]> git.lyx.org Git - features.git/blobdiff - src/frontends/controllers/biblio.C
change "support/std_sstream.h" to <sstream>
[features.git] / src / frontends / controllers / biblio.C
index 99d643b5c1faf5031ce85c0894bb57fb2b92748d..8522f1fc663caa81a2d915c4a7e35ef45e7d7772 100644 (file)
  * Licence details can be found in the file COPYING.
  *
  * \author Angus Leeming
- * \author Herbert Voss
+ * \author Herbert Voß
  *
- * Full author contact details are available in file CREDITS
+ * Full author contact details are available in file CREDITS.
  */
 
 #include <config.h>
 
-#ifdef __GNUG__
-#pragma implementation
-#endif
-
-#include "LString.h"
 #include "biblio.h"
-#include "gettext.h" // for _()
-#include "helper_funcs.h"
+
+#include "buffer.h"
+#include "bufferparams.h"
+#include "gettext.h"
+
 #include "support/lstrings.h"
-#include "support/LAssert.h"
 
 #include <boost/regex.hpp>
 
 #include <algorithm>
+#include <sstream>
 
+using std::string;
+using std::ostringstream;
 using std::vector;
 
+
+namespace lyx {
+
+using support::ascii_lowercase;
+using support::bformat;
+using support::compare_ascii_no_case;
+using support::contains;
+using support::getVectorFromString;
+using support::ltrim;
+using support::prefixIs;
+using support::rtrim;
+using support::split;
+using support::subst;
+using support::token;
+using support::trim;
+
 namespace biblio {
 
+namespace {
+
+vector<string> const init_possible_cite_commands()
+{
+       char const * const pos[] = {
+               "cite",
+               "citet", "citep", "citealt", "citealp",
+               "citeauthor", "citeyear", "citeyearpar",
+               "citet*", "citep*", "citealt*", "citealp*", "citeauthor*",
+               "Citet",  "Citep",  "Citealt",  "Citealp",  "Citeauthor",
+               "Citet*", "Citep*", "Citealt*", "Citealp*", "Citeauthor*",
+               "fullcite",
+               "footcite", "footcitet", "footcitep", "footcitealt",
+               "footcitealp", "footciteauthor", "footciteyear",
+               "footciteyearpar",
+               "citefield",
+               "citetitle",
+               "cite*"
+       };
+       size_t const size_pos = sizeof(pos) / sizeof(pos[0]);
+
+       return vector<string>(pos, pos + size_pos);
+}
+
+
+vector<string> const & possible_cite_commands()
+{
+       static vector<string> const pos = init_possible_cite_commands();
+       return pos;
+}
+
+
+bool is_possible_cite_command(string const & input)
+{
+       vector<string> const & possibles = possible_cite_commands();
+       vector<string>::const_iterator const end = possibles.end();
+       return std::find(possibles.begin(), end, input) != end;
+}
+
+
+string const default_cite_command(CiteEngine engine)
+{
+       string str;
+       switch (engine) {
+       case ENGINE_BASIC:
+               str = "cite";
+               break;
+       case ENGINE_NATBIB_AUTHORYEAR:
+               str = "citet";
+               break;
+       case ENGINE_NATBIB_NUMERICAL:
+               str = "citep";
+               break;
+       case ENGINE_JURABIB:
+               str = "cite";
+               break;
+       }
+       return str;
+}
+
+} // namespace anon
+
+
+string const asValidLatexCommand(string const & input,
+                                CiteEngine_enum const & engine)
+{
+       string const default_str = default_cite_command(engine);
+       if (!is_possible_cite_command(input))
+               return default_str;
+
+       string output;
+       switch (engine) {
+       case ENGINE_BASIC:
+               output = default_str;
+               break;
+
+       case ENGINE_NATBIB_AUTHORYEAR:
+       case ENGINE_NATBIB_NUMERICAL:
+               if (input == "cite" || input == "citefield" ||
+                   input == "citetitle" || input == "cite*")
+                       output = default_str;
+               else if (prefixIs(input, "foot"))
+                       output = input.substr(4);
+               else
+                       output = input;
+               break;
+
+       case ENGINE_JURABIB: {
+               // Jurabib does not support the 'uppercase' natbib style.
+               if (input[0] == 'C')
+                       output = string(1, 'c') + input.substr(1);
+               else
+                       output = input;
+
+               // Jurabib does not support the 'full' natbib style.
+                string::size_type const n = output.size() - 1;
+                if (output != "cite*" && output[n] == '*')
+                        output = output.substr(0, n);
+
+               break;
+       }
+       }
+
+       return output;
+}
+
+
 string const familyName(string const & name)
 {
        // Very simple parser
@@ -40,10 +163,10 @@ string const familyName(string const & name)
        // "Surname, F."
        // "FirstName Surname"
        // "F. Surname"
-       string::size_type idx = fname.find(",");
+       string::size_type idx = fname.find(',');
        if (idx != string::npos)
                return ltrim(fname.substr(0, idx));
-       idx = fname.rfind(".");
+       idx = fname.rfind('.');
        if (idx != string::npos)
                fname = ltrim(fname.substr(idx + 1));
        // test if we have a LaTeX Space in front
@@ -56,7 +179,7 @@ string const familyName(string const & name)
 
 string const getAbbreviatedAuthor(InfoMap const & map, string const & key)
 {
-       lyx::Assert(!map.empty());
+       BOOST_ASSERT(!map.empty());
 
