X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2Finsets%2FInsetBibtex.cpp;h=876adea54b1770ad15a140a3a8d50eb2890edff5;hb=48b1e8a0aca2f3f3faa8f1f800568e47792ba9a0;hp=13c91b0739e6b5a25cb5b853bd6b6991fbf76cd4;hpb=670efa8f646218f2a378f0cc614c4c37a9f6b89a;p=lyx.git diff --git a/src/insets/InsetBibtex.cpp b/src/insets/InsetBibtex.cpp index 13c91b0739..876adea54b 100644 --- a/src/insets/InsetBibtex.cpp +++ b/src/insets/InsetBibtex.cpp @@ -5,6 +5,7 @@ * * \author Alejandro Aguilar Sierra * \author Richard Heck (BibTeX parser improvements) + * \author Jürgen Spitzmüller * * Full author contact details are available in file CREDITS. */ @@ -16,6 +17,7 @@ #include "BiblioInfo.h" #include "Buffer.h" #include "BufferParams.h" +#include "CiteEnginesList.h" #include "Cursor.h" #include "DispatchResult.h" #include "Encoding.h" @@ -24,20 +26,24 @@ #include "FuncRequest.h" #include "FuncStatus.h" #include "LaTeXFeatures.h" -#include "output_xhtml.h" +#include "output_latex.h" +#include "xml.h" #include "OutputParams.h" #include "PDFOptions.h" #include "texstream.h" #include "TextClass.h" +#include "TocBackend.h" #include "frontends/alert.h" #include "support/convert.h" #include "support/debug.h" #include "support/docstream.h" +#include "support/docstring_list.h" #include "support/ExceptionMessage.h" #include "support/FileNameList.h" #include "support/filetools.h" +#include "support/regex.h" #include "support/gettext.h" #include "support/lstrings.h" #include "support/os.h" @@ -45,6 +51,10 @@ #include "support/textutils.h" #include +#include +#include + +#include using namespace std; using namespace lyx::support; @@ -57,19 +67,7 @@ namespace os = support::os; InsetBibtex::InsetBibtex(Buffer * buf, InsetCommandParams const & p) : InsetCommand(buf, p) -{ - buffer().invalidateBibfileCache(); - buffer().removeBiblioTempFiles(); -} - - -InsetBibtex::~InsetBibtex() -{ - if (isBufferLoaded()) { - buffer().invalidateBibfileCache(); - buffer().removeBiblioTempFiles(); - } -} +{} ParamInfo const & InsetBibtex::findInfo(string const & /* cmdName */) @@ -79,6 +77,9 @@ ParamInfo const & InsetBibtex::findInfo(string const & /* cmdName */) param_info_.add("btprint", ParamInfo::LATEX_OPTIONAL); param_info_.add("bibfiles", ParamInfo::LATEX_REQUIRED); param_info_.add("options", ParamInfo::LYX_INTERNAL); + param_info_.add("encoding", ParamInfo::LYX_INTERNAL); + param_info_.add("file_encodings", ParamInfo::LYX_INTERNAL); + param_info_.add("biblatexopts", ParamInfo::LATEX_OPTIONAL); } return param_info_; } @@ -89,7 +90,7 @@ void InsetBibtex::doDispatch(Cursor & cur, FuncRequest & cmd) switch (cmd.action()) { case LFUN_INSET_EDIT: - editDatabases(); + editDatabases(cmd.argument()); break; case LFUN_INSET_MODIFY: { @@ -110,8 +111,7 @@ void InsetBibtex::doDispatch(Cursor & cur, FuncRequest & cmd) cur.recordUndo(); setParams(p); - buffer().invalidateBibfileCache(); - buffer().removeBiblioTempFiles(); + cur.buffer()->clearBibFileCache(); cur.forceBufferUpdate(); break; } @@ -137,18 +137,19 @@ bool InsetBibtex::getStatus(Cursor & cur, FuncRequest const & cmd, } -void InsetBibtex::editDatabases() const +void InsetBibtex::editDatabases(docstring const & db) const { vector bibfilelist = getVectorFromString(getParam("bibfiles")); if (bibfilelist.empty()) return; - int nr_databases = bibfilelist.size(); - if (nr_databases > 1) { - docstring message = bformat(_("The BibTeX inset includes %1$s databases.\n" + size_t nr_databases = bibfilelist.size(); + if (nr_databases > 1 && db.empty()) { + docstring const engine = usingBiblatex() ? _("Biblatex") : _("BibTeX"); + docstring message = bformat(_("The %1$s[[BibTeX/Biblatex]] inset includes %2$s databases.\n" "If you proceed, all of them will be opened."), - convert(nr_databases)); + engine, convert(nr_databases)); int const ret = Alert::prompt(_("Open Databases?"), message, 0, 1, _("&Cancel"), _("&Proceed")); @@ -159,16 +160,25 @@ void InsetBibtex::editDatabases() const vector::const_iterator it = bibfilelist.begin(); vector::const_iterator en = bibfilelist.end(); for (; it != en; ++it) { - FileName const bibfile = getBibTeXPath(*it, buffer()); - formats.edit(buffer(), bibfile, - formats.getFormatFromFile(bibfile)); + if (!db.empty() && db != *it) + continue; + FileName const bibfile = buffer().getBibfilePath(*it); + theFormats().edit(buffer(), bibfile, + theFormats().getFormatFromFile(bibfile)); } } +bool InsetBibtex::usingBiblatex() const +{ + return buffer().masterParams().useBiblatex(); +} + + docstring InsetBibtex::screenLabel() const { - return _("BibTeX Generated Bibliography"); + return usingBiblatex() ? _("Biblatex Generated Bibliography") + : _("BibTeX Generated Bibliography"); } @@ -195,130 +205,81 @@ docstring InsetBibtex::toolTip(BufferView const & /*bv*/, int /*x*/, int /*y*/) style = split(style, bibtotoc, char_type(',')); } - tip += _("Style File:"); - tip += "
