*
* \author Angus Leeming
* \author Herbert Voß
- * \author Richard Heck
+ * \author Richard Kimberly Heck
* \author Julien Rioux
* \author Jürgen Spitzmüller
*
#include <config.h>
#include "BiblioInfo.h"
+
#include "Buffer.h"
#include "BufferParams.h"
-#include "buffer_funcs.h"
#include "Citation.h"
#include "Encoding.h"
-#include "InsetIterator.h"
#include "Language.h"
-#include "output_xhtml.h"
-#include "Paragraph.h"
#include "TextClass.h"
#include "TocBackend.h"
+#include "xml.h"
#include "support/convert.h"
#include "support/debug.h"
#include "support/docstream.h"
+#include "support/FileName.h"
#include "support/gettext.h"
#include "support/lassert.h"
#include "support/lstrings.h"
-#include "support/regex.h"
#include "support/textutils.h"
#include <map>
+#include <regex>
#include <set>
using namespace std;
namespace {
-// gets the "prename" and "family name" from an author-type string
-pair<docstring, docstring> nameParts(docstring const & name)
+// Remove placeholders from names
+docstring renormalize(docstring const & input)
+{
+ docstring res = subst(input, from_ascii("$$space!"), from_ascii(" "));
+ return subst(res, from_ascii("$$comma!"), from_ascii(","));
+}
+
+
+// Split the surname into prefix ("von-part") and family name
+pair<docstring, docstring> parseSurname(docstring const & sname)
+{
+ // Split the surname into its tokens
+ vector<docstring> pieces = getVectorFromString(sname, from_ascii(" "));
+ if (pieces.size() < 2)
+ return make_pair(docstring(), sname);
+
+ // Now we look for pieces that begin with a lower case letter.
+ // All except for the very last token constitute the "von-part".
+ docstring prefix;
+ vector<docstring>::const_iterator it = pieces.begin();
+ vector<docstring>::const_iterator const en = pieces.end();
+ bool first = true;
+ for (; it != en; ++it) {
+ if ((*it).empty())
+ continue;
+ // If this is the last piece, then what we now have is
+ // the family name, notwithstanding the casing.
+ if (it + 1 == en)
+ break;
+ char_type const c = (*it)[0];
+ // If the piece starts with a upper case char, we assume
+ // this is part of the surname.
+ if (!isLower(c))
+ break;
+ // Nothing of the former, so add this piece to the prename
+ if (!first)
+ prefix += " ";
+ else
+ first = false;
+ prefix += *it;
+ }
+
+ // Reconstruct the family name.
+ // Note that if we left the loop with because it + 1 == en,
+ // then this will still do the right thing, i.e., make surname
+ // just be the last piece.
+ docstring surname;
+ first = true;
+ for (; it != en; ++it) {
+ if (!first)
+ surname += " ";
+ else
+ first = false;
+ surname += *it;
+ }
+ return make_pair(prefix, surname);
+}
+
+
+struct name_parts {
+ docstring surname;
+ docstring prename;
+ docstring suffix;
+ docstring prefix;
+};
+
+
+// gets the name parts (prename, surname, prefix, suffix) from an author-type string
+name_parts nameParts(docstring const & iname)
{
- if (name.empty())
- return make_pair(docstring(), docstring());
+ name_parts res;
+ if (iname.empty())
+ return res;
+
+ // First we check for goupings (via {...}) and replace blanks and
+ // commas inside groups with temporary placeholders
+ docstring name;
+ int gl = 0;
+ docstring::const_iterator p = iname.begin();
+ while (p != iname.end()) {
+ // count grouping level
+ if (*p == '{')
+ ++gl;
+ else if (*p == '}')
+ --gl;
+ // generate string with probable placeholders
+ if (*p == ' ' && gl > 0)
+ name += from_ascii("$$space!");
+ else if (*p == ',' && gl > 0)
+ name += from_ascii("$$comma!");
+ else
+ name += *p;
+ ++p;
+ }
- // first we look for a comma, and take the last name to be everything
- // preceding the right-most one, so that we also get the "jr" part.
+ // Now we look for a comma, and take the last name to be everything
+ // preceding the right-most one, so that we also get the name suffix
+ // (aka "jr" part).
vector<docstring> pieces = getVectorFromString(name);
- if (pieces.size() > 1)
- // whether we have a jr. part or not, it's always
- // the first and last item (reversed)
- return make_pair(pieces.back(), pieces.front());
+ if (pieces.size() > 1) {
+ // Whether we have a name suffix or not, the prename is
+ // always last item
+ res.prename = renormalize(pieces.back());
+ // The family name, conversely, is always the first item.
+ // However, it might contain a prefix (aka "von" part)
+ docstring const sname = pieces.front();
+ res.prefix = renormalize(parseSurname(sname).first);
+ res.surname = renormalize(parseSurname(sname).second);
+ // If we have three pieces (the maximum allowed by BibTeX),
+ // the second one is the name suffix.
+ if (pieces.size() > 2)
+ res.suffix = renormalize(pieces.at(1));
+ return res;
+ }
- // OK, so now we want to look for the last name. We're going to
- // include the "von" part. This isn't perfect.
+ // OK, so now we want to look for the last name.
// Split on spaces, to get various tokens.
pieces = getVectorFromString(name, from_ascii(" "));
- // If we only get two, assume the last one is the last name
- if (pieces.size() <= 2)
- return make_pair(pieces.front(), pieces.back());
+ // No space: Only a family name given
+ if (pieces.size() < 2) {
+ res.surname = renormalize(pieces.back());
+ return res;
+ }
+ // If we get two pieces, assume "prename surname"
+ if (pieces.size() == 2) {
+ res.prename = renormalize(pieces.front());
+ res.surname = renormalize(pieces.back());
+ return res;
+ }
- // Now we look for the first token that begins with
- // a lower case letter or an opening group {.