        InfoMap::const_iterator it = map.find(key);
        if (it == map.end())
@@ -80,6 +203,7 @@ string const getAbbreviatedAuthor(InfoMap const & map, string const & key)
        }
 
        string author = parseBibTeX(data, "author");
+
        if (author.empty())
                author = parseBibTeX(data, "editor");
 
@@ -94,19 +218,20 @@ string const getAbbreviatedAuthor(InfoMap const & map, string const & key)
        if (authors.empty())
                return author;
 
-       author = familyName(authors[0]);
        if (authors.size() == 2)
-               author += _(" and ") + familyName(authors[1]);
-       else if (authors.size() > 2)
-               author += _(" et al.");
+               return bformat(_("%1$s and %2$s"),
+                       familyName(authors[0]), familyName(authors[1]));
+
+       if (authors.size() > 2)
+               return bformat(_("%1$s et al."), familyName(authors[0]));
 
-       return author;
+       return familyName(authors[0]);
 }
 
 
 string const getYear(InfoMap const & map, string const & key)
 {
-       lyx::Assert(!map.empty());
+       BOOST_ASSERT(!map.empty());
 
        InfoMap::const_iterator it = map.find(key);
        if (it == map.end())
@@ -150,7 +275,7 @@ struct compareNoCase: public std::binary_function<string, string, bool>
                return compare_ascii_no_case(s1, s2) < 0;
        }
 };
+
 } // namespace anon
 
 
@@ -170,7 +295,7 @@ vector<string> const getKeys(InfoMap const & map)
 
 string const getInfo(InfoMap const & map, string const & key)
 {
-       lyx::Assert(!map.empty());
+       BOOST_ASSERT(!map.empty());
 
        InfoMap::const_iterator it = map.find(key);
        if (it == map.end())
@@ -227,7 +352,7 @@ string const getInfo(InfoMap const & map, string const & key)
        if (!year.empty())
                result << ", " << year;
 
-       string const result_str = rtrim(STRCONV(result.str()));
+       string const result_str = rtrim(result.str());
        if (!result_str.empty())
                return result_str;
 
@@ -238,81 +363,56 @@ string const getInfo(InfoMap const & map, string const & key)
 
 namespace {
 
-// The functions doing the dirty work for the search.
-vector<string>::const_iterator
-simpleSearch(InfoMap const & theMap,
-            vector<string> const & keys,
-            string const & expr,
-            vector<string>::const_iterator start,
-            Direction dir,
-            bool caseSensitive)
+// Escape special chars.
+// All characters are literals except: '.|*?+(){}[]^$\'
+// These characters are literals when preceded by a "\", which is done here
+string const escape_special_chars(string const & expr)
 {
-       string tmp = expr;
-       if (!caseSensitive)
-               tmp = lowercase(tmp);
-
-       vector<string> searchwords = getVectorFromString(tmp, " ");
-
-       // Loop over all keys from start...
-       for (vector<string>::const_iterator it = start;
-            // End condition is direction-dependent.
-            (dir == FORWARD) ? (it<keys.end()) : (it>=keys.begin());
-            // increment is direction-dependent.
-            (dir == FORWARD) ? (++it) : (--it)) {
-
-               string data = (*it);
-               InfoMap::const_iterator info = theMap.find(*it);
-               if (info != theMap.end())
-                       data += " " + info->second;
-               if (!caseSensitive)
-                       data = lowercase(data);
-
-               bool found = true;
-
-               // Loop over all search words...
-               for (vector<string>::const_iterator sit = searchwords.begin();
-                    sit != searchwords.end(); ++sit) {
-                       if (data.find(*sit) == string::npos) {
-                               found = false;
-                               break;
-                       }
-               }
-
-               if (found) return it;
-       }
-
-       return keys.end();
+       // Search for all chars '.|*?+(){}[^$]\'
+       // Note that '[' and '\' must be escaped.
+       // This is a limitation of boost::regex, but all other chars in BREs
+       // are assumed literal.
+       boost::RegEx reg("[].|*?+(){}^$\\[\\\\]");
+
+       // $& is a perl-like expression that expands to all of the current match
+       // The '$' must be prefixed with the escape character '\' for
+       // boost to treat it as a literal.
+       // Thus, to prefix a matched expression with '\', we use:
+       return reg.Merge(expr, "\\\\$&");
 }
 