  • " + (style.empty() ? _("none") : style) + "
"; + docstring const btprint = getParam("btprint"); + if (!usingBiblatex()) { + tip += _("Style File:"); + tip += "
  • " + (style.empty() ? _("none") : style) + "
"; - tip += _("Lists:") + " "; - docstring btprint = getParam("btprint"); + tip += _("Lists:") + " "; if (btprint == "btPrintAll") tip += _("all references"); else if (btprint == "btPrintNotCited") tip += _("all uncited references"); else tip += _("all cited references"); - - if (toc) { - tip += ", "; - tip += _("included in TOC"); + if (toc) { + tip += ", "; + tip += _("included in TOC"); + } + if (!buffer().parent() + && buffer().params().multibib == "child") { + tip += "
"; + tip += _("Note: This bibliography is not output, since bibliographies in the master file " + "are not allowed with the setting 'Multiple bibliographies per child document'"); + } + } else { + tip += _("Lists:") + " "; + if (btprint == "bibbysection") + tip += _("all reference units"); + else if (btprint == "btPrintAll") + tip += _("all references"); + else + tip += _("all cited references"); + if (toc) { + tip += ", "; + tip += _("included in TOC"); + } + if (!getParam("biblatexopts").empty()) { + tip += "
"; + tip += _("Options: ") + getParam("biblatexopts"); + } } return tip; } -static string normalizeName(Buffer const & buffer, - OutputParams const & runparams, string const & name, string const & ext) -{ - string const fname = makeAbsPath(name, buffer.filePath()).absFileName(); - if (FileName::isAbsolute(name) || !FileName(fname + ext).isReadableFile()) - return name; - if (!runparams.nice) - return fname; - - // FIXME UNICODE - return to_utf8(makeRelPath(from_utf8(fname), - from_utf8(buffer.masterBuffer()->filePath()))); -} - - void InsetBibtex::latex(otexstream & os, OutputParams const & runparams) const { - // the sequence of the commands: + // The sequence of the commands: + // With normal BibTeX: // 1. \bibliographystyle{style} // 2. \addcontentsline{...} - if option bibtotoc set // 3. \bibliography{database} - // and with bibtopic: + // With bibtopic: // 1. \bibliographystyle{style} // 2. \begin{btSect}{database} // 3. \btPrint{Cited|NotCited|All} // 4. \end{btSect} + // With Biblatex: + // \printbibliography[biblatexopts] + // or + // \bibbysection[biblatexopts] - if btprint is "bibbysection" + + // chapterbib does not allow bibliographies in the master + if (!usingBiblatex() && !runparams.is_child + && buffer().params().multibib == "child") + return; - // Database(s) - // If we are processing the LaTeX file in a temp directory then - // copy the .bib databases to this temp directory, mangling their - // names in the process. Store this mangled name in the list of - // all databases. - // (We need to do all this because BibTeX *really*, *really* - // can't handle "files with spaces" and Windows users tend to - // use such filenames.) - // Otherwise, store the (maybe absolute) path to the original, - // unmangled database name. - vector bibfilelist = getVectorFromString(getParam("bibfiles")); - vector::const_iterator it = bibfilelist.begin(); - vector::const_iterator en = bibfilelist.end(); - odocstringstream dbs; - bool didone = false; - - // determine the export format - string const tex_format = flavor2format(runparams.flavor); - - for (; it != en; ++it) { - string utf8input = to_utf8(*it); - string database = - normalizeName(buffer(), runparams, utf8input, ".bib"); - FileName const try_in_file = - makeAbsPath(database + ".bib", buffer().filePath()); - bool const not_from_texmf = try_in_file.isReadableFile(); - - if (!runparams.inComment && !runparams.dryrun && !runparams.nice && - not_from_texmf) { - // mangledFileName() needs the extension - DocFileName const in_file = DocFileName(try_in_file); - database = removeExtension(in_file.mangledFileName()); - FileName const out_file = makeAbsPath(database + ".bib", - buffer().masterBuffer()->temppath()); - bool const success = in_file.copyTo(out_file); - if (!success) { - lyxerr << "Failed to copy '" << in_file - << "' to '" << out_file << "'" - << endl; - } - } else if (!runparams.inComment && runparams.nice && not_from_texmf) { - runparams.exportdata->addExternalFile(tex_format, try_in_file, database + ".bib"); - if (!isValidLaTeXFileName(database)) { - frontend::Alert::warning(_("Invalid filename"), - _("The following filename will cause troubles " - "when running the exported file through LaTeX: ") + - from_utf8(database)); - } - if (!isValidDVIFileName(database)) { - frontend::Alert::warning(_("Problematic filename for DVI"), - _("The following filename can cause troubles " - "when running the exported file through LaTeX " - "and opening the resulting DVI: ") + - from_utf8(database), true); - } - } - - if (didone) - dbs << ','; - else - didone = true; - // FIXME UNICODE - dbs << from_utf8(latex_path(database)); - } - docstring const db_out = dbs.str(); - - // Post this warning only once. - static bool warned_about_spaces = false; - if (!warned_about_spaces && - runparams.nice && db_out.find(' ') != docstring::npos) { - warned_about_spaces = true; - Alert::warning(_("Export Warning!"), - _("There are spaces in the paths to your BibTeX databases.