#include "Buffer.h"
#include "BufferParams.h"
#include "buffer_funcs.h"
+#include "Encoding.h"
#include "InsetIterator.h"
#include "Paragraph.h"
#include "support/gettext.h"
#include "support/lassert.h"
#include "support/lstrings.h"
+#include "support/textutils.h"
#include "boost/regex.hpp"
namespace lyx {
+namespace {
+
+// gets the "family name" from an author-type string
+docstring familyName(docstring const & name)
+{
+ if (name.empty())
+ return docstring();
+
+ // 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.
+ docstring::size_type idx = name.rfind(',');
+ if (idx != docstring::npos)
+ return ltrim(name.substr(0, idx));
+
+ // OK, so now we want to look for the last name. We're going to
+ // include the "von" part. This isn't perfect.
+ // Split on spaces, to get various tokens.
+ vector<docstring> pieces = getVectorFromString(name, from_ascii(" "));
+ // If we only get two, assume the last one is the last name
+ if (pieces.size() <= 2)
+ return pieces.back();
+
+ // Now we look for the first token that begins with a lower case letter.
+ vector<docstring>::const_iterator it = pieces.begin();
+ vector<docstring>::const_iterator en = pieces.end();
+ for (; it != en; ++it) {
+ if ((*it).size() == 0)
+ continue;
+ char_type const c = (*it)[0];
+ if (isLower(c))
+ break;
+ }
+
+ if (it == en) // we never found a "von"
+ return pieces.back();
+
+ // reconstruct what we need to return
+ docstring retval;
+ bool first = true;
+ for (; it != en; ++it) {
+ if (!first)
+ retval += " ";
+ else
+ first = false;
+ retval += *it;
+ }
+ return retval;
+}
+
+// converts a string containing LaTeX commands into unicode
+// for display.
+docstring convertLaTeXCommands(docstring const & str)
+{
+ docstring val = str;
+ docstring ret;
+
+ bool scanning_cmd = false;
+ bool scanning_math = false;
+ bool escaped = false; // used to catch \$, etc.
+ while (val.size()) {
+ char_type const ch = val[0];
+
+ // if we're scanning math, we output everything until we
+ // find an unescaped $, at which point we break out.
+ if (scanning_math) {
+ if (escaped)
+ escaped = false;
+ else if (ch == '\\')
+ escaped = true;
+ else if (ch == '$')
+ scanning_math = false;
+ ret += ch;
+ val = val.substr(1);
+ continue;
+ }
+
+ // if we're scanning a command name, then we just
+ // discard characters until we hit something that
+ // isn't alpha.
+ if (scanning_cmd) {
+ if (isAlphaASCII(ch)) {
+ val = val.substr(1);
+ escaped = false;
+ continue;
+ }
+ // so we're done with this command.
+ // now we fall through and check this character.
+ scanning_cmd = false;
+ }
+
+ // was the last character a \? If so, then this is something like: \\,
+ // or \$, so we'll just output it. That's probably not always right...
+ if (escaped) {
+ ret += ch;
+ val = val.substr(1);
+ escaped = false;
+ continue;
+ }
+
+ if (ch == '$') {
+ ret += ch;
+ val = val.substr(1);
+ scanning_math = true;
+ continue;
+ }
+
+ // we just ignore braces
+ if (ch == '{' || ch == '}') {
+ val = val.substr(1);
+ continue;
+ }
+
+ // we're going to check things that look like commands, so if
+ // this doesn't, just output it.
+ if (ch != '\\') {
+ ret += ch;
+ val = val.substr(1);
+ continue;
+ }
+
+ // ok, could be a command of some sort
+ // let's see if it corresponds to some unicode
+ // unicodesymbols has things in the form: \"{u},
+ // whereas we may see things like: \"u. So we'll
+ // look for that and change it, if necessary.
+ static boost::regex const reg("^\\\\\\W\\w");
+ if (boost::regex_search(to_utf8(val), reg)) {
+ val.insert(3, from_ascii("}"));
+ val.insert(2, from_ascii("{"));
+ }
+ docstring rem;
+ docstring const cnvtd = Encodings::fromLaTeXCommand(val, rem);
+ if (!cnvtd.empty()) {
+ // it did, so we'll take that bit and proceed with what's left
+ ret += cnvtd;
+ val = rem;
+ continue;
+ }
+ // it's a command of some sort
+ scanning_cmd = true;
+ escaped = true;
+ val = val.substr(1);
+ }
+ return ret;
+}
+
+} // anon namespace
+
+
//////////////////////////////////////////////////////////////////////
//
// BibTeXInfo
//////////////////////////////////////////////////////////////////////
BibTeXInfo::BibTeXInfo(docstring const & key, docstring const & type)
- : is_bibtex_(true), bib_key_(key), entry_type_(type)
+ : is_bibtex_(true), bib_key_(key), entry_type_(type), info_()
{}
}
-docstring const & BibTeXInfo::getValueForField(docstring const & field) const
-{
- BibTeXInfo::const_iterator it = find(field);
- if (it != end())
- return it->second;
- static docstring const empty_value = docstring();
- return empty_value;
-}
-
-
-docstring const & BibTeXInfo::getValueForField(string const & field) const
-{
- return getValueForField(from_ascii(field));
-}
-
-
-docstring familyName(docstring const & name)
-{
- if (name.empty())
- return docstring();
-
- // Very simple parser
- docstring fname = name;
-
- // possible authorname combinations are:
- // "Surname, FirstName"
- // "Surname, F."