 
-vector<string>::const_iterator
-regexSearch(InfoMap const & theMap,
-           vector<string> const & keys,
-           string const & expr,
-           vector<string>::const_iterator start,
-           Direction dir)
+// A functor for use with std::find_if, used to ascertain whether a
+// data entry matches the required regex_
+struct RegexMatch : public std::unary_function<string, bool>
 {
-       boost::regex reg(STRCONV(expr));
-
-       for (vector<string>::const_iterator it = start;
-            // End condition is direction-dependent.
-            (dir == FORWARD) ? (it < keys.end()) : (it >= keys.begin());
-            // increment is direction-dependent.
-            (dir == FORWARD) ? (++it) : (--it)) {
-
-               string data = (*it);
-               InfoMap::const_iterator info = theMap.find(*it);
-               if (info != theMap.end())
-                       data += " " + info->second;
-
-               if (boost::regex_match(STRCONV(data), reg)) {
-                       return it;
-               }
+       // re and icase are used to construct an instance of boost::RegEx.
+       // if icase is true, then matching is insensitive to case
+       RegexMatch(InfoMap const & m, string const & re, bool icase)
+               : map_(m), regex_(re, icase) {}
+
+       bool operator()(string const & key) const {
+               if (!validRE())
+                       return false;
+
+               // the data searched is the key + its associated BibTeX/biblio
+               // fields
+               string data = key;
+               InfoMap::const_iterator info = map_.find(key);
+               if (info != map_.end())
+                       data += ' ' + info->second;
+
+               // Attempts to find a match for the current RE
+               // somewhere in data.
+               return regex_.Search(data);
        }
 
-       return keys.end();
-}
+       bool validRE() const { return regex_.error_code() == 0; }
 
+private:
+       InfoMap const map_;
+       mutable boost::RegEx regex_;
+};
 
 } // namespace anon
 
@@ -320,7 +420,7 @@ regexSearch(InfoMap const & theMap,
 vector<string>::const_iterator
 searchKeys(InfoMap const & theMap,
           vector<string> const & keys,
-          string const & expr,
+          string const & search_expr,
           vector<string>::const_iterator start,
           Search type,
           Direction dir,
@@ -330,15 +430,34 @@ searchKeys(InfoMap const & theMap,
        if (start < keys.begin() || start >= keys.end())
                return keys.end();
 
-       string search_expr = trim(expr);
-       if (search_expr.empty())
+       string expr = trim(search_expr);
+       if (expr.empty())
                return keys.end();
 
        if (type == SIMPLE)
-               return simpleSearch(theMap, keys, search_expr, start, dir,
-                                   caseSensitive);
+               // We must escape special chars in the search_expr so that
+               // it is treated as a simple string by boost::regex.
+               expr = escape_special_chars(expr);
+
+       // Build the functor that will be passed to find_if.
+       RegexMatch const match(theMap, expr, !caseSensitive);
+       if (!match.validRE())
+               return keys.end();
+
+       // Search the vector of 'keys' from 'start' for one that matches the
+       // predicate 'match'. Searching can be forward or backward from start.
+       if (dir == FORWARD)
+               return std::find_if(start, keys.end(), match);
 
-       return regexSearch(theMap, keys, search_expr, start, dir);
+       vector<string>::const_reverse_iterator rit(start);
+       vector<string>::const_reverse_iterator rend = keys.rend();
+       rit = std::find_if(rit, rend, match);
+
+       if (rit == rend)
+               return keys.end();
+       // This is correct and always safe.
+       // (See Meyer's Effective STL, Item 28.)
+       return (++rit).base();
 }
 
 
@@ -365,8 +484,8 @@ string const parseBibTeX(string data, string const & findkey)
                // the \n and in the second we replace it
                // with a space
                if (!dummy.empty()) {
-                       if (!contains(dummy, "="))
-                               data_ += (' ' + dummy);
+                       if (!contains(dummy, '='))
+                               data_ += ' ' + dummy;
                        else
                                data_ += dummy;
                }
@@ -408,7 +527,7 @@ string const parseBibTeX(string data, string const & findkey)
        keyvalue = dummy;
        dummy = token(data, ',', Entries++);
        while (!contains(dummy, '=') && !dummy.empty()) {
-               keyvalue += (',' + dummy);
+               keyvalue += ',' + dummy;
                dummy = token(data, ',', Entries++);
        }
 
@@ -474,7 +593,6 @@ string const parseBibTeX(string data, string const & findkey)
 
 namespace {
 
-using namespace biblio;
 
 char const * const citeCommands[] = {
        "cite", "citet", "citep", "citealt", "citealp", "citeauthor",
@@ -505,21 +623,21 @@ unsigned int const nCiteStylesUCase =
 } // namespace anon
 