\n" - "BibTeX will be unable to find them.")); + if (runparams.inDeletedInset) { + // We cannot strike-out bibligraphies, + // so we just output a note. + os << "\\textbf{" + << buffer().B_("[BIBLIOGRAPHY DELETED!]") + << "}"; + return; } - // Style-Options + string style = to_utf8(getParam("options")); // maybe empty! and with bibtotoc string bibtotoc; if (prefixIs(style, "bibtotoc")) { @@ -327,102 +288,127 @@ void InsetBibtex::latex(otexstream & os, OutputParams const & runparams) const style = split(style, bibtotoc, ','); } - if (style == "default") - style = buffer().params().biblio_style; - - if (!style.empty() && !buffer().params().use_bibtopic) { - string base = normalizeName(buffer(), runparams, style, ".bst"); - FileName const try_in_file = - makeAbsPath(base + ".bst", buffer().filePath()); - bool const not_from_texmf = try_in_file.isReadableFile(); - // If this style does not come from texmf and we are not - // exporting to .tex copy it to the tmp directory. - // This prevents problems with spaces and 8bit charcaters - // in the file name. - if (!runparams.inComment && !runparams.dryrun && !runparams.nice && - not_from_texmf) { - // use new style name - DocFileName const in_file = DocFileName(try_in_file); - base = removeExtension(in_file.mangledFileName()); - FileName const out_file = makeAbsPath(base + ".bst", - buffer().masterBuffer()->temppath()); - bool const success = in_file.copyTo(out_file); - if (!success) { - lyxerr << "Failed to copy '" << in_file - << "' to '" << out_file << "'" - << endl; - } - } - // FIXME UNICODE - os << "\\bibliographystyle{" - << from_utf8(latex_path(normalizeName(buffer(), runparams, base, ".bst"))) - << "}\n"; - } - - // Post this warning only once. - static bool warned_about_bst_spaces = false; - if (!warned_about_bst_spaces && runparams.nice && contains(style, ' ')) { - warned_about_bst_spaces = true; - Alert::warning(_("Export Warning!"), - _("There are spaces in the path to your BibTeX style file.\n" - "BibTeX will be unable to find it.")); - } - - if (!db_out.empty() && buffer().params().use_bibtopic) { - os << "\\begin{btSect}"; - if (!style.empty()) - os << "[" << style << "]"; - os << "{" << db_out << "}\n"; - docstring btprint = getParam("btprint"); - if (btprint.empty()) - // default - btprint = from_ascii("btPrintCited"); - os << "\\" << btprint << "\n" - << "\\end{btSect}\n"; - } - - // bibtotoc-Option - if (!bibtotoc.empty() && !buffer().params().use_bibtopic) { - // set label for hyperref, see http://www.lyx.org/trac/ticket/6470 - if (buffer().params().pdfoptions().use_hyperref) - os << "\\phantomsection"; - if (buffer().params().documentClass().hasLaTeXLayout("chapter")) - os << "\\addcontentsline{toc}{chapter}{\\bibname}"; - else if (buffer().params().documentClass().hasLaTeXLayout("section")) - os << "\\addcontentsline{toc}{section}{\\refname}"; - } - - if (!db_out.empty() && !buffer().params().use_bibtopic) { + if (usingBiblatex()) { + // Options + string opts = to_utf8(getParam("biblatexopts")); + // bibtotoc-Option + if (!bibtotoc.empty()) + opts = opts.empty() ? "heading=bibintoc" : "heading=bibintoc," + opts; + // The bibliography command docstring btprint = getParam("btprint"); - if (btprint == "btPrintAll") { + if (btprint == "btPrintAll") os << "\\nocite{*}\n"; + if (btprint == "bibbysection" && !buffer().masterParams().multibib.empty()) + os << "\\bibbysection"; + else + os << "\\printbibliography"; + if (!opts.empty()) + os << "[" << opts << "]"; + os << "\n"; + } else {// using BibTeX + // Database(s) + vector> const dbs = + buffer().prepareBibFilePaths(runparams, getBibFiles(), false); + vector db_out; + for (pair const & db : dbs) + db_out.push_back(db.first); + // Style options + if (style == "default") + style = buffer().masterParams().defaultBiblioStyle(); + if (!style.empty() && !buffer().masterParams().useBibtopic()) { + string base = buffer().masterBuffer()->prepareFileNameForLaTeX(style, ".bst", runparams.nice); + FileName const try_in_file = + makeAbsPath(base + ".bst", buffer().filePath()); + bool const not_from_texmf = try_in_file.isReadableFile(); + // If this style does not come from texmf and we are not + // exporting to .tex copy it to the tmp directory. + // This prevents problems with spaces and 8bit characters + // in the file name. + if (!runparams.inComment && !runparams.dryrun && !runparams.nice && + not_from_texmf) { + // use new style name + DocFileName const in_file = DocFileName(try_in_file); + base = removeExtension(in_file.mangledFileName()); + FileName const out_file = makeAbsPath(base + ".bst", + buffer().masterBuffer()->temppath()); + bool const success = in_file.copyTo(out_file); + if (!success) { + LYXERR0("Failed to copy '" << in_file + << "' to '" << out_file << "'"); + } + } + // FIXME UNICODE + os << "\\bibliographystyle{" + << from_utf8(latex_path(buffer().prepareFileNameForLaTeX(base, ".bst", runparams.nice))) + << "}\n"; + } + // Warn about spaces in bst path. Warn only once. + static bool warned_about_bst_spaces = false; + if (!warned_about_bst_spaces && runparams.nice && contains(style, ' ')) { + warned_about_bst_spaces = true; + Alert::warning(_("Export Warning!"), + _("There are spaces in the path to your BibTeX style file.\n" + "BibTeX will be unable to find it.")); + } + // Encoding + bool encoding_switched = false; + Encoding const * const save_enc = runparams.encoding; + docstring const encoding = getParam("encoding"); + if (!encoding.empty() && encoding != from_ascii("default")) { + Encoding const * const enc = encodings.fromLyXName(to_ascii(encoding)); + if (enc != runparams.encoding) { + os << "\\bgroup"; + switchEncoding(os.os(), buffer().params(), runparams, *enc, true); + runparams.encoding = enc; + encoding_switched = true; + } + } + // Handle the bibtopic case + if (!db_out.empty() && buffer().masterParams().useBibtopic()) { + os << "\\begin{btSect}"; + if (!style.empty()) + os << "[" << style << "]"; + os << "{" << getStringFromVector(db_out) << "}\n"; + docstring btprint = getParam("btprint"); + if (btprint.empty()) + // default + btprint = from_ascii("btPrintCited"); + os << "\\" << btprint << "\n" + << "\\end{btSect}\n"; + } + // bibtotoc option + if (!bibtotoc.empty() && !buffer().masterParams().useBibtopic() + && !buffer().masterParams().documentClass().bibInToc()) { + // set label for hyperref, see http://www.lyx.org/trac/ticket/6470 + if (buffer().masterParams().pdfoptions().use_hyperref) + os << "\\phantomsection"; + if (buffer().masterParams().documentClass().hasLaTeXLayout("chapter")) + os << "\\addcontentsline{toc}{chapter}{\\bibname}"; + else if (buffer().masterParams().documentClass().hasLaTeXLayout("section")) + os << "\\addcontentsline{toc}{section}{\\refname}"; + } + // The bibliography command + if (!db_out.empty() && !buffer().masterParams().useBibtopic()) { + docstring btprint = getParam("btprint"); + if (btprint == "btPrintAll") { + os << "\\nocite{*}\n"; + } + os << "\\bibliography{" << getStringFromVector(db_out) << "}\n"; + } + if (encoding_switched){ + // Switch back + switchEncoding(os.os(), buffer().params(), + runparams, *save_enc, true, true); + os << "\\egroup" << breakln; + runparams.encoding = save_enc; } - os << "\\bibliography{" << db_out << "}\n"; } } -support::FileNameList InsetBibtex::getBibFiles() const +docstring_list InsetBibtex::getBibFiles() const { - FileName path(buffer().filePath()); - support::PathChanger p(path); - - support::FileNameList vec; - - vector bibfilelist = getVectorFromString(getParam("bibfiles")); - vector::const_iterator it = bibfilelist.begin(); - vector::const_iterator en = bibfilelist.end(); - for (; it != en; ++it) { - FileName const file = getBibTeXPath(*it, buffer()); - - if (!file.empty()) - vec.push_back(file); - else - LYXERR0("Couldn't find " + to_utf8(*it) + " in InsetBibtex::getBibFiles()!"); - } - - return vec; - + return getVectorFromString(getParam("bibfiles")); } namespace { @@ -472,7 +458,7 @@ namespace { keepCase }; - /// remove whitespace characters, read characer sequence + /// remove whitespace characters, read character sequence /// not containing whitespace characters or characters in /// delimChars, and remove further whitespace characters. /// @@ -498,11 +484,9 @@ namespace { return false; // read value - bool legalChar = true; while (ifs && !isSpace(ch) && - delimChars.find(ch) == docstring::npos && - (legalChar = (illegalChars.find(ch) == docstring::npos)) - ) + delimChars.find(ch) == docstring::npos && + illegalChars.find(ch) == docstring::npos) { if (chCase == makeLowerCase) val += lowercase(ch); @@ -511,7 +495,7 @@ namespace { ifs.get(ch); } - if (!legalChar) { + if (illegalChars.find(ch) != docstring::npos) { ifs.putback(ch); return false; } @@ -531,7 +515,7 @@ namespace { /// read subsequent bibtex values that are delimited with a #-character. /// Concatenate all parts and replace names with the associated string in /// the variable strings. - /// @return true if reading was successfull (all single parts were delimited + /// @return true if reading was successful (all single parts were delimited /// correctly) bool readValue(docstring & val, ifdocstream & ifs, const VarMap & strings) { @@ -663,16 +647,16 @@ namespace { return true; } -} +} // namespace -void InsetBibtex::collectBibKeys(InsetIterator const & /*di*/) const +void