+ // More than 3 pieces: A name prefix (aka "von" part) might be included.
+ // We look for the first piece that begins with a lower case letter
+ // (which is the name prefix, if it is not the last token) or the last token.
docstring prename;
vector<docstring>::const_iterator it = pieces.begin();
- vector<docstring>::const_iterator en = pieces.end();
+ vector<docstring>::const_iterator const en = pieces.end();
bool first = true;
for (; it != en; ++it) {
if ((*it).empty())
continue;
char_type const c = (*it)[0];
- if (isLower(c) || c == '{')
+ // If the piece starts with a lower case char, we assume
+ // this is the name prefix and thus prename is complete.
+ if (isLower(c))
+ break;
+ // Same if this is the last piece, which is always the surname.
+ if (it + 1 == en)
break;
+ // Nothing of the former, so add this piece to the prename
if (!first)
prename += " ";
else
prename += *it;
}
- if (it == en) // we never found a "von" or group
- return make_pair(prename, pieces.back());
-
- // reconstruct the family name
+ // Now reconstruct the family name and strip the prefix.
+ // Note that if we left the loop because it + 1 == en,
+ // then this will still do the right thing, i.e., make surname
+ // just be the last piece.
docstring surname;
first = true;
for (; it != en; ++it) {
first = false;
surname += *it;
}
- return make_pair(prename, surname);
+ res.prename = renormalize(prename);
+ res.prefix = renormalize(parseSurname(surname).first);
+ res.surname = renormalize(parseSurname(surname).second);
+ return res;
}
-docstring constructName(docstring const & name, string const scheme)
+docstring constructName(docstring const & name, string const & scheme)
{
// re-constructs a name from name parts according
// to a given scheme
- docstring const prename = nameParts(name).first;
- docstring const surname = nameParts(name).second;
- docstring result = from_ascii(scheme);
+ docstring const prename = nameParts(name).prename;
+ docstring const surname = nameParts(name).surname;
+ docstring const prefix = nameParts(name).prefix;
+ docstring const suffix = nameParts(name).suffix;
+ string res = scheme;
+ static regex const reg1("(.*)(\\{%prename%\\[\\[)([^\\]]+)(\\]\\]\\})(.*)");
+ static regex const reg2("(.*)(\\{%suffix%\\[\\[)([^\\]]+)(\\]\\]\\})(.*)");
+ static regex const reg3("(.*)(\\{%prefix%\\[\\[)([^\\]]+)(\\]\\]\\})(.*)");
+ smatch sub;
+ // Changing the first parameter of regex_match() may corrupt the
+ // second one. In this case we use the temporary string tmp.
+ if (regex_match(scheme, sub, reg1)) {
+ res = sub.str(1);
+ if (!prename.empty())
+ res += sub.str(3);
+ res += sub.str(5);
+ }
+ if (regex_match(res, sub, reg2)) {
+ string tmp = sub.str(1);
+ if (!suffix.empty())
+ tmp += sub.str(3);
+ res = tmp + sub.str(5);
+ }
+ if (regex_match(res, sub, reg3)) {
+ string tmp = sub.str(1);
+ if (!prefix.empty())
+ tmp += sub.str(3);
+ res = tmp + sub.str(5);
+ }
+ docstring result = from_ascii(res);
result = subst(result, from_ascii("%prename%"), prename);
result = subst(result, from_ascii("%surname%"), surname);
+ result = subst(result, from_ascii("%prefix%"), prefix);
+ result = subst(result, from_ascii("%suffix%"), suffix);
return result;
}
-bool multipleAuthors(docstring const author)
+vector<docstring> const getAuthors(docstring const & author)
{
- vector<docstring> const authors =
- getVectorFromString(author, from_ascii(" and "));
- return authors.size() > 1;
+ // We check for goupings (via {...}) and only consider " and "
+ // outside groups as author separator. This is to account
+ // for cases such as {{Barnes and Noble, Inc.}}, which
+ // need to be treated as one single family name.
+ // We use temporary placeholders in order to differentiate the
+ // diverse " and " cases.
+
+ // First, we temporarily replace all ampersands. It is rather unusual
+ // in author names, but can happen (consider cases such as "C \& A Corp.").
+ docstring iname = subst(author, from_ascii("&"), from_ascii("$$amp!"));
+ // Then, we temporarily make all " and " strings to ampersands in order
+ // to handle them later on a per-char level. Note that arbitrary casing
+ // ("And", "AND", "aNd", ...) is allowed in bibtex (#10465).
+ static regex const and_reg("(.* )([aA][nN][dD])( .*)");
+ smatch sub;
+ string res = to_utf8(iname);
+ while (regex_match(res, sub, and_reg))
+ res = sub.str(1) + "&" + sub.str(3);
+ iname = from_utf8(res);
+ // Now we traverse through the string and replace the "&" by the proper
+ // output in- and outside groups
+ docstring name;
+ int gl = 0;
+ docstring::const_iterator p = iname.begin();
+ while (p != iname.end()) {
+ // count grouping level
+ if (*p == '{')
+ ++gl;
+ else if (*p == '}')
+ --gl;
+ // generate string with probable placeholders
+ if (*p == '&') {
+ if (gl > 0)
+ // Inside groups, we output "and"
+ name += from_ascii("and");
+ else
+ // Outside groups, we output a separator
+ name += from_ascii("$$namesep!");
+ }
+ else
+ name += *p;
+ ++p;
+ }
+
+ // re-insert the literal ampersands
+ name = subst(name, from_ascii("$$amp!"), from_ascii("&"));
+
+ // Now construct the actual vector
+ return getVectorFromString(name, from_ascii(" $$namesep! "));
+}
+
+
+bool multipleAuthors(docstring const & author)
+{
+ return getAuthors(author).size() > 1;
}
bool scanning_cmd = false;
bool scanning_math = false;
+ bool is_section = false;
bool escaped = false; // used to catch \$, etc.
