X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FBiblioInfo.cpp;h=420eaa2177c87f9d1d05f92d939565fd60313136;hb=f5123ccfa82215030c70561af881175d92244492;hp=81234b89d873aea7bc1e8630cd4468b043307d4d;hpb=782db81f771cb12977f19b050fb2cbaa8511e16e;p=lyx.git diff --git a/src/BiblioInfo.cpp b/src/BiblioInfo.cpp index 81234b89d8..420eaa2177 100644 --- a/src/BiblioInfo.cpp +++ b/src/BiblioInfo.cpp @@ -18,7 +18,9 @@ #include "buffer_funcs.h" #include "Encoding.h" #include "InsetIterator.h" +#include "Language.h" #include "Paragraph.h" +#include "TextClass.h" #include "TocBackend.h" #include "insets/Inset.h" @@ -27,13 +29,14 @@ #include "insets/InsetInclude.h" #include "support/convert.h" +#include "support/debug.h" #include "support/docstream.h" #include "support/gettext.h" #include "support/lassert.h" #include "support/lstrings.h" #include "support/textutils.h" -#include "boost/regex.hpp" +#include "support/regex.h" #include @@ -172,8 +175,8 @@ docstring convertLaTeXCommands(docstring const & str) // 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)) { + static lyx::regex const reg("^\\\\\\W\\w"); + if (lyx::regex_search(to_utf8(val), reg)) { val.insert(3, from_ascii("}")); val.insert(2, from_ascii("{")); } @@ -217,7 +220,11 @@ docstring const BibTeXInfo::getAbbreviatedAuthor() const return docstring(); docstring authors; - split(opt, authors, '('); + docstring const remainder = trim(split(opt, authors, '(')); + if (remainder.empty()) + // in this case, we didn't find a "(", + // so we don't have author (year) + return docstring(); return authors; } @@ -228,6 +235,9 @@ docstring const BibTeXInfo::getAbbreviatedAuthor() const return bib_key_; } + // 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 const authors = @@ -254,9 +264,12 @@ docstring const BibTeXInfo::getYear() const return docstring(); docstring authors; - docstring const tmp = split(opt, authors, '('); + docstring tmp = split(opt, authors, '('); + if (tmp.empty()) + // we don't have author (year) + return docstring(); docstring year; - split(tmp, year, ')'); + tmp = split(tmp, year, ')'); return year; } @@ -269,7 +282,247 @@ docstring const BibTeXInfo::getXRef() const } -docstring const & BibTeXInfo::getInfo(BibTeXInfo const * const xref) const +namespace { + string parseOptions(string const & format, string & optkey, + string & ifpart, string & elsepart); + + // Calls parseOptions to deal with an embedded option, such as: + // {%number%[[, no.~%number%]]} + // which must appear at the start of format. ifelsepart gets the + // whole of the option, and we return what's left after the option. + // we return format if there is an error. + string parseEmbeddedOption(string const & format, string & ifelsepart) + { + LASSERT(format[0] == '{' && format[1] == '%', return format); + string optkey; + string ifpart; + string elsepart; + string const rest = parseOptions(format, optkey, ifpart, elsepart); + if (format == rest) { // parse error + LYXERR0("ERROR! Couldn't parse `" << format <<"'."); + return format; + } + LASSERT(rest.size() <= format.size(), /* */); + ifelsepart = format.substr(0, format.size() - rest.size()); + return rest; + } + + + // Gets a "clause" from a format string, where the clause is + // delimited by '[[' and ']]'. Returns what is left after the + // clause is removed, and returns format if there is an error. + string getClause(string const & format, string & clause) + { + string fmt = format; + // remove '[[' + fmt = fmt.substr(2); + // we'll remove characters from the front of fmt as we + // deal with them + while (fmt.size()) { + if (fmt[0] == ']' && fmt.size() > 1 && fmt[1] == ']') { + // that's the end + fmt = fmt.substr(2); + break; + } + // check for an embedded option + if (fmt[0] == '{' && fmt.size() > 1 && fmt[1] == '%') { + string part; + string const rest = parseEmbeddedOption(fmt, part); + if (fmt == rest) { + LYXERR0("ERROR! Couldn't parse embedded option in `" << format <<"'."); + return format; + } + clause += part; + fmt = rest; + } else { // it's just a normal character + clause += fmt[0]; + fmt = fmt.substr(1); + } + } + return fmt; + } + + + // parse an options string, which must appear at the start of the + // format parameter. puts the parsed bits in optkey, ifpart, and + // elsepart and returns what's left after the option is removed. + // if there's an error, it returns format itself. + string parseOptions(string const & format, string & optkey, + string & ifpart, string & elsepart) + { + LASSERT(format[0] == '{' && format[1] == '%', return format); + // strip '{%' + string fmt = format.substr(2); + size_t pos = fmt.find('%'); // end of key + if (pos == string::npos) { + LYXERR0("Error parsing `" << format <<"'. Can't find end of key."); + return format; + } + optkey = fmt.substr(0,pos); + fmt = fmt.substr(pos + 1); + // [[format]] should be next + if (fmt[0] != '[' || fmt[1] != '[') { + LYXERR0("Error parsing `" << format <<"'. Can't find '[[' after key."); + return format; + } + + string curfmt = fmt; + fmt = getClause(curfmt, ifpart); + if (fmt == curfmt) { + LYXERR0("Error parsing `" << format <<"'. Couldn't get if clause."); + return format; + } + + if (fmt[0] == '}') // we're done, no else clause + return fmt.substr(1); + + // else part should follow + if (fmt[0] != '[' || fmt[1] != '[') { + LYXERR0("Error parsing `" << format <<"'. Can't find else clause."); + return format; + } + + curfmt = fmt; + fmt = getClause(curfmt, elsepart); + // we should be done + if (fmt == curfmt || fmt[0] != '}') { + LYXERR0("Error parsing `" << format <<"'. Can't find end of option."); + return format; + } + return fmt.substr(1); +} + +} // anon namespace + + +docstring BibTeXInfo::expandFormat(string const & format, + BibTeXInfo const * const xref, int & counter, Buffer const & buf, + bool richtext) const +{ + // incorrect use of macros could put us in an infinite loop + static int max_passes = 5000; + docstring ret; // return value + string key; + bool scanning_key = false; + bool scanning_rich = false; + + string fmt = format; + // we'll remove characters from the front of fmt as we + // deal with them + while (fmt.size()) { + if (counter++ > max_passes) { + LYXERR0("Recursion limit reached while parsing `" + << format << "'."); + return _("ERROR!"); + } + + char_type thischar = fmt[0]; + if (thischar == '%') { + // beginning or end of key + if (scanning_key) { + // end of key + scanning_key = false; + // so we replace the key with its value, which may be empty + if (key[0] == '!') { + // macro + string const val = + buf.params().documentClass().getCiteMacro(key); + fmt = val + fmt.substr(1); + continue; + } else if (key[0] == '_') { + // a translatable bit + string const val = + buf.params().documentClass().getCiteMacro(key); + docstring const trans = + translateIfPossible(from_utf8(val), buf.params().language->code()); + ret += trans; + } else { + docstring const val = getValueForKey(key, xref); + ret += val; + } + } else { + // beginning of key + key.clear(); + scanning_key = true; + } + } + else if (thischar == '{') { + // beginning of option? + if (scanning_key) { + LYXERR0("ERROR: Found `{' when scanning key in `" << format << "'."); + return _("ERROR!"); + } + if (fmt.size() > 1) { + if (fmt[1] == '%') { + // it is the beginning of an optional format + string optkey; + string ifpart; + string elsepart; + string const newfmt = + parseOptions(fmt, optkey, ifpart, elsepart); + if (newfmt == fmt) // parse error + return _("ERROR!"); + fmt = newfmt; + docstring const val = getValueForKey(optkey, xref); + if (!val.empty()) + ret += expandFormat(ifpart, xref, counter, buf, richtext); + else if (!elsepart.empty()) + ret += expandFormat(elsepart, xref, counter, buf, richtext); + // fmt will have been shortened for us