released 3.12.7

Like GNU grep, unpaired ) in regex are matched literally when option -Y is used and by the grep and egrep aliases of ugrep
Genivia · Aug 14, 2023 · c361d55 · c361d55
1 parent 8f2895d
commit c361d55
Show file tree

Hide file tree

Showing 12 changed files with 75 additions and 51 deletions.
diff --git a/README.md b/README.md
@@ -22,12 +22,12 @@ The ugrep tools include the following powerful commands:
 Development roadmap
 -------------------
 
-- my highest priority is testing and quality assurance to continue to make sure ugrep has no bugs and is 100% reliable, my nightmare would be something like ripgrep's bugs (that I actually uncovered in benchmarking ugrep)
-- listen to users to continue to improve ugrep with new and updated features
+- my highest priority is testing and quality assurance to continue to make sure ugrep has no bugs and is reliable
 - make ugrep even faster, see [my latest blog article](https://www.genivia.com/ugrep.html) demonstrating with a reproducible benchmark that ugrep beats GNU grep and ripgrep in terms of raw performance
+- listen to users to continue to improve ugrep with new and updated features
 - share reproducible performance data with the community
 - improve the interactive TUI with a split screen
-- add file indexing to accelerate cold search performance, see the [ugrep-indexer](https://github.com/Genivia/ugrep-indexer) for details on a new indexing method that I call a *monotonic indexer*
+- add file indexing to speed up cold search performance, see [ugrep-indexer](https://github.com/Genivia/ugrep-indexer)
 
 Overview
 --------
@@ -5358,7 +5358,7 @@ in markdown:
 
 
 
-    ugrep 3.12.6                     August 6, 2023                         UGREP(1)
+    ugrep 3.12.7                     August 14, 2023                        UGREP(1)
 
 🔝 [Back to table of contents](#toc)
 

diff --git a/bin/win32/ugrep.exe b/bin/win32/ugrep.exe
diff --git a/bin/win64/ugrep.exe b/bin/win64/ugrep.exe
diff --git a/include/reflex/convert.h b/include/reflex/convert.h
@@ -68,7 +68,8 @@ namespace convert_flag {
   const convert_flag_type dotall     = 0x0080; ///< convert `.` (dot) to match all, same as `(?s)`
   const convert_flag_type freespace  = 0x0100; ///< convert regex by removing spacing, same as `(?x)`
   const convert_flag_type notnewline = 0x0200; ///< inverted character classes and \s do not match newline `\n`
-  const convert_flag_type permissive = 0x0400; ///< convert Unicode to compact UTF-8 patterns and DFA, permits some invalid UTF-8 sequences
+  const convert_flag_type permissive = 0x0400; ///< convert Unicode to compact UTF-8 patterns, permits some invalid UTF-8 sequences
+  const convert_flag_type closing    = 0x8000; ///< permit matching ) when it has no opening (
 }
 
 /// @brief Returns the converted regex string given a regex library signature and conversion flags, throws regex_error.

diff --git a/include/reflex/pattern.h b/include/reflex/pattern.h
@@ -803,15 +803,15 @@ class Pattern {
     Option() : b(), h(), e(), f(), i(), m(), n(), o(), p(), q(), r(), s(), w(), x(), z() { }
     bool                     b; ///< disable escapes in bracket lists
     bool                     h; ///< construct indexing hash finite state automaton
-    Char                     e; ///< escape character, or > 255 for none, '\\' default
-    std::vector<std::string> f; ///< output to files
+    Char                     e; ///< escape character, or > 255 for none, a backslash by default
+    std::vector<std::string> f; ///< output the patterns and/or DFA to files(s)
     bool                     i; ///< case insensitive mode, also `(?i:X)`
     bool                     m; ///< multi-line mode, also `(?m:X)`
     std::string              n; ///< pattern name (for use in generated code)
     bool                     o; ///< generate optimized FSM code for option f
     bool                     p; ///< with option f also output predict match array for fast search with find()
     bool                     q; ///< enable "X" quotation of verbatim content, also `(?q:X)`
-    bool                     r; ///< raise syntax errors
+    bool                     r; ///< raise syntax errors as exceptions
     bool                     s; ///< single-line mode (dotall mode), also `(?s:X)`
     bool                     w; ///< write error message to stderr
     bool                     x; ///< free-spacing mode, also `(?x:X)`

diff --git a/lib/convert.cpp b/lib/convert.cpp
@@ -1750,19 +1750,35 @@ std::string convert(const char *pattern, const char *signature, convert_flag_typ
         else
         {
           if (lev == 1)
-            throw regex_error(regex_error::mismatched_parens, pattern, pos);
-          if (beg)
+          {
+            if (!(flags & convert_flag::closing))
+              throw regex_error(regex_error::mismatched_parens, pattern, pos);
+            if (!(flags & convert_flag::basic) || bre)
+            {
+              // translate a closing ) to \) when it has no opening (
+              regex.append(&pattern[loc], pos - loc).push_back('\\');
+              loc = pos;
+              anc = false;
+              beg = false;
+            }
+          }
+          else if (beg)
+          {
             throw regex_error(regex_error::empty_expression, pattern, pos);
-          if (lap == lev)
+          }
+          else
           {
-            // lex lookahead: translate ) to ))
-            regex.append(&pattern[loc], pos - loc).push_back(')');
-            loc = pos;
-            lap = 0;
+            if (lap == lev)
+            {
+              // lex lookahead: translate ) to ))
+              regex.append(&pattern[loc], pos - loc).push_back(')');
+              loc = pos;
+              lap = 0;
+            }
+            // terminate (?isx:...)
+            mod[lev].clear();
+            --lev;
           }
-          // terminate (?isx:...)
-          mod[lev].clear();
-          --lev;
         }
         break;
       case '|':
@@ -2005,10 +2021,6 @@ std::string convert(const char *pattern, const char *signature, convert_flag_typ
           anc = false;
           beg = false;
         }
-        else
-        {
-          throw regex_error(regex_error::mismatched_braces, pattern, pos);
-        }
         break;
       case '#':
         if ((flags & convert_flag::lex) && (flags & convert_flag::freespace))