while (!val.empty()) {
char_type const ch = val[0];
// discard characters until we hit something that
// isn't alpha.
if (scanning_cmd) {
+ if (!is_section && ch == 'S') {
+ is_section = true;
+ val = val.substr(1);
+ continue;
+ }
if (isAlphaASCII(ch)) {
+ is_section = false;
val = val.substr(1);
escaped = false;
continue;
+ } else if (is_section) {
+ ret.push_back(0x00a7);
+ is_section = false;
+ continue;
}
// so we're done with this command.
// now we fall through and check this character.
+ is_section = false;
scanning_cmd = false;
}
continue;
}
+ if (ch == '~') {
+ ret += char_type(0x00a0);
+ val = val.substr(1);
+ continue;
+ }
+
if (ch == '$') {
ret += ch;
val = val.substr(1);
continue;
}
- // we just ignore braces
+ // Change text mode accents in the form
+ // {\v a} to \v{a} (see #9340).
+ // FIXME: This is a sort of mini-tex2lyx.
+ // Use the real tex2lyx instead!
+ static regex const tma_reg("^\\{\\\\[bcCdfGhHkrtuUv]\\s\\w\\}");
+ if (regex_search(to_utf8(val), tma_reg)) {
+ val = val.substr(1);
+ val.replace(2, 1, from_ascii("{"));
+ continue;
+ }
+
+ // Apart from the above, we just ignore braces
if (ch == '{' || ch == '}') {
val = val.substr(1);
continue;
// look for that and change it, if necessary.
// FIXME: This is a sort of mini-tex2lyx.
// Use the real tex2lyx instead!
- static lyx::regex const reg("^\\\\\\W\\w");
- if (lyx::regex_search(to_utf8(val), reg)) {
+ static regex const reg("^\\\\\\W\\w");
+ if (regex_search(to_utf8(val), reg)) {
val.insert(3, from_ascii("}"));
val.insert(2, from_ascii("{"));
}
return ret;
}
-} // anon namespace
+} // namespace
//////////////////////////////////////////////////////////////////////
//////////////////////////////////////////////////////////////////////
BibTeXInfo::BibTeXInfo(docstring const & key, docstring const & type)
- : is_bibtex_(true), bib_key_(key), entry_type_(type), info_(),
- modifier_(0)
+ : is_bibtex_(true), bib_key_(key), num_bib_key_(0), entry_type_(type),
+ info_(), format_(), modifier_(0)
{}
+
docstring const BibTeXInfo::getAuthorOrEditorList(Buffer const * buf,
bool full, bool forceshort) const
{
}
-docstring const BibTeXInfo::getAuthorList(Buffer const * buf, docstring author,
- bool full, bool forceshort, bool allnames,
- bool beginning) const
+docstring const BibTeXInfo::getAuthorList(Buffer const * buf,
+ docstring const & author, bool const full, bool const forceshort,
+ bool const allnames, bool const beginning) const
{
// Maxnames treshold depend on engine
size_t maxnames = buf ?
// in this case, we didn't find a "(",
// so we don't have author (year)
return docstring();
+ if (full) {
+ // Natbib syntax is "Jones et al.(1990)Jones, Baker, and Williams"
+ docstring const fullauthors = trim(rsplit(remainder, ')'));
+ if (!fullauthors.empty())
+ return fullauthors;
+ }
return authors;
}
if (author.empty())
return author;
- // FIXME Move this to a separate routine that can
- // be called from elsewhere.
- //
// OK, we've got some names. Let's format them.
- // Try to split the author list on " and "
- vector<docstring> const authors =
- getVectorFromString(author, from_ascii(" and "));
+ // Try to split the author list
+ vector<docstring> const authors = getAuthors(author);
docstring retval;
// These are defined in the styles
string const etal =
- buf ? buf->params().documentClass().getCiteMacro(engine_type, "_etal")
+ buf ? buf->params().documentClass().getCiteMacro(engine_type, "B_etal")
: " et al.";
string const namesep =
- buf ? buf->params().documentClass().getCiteMacro(engine_type, "_namesep")
+ buf ? buf->params().documentClass().getCiteMacro(engine_type, "B_namesep")
: ", ";
string const lastnamesep =
- buf ? buf->params().documentClass().getCiteMacro(engine_type, "_lastnamesep")
+ buf ? buf->params().documentClass().getCiteMacro(engine_type, "B_lastnamesep")
: ", and ";
string const pairnamesep =
- buf ? buf->params().documentClass().getCiteMacro(engine_type, "_pairnamesep")
+ buf ? buf->params().documentClass().getCiteMacro(engine_type, "B_pairnamesep")
: " and ";
string firstnameform =
buf ? buf->params().documentClass().getCiteMacro(engine_type, "!firstnameform")
- : "%surname%, %prename%";
+ : "{%prefix%[[%prefix% ]]}%surname%{%suffix%[[, %suffix%]]}{%prename%[[, %prename%]]}";
if (!beginning)
firstnameform = buf ? buf->params().documentClass().getCiteMacro(engine_type, "!firstbynameform")
- : "%prename% %surname%";
+ : "%prename% {%prefix%[[%prefix% ]]}%surname%{%suffix%[[, %suffix%]]}";
string othernameform = buf ? buf->params().documentClass().getCiteMacro(engine_type, "!othernameform")
- : "%surname%, %prename%";
+ : "{%prefix%[[%prefix% ]]}%surname%{%suffix%[[, %suffix%]]}{%prename%[[, %prename%]]}";
if (!beginning)
othernameform = buf ? buf->params().documentClass().getCiteMacro(engine_type, "!otherbynameform")
- : "%prename% %surname%";
+ : "%prename% {%prefix%[[%prefix% ]]}%surname%{%suffix%[[, %suffix%]]}";
+ string citenameform = buf ? buf->params().documentClass().getCiteMacro(engine_type, "!citenameform")
+ : "{%prefix%[[%prefix% ]]}%surname%";
// Shorten the list (with et al.) if forceshort is set
- // and the list can actually be shorten, else if maxcitenames
+ // and the list can actually be shortened, else if maxcitenames
// is passed and full is not set.