already + continue; + } + if (fmt[1] == '!') { + // beginning of rich text + scanning_rich = true; + fmt = fmt.substr(2); + continue; + } + } + // we are here if '{' was not followed by % or !. + // So it's just a character. + ret += thischar; + } + else if (scanning_rich && thischar == '!' + && fmt.size() > 1 && fmt[1] == '}') { + // end of rich text + scanning_rich = false; + fmt = fmt.substr(2); + continue; + } + else if (scanning_key) + key += char(thischar); + else if (richtext) { + if (scanning_rich) + ret += thischar; + else { + // we need to escape '<' and '>' + if (thischar == '<') + ret += "<"; + else if (thischar == '>') + ret += ">"; + else + ret += thischar; + } + } else if (!scanning_rich /* && !richtext */) + ret += thischar; + // else the character is discarded, which will happen only if + // richtext == false and we are scanning rich text + fmt = fmt.substr(1); + } // for loop + if (scanning_key) { + LYXERR0("Never found end of key in `" << format << "'!"); + return _("ERROR!"); + } + if (scanning_rich) { + LYXERR0("Never found end of rich text in `" << format << "'!"); + return _("ERROR!"); + } + return ret; +} + + +docstring const & BibTeXInfo::getInfo(BibTeXInfo const * const xref, + Buffer const & buf, bool richtext) const { if (!info_.empty()) return info_; @@ -279,59 +532,15 @@ docstring const & BibTeXInfo::getInfo(BibTeXInfo const * const xref) const info_ = it->second; return info_; } - - // FIXME - // 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 = getValueForKey("author", xref); - if (author.empty()) - author = getValueForKey("editor", xref); - - docstring year = getValueForKey("year", xref); - docstring title = getValueForKey("title", xref); - docstring docLoc = getValueForKey("pages", xref); - if (docLoc.empty()) { - docLoc = getValueForKey("chapter", xref); - if (!docLoc.empty()) - docLoc = _("Ch. ") + docLoc; - } else { - docLoc = _("pp. ") + docLoc; - } - docstring media = getValueForKey("journal", xref); - if (media.empty()) { - media = getValueForKey("publisher", xref); - if (media.empty()) { - media = getValueForKey("school", xref); - if (media.empty()) - media = getValueForKey("institution"); - } - } - docstring volume = getValueForKey("volume", xref); - - odocstringstream result; - if (!author.empty()) - result << author << ", "; - if (!title.empty()) - result << title; - if (!media.empty()) - result << ", " << media; - if (!year.empty()) - result << " (" << year << ")"; - if (!docLoc.empty()) - result << ", " << docLoc; - - docstring const result_str = rtrim(result.str()); - if (!result_str.empty()) { - info_ = convertLaTeXCommands(result_str); - return info_; - } + DocumentClass const & dc = buf.params().documentClass(); + string const & format = dc.getCiteFormat(to_utf8(entry_type_)); + int counter = 0; + info_ = expandFormat(format, xref, counter, buf, richtext); - // This should never happen (or at least be very unusual!) - static docstring e = docstring(); - return e; + if (!info_.empty()) + info_ = convertLaTeXCommands(info_); + return info_; } @@ -458,7 +667,8 @@ docstring const BiblioInfo::getYear(docstring const & key, bool use_modifier) co } -docstring const BiblioInfo::getInfo(docstring const & key) const +docstring const BiblioInfo::getInfo(docstring const & key, + Buffer const & buf, bool richtext) const { BiblioInfo::const_iterator it = find(key); if (it == end()) @@ -471,10 +681,20 @@ docstring const BiblioInfo::getInfo(docstring const & key) const if (xrefit != end()) xrefptr = &(xrefit->second); } - return data.getInfo(xrefptr); + return data.getInfo(xrefptr, buf, richtext); } +bool BiblioInfo::isBibtex(docstring const & key) const +{ + BiblioInfo::const_iterator it = find(key); + if (it == end()) + return false; + return it->second.isBibTeX(); +} + + + vector const BiblioInfo::getCiteStrings( docstring const & key, Buffer const & buf) const {