InsetBibtex::collectBibKeys(InsetIterator const & /*di*/, FileNameList & checkedFiles) const { - parseBibTeXFiles(); + parseBibTeXFiles(checkedFiles); } -void InsetBibtex::parseBibTeXFiles() const +void InsetBibtex::parseBibTeXFiles(FileNameList & checkedFiles) const { // This bibtex parser is a first step to parse bibtex files // more precisely. @@ -686,27 +670,42 @@ void InsetBibtex::parseBibTeXFiles() const // bibtex does. // // Officially bibtex does only support ASCII, but in practice - // you can use the encoding of the main document as long as - // some elements like keys and names are pure ASCII. Therefore - // we convert the file from the buffer encoding. + // you can use any encoding as long as some elements like keys + // and names are pure ASCII. We support specifying an encoding, + // and we convert the file from that (default is buffer encoding). // We don't restrict keys to ASCII in LyX, since our own // InsetBibitem can generate non-ASCII keys, and nonstandard // 8bit clean bibtex forks exist. BiblioInfo keylist; - support::FileNameList const files = getBibFiles(); - support::FileNameList::const_iterator it = files.begin(); - support::FileNameList::const_iterator en = files.end(); - for (; it != en; ++ it) { - ifdocstream ifs(it->toFilesystemEncoding().c_str(), - ios_base::in, buffer().params().encoding().iconvName()); + docstring_list const files = getBibFiles(); + for (auto const & bf : files) { + FileName const bibfile = buffer().getBibfilePath(bf); + if (bibfile.empty()) { + LYXERR0("Unable to find path for " << bf << "!"); + continue; + } + if (find(checkedFiles.begin(), checkedFiles.end(), bibfile) != checkedFiles.end()) + // already checked this one. Skip. + continue; + else + // record that we check this. + checkedFiles.push_back(bibfile); + string encoding = buffer().masterParams().encoding().iconvName(); + string ienc = buffer().masterParams().bibFileEncoding(to_utf8(bf)); + if (ienc.empty() || ienc == "general") + ienc = to_ascii(params()["encoding"]); + + if (!ienc.empty() && ienc != "auto-legacy-plain" && ienc != "auto-legacy" && encodings.fromLyXName(ienc)) + encoding = encodings.fromLyXName(ienc)->iconvName(); + ifdocstream ifs(bibfile.toFilesystemEncoding().c_str(), + ios_base::in, encoding); char_type ch; VarMap strings; while (ifs) { - ifs.get(ch); if (!ifs) break; @@ -865,18 +864,6 @@ void InsetBibtex::parseBibTeXFiles() const } -FileName InsetBibtex::getBibTeXPath(docstring const & filename, Buffer const & buf) -{ - string texfile = changeExtension(to_utf8(filename), "bib"); - // note that, if the filename can be found directly from the path, - // findtexfile will just return a FileName object for that path. - FileName file(findtexfile(texfile, "bib")); - if (file.empty()) - file = FileName(makeAbsPath(texfile, buf.filePath())); - return file; -} - - bool InsetBibtex::addDatabase(docstring const & db) { docstring bibfiles = getParam("bibfiles"); @@ -911,8 +898,11 @@ bool InsetBibtex::delDatabase(docstring const & db) void InsetBibtex::validate(LaTeXFeatures & features) const { - if (features.bufferParams().use_bibtopic) + BufferParams const & mparams = features.buffer().masterParams(); + if (mparams.useBibtopic()) features.require("bibtopic"); + else if (!mparams.useBiblatex() && mparams.multibib == "child") + features.require("chapterbib"); // FIXME XHTML // It'd be better to be able to get this from an InsetLayout, but at present // InsetLayouts do not seem really to work for things that aren't InsetTexts. @@ -923,10 +913,72 @@ void InsetBibtex::validate(LaTeXFeatures & features) const } +void InsetBibtex::updateBuffer(ParIterator const &, UpdateType, bool const /*deleted*/) +{ + buffer().registerBibfiles(getBibFiles()); + // record encoding of bib files for biblatex + string const enc = (params()["encoding"] == from_ascii("default")) ? + string() : to_ascii(params()["encoding"]); + bool invalidate = false; + if (buffer().params().bibEncoding() != enc) { + buffer().params().setBibEncoding(enc); + invalidate = true; + } + map encs = getFileEncodings(); + map::const_iterator it = encs.begin(); + for (; it != encs.end(); ++it) { + if (buffer().params().bibFileEncoding(it->first) != it->second) { + buffer().params().setBibFileEncoding(it->first, it->second); + invalidate = true; + } + } + if (invalidate) + buffer().invalidateBibinfoCache(); +} + + +map InsetBibtex::getFileEncodings() const +{ + vector ps = + getVectorFromString(to_utf8(getParam("file_encodings")), "\t"); + std::map res; + for (string const & s: ps) { + string key; + string val = split(s, key, ' '); + res[key] = val; + } + return res; +} + + +docstring InsetBibtex::getRefLabel() const +{ + if (buffer().masterParams().documentClass().hasLaTeXLayout("chapter")) + return buffer().B_("Bibliography"); + return buffer().B_("References"); +} + + +void InsetBibtex::addToToc(DocIterator const & cpit, bool output_active, + UpdateType, TocBackend & backend) const +{ + if (!prefixIs(to_utf8(getParam("options")), "bibtotoc")) + return; + + docstring const str = getRefLabel(); + shared_ptr toc = backend.toc("tableofcontents"); + // Assign to appropriate level + int const item_depth = + (buffer().masterParams().documentClass().hasLaTeXLayout("chapter")) + ? 