bool shorten = forceshort && authors.size() > 1;
vector<docstring>::const_iterator it = authors.begin();
retval += (i == 0) ? constructName(*it, firstnameform)
: constructName(*it, othernameform);
else
- retval += nameParts(*it).second;
+ retval += constructName(*it, citenameform);
}
if (shorten) {
if (allnames)
retval = constructName(authors[0], firstnameform) + (buf ? buf->B_(etal) : from_ascii(etal));
else
- retval = nameParts(authors[0]).second + (buf ? buf->B_(etal) : from_ascii(etal));
+ retval = constructName(authors[0], citenameform) + (buf ? buf->B_(etal) : from_ascii(etal));
}
return convertLaTeXCommands(retval);
static regex const ereg(".*/[-]?([\\d]{4}).*");
smatch sm;
string const date = to_utf8(year);
- regex_match(date, sm, yreg);
+ if (!regex_match(date, sm, yreg))
+ // cannot parse year.
+ return docstring();
year = from_ascii(sm[1]);
// check for an endyear
if (regex_match(date, sm, ereg))
}
+void BibTeXInfo::getLocators(docstring & doi, docstring & url, docstring & file) const
+{
+ if (is_bibtex_) {
+ // get "doi" entry from citation record
+ doi = operator[]("doi");
+ if (!doi.empty() && !prefixIs(doi,from_ascii("http")))
+ doi = "https://doi.org/" + doi;
+ // get "url" entry from citation record
+ url = operator[]("url");
+ // get "file" entry from citation record
+ file = operator[]("file");
+
+ // Jabref case, "file" field has a format (depending on exporter):
+ // Description:Location:Filetype;Description:Location:Filetype...
+ // or simply:
+ // Location;Location;...
+ // We will strip out the locations and return an \n-separated list
+ if (!file.empty()) {
+ docstring filelist;
+ vector<docstring> files = getVectorFromString(file, from_ascii(";"));
+ for (auto const & f : files) {
+ // first try if we have Description:Location:Filetype
+ docstring ret, filedest, tmp;
+ ret = split(f, tmp, ':');
+ tmp = split(ret, filedest, ':');
+ if (filedest.empty())
+ // we haven't, so use the whole string
+ filedest = f;
+ // TODO howto deal with relative directories?
+ FileName fn(to_utf8(filedest));
+ if (fn.exists()) {
+ if (!filelist.empty())
+ filelist += '\n';
+ filelist += "file:///" + filedest;
+ }
+ }
+ if (!filelist.empty())
+ file = filelist;
+ }
+
+ // kbibtex case, "localfile" field with format:
+ // file1.pdf;file2.pdf
+ // We will strip out the locations and return an \n-separated list
+ docstring kfile;
+ if (file.empty())
+ kfile = operator[]("localfile");
+ if (!kfile.empty()) {
+ docstring filelist;
+ vector<docstring> files = getVectorFromString(kfile, from_ascii(";"));
+ for (auto const & f : files) {
+ // TODO howto deal with relative directories?
+ FileName fn(to_utf8(f));
+ if (fn.exists()) {
+ if (!filelist.empty())
+ filelist += '\n';
+ filelist = "file:///" + f;
+ }
+ }
+ if (!filelist.empty())
+ file = filelist;
+ }
+
+ if (!url.empty())
+ return;
+
+ // try biblatex specific fields, see its manual
+ // 3.13.7 "Electronic Publishing Informationl"
+ docstring eprinttype = operator[]("eprinttype");
+ docstring eprint = operator[]("eprint");
+ if (eprint.empty())
+ return;
+
+ if (eprinttype == "arxiv")
+ url = "https://arxiv.org/abs/" + eprint;
+ if (eprinttype == "jstor")
+ url = "https://www.jstor.org/stable/" + eprint;
+ if (eprinttype == "pubmed")
+ url = "http://www.ncbi.nlm.nih.gov/pubmed/" + eprint;
+ if (eprinttype == "hdl")
+ url = "https://hdl.handle.net/" + eprint;
+ if (eprinttype == "googlebooks")
+ url = "http://books.google.com/books?id=" + eprint;
+
+ return;
+ }
+
+ // Here can be handled the bibliography environment. All one could do
+ // here is let LyX scan the entry for URL or HRef insets.
+}
+
+
namespace {
docstring parseOptions(docstring const & format, string & optkey,
}
-} // anon namespace
+} // namespace
/* FIXME
Bug #9131 revealed an oddity in how we are generating citation information
-when more than one key is given. We end up building a longer and longer format
+when more than one key is given. We end up building a longer and longer format
string as we go, which we then have to re-parse, over and over and over again,
rather than generating the information for the individual keys and then putting
all of that together. We do that to deal with the way separators work, from what
bit of work, however.