- // "FirstName Surname"
- // "F. Surname"
- docstring::size_type idx = fname.find(',');
- if (idx != docstring::npos)
- return ltrim(fname.substr(0, idx));
- idx = fname.rfind('.');
- if (idx != docstring::npos && idx + 1 < fname.size())
- fname = ltrim(fname.substr(idx + 1));
- // test if we have a LaTeX Space in front
- if (fname[0] == '\\')
- return fname.substr(2);
- return rtrim(fname);
-}
-
docstring const BibTeXInfo::getAbbreviatedAuthor() const
{
if (!is_bibtex_) {
- docstring const opt = trim(getValueForField("label"));
+ docstring const opt = trim(operator[]("label"));
if (opt.empty())
return docstring();
return authors;
}
- docstring author = getValueForField("author");
+ docstring author = operator[]("author");
if (author.empty()) {
- author = getValueForField("editor");
+ author = operator[]("editor");
if (author.empty())
return bib_key_;
}
docstring const BibTeXInfo::getYear() const
{
- if (!is_bibtex_) {
- docstring const opt = trim(getValueForField("label"));
- if (opt.empty())
- return docstring();
+ if (is_bibtex_)
+ return operator[]("year");
- docstring authors;
- docstring const tmp = split(opt, authors, '(');
- docstring year;
- split(tmp, year, ')');
- return year;
- }
+ docstring const opt = trim(operator[]("label"));
+ if (opt.empty())
+ return docstring();
- docstring year = getValueForField("year");
- if (year.empty())
- year = _("No year");
+ docstring authors;
+ docstring const tmp = split(opt, authors, '(');
+ docstring year;
+ split(tmp, year, ')');
return year;
}
-docstring const BibTeXInfo::getInfo() const
+docstring const BibTeXInfo::getXRef() const
+{
+ if (!is_bibtex_)
+ return docstring();
+ return operator[]("crossref");
+}
+
+
+docstring const & BibTeXInfo::getInfo(BibTeXInfo const * const xref) const
{
+ if (!info_.empty())
+ return info_;
+
if (!is_bibtex_) {
BibTeXInfo::const_iterator it = find(from_ascii("ref"));
- return it->second;
+ info_ = it->second;
+ return info_;
}
// FIXME
- // This could be made a lot better using the entryType
+ // This could be made a lot better using the entry_type_
// field to customize the output based upon entry type.
// Search for all possible "required" fields
- docstring author = getValueForField("author");
+ docstring author = getValueForKey("author", xref);
if (author.empty())
- author = getValueForField("editor");
+ author = getValueForKey("editor", xref);
- docstring year = getValueForField("year");
- docstring title = getValueForField("title");
- docstring docLoc = getValueForField("pages");
+ docstring year = getValueForKey("year", xref);
+ docstring title = getValueForKey("title", xref);
+ docstring docLoc = getValueForKey("pages", xref);
if (docLoc.empty()) {
- docLoc = getValueForField("chapter");
+ docLoc = getValueForKey("chapter", xref);
if (!docLoc.empty())
- docLoc = from_ascii("Ch. ") + docLoc;
+ docLoc = _("Ch. ") + docLoc;
} else {
- docLoc = from_ascii("pp. ") + docLoc;
+ docLoc = _("pp. ") + docLoc;
}
- docstring media = getValueForField("journal");
+ docstring media = getValueForKey("journal", xref);
if (media.empty()) {
- media = getValueForField("publisher");
+ media = getValueForKey("publisher", xref);
if (media.empty()) {
- media = getValueForField("school");
+ media = getValueForKey("school", xref);
if (media.empty())
- media = getValueForField("institution");
+ media = getValueForKey("institution");
}
}
- docstring volume = getValueForField("volume");
+ docstring volume = getValueForKey("volume", xref);
odocstringstream result;
if (!author.empty())
if (!media.empty())
result << ", " << media;
if (!year.empty())
- result << ", " << year;
+ result << " (" << year << ")";
if (!docLoc.empty())
result << ", " << docLoc;
docstring const result_str = rtrim(result.str());
- if (!result_str.empty())
- return result_str;
+ if (!result_str.empty()) {
+ info_ = convertLaTeXCommands(result_str);
+ return info_;
+ }
// This should never happen (or at least be very unusual!)
- return docstring();
+ static docstring e = docstring();
+ return e;
+}
+
+
+docstring const & BibTeXInfo::operator[](docstring const & field) const
+{
+ BibTeXInfo::const_iterator it = find(field);
+ if (it != end())
+ return it->second;
+ static docstring const empty_value = docstring();
+ return empty_value;
+}
+
+
+docstring const & BibTeXInfo::operator[](string const & field) const
+{
+ return operator[](from_ascii(field));
+}
+
+
+docstring BibTeXInfo::getValueForKey(string const & key,
+ BibTeXInfo const * const xref) const
+{
+ docstring const ret = operator[](key);
+ if (!ret.empty() || !xref)
+ return ret;
+ return (*xref)[key];
}
if (it == end())
return docstring();
BibTeXInfo const & data = it->second;
+ docstring year = data.getYear();
+ if (!year.empty())
+ return year;
+ // let's try the crossref
+ docstring const xref = data.getXRef();
+ if (xref.empty())
+ return _("No year"); // no luck
+ BiblioInfo::const_iterator const xrefit = find(xref);
+ if (xrefit == end())
+ return _("No year"); // no luck again
+ BibTeXInfo const & xref_data = xrefit->second;
+ return xref_data.getYear();
return data.getYear();
}
if (it == end())
return docstring();
BibTeXInfo const & data = it->second;
- return data.getInfo();
+ BibTeXInfo const * xrefptr = 0;
+ docstring const xref = data.getXRef();
+ if (!xref.empty()) {
+ BiblioInfo::const_iterator const xrefit = find(xref);
+ if (xrefit != end())
+ xrefptr = &(xrefit->second);
+ }
+ return data.getInfo(xrefptr);
}