1 : 2; + toc->push_back(TocItem(cpit, item_depth, str, output_active)); +} + + int InsetBibtex::plaintext(odocstringstream & os, OutputParams const & op, size_t max_length) const { - docstring const reflabel = buffer().B_("References"); + docstring const reflabel = getRefLabel(); // We could output more information here, e.g., what databases are included // and information about options. But I don't necessarily see any reason to @@ -944,6 +996,10 @@ int InsetBibtex::plaintext(odocstringstream & os, docstring refoutput; refoutput += reflabel + "\n\n"; + // Tell BiblioInfo our purpose + CiteItem ci; + ci.context = CiteItem::Export; + // Now we loop over the entries vector::const_iterator vit = cites.begin(); vector::const_iterator const ven = cites.end(); @@ -957,30 +1013,36 @@ int InsetBibtex::plaintext(odocstringstream & os, refoutput += "[" + entry.label() + "] "; // FIXME Right now, we are calling BibInfo::getInfo on the key, // which will give us all the cross-referenced info. But for every - // entry, so there's a lot of repitition. This should be fixed. - refoutput += bibinfo.getInfo(entry.key(), buffer(), false) + "\n\n"; + // entry, so there's a lot of repetition. This should be fixed. + refoutput += bibinfo.getInfo(entry.key(), buffer(), ci) + "\n\n"; } os << refoutput; - return refoutput.size(); + return int(refoutput.size()); } // FIXME // docstring InsetBibtex::entriesAsXHTML(vector const & entries) // And then here just: entriesAsXHTML(buffer().masterBibInfo().citedEntries()) -docstring InsetBibtex::xhtml(XHTMLStream & xs, OutputParams const &) const +docstring InsetBibtex::xhtml(XMLStream & xs, OutputParams const &) const { BiblioInfo const & bibinfo = buffer().masterBibInfo(); bool const all_entries = getParam("btprint") == "btPrintAll"; - vector const & cites = + vector const & cites = all_entries ? bibinfo.getKeys() : bibinfo.citedEntries(); docstring const reflabel = buffer().B_("References"); - xs << html::StartTag("h2", "class='bibtex'") + // tell BiblioInfo our purpose + CiteItem ci; + ci.context = CiteItem::Export; + ci.richtext = true; + ci.max_key_size = UINT_MAX; + + xs << xml::StartTag("h2", "class='bibtex'") << reflabel - << html::EndTag("h2") - << html::StartTag("div", "class='bibtex'"); + << xml::EndTag("h2") + << xml::StartTag("div", "class='bibtex'"); // Now we loop over the entries vector::const_iterator vit = cites.begin(); @@ -991,32 +1053,366 @@ docstring InsetBibtex::xhtml(XHTMLStream & xs, OutputParams const &) const continue; BibTeXInfo const & entry = biit->second; - string const attr = "class='bibtexentry' id='LyXCite-" - + to_utf8(html::cleanAttr(entry.key())) + "'"; - xs << html::StartTag("div", attr); - + string const attr = "class='bibtexentry' id='LyXCite-" + + to_utf8(xml::cleanAttr(entry.key())) + "'"; + xs << xml::StartTag("div", attr); + // don't print labels if we're outputting all entries if (!all_entries) { - xs << html::StartTag("span", "class='bibtexlabel'") + xs << xml::StartTag("span", "class='bibtexlabel'") << entry.label() - << html::EndTag("span"); + << xml::EndTag("span"); } - + // FIXME Right now, we are calling BibInfo::getInfo on the key, // which will give us all the cross-referenced info. But for every - // entry, so there's a lot of repitition. This should be fixed. - xs << html::StartTag("span", "class='bibtexinfo'") - << XHTMLStream::ESCAPE_AND - << bibinfo.getInfo(entry.key(), buffer(), true) - << html::EndTag("span") - << html::EndTag("div") - << html::CR(); + // entry, so there's a lot of repetition. This should be fixed. + xs << xml::StartTag("span", "class='bibtexinfo'") + << XMLStream::ESCAPE_AND + << bibinfo.getInfo(entry.key(), buffer(), ci) + << xml::EndTag("span") + << xml::EndTag("div") + << xml::CR(); } - xs << html::EndTag("div"); + xs << xml::EndTag("div"); return docstring(); } +void InsetBibtex::docbook(XMLStream & xs, OutputParams const &) const +{ + BiblioInfo const & bibinfo = buffer().masterBibInfo(); + bool const all_entries = getParam("btprint") == "btPrintAll"; + vector const & cites = + all_entries ? bibinfo.getKeys() : bibinfo.citedEntries(); + + docstring const reflabel = buffer().B_("References"); + + // Tell BiblioInfo our purpose (i.e. generate HTML rich text). + CiteItem ci; + ci.context = CiteItem::Export; + ci.richtext = true; + ci.max_key_size = UINT_MAX; + + // Header for bibliography (title required). + xs << xml::StartTag("bibliography"); + xs << xml::CR(); + xs << xml::StartTag("title"); + xs << reflabel; + xs << xml::EndTag("title") << xml::CR(); + + // Translation between keys in each entry and DocBook tags. + // IDs for publications; list: http://tdg.docbook.org/tdg/5.2/biblioid.html. + vector> biblioId = { // + make_pair("doi", "doi"), + make_pair("isbn", "isbn"), + make_pair("issn", "issn"), + make_pair("isrn", "isrn"), + make_pair("istc", "istc"), + make_pair("lccn", "libraryofcongress"), + make_pair("number", "pubsnumber"), + make_pair("url", "uri") + }; + // Relations between documents. + vector> relations = { // + make_pair("journal", "journal"), + make_pair("booktitle", "book"), + make_pair("series", "series") + }; + // Various things that do not fit DocBook. + vector misc = { "language", "school", "note" }; + + // Store the mapping between BibTeX and DocBook. + map toDocBookTag; + toDocBookTag["fullnames:author"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["publisher"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["address"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["editor"] = "editor"; + toDocBookTag["institution"] = "SPECIFIC"; // No direct translation to DocBook: . + + toDocBookTag["title"] = "title"; + toDocBookTag["volume"] = "volumenum"; + toDocBookTag["edition"] = "edition"; + toDocBookTag["pages"] = "artpagenums"; + + toDocBookTag["abstract"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["keywords"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["year"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["month"] = "SPECIFIC"; // No direct translation to DocBook: . + + toDocBookTag["journal"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["booktitle"] = "SPECIFIC"; // No direct translation to DocBook: . + toDocBookTag["series"] = "SPECIFIC"; // No direct translation to DocBook: . + + for (auto const & id: biblioId) + toDocBookTag[id.first] = "SPECIFIC"; // No direct translation to DocBook: . + for (auto const & id: relations) + toDocBookTag[id.first] = "SPECIFIC"; // No direct translation to DocBook: . + for (auto const & id: misc) + toDocBookTag[id] = "SPECIFIC"; // No direct translation to DocBook: . + + // Loop over the entries. If there are no entries, add a comment to say so. + auto vit = cites.begin(); + auto ven = cites.end(); + + if (vit == ven) { + xs << XMLStream::ESCAPE_NONE << ""; + } + + for (; vit != ven; ++vit) { + BiblioInfo::const_iterator const biit = bibinfo.find(*vit); + if (biit == bibinfo.end()) + continue; + + BibTeXInfo const & entry = biit->second; + string const attr = "xml:id=\"" + to_utf8(xml::cleanID(entry.key())) + "\""; + xs << xml::StartTag("biblioentry", attr); + xs << xml::CR(); + + // FIXME Right now, we are calling BibInfo::getInfo on the key, + // which will give us all the cross-referenced info. But for every + // entry, so there's a lot of repetition. This should be fixed. + + // Parse the results of getInfo and emit the corresponding DocBook tags. Interesting pieces have the form + // "STH", the rest of the text may be discarded. + // Could have written a DocBook version of expandFormat (that parses a citation into HTML), but it implements + // some kind of recursion. Still, a (static) conversion step between the citation format and DocBook would have + // been required. All in all, both codes approaches would have been similar, but this parsing allows relying + // on existing building blocks. + + string html = to_utf8(bibinfo.getInfo(entry.key(), buffer(), ci)); + regex tagRegex("([^<]*)"); + smatch match; + auto tagIt = lyx::sregex_iterator(html.cbegin(), html.cend(), tagRegex, regex_constants::match_default); + auto tagEnd = lyx::sregex_iterator(); + map delayedTags; + + // Read all tags from HTML and convert those that have a 1:1 matching. + while (tagIt != tagEnd) { + string tag = tagIt->str(); // regex_match cannot work with temporary strings. + ++tagIt; + + if (regex_match(tag, match, tagRegex)) { + if (toDocBookTag[match[1]] == "SPECIFIC") { + delayedTags[match[1]] = match[2]; + } else { + xs << xml::StartTag(toDocBookTag[match[1]]); + xs << from_utf8(match[2].str()); + xs << xml::EndTag(toDocBookTag[match[1]]); + } + } else { + LYXERR0("The BibTeX field " << match[1].str() << " is unknown."); + xs << XMLStream::ESCAPE_NONE << from_utf8("\n"); + } + } + + // Type of document (book, journal paper, etc.). + xs << xml::StartTag("bibliomisc", "role=\"type\""); + xs << entry.entryType(); + xs << xml::EndTag("bibliomisc"); + xs << xml::CR(); + + // Handle tags that have complex transformations. + if (! delayedTags.empty()) { + unsigned long remainingTags = delayedTags.size(); // Used as a workaround. With GCC 7, when erasing all + // elements one by one, some elements may still pop in later on (even though they were deleted previously). + auto hasTag = [&delayedTags](string key) { return delayedTags.find(key) != delayedTags.end(); }; + auto