*/
docstring BibTeXInfo::expandFormat(docstring const & format,
- BibTeXInfoList const xrefs, int & counter, Buffer const & buf,
+ BibTeXInfoList const & xrefs, int & counter, Buffer const & buf,
CiteItem const & ci, bool next, bool second) const
{
// incorrect use of macros could put us in an infinite loop
fmt = from_utf8(val) + fmt.substr(1);
counter += 1;
continue;
- } else if (key[0] == '_') {
- // a translatable bit
+ } else if (prefixIs(key, "B_")) {
+ // a translatable bit (to the Buffer language)
string const val =
buf.params().documentClass().getCiteMacro(engine_type, key);
docstring const trans =
translateIfPossible(from_utf8(val), buf.params().language->code());
ret << trans;
+ } else if (key[0] == '_') {
+ // a translatable bit (to the GUI language)
+ string const val =
+ buf.params().documentClass().getCiteMacro(engine_type, key);
+ docstring const trans =
+ translateIfPossible(from_utf8(val));
+ ret << trans;
} else {
docstring const val =
getValueForKey(key, buf, ci, xrefs, max_keysize);
}
-docstring const & BibTeXInfo::getInfo(BibTeXInfoList const xrefs,
- Buffer const & buf, CiteItem const & ci) const
+docstring const & BibTeXInfo::getInfo(BibTeXInfoList const & xrefs,
+ Buffer const & buf, CiteItem const & ci, docstring const & format_in) const
{
bool const richtext = ci.richtext;
- if (!richtext && !info_.empty())
+ CiteEngineType const engine_type = buf.params().citeEngineType();
+ DocumentClass const & dc = buf.params().documentClass();
+ docstring const & format = format_in.empty()?
+ from_utf8(dc.getCiteFormat(engine_type, to_utf8(entry_type_)))
+ : format_in;
+
+ if (format != format_) {
+ // clear caches since format changed
+ info_.clear();
+ info_richtext_.clear();
+ format_ = format;
+ }
+
+ if (!richtext && !info_.empty()) {
+ info_ = convertLaTeXCommands(processRichtext(info_, false));
return info_;
+ }
if (richtext && !info_richtext_.empty())
return info_richtext_;
return info_;
}
- CiteEngineType const engine_type = buf.params().citeEngineType();
- DocumentClass const & dc = buf.params().documentClass();
- docstring const & format =
- from_utf8(dc.getCiteFormat(engine_type, to_utf8(entry_type_)));
int counter = 0;
info_ = expandFormat(format, xrefs, counter, buf,
ci, false, false);
}
-docstring const BibTeXInfo::getLabel(BibTeXInfoList const xrefs,
+docstring const BibTeXInfo::getLabel(BibTeXInfoList const & xrefs,
Buffer const & buf, docstring const & format,
CiteItem const & ci, bool next, bool second) const
{
docstring BibTeXInfo::getValueForKey(string const & oldkey, Buffer const & buf,
- CiteItem const & ci, BibTeXInfoList const xrefs, size_t maxsize) const
+ CiteItem const & ci, BibTeXInfoList const & xrefs, size_t maxsize) const
{
// anything less is pointless
LASSERT(maxsize >= 16, maxsize = 16);
}
docstring ret = operator[](key);
- if (ret.empty() && !xrefs.empty()) {
- vector<BibTeXInfo const *>::const_iterator it = xrefs.begin();
- vector<BibTeXInfo const *>::const_iterator en = xrefs.end();
- for (; it != en; ++it) {
- if (*it && !(**it)[key].empty()) {
- ret = (**it)[key];
- break;
- }
- }
- }
if (ret.empty()) {
+ docstring subtype;
+ if (contains(key, ':'))
+ subtype = from_ascii(token(key, ':', 1));
// some special keys
// FIXME: dialog, textbefore and textafter have nothing to do with this
if (key == "dialog" && ci.context == CiteItem::Dialog)
ret = cite_number_;
else if (prefixIs(key, "ifmultiple:")) {
// Return whether we have multiple authors
- docstring const kind = operator[](from_ascii(key.substr(11)));
+ docstring const kind = operator[](subtype);
if (multipleAuthors(kind))
ret = from_ascii("x"); // any non-empty string will do
}
// Special key to provide abbreviated name list,
// with respect to maxcitenames. Suitable for Bibliography
// beginnings.
- docstring const kind = operator[](from_ascii(key.substr(11)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, false, false, true);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
} else if (prefixIs(key, "fullnames:")) {
// Return a full name list. Suitable for Bibliography
// beginnings.
- docstring const kind = operator[](from_ascii(key.substr(10)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, true, false, true);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
// Special key to provide abbreviated name lists,
// irrespective of maxcitenames. Suitable for Bibliography
// beginnings.
- docstring const kind = operator[](from_ascii(key.substr(15)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, false, true, true);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
// Special key to provide abbreviated name list,
// with respect to maxcitenames. Suitable for further names inside a
// bibliography item // (such as "ed. by ...")
- docstring const kind = operator[](from_ascii(key.substr(11)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, false, false, true, false);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
} else if (prefixIs(key, "fullbynames:")) {
// Return a full name list. Suitable for further names inside a
// bibliography item // (such as "ed. by ...")
- docstring const kind = operator[](from_ascii(key.substr(10)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, true, false, true, false);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
// Special key to provide abbreviated name lists,
// irrespective of maxcitenames. Suitable for further names inside a
// bibliography item // (such as "ed. by ...")
