diff --git a/CHANGELOG.md b/CHANGELOG.md --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -4,3 +4,20 @@ * Initial release. +## 0.1.0.1 + + * Fix sorting when no `` element given. The spec says: + "In the absence of cs:sort, cites and bibliographic entries appear in + the order in which they are cited." This affects IEEE in particular. See + jgm/pandoc#6741. + + * Improve `sameNames` and citation grouping. Preivously if a citation + item had a prefix, it would not be grouped with following citations. + See jgm/pandoc#6722 for discussion. + + * Remove unneeded `hasNoSuffix` check in `sameNames`. + + * Remove unneeded import + + * `citeproc` executable: strip BOM before parsing style (#18). + diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -3,12 +3,7 @@ [![BSD2 license](https://img.shields.io/badge/license-BSD2-blue.svg)](LICENSE) [![CI tests](https://github.com/jgm/citeproc/workflows/CI%20tests/badge.svg)](https://github.com/jgm/citeproc/actions) - - This library generates citations and bibliography formatted according to a [CSL] style. Currently version 1.0.2 of the CSL @@ -94,7 +89,7 @@ Although this library is much more accurate in implementing the CSL spec than pandoc-citeproc was, it still fails some of the -tests from the CSL test suite (59/845). However, most of the +tests from the CSL test suite (58/845). However, most of the failures are on minor corner cases, and in many cases the expected behavior goes beyond what is required by the CSL spec. diff --git a/app/Main.hs b/app/Main.hs --- a/app/Main.hs +++ b/app/Main.hs @@ -45,7 +45,8 @@ Left e -> err e Right (inp :: Inputs (CslJson Text)) -> do stylesheet <- case optStyle opt of - Just fp -> TIO.readFile fp + Just fp -> T.dropWhile (=='\xFEFF') <$> -- drop BOM + TIO.readFile fp Nothing -> case inputsStyle inp of Just s -> return s diff --git a/citeproc.cabal b/citeproc.cabal --- a/citeproc.cabal +++ b/citeproc.cabal @@ -1,6 +1,6 @@ cabal-version: 2.2 name: citeproc -version: 0.1 +version: 0.1.0.1 synopsis: Generates citations and bibliography from CSL styles. description: citeproc parses CSL style files and uses them to generate a list of formatted citations and bibliography @@ -20,6 +20,7 @@ locales/*.xml locales/locales.json man/citeproc.1.md + man/citeproc.1 test/NOTES.md test/csl/*.txt test/extra/*.txt diff --git a/man/citeproc.1 b/man/citeproc.1 new file mode 100644 --- /dev/null +++ b/man/citeproc.1 @@ -0,0 +1,199 @@ +.\" Automatically generated by Pandoc 2.11 +.\" +.TH "citeproc" "1" "" "citeproc 0.1.0.1" "" +.hy +.SH NAME +.PP +citeproc - process citations using a CSL stylesheet. +.SH SYNOPSIS +.PP +\f[C]citeproc [options] [file]\f[R] +.SH DESCRIPTION +.PP +\f[C]citeproc\f[R] reads a JSON-encoded \f[C]Inputs\f[R] object from +\f[C]stdin\f[R] (or from a file if a filename is provided) and writes a +JSON-encoded \f[C]Result\f[R] object to \f[C]stdout\f[R]. +This executable can be used to add citation processing to non-Haskell +projects. +.SH OPTIONS +.TP +\f[B]\f[CB]-s\f[B]\f[R] \f[I]FILE\f[R], \f[B]\f[CB]--style=\f[B]\f[R]\f[I]FILE\f[R] +Specify a CSL style to be used. +.TP +\f[B]\f[CB]-r\f[B]\f[R] \f[I]FILE\f[R], \f[B]\f[CB]--references=\f[B]\f[R]\f[I]FILE\f[R] +Specify a CSL JSON bibliography to be used as a source for references. +.TP +\f[B]\f[CB]a\f[B]\f[R] \f[I]FILE\f[R], \f[B]\f[CB]--abbreviations=\f[B]\f[R]\f[I]FILE\f[R] +Specify a CSL abbreviations file. +.TP +\f[B]\f[CB]l\f[B]\f[R] \f[I]LANG\f[R], \f[B]\f[CB]--lang=\f[B]\f[R]\f[I]LANG\f[R] +Specify a locale to override the style\[cq]s default (IETF language +code). +.TP +\f[B]\f[CB]f\f[B]\f[R] \f[I]html|json\f[R], \f[B]\f[CB]--format=\f[B]\f[R]\f[I]html|json\f[R] +Specify the format to be used for the entries. +\f[C]html\f[R] (the default) uses HTML tags and entities for formatting. +\f[C]json\f[R] creates a structured JSON representation of a formatted +document. +.TP +\f[B]\f[CB]-h, --help\f[B]\f[R] +Print usage information. +.TP +\f[B]\f[CB]-V, --version\f[B]\f[R] +Print version. +.SH NOTES +.PP +The input JSON should have the structure: +.IP +.nf +\f[C] +{ \[dq]citations\[dq]: [ ...list of citations... ], + \[dq]references\[dq]: [ ...list of references... ], + \[dq]style\[dq]: \[dq]\[dq], + \[dq]abbreviations\[dq]: { ...abbreviations... }, + \[dq]lang\[dq]: \[dq]fr-FR\[dq] } +\f[R] +.fi +.PP +None of these fields is mandatory. +Instead of providing \f[C]references\f[R] in the input JSON, one can +specify a file containing a CSL JSON bibliography, using the +\f[C]--references\f[R] option on the command line. +Instead of providing a CSL stylesheet in the JSON, one can specify a +file using the \f[C]--style\f[R] option. +\f[C]--abbreviations\f[R] and \f[C]--lang\f[R] may also be used on the +command line to specify an abbreviations file or a locale. +(Command-line options always override any values given in the JSON +file.) +.PP +A citation is structured like this: +.IP +.nf +\f[C] +{ \[dq]citationID\[dq]: \[dq]foo\[dq], + \[dq]citationItems\[dq]: [ ...list of citationItems... ], + \[dq]citationNoteNumber\[dq]: 3 } +\f[R] +.fi +.PP +Only \f[C]citationItems\f[R] is necessary. +Alternatively, instead of +.IP +.nf +\f[C] +{ \[dq]citationItems\[dq]: [ ... ] } +\f[R] +.fi +.PP +one can just specify an array of items directly: +.IP +.nf +\f[C] +[ ... ] +\f[R] +.fi +.PP +A citation item is structured like this: +.IP +.nf +\f[C] +{ \[dq]id\[dq]: \[dq]foo\[dq], + \[dq]type\[dq]: \[dq]suppress-author\[dq], + \[dq]label\[dq]: \[dq]page\[dq], + \[dq]locator\[dq]: \[dq]45\[dq], + \[dq]prefix\[dq]: \[dq]see \[dq], + \[dq]suffix\[dq]: \[dq] and others\[dq] } +\f[R] +.fi +.PP +Only \f[C]id\f[R] is mandatory. +If \f[C]type\f[R] is omitted, it will be assumed to be +\f[C]normal-cite\f[R] (other values are \f[C]suppress-author\f[R] and +\f[C]author-only\f[R]). +.PP +A reference is structured like this: +.IP +.nf +\f[C] +{ + \[dq]author\[dq]: [ + { + \[dq]family\[dq]: \[dq]Aristotle\[dq] + } + ], + \[dq]id\[dq]: \[dq]aristotle:prior\[dq], + \[dq]issued\[dq]: { + \[dq]date-parts\[dq]: [ + [ + 1989 + ] + ] + }, + \[dq]publisher\[dq]: \[dq]Hackett\[dq], + \[dq]publisher-place\[dq]: \[dq]Indianapolis\[dq], + \[dq]title\[dq]: \[dq]Prior analytics\[dq], + \[dq]translator\[dq]: [ + { + \[dq]family\[dq]: \[dq]Smith\[dq], + \[dq]given\[dq]: \[dq]Robin\[dq] + } + ], + \[dq]type\[dq]: \[dq]book\[dq] +} +\f[R] +.fi +.PP +An abbreviations object has this form: +.IP +.nf +\f[C] +{ \[dq]default\[dq]: { + \[dq]container-title\[dq]: { + \[dq]Lloyd\[aq]s Law Reports\[dq]: \[dq]Lloyd\[aq]s Rep\[dq], + \[dq]Estates Gazette\[dq]: \[dq]EG\[dq], + \[dq]Scots Law Times\[dq]: \[dq]SLT\[dq] + } + } +} +\f[R] +.fi +.PP +The abbreviations will be substituted in the output only when the +\f[C]form\f[R] attribute for the style element that renders the variable +is set to \f[C]short\f[R]. +.PP +The output JSON will have the structure: +.IP +.nf +\f[C] +{ \[dq]citations\[dq]: [ ...list of strings... ], + \[dq]bibliography\[dq]: [ ...list of arrays: item id and a string... ], + \[dq]warnings\[dq]: [ ...list of warnings... ] +} +\f[R] +.fi +.PP +The contents of the entries will be HTML by default, but if +\f[C]--format=json\f[R] is specified, it will be a structured +representation of formatted text. +.IP +.nf +\f[C] +[ + \[dq]\[em]\[em]\[em]. 1983b. \[lq]The Concept of Truth in Formalized Languages.\[rq] In \[dq], + { + \[dq]format\[dq]: \[dq]italics\[dq], + \[dq]contents\[dq]: [ + \[dq]Logic, Semantics, Metamathematics\[dq] + ] + }, + \[dq], edited by John Corcoran, 152\[en]278. Indianapolis: Hackett.