 
-CitationStyle const getCitationStyle(string const & command)
+CitationStyle::CitationStyle(string const & command)
+       : style(CITE), full(false), forceUCase(false)
 {
-       if (command.empty()) return CitationStyle();
+       if (command.empty())
+               return;
 
-       CitationStyle cs;
        string cmd = command;
-
        if (cmd[0] == 'C') {
-               cs.forceUCase = true;
+               forceUCase = true;
                cmd[0] = 'c';
        }
 
-       size_t n = cmd.size() - 1;
-       if (cmd[n] == '*') {
-               cs.full = true;
+       string::size_type const n = cmd.size() - 1;
+       if (cmd != "cite" && cmd[n] == '*') {
+               full = true;
                cmd = cmd.substr(0,n);
        }
 
@@ -528,25 +646,23 @@ CitationStyle const getCitationStyle(string const & command)
 
        if (ptr != last) {
                size_t idx = ptr - citeCommands;
-               cs.style = citeStyles[idx];
+               style = citeStyles[idx];
        }
-
-       return cs;
 }
 
 
-string const getCiteCommand(CiteStyle command, bool full, bool forceUCase)
+string const CitationStyle::asLatexStr() const
 {
-       string cite = citeCommands[command];
+       string cite = citeCommands[style];
        if (full) {
                CiteStyle const * last = citeStylesFull + nCiteStylesFull;
-               if (std::find(citeStylesFull, last, command) != last)
-                       cite += "*";
+               if (std::find(citeStylesFull, last, style) != last)
+                       cite += '*';
        }
 
        if (forceUCase) {
                CiteStyle const * last = citeStylesUCase + nCiteStylesUCase;
-               if (std::find(citeStylesUCase, last, command) != last)
+               if (std::find(citeStylesUCase, last, style) != last)
                        cite[0] = 'C';
        }
 
@@ -554,22 +670,40 @@ string const getCiteCommand(CiteStyle command, bool full, bool forceUCase)
 }
 
 
-vector<CiteStyle> const getCiteStyles(bool usingNatbib)
+CiteEngine_enum getEngine(Buffer const & buffer)
 {
-       unsigned int nStyles = 1;
+       return buffer.params().cite_engine;
+}
+
+
+vector<CiteStyle> const getCiteStyles(CiteEngine_enum const & engine)
+{
+       unsigned int nStyles = 0;
        unsigned int start = 0;
-       if (usingNatbib) {
+
+       switch (engine) {
+       case ENGINE_BASIC:
+               nStyles = 1;
+               start = 0;
+               break;
+       case ENGINE_NATBIB_AUTHORYEAR:
+       case ENGINE_NATBIB_NUMERICAL:
                nStyles = nCiteStyles - 1;
                start = 1;
+               break;
+       case ENGINE_JURABIB:
+               nStyles = nCiteStyles;
+               start = 0;
+               break;
        }
 
-       vector<CiteStyle> styles(nStyles);
+       typedef vector<CiteStyle> cite_vec;
 
-       vector<CiteStyle>::size_type i = 0;
+       cite_vec styles(nStyles);
+       cite_vec::size_type i = 0;
        int j = start;
-       for (; i != styles.size(); ++i, ++j) {
+       for (; i != styles.size(); ++i, ++j)
                styles[i] = citeStyles[j];
-       }
 
        return styles;
 }
@@ -619,7 +753,7 @@ getNumericalStrings(string const & key,
                        break;
 
                case CITEYEARPAR:
-                       str = "(" + year + ")";
+                       str = '(' + year + ')';
                        break;
                }
 
@@ -649,16 +783,21 @@ getAuthorYearStrings(string const & key,
 
                switch (styles[i]) {
                case CITE:
+                       // jurabib only: Author/Annotator
+                       // (i.e. the "before" field, 2nd opt arg)
+                       str = author + "/<" + _("before") + '>';
+                       break;
+
                case CITET:
-                       str = author + " (" + year + ")";
+                       str = author + " (" + year + ')';
                        break;
 
                case CITEP:
-                       str = "(" + author + ", " + year + ")";
+                       str = '(' + author + ", " + year + ')';
                        break;
 
                case CITEALT:
-                       str = author + " " + year ;
+                       str = author + ' ' + year ;
                        break;
 
                case CITEALP:
@@ -674,7 +813,7 @@ getAuthorYearStrings(string const & key,
                        break;
 
                case CITEYEARPAR:
-                       str = "(" + year + ")";
+                       str = '(' + year + ')';
                        break;
                }
 
@@ -685,3 +824,4 @@ getAuthorYearStrings(string const & key,
 }
 
 } // namespace biblio
+} // namespace lyx