Extract page numbers.

src/main.cpp

@@ -119,7 +119,7 @@ int main(int argc, char *argv[])
                     }
                     if (!match.page.empty())
                     {
-                        cout << ", " << match.page;
+                        cout << ", page " << match.page;
                     }
                     cout << ": " << match.context.first << match.text
                          << match.context.second << '\n';

src/search.cpp

@@ -20,7 +20,6 @@
 #include "zip.hpp"
 
 #include <boost/regex.hpp>
-#include <boost/regex/v4/regex_match.hpp>
 
 #include <algorithm>
 #include <string>
@@ -73,6 +72,7 @@ std::vector<match> search(const fs::path &filepath, std::string_view regex,
         std::string::const_iterator end{document.end()};
         boost::match_results<std::string::const_iterator> match_result;
         std::string last_headline;
+        std::string last_page;
 
         while (boost::regex_search(begin, end, match_result, re,
                                    boost::match_default))
@@ -87,6 +87,12 @@ std::vector<match> search(const fs::path &filepath, std::string_view regex,
                 last_headline = current_headline;
             }
             match.headline = last_headline;
+            const auto current_page{page(match_result.prefix().str())};
+            if (!current_page.empty())
+            {
+                last_page = current_page;
+            }
+            match.page = last_page;
 
             matches.emplace_back(match);
             begin = match_result[0].second;
@@ -111,6 +117,18 @@ void cleanup_text(std::string &text)
         {
             replacement = "</H>";
         }
+        else if (text.substr(pos, 5) == "<span")
+        {
+            auto endpos{text.find('>')};
+            boost::match_results<const char *> match;
+            const boost::regex re_pagebreak{".+pagebreak.+(title|aria-label)"
+                                            "=\"([[:alnum:]]+)\".*"};
+            if (boost::regex_search(text.substr(pos, endpos).data(), match,
+                                    re_pagebreak))
+            {
+                replacement = "<PAGE " + match[2] + ">";
+            }
+        }
         text.replace(pos, text.find('>', pos) + 1 - pos, replacement);
         pos += replacement.length();
     }
@@ -158,7 +176,7 @@ context(const boost::match_results<std::string::const_iterator> &match,
     return {prefix.substr(pos_before + 2), suffix.substr(0, pos_after - 1)};
 }
 
-[[nodiscard]] std::string headline(const std::string_view prefix)
+std::string headline(const std::string_view prefix)
 {
     size_t pos{prefix.length()};
     while ((pos = prefix.rfind("<H>", pos)) != std::string_view::npos)
@@ -170,4 +188,16 @@ context(const boost::match_results<std::string::const_iterator> &match,
     return {};
 }
 
+std::string page(const std::string_view prefix)
+{
+    size_t pos{prefix.length()};
+    while ((pos = prefix.rfind("<PAGE ", pos)) != std::string_view::npos)
+    {
+        pos += 6;
+        return std::string{prefix.substr(pos, prefix.find('>', pos) - pos)};
+    }
+
+    return {};
+}
+
 } // namespace epubgrep::search

src/search.hpp

@@ -73,6 +73,9 @@ context(const boost::match_results<std::string::const_iterator> &match,
 //! Return last headline if possible.
 [[nodiscard]] std::string headline(std::string_view prefix);
 
+//! Return current page if possible.
+[[nodiscard]] std::string page(std::string_view prefix);
+
 } // namespace epubgrep::search
 
 #endif // EPUBGREP_SEARCH_HPP