- docstring const kind = operator[](from_ascii(key.substr(15)));
+ docstring const kind = operator[](subtype);
ret = getAuthorList(&buf, kind, false, true, true, false);
if (ci.forceUpperCase && isLowerCase(ret[0]))
ret[0] = uppercase(ret[0]);
ret = ci.textBefore;
else if (key == "textafter")
ret = ci.textAfter;
- else if (key == "curpretext")
- ret = ci.getPretexts()[bib_key_];
- else if (key == "curposttext")
- ret = ci.getPosttexts()[bib_key_];
- else if (key == "year")
+ else if (key == "curpretext") {
+ vector<pair<docstring, docstring>> pres = ci.getPretexts();
+ vector<pair<docstring, docstring>>::iterator it = pres.begin();
+ int numkey = 1;
+ for (; it != pres.end() ; ++it) {
+ if ((*it).first == bib_key_ && numkey == num_bib_key_) {
+ ret = (*it).second;
+ pres.erase(it);
+ break;
+ }
+ if ((*it).first == bib_key_)
+ ++numkey;
+ }
+ } else if (key == "curposttext") {
+ vector<pair<docstring, docstring>> posts = ci.getPosttexts();
+ vector<pair<docstring, docstring>>::iterator it = posts.begin();
+ int numkey = 1;
+ for (; it != posts.end() ; ++it) {
+ if ((*it).first == bib_key_ && numkey == num_bib_key_) {
+ ret = (*it).second;
+ posts.erase(it);
+ break;
+ }
+ if ((*it).first == bib_key_)
+ ++numkey;
+ }
+ } else if (key == "year")
ret = getYear();
}
+ // If we have no result, check in the cross-ref'ed entries
+ if (ret.empty() && !xrefs.empty()) {
+ bool const biblatex =
+ buf.params().documentClass().citeFramework() == "biblatex";
+ // xr is a (reference to a) BibTeXInfo const *
+ for (auto const & xr : xrefs) {
+ if (!xr)
+ continue;
+ // use empty BibTeXInfoList to avoid loops
+ BibTeXInfoList xr_dummy;
+ ret = xr->getValueForKey(oldkey, buf, ci, xr_dummy, maxsize);
+ if (!ret.empty())
+ // success!
+ break;
+ // in biblatex, cross-ref'ed titles are mapped
+ // to booktitle. Same for subtitle etc.
+ if (biblatex && prefixIs(key, "book"))
+ ret = (*xr)[key.substr(4)];
+ // likewise, author is maped onto bookauthor
+ else if (biblatex && contains(key, ":bookauthor"))
+ ret = xr->getValueForKey(subst(key, "bookauthor", "author"),
+ buf, ci, xr_dummy, maxsize);
+ if (!ret.empty())
+ // success!
+ break;
+ }
+ }
+
if (cleanit)
- ret = html::cleanAttr(ret);
+ ret = xml::cleanAttr(ret);
// make sure it is not too big
support::truncateWithEllipsis(ret, maxsize);
namespace {
// A functor for use with sort, leading to case insensitive sorting
-class compareNoCase: public binary_function<docstring, docstring, bool>
-{
-public:
- bool operator()(docstring const & s1, docstring const & s2) const {
- return compare_no_case(s1, s2) < 0;
- }
-};
+bool compareNoCase(const docstring & a, const docstring & b) {
+ return compare_no_case(a, b) < 0;
+}
-} // namespace anon
+} // namespace
vector<docstring> const BiblioInfo::getXRefs(BibTeXInfo const & data, bool const nested) const
// XData field can consist of a comma-separated list of keys
vector<docstring> const xdatakeys = getVectorFromString(data["xdata"]);
if (!xdatakeys.empty()) {
- vector<docstring>::const_iterator xit = xdatakeys.begin();
- vector<docstring>::const_iterator xen = xdatakeys.end();
- for (; xit != xen; ++xit) {
- docstring const xdatakey = *xit;
+ for (auto const & xdatakey : xdatakeys) {
result.push_back(xdatakey);
BiblioInfo::const_iterator it = find(xdatakey);
if (it != end()) {
vector<docstring> const BiblioInfo::getKeys() const
{
vector<docstring> bibkeys;
- BiblioInfo::const_iterator it = begin();
- for (; it != end(); ++it)
- bibkeys.push_back(it->first);
- sort(bibkeys.begin(), bibkeys.end(), compareNoCase());
+ for (auto const & bi : *this)
+ bibkeys.push_back(bi.first);
+ sort(bibkeys.begin(), bibkeys.end(), &compareNoCase);
return bibkeys;
}
vector<docstring> const BiblioInfo::getFields() const
{
vector<docstring> bibfields;
- set<docstring>::const_iterator it = field_names_.begin();
- set<docstring>::const_iterator end = field_names_.end();
- for (; it != end; ++it)
- bibfields.push_back(*it);
+ for (auto const & fn : field_names_)
+ bibfields.push_back(fn);
sort(bibfields.begin(), bibfields.end());
return bibfields;
}
vector<docstring> const BiblioInfo::getEntries() const
{
vector<docstring> bibentries;
- set<docstring>::const_iterator it = entry_types_.begin();
- set<docstring>::const_iterator end = entry_types_.end();
- for (; it != end; ++it)
- bibentries.push_back(*it);
+ for (auto const & et : entry_types_)
+ bibentries.push_back(et);
sort(bibentries.begin(), bibentries.end());
return bibentries;
}
return data.citeNumber();
}
+void BiblioInfo::getLocators(docstring const & key, docstring & doi, docstring & url, docstring & file) const
+{
+ BiblioInfo::const_iterator it = find(key);
+ if (it == end())
+ return;
+ BibTeXInfo const & data = it->second;
+ data.getLocators(doi,url,file);
+}
+
docstring const BiblioInfo::getYear(docstring const & key, bool use_modifier) const