\[dq] +] +\f[R] +.fi +.PP +This representation can be used if you want to convert the result to a +format other than HTML. +.SH AUTHORS +.PP +John MacFarlane diff --git a/src/Citeproc.hs b/src/Citeproc.hs --- a/src/Citeproc.hs +++ b/src/Citeproc.hs @@ -10,7 +10,6 @@ , citeproc , Result(..) ) where -import Data.Text (Text) import Data.Semigroup import qualified Data.Text as T import qualified Data.Set as Set diff --git a/src/Citeproc/Eval.hs b/src/Citeproc/Eval.hs --- a/src/Citeproc/Eval.hs +++ b/src/Citeproc/Eval.hs @@ -28,7 +28,7 @@ import Control.Applicative import Data.Generics.Uniplate.Operations (universe, transform) --- import Debug.Trace (trace) +-- import Debug.Trace (traceShowId) -- import Text.Show.Pretty (ppShow) -- ppTrace :: Show a => a -> a -- ppTrace x = trace (ppShow x) x @@ -127,7 +127,14 @@ addIfMissing _ (Just x) = Just x go = do - assignCitationNumbers (map referenceId refs) + -- list of citationItemIds that are actually cited + let citationOrder = M.fromList $ reverse $ zip + (concatMap (map citationItemId . citationItems) citations) + [(1 :: Int)..] + let citeIds = M.keysSet citationOrder + let sortedCiteIds = sortOn (`M.lookup` citationOrder) + (map referenceId refs) + assignCitationNumbers sortedCiteIds -- sorting of bibliography, insertion of citation-number (bibCitations, bibSortKeyMap) <- case styleBibliography style of @@ -141,9 +148,9 @@ refs let sortedIds = if null (layoutSortKeys biblayout) - then map referenceId refs + then sortedCiteIds else sortOn (`M.lookup` bibSortKeyMap) - (map referenceId refs) + (map referenceId refs) assignCitationNumbers $ case layoutSortKeys biblayout of (SortKeyVariable Descending "citation-number":_) @@ -160,9 +167,6 @@ [CitationItem ident Nothing Nothing NormalCite Nothing Nothing]) sortedIds return (bibCitations, bibSortKeyMap) - -- list of citationItemIds that are actually cited - let citeIds = foldr (Set.insert . citationItemId) mempty - (concatMap citationItems citations) -- styling of citations sortKeyMap <- foldM (\m citeId -> do @@ -777,29 +781,27 @@ (Tagged (TagItem _ _) (Tagged (TagCitationNumber n2) _xs2)) = n2 == n1 + 1 isAdjacentCitationNumber _ _ = False - sameNames (Tagged (TagItem ty1 _id1) x1) - (Tagged (TagItem ty2 _id2) x2) - | hasNoSuffix x1 - , hasNoSuffix x2 - , ty1 /= AuthorOnly - , ty2 /= AuthorOnly - = case (unFormat x1, unFormat x2) of - (Tagged (TagNames t1 _nf1 ns1) ws1, - Tagged (TagNames t2 _nf2 ns2) ws2) - -> t1 == t2 && (if ns1 == ns2 - then not (null ns1) || ws1 == ws2 - else ws1 == ws2) - -- case where it's just a date with no name or anything; - -- we treat this as same name e.g. (1955a,b) - (Tagged TagDate{} _, Tagged TagDate{} _) + sameNames x y = + case (extractTags x, extractTags y) of + (Just (Tagged (TagNames t1 _nf1 ns1) ws1), + Just (Tagged (TagNames t2 _nf2 ns2) ws2)) + -> t1 == t2 && (if ns1 == ns2 + then not (null ns1) || ws1 == ws2 + else ws1 == ws2) + -- case where it's just a date with no name or anything; + -- we treat this as same name e.g. (1955a,b) + (Just (Tagged TagDate{} _), Just (Tagged TagDate{} _)) -> True -- case where title is substituted _ -> False - sameNames _ _ = False - hasNoSuffix (Formatted f' _) = isNothing (formatSuffix f') - hasNoSuffix _ = True - unFormat (Formatted _ (z:_)) = unFormat z - unFormat x = x + extractTags x = + let items = [y | y@(Tagged (TagItem ty _) _) <- universe x + , ty /= AuthorOnly] + names = [y | y@(Tagged TagNames{} _) <- concatMap universe items] + dates = [y | y@(Tagged TagDate{} _) <- concatMap universe items] + in if null items + then Nothing + else listToMaybe names <|> listToMaybe dates groupAndCollapseCitations _ _ _ _ x = x takeSeq :: Show a => (a -> a -> Bool) -> [a] -> ([a], [a]) diff --git a/test/Spec.hs b/test/Spec.hs --- a/test/Spec.hs +++ b/test/Spec.hs @@ -319,7 +319,7 @@ (length (skipped counts)) case length (failed counts) + length (errored counts) of 0 -> exitWith ExitSuccess - n | n <= 59 -> do + n | n <= 58 -> do putStrLn "We have passed all the CSL tests we expect to..." exitWith ExitSuccess | otherwise -> exitWith $ ExitFailure n