clean_str(paste0(c(toupper(letts[1]), letts[-1]), collapse = ""))

tobibentry$howpublished <- howpublished
}
tobibentry$howpublished <- make_howpublised(x)

# institution/organization ----

Expand All @@ -320,41 +236,6 @@ cff_bibtex_parser <- function(x) {
# journal----
tobibentry$journal <- x$journal

# key: First two given of author and year----
# Bear in mind institutions has only given
# Use the first two authors
aut_sur <- lapply(tobibentry$author[1:2], function(z) {
unz <- unlist(z)
if ("family" %in% names(unz)) {
r <- unz["family"]
return(clean_str(r))
}

r <- unz["given"]
return(clean_str(r))
})


aut_sur <- tolower(paste0(unlist(aut_sur), collapse = ""))
aut_sur <- gsub("\\s*", "", aut_sur)

# Try hard to remove accents
# First with iconv
aut_sur <- iconv(aut_sur,
from = "UTF-8", to = "ASCII//TRANSLIT",
sub = "?"
)

# Next to latex
aut_sur <- encoded_utf_to_latex(aut_sur)

# Finally keep only a-z letters for key
aut_sur <- gsub("[^_a-z]", "", aut_sur)

y <- x$year

tobibentry$key <- paste(c(aut_sur, y), collapse = ":")

# month----
m <- x$month

Expand Down Expand Up @@ -446,6 +327,10 @@ cff_bibtex_parser <- function(x) {
tobibentry$bibtype <- "inbook"
}

# key: First two given of author and year----
tobibentry$key <- make_bibkey(tobibentry)


# Handle anonymous author----
# If anonymous and not needed, then not use it

Expand Down Expand Up @@ -525,3 +410,130 @@ cff_bibtex_parser <- function(x) {

return(bib)
}


guess_bibtype <- function(x) {
init_guess <- switch(tolower(x$type),
"article" = "article",
"book" = "book",
"manual" = "manual",
"unpublished" = "unpublished",
"conference" = "inproceedings",
"conference-paper" = "inproceedings",
"proceedings" = "proceedings",
"magazine-article" = "article",
"newspaper-article" = "article",
"pamphlet" = "booklet",
"report" = "techreport",
"thesis" = "mastersthesis",
# We would need to guess
"misc"
)


# Try guess thesis
ttype <- clean_str(gsub("[[:punct:]]", "",
x$`thesis-type`,
perl = TRUE
))

if (!is.null(ttype) && x$type == "thesis") {
if (grepl("Phd", ttype, ignore.case = TRUE)) {
init_guess <- "phdthesis"
}
}

# Check if it may be an incollection
# Hint: is misc with collection-title and publisher

if (all(
init_guess == "misc", !is.null(x$`collection-title`),
!is.null(x$publisher), !is.null(x$year)
)) {
init_guess <- "incollection"
}

init_guess
}

guess_address <- function(x) {
# BibTeX 'address' is taken from the publisher (book, others) or the
# conference (inproceedings).
# Set logic: conference > institution > publisher
if (!is.null(x$conference)) {
addr_search <- x$conference
} else if (!is.null(x$institution)) {
addr_search <- x$institution
} else {
addr_search <- x$publisher
}


address <- clean_str(paste(
c(
addr_search$address,
addr_search$city,
addr_search$region,
addr_search$country
),
collapse = ", "
))

# As a fallback, use also location
if (is.null(address) && !is.null(x$location)) {
address <- clean_str(x$location$name)
}

address
}


make_bibkey <- function(tobibentry) {
# Bear in mind institutions has only given
# Use the first two authors
aut_sur <- lapply(tobibentry$author[1:2], function(z) {
unz <- unlist(z)
if ("family" %in% names(unz)) {
r <- unz["family"]
return(clean_str(r))
}

r <- unz["given"]
return(clean_str(r))
})


aut_sur <- tolower(paste0(unlist(aut_sur), collapse = ""))
aut_sur <- gsub("\\s*", "", aut_sur)

# Try hard to remove accents
# First with iconv
aut_sur <- iconv(aut_sur,
from = "UTF-8", to = "ASCII//TRANSLIT",
sub = "?"
)

# Next to latex
aut_sur <- encoded_utf_to_latex(aut_sur)

# Finally keep only a-z letters for key
aut_sur <- gsub("[^_a-z]", "", aut_sur)

y <- tobibentry$year

key <- paste(c(aut_sur, y), collapse = ":")
key
}

make_howpublised <- function(x) {
howpublished <- x$medium

if (!is.null(howpublished)) {
# Capitalize first letter
letts <- unlist(strsplit(howpublished, "|"))
howpublished <-
clean_str(paste0(c(toupper(letts[1]), letts[-1]), collapse = ""))
}

clean_str(howpublished)
}
Loading

0 comments on commit 8c751c9

Please sign in to comment.