{
if (xrefs.empty())
// no luck
return docstring();
- vector<docstring>::const_iterator it = xrefs.begin();
- vector<docstring>::const_iterator en = xrefs.end();
- for (; it != en; ++it) {
- BiblioInfo::const_iterator const xrefit = find(*it);
+ for (docstring const & xref : xrefs) {
+ BiblioInfo::const_iterator const xrefit = find(xref);
if (xrefit == end())
continue;
BibTeXInfo const & xref_data = xrefit->second;
docstring const BiblioInfo::getInfo(docstring const & key,
- Buffer const & buf, CiteItem const & ci) const
+ Buffer const & buf, CiteItem const & ci, docstring const & format) const
{
BiblioInfo::const_iterator it = find(key);
if (it == end())
- return docstring(_("Bibliography entry not found!"));
+ return _("Bibliography entry not found!");
BibTeXInfo const & data = it->second;
BibTeXInfoList xrefptrs;
- vector<docstring> const xrefs = getXRefs(data);
- if (!xrefs.empty()) {
- vector<docstring>::const_iterator it = xrefs.begin();
- vector<docstring>::const_iterator en = xrefs.end();
- for (; it != en; ++it) {
- BiblioInfo::const_iterator const xrefit = find(*it);
- if (xrefit != end())
- xrefptrs.push_back(&(xrefit->second));
- }
+ for (docstring const & xref : getXRefs(data)) {
+ BiblioInfo::const_iterator const xrefit = find(xref);
+ if (xrefit != end())
+ xrefptrs.push_back(&(xrefit->second));
}
- return data.getInfo(xrefptrs, buf, ci);
+ return data.getInfo(xrefptrs, buf, ci, format);
}
LASSERT(max_size >= 16, max_size = 16);
// we can't display more than 10 of these, anyway
+ // but since we truncate in the middle,
+ // we need to split into two halfs.
bool const too_many_keys = keys.size() > 10;
- if (too_many_keys)
- keys.resize(10);
+ vector<docstring> lkeys;
+ if (too_many_keys) {
+ lkeys.insert(lkeys.end(), keys.end() - 5, keys.end());
+ keys.resize(5);
+ keys.insert(keys.end(), lkeys.begin(), lkeys.end());
+ }
CiteEngineType const engine_type = buf.params().citeEngineType();
DocumentClass const & dc = buf.params().documentClass();
docstring ret = format;
vector<docstring>::const_iterator key = keys.begin();
vector<docstring>::const_iterator ken = keys.end();
+ vector<docstring> handled_keys;
for (int i = 0; key != ken; ++key, ++i) {
+ handled_keys.push_back(*key);
+ int n = 0;
+ for (auto const & k : handled_keys) {
+ if (k == *key)
+ ++n;
+ }
BiblioInfo::const_iterator it = find(*key);
BibTeXInfo empty_data;
empty_data.key(*key);
vector<BibTeXInfo const *> xrefptrs;
if (it != end()) {
data = it->second;
- vector<docstring> const xrefs = getXRefs(data);
- if (!xrefs.empty()) {
- vector<docstring>::const_iterator it = xrefs.begin();
- vector<docstring>::const_iterator en = xrefs.end();
- for (; it != en; ++it) {
- BiblioInfo::const_iterator const xrefit = find(*it);
- if (xrefit != end())
- xrefptrs.push_back(&(xrefit->second));
- }
+ for (docstring const & xref : getXRefs(data)) {
+ BiblioInfo::const_iterator const xrefit = find(xref);
+ if (xrefit != end())
+ xrefptrs.push_back(&(xrefit->second));
}
}
+ data.numKey(n);
ret = data.getLabel(xrefptrs, buf, ret, ci, key + 1 != ken, i == 1);
}
- if (too_many_keys)
- ret.push_back(0x2026);//HORIZONTAL ELLIPSIS
- support::truncateWithEllipsis(ret, max_size);
+ support::truncateWithEllipsis(ret, max_size, true);
+
return ret;
}
}
-vector<docstring> const BiblioInfo::getCiteStrings(
+BiblioInfo::CiteStringMap const BiblioInfo::getCiteStrings(
vector<docstring> const & keys, vector<CitationStyle> const & styles,
Buffer const & buf, CiteItem const & ci) const
{
if (empty())
- return vector<docstring>();
+ return vector<pair<docstring,docstring>>();
string style;
- vector<docstring> vec(styles.size());
- for (size_t i = 0; i != vec.size(); ++i) {
+ CiteStringMap csm(styles.size());
+ for (size_t i = 0; i != csm.size(); ++i) {
style = styles[i].name;
- vec[i] = getLabel(keys, buf, style, ci);
+ csm[i] = make_pair(from_ascii(style), getLabel(keys, buf, style, ci));
}
- return vec;
+ return csm;
}
|| (lauth == rauth && lyear == ryear && ltitl < rtitl);
}
-}
+} // namespace
void BiblioInfo::collectCitedEntries(Buffer const & buf)
// FIXME We may want to collect these differently, in the first case,
// so that we might have them in order of appearance.
set<docstring> citekeys;
- shared_ptr<Toc const> toc = buf.tocBackend().toc("citation");
- Toc::const_iterator it = toc->begin();
- Toc::const_iterator const en = toc->end();
- for (; it != en; ++it) {
- if (it->str().empty())
+ Toc const & toc = *buf.tocBackend().toc("citation");
+ for (auto const & t : toc) {
+ if (t.str().empty())
continue;
- vector<docstring> const keys = getVectorFromString(it->str());
+ vector<docstring> const keys = getVectorFromString(t.str());
citekeys.insert(keys.begin(), keys.end());
}
if (citekeys.empty())
// We will now convert it to a list of the BibTeXInfo objects used in
// this document...