getTag = [&delayedTags](string key) { return from_utf8(delayedTags[key]); }; + auto eraseTag = [&delayedTags, &remainingTags](string key) { + remainingTags -= 1; + delayedTags.erase(key); + }; + + // Notes on order of checks. + // - address goes with publisher if there is one, so check this first. Otherwise, the address goes with + // the entry without other details. + + // + if (hasTag("publisher")) { + xs << xml::StartTag("publisher"); + xs << xml::CR(); + xs << xml::StartTag("publishername"); + xs << getTag("publisher"); + xs << xml::EndTag("publishername"); + xs << xml::CR(); + + if (hasTag("address")) { + xs << xml::StartTag("address"); + xs << getTag("address"); + xs << xml::EndTag("address"); + eraseTag("address"); + } + + xs << xml::EndTag("publisher"); + xs << xml::CR(); + eraseTag("publisher"); + } + + if (hasTag("address")) { + xs << xml::StartTag("address"); + xs << getTag("address"); + xs << xml::EndTag("address"); + eraseTag("address"); + } + + // + if (hasTag("keywords")) { + // Split the keywords on comma. + docstring keywordSet = getTag("keywords"); + vector keywords; + if (keywordSet.find(from_utf8(",")) == string::npos) { + keywords = { keywordSet }; + } else { + size_t pos = 0; + while ((pos = keywordSet.find(from_utf8(","))) != string::npos) { + keywords.push_back(keywordSet.substr(0, pos)); + keywordSet.erase(0, pos + 1); + } + keywords.push_back(keywordSet); + } + + xs << xml::StartTag("keywordset") << xml::CR(); + for (auto & kw: keywords) { + kw.erase(kw.begin(), std::find_if(kw.begin(), kw.end(), + [](int c) {return !std::isspace(c);})); + xs << xml::StartTag("keyword"); + xs << kw; + xs << xml::EndTag("keyword"); + xs << xml::CR(); + } + xs << xml::EndTag("keywordset") << xml::CR(); + eraseTag("keywords"); + } + + // + // Example: http://tdg.docbook.org/tdg/5.1/biblioset.html + if (hasTag("year")) { + docstring value = getTag("year"); + eraseTag("year"); + + // Follow xsd:gYearMonth format (http://books.xmlschemata.org/relaxng/ch19-77135.html). + if (hasTag("month")) { + value += "-" + getTag("month"); + eraseTag("month"); + } + + xs << xml::StartTag("pubdate"); + xs << value; + xs << xml::EndTag("pubdate"); + xs << xml::CR(); + } + + // + if (hasTag("institution")) { + xs << xml::StartTag("org"); + xs << xml::CR(); + xs << xml::StartTag("orgname"); + xs << getTag("institution"); + xs << xml::EndTag("orgname"); + xs << xml::CR(); + xs << xml::EndTag("org"); + xs << xml::CR(); + eraseTag("institution"); + } + + // + // Example: http://tdg.docbook.org/tdg/5.1/biblioset.html + for (auto const & id: relations) { + if (hasTag(id.first)) { + xs << xml::StartTag("biblioset", "relation=\"" + id.second + "\""); + xs << xml::CR(); + xs << xml::StartTag("title"); + xs << getTag(id.first); + xs << xml::EndTag("title"); + xs << xml::CR(); + xs << xml::EndTag("biblioset"); + xs << xml::CR(); + eraseTag(id.first); + } + } + + // + // Example: http://tdg.docbook.org/tdg/5.1/authorgroup.html + if (hasTag("fullnames:author")) { + // Perform full parsing of the BibTeX string, dealing with the many corner cases that might + // be encountered. + authorsToDocBookAuthorGroup(getTag("fullnames:author"), xs, buffer()); + eraseTag("fullnames:author"); + } + + // + if (hasTag("abstract")) { + // Split the paragraphs on new line. + docstring abstract = getTag("abstract"); + vector paragraphs; + if (abstract.find(from_utf8("\n")) == string::npos) { + paragraphs = { abstract }; + } else { + size_t pos = 0; + while ((pos = abstract.find(from_utf8(","))) != string::npos) { + paragraphs.push_back(abstract.substr(0, pos)); + abstract.erase(0, pos + 1); + } + paragraphs.push_back(abstract); + } + + xs << xml::StartTag("abstract"); + xs << xml::CR(); + for (auto const & para: paragraphs) { + if (para.empty()) + continue; + xs << xml::StartTag("para"); + xs << para; + xs << xml::EndTag("para"); + } + xs << xml::CR(); + xs << xml::EndTag("abstract"); + xs << xml::CR(); + eraseTag("abstract"); + } + + // + for (auto const & id: biblioId) { + if (hasTag(id.first)) { + xs << xml::StartTag("biblioid", "class=\"" + id.second + "\""); + xs << getTag(id.first); + xs << xml::EndTag("biblioid"); + xs << xml::CR(); + eraseTag(id.first); + } + } + + // + for (auto const & id: misc) { + if (hasTag(id)) { + xs << xml::StartTag("bibliomisc", "role=\"" + id + "\""); + xs << getTag(id); + xs << xml::EndTag("bibliomisc"); + xs << xml::CR(); + eraseTag(id); + } + } + + // After all tags are processed, check for errors. + if (remainingTags > 0) { + LYXERR0("Still delayed tags not yet handled."); + xs << XMLStream::ESCAPE_NONE << from_utf8("\n"); + } + } + + xs << xml::EndTag("biblioentry"); + xs << xml::CR(); + } + + // Footer for bibliography. + xs << xml::EndTag("bibliography"); +} + + void InsetBibtex::write(ostream & os) const { params().Write(os, &buffer());