vector<BibTeXInfo const *> bi;
- set<docstring>::const_iterator cit = citekeys.begin();
- set<docstring>::const_iterator const cen = citekeys.end();
- for (; cit != cen; ++cit) {
- BiblioInfo::const_iterator const bt = find(*cit);
+ for (auto const & ck : citekeys) {
+ BiblioInfo::const_iterator const bt = find(ck);
if (bt == end() || !bt->second.isBibTeX())
continue;
bi.push_back(&(bt->second));
sort(bi.begin(), bi.end(), lSorter);
// Now we can write the sorted keys
- vector<BibTeXInfo const *>::const_iterator bit = bi.begin();
- vector<BibTeXInfo const *>::const_iterator ben = bi.end();
- for (; bit != ben; ++bit)
- cited_entries_.push_back((*bit)->key());
+ // b is a BibTeXInfo const *
+ for (auto const & b : bi)
+ cited_entries_.push_back(b->key());
}
// modifiers, like "1984a"
map<docstring, BibTeXInfo>::iterator last = bimap_.end();
- vector<docstring>::const_iterator it = cited_entries_.begin();
- vector<docstring>::const_iterator const en = cited_entries_.end();
- for (; it != en; ++it) {
- map<docstring, BibTeXInfo>::iterator const biit = bimap_.find(*it);
+ // add letters to years
+ for (auto const & ce : cited_entries_) {
+ map<docstring, BibTeXInfo>::iterator const biit = bimap_.find(ce);
// this shouldn't happen, but...
if (biit == bimap_.end())
// ...fail gracefully, anyway.
}
}
// Set the labels
- it = cited_entries_.begin();
- for (; it != en; ++it) {
- map<docstring, BibTeXInfo>::iterator const biit = bimap_.find(*it);
+ for (auto const & ce : cited_entries_) {
+ map<docstring, BibTeXInfo>::iterator const biit = bimap_.find(ce);
// this shouldn't happen, but...
if (biit == bimap_.end())
// ...fail gracefully, anyway.
docstring const auth = entry.getAuthorOrEditorList(&buf, false);
// we do it this way so as to access the xref, if necessary
// note that this also gives us the modifier
- docstring const year = getYear(*it, buf, true);
+ docstring const year = getYear(ce, buf, true);
if (!auth.empty() && !year.empty())
entry.label(auth + ' ' + year);
else
return cmd;
}
+
+void authorsToDocBookAuthorGroup(docstring const & authorsString, XMLStream & xs, Buffer const & buf,
+ const std::string type)
+{
+ // This function closely mimics getAuthorList, but produces DocBook instead of text.
+ // It has been greatly simplified, as the complete list of authors is always produced. No separators are required,
+ // as the output has a database-like shape.
+ // constructName has also been merged within, as it becomes really simple and leads to no copy-paste.
+
+ if (! type.empty() && (type != "author" && type != "book")) {
+ LYXERR0("ERROR! Unexpected author contribution `" << type <<"'.");
+ return;
+ }
+
+ if (authorsString.empty()) {
+ return;
+ }
+
+ // Split the input list of authors into individual authors.
+ vector<docstring> const authors = getAuthors(authorsString);
+
+ // Retrieve the "et al." variation.
+ string const etal = buf.params().documentClass().getCiteMacro(buf.params().citeEngineType(), "_etal");
+
+ // Output the list of authors.
+ xs << xml::StartTag("authorgroup");
+ xs << xml::CR();
+
+ auto it = authors.cbegin();
+ auto en = authors.cend();
+ for (size_t i = 0; it != en; ++it, ++i) {
+ const std::string tag = (type.empty() || type == "author") ? "author" : "othercredit";
+ const std::string attr = (type == "book") ? R"(class="other" otherclass="bookauthor")" : "";
+
+ xs << xml::StartTag(tag, attr);
+ xs << xml::CR();
+ xs << xml::StartTag("personname");
+ xs << xml::CR();
+ const docstring name = *it;
+
+ // All authors go in a <personname>. If more structure is known, use it; otherwise (just "et al."),
+ // print it as such.
+ if (name == "others") {
+ xs << buf.B_(etal);
+ } else {
+ name_parts parts = nameParts(name);
+ if (! parts.prefix.empty()) {
+ xs << xml::StartTag("honorific");
+ xs << parts.prefix;
+ xs << xml::EndTag("honorific");
+ xs << xml::CR();
+ }
+ if (! parts.prename.empty()) {
+ xs << xml::StartTag("firstname");
+ xs << parts.prename;
+ xs << xml::EndTag("firstname");
+ xs << xml::CR();
+ }
+ if (! parts.surname.empty()) {
+ xs << xml::StartTag("surname");
+ xs << parts.surname;
+ xs << xml::EndTag("surname");
+ xs << xml::CR();
+ }
+ if (! parts.suffix.empty()) {
+ xs << xml::StartTag("othername", "role=\"suffix\"");
+ xs << parts.suffix;
+ xs << xml::EndTag("othername");
+ xs << xml::CR();
+ }
+ }
+
+ xs << xml::EndTag("personname");
+ xs << xml::CR();
+ xs << xml::EndTag(tag);
+ xs << xml::CR();
+
+ // Could add an affiliation after <personname>, but not stored in BibTeX.
+ }
+ xs << xml::EndTag("authorgroup");
+ xs << xml::CR();
+}
+
} // namespace lyx
projects / lyx.git / blobdiff