diff --git a/lib/pattern.cpp b/lib/pattern.cpp
@@ -493,17 +493,17 @@ void Pattern::init_options(const char *options)
           opt_.x = true;
           break;
         case 'z':
-            for (const char *t = s += (s[1] == '='); *s != ';' && *s != '\0'; ++t)
+          for (const char *t = s += (s[1] == '='); *s != ';' && *s != '\0'; ++t)
+          {
+            if (std::isspace(*t) || *t == ';' || *t == '\0')
             {
-              if (std::isspace(*t) || *t == ';' || *t == '\0')
-              {
-                if (t > s + 1)
-                  opt_.z = std::string(s + 1, t - s - 1);
-                s = t;
-              }
+              if (t > s + 1)
+                opt_.z = std::string(s + 1, t - s - 1);
+              s = t;
             }
-            --s;
-            break;
+          }
+          --s;
+          break;
         case 'f':
         case 'n':
           for (const char *t = s += (s[1] == '='); *s != ';' && *s != '\0'; ++t)
@@ -596,7 +596,10 @@ void Pattern::parse(
         Char c = at(end);
         if (c == '\0' || c == '|')
           break;
-        if (c == '.' || c == '^' || c == '$' || c == '(' || c == ')' || c == '[' || c == '{' || c == '?' || c == '*' || c == '+')
+        if (c == '.' || c == '^' || c == '$' ||
+            c == '(' || c == '[' || c == '{' ||
+            c == '?' || c == '*' || c == '+' ||
+            c == ')')
         {
           end = loc;
           break;
@@ -1115,6 +1118,10 @@ void Pattern::parse3(
           lazyset.clear();
         }
       }
+      else if (at(loc) == '\0')
+      {
+        error(regex_error::mismatched_braces, loc);
+      }
       else
       {
         error(regex_error::invalid_repeat, loc);
@@ -1381,17 +1388,13 @@ void Pattern::parse4(
   {
     error(begin ? regex_error::empty_expression : regex_error::mismatched_parens, loc++);
   }
-  else if (c == '}')
-  {
-    error(regex_error::mismatched_braces, loc++);
-  }
   else if (c != '\0' && c != '|' && c != '?' && c != '*' && c != '+')
   {
     pos_add(firstpos, loc);
     pos_add(lastpos, loc);
     nullable = false;
     if (c == opt_.e)
-      (void)parse_esc(loc);
+      c = parse_esc(loc);
     else
       ++loc;
   }
@@ -4060,8 +4063,8 @@ bool Pattern::match_hfa(const uint8_t *indexed, size_t size) const
 {
   if (!has_hfa())
     return false;
-  HFA::VisitSet visit[2]; // we alternate and swap two visit bitsets, to produce a new one from the previous
-  bool accept = false; // a flag to indicate that we reached an accept (or dead) state, i.e. a possible match is found
+  HFA::VisitSet visit[2]; // we alternate two state visit bitsets, to produce a new one from the previous
+  bool accept = false; // a flag to indicate that we reached an accept (= dead) state, i.e. a possible match is found
   for (size_t level = 0; level < HFA::MAX_DEPTH && !accept; ++level)
     if (!match_hfa_transitions(level, hfa_.hashes[level], indexed, size, visit[level & 1], visit[~level & 1], accept))
       return false;
@@ -4098,7 +4101,7 @@ bool Pattern::match_hfa_transitions(size_t level, const HFA::Hashes& hashes, con
         {
           HFA::States::const_iterator state = hfa_.states.find(next->first);
           if (state == hfa_.states.end() || state->second.empty())
-            return accept = true; // reached an accepting (dead) state (dead means accept in HFA)
+            return accept = true; // reached an accepting (= dead) state (dead means accept in HFA)
           const HFA::StateSet::const_iterator index_end = state->second.end();
           for (HFA::StateSet::const_iterator index = state->second.begin(); index != index_end; ++index)
             next_visit.set(*index, true);

diff --git a/man/ugrep.1 b/man/ugrep.1
@@ -1,4 +1,4 @@
-.TH UGREP "1" "August 06, 2023" "ugrep 3.12.6" "User Commands"
+.TH UGREP "1" "August 14, 2023" "ugrep 3.12.7" "User Commands"
 .SH NAME
 \fBugrep\fR, \fBug\fR -- file pattern searcher
 .SH SYNOPSIS

diff --git a/src/query.cpp b/src/query.cpp
@@ -452,14 +452,14 @@ void Query::redraw()
     Screen::put( 9, 0, "\033[7mM-Left\033[m \033[7mM-Right\033[m pan ½ page");
     Screen::put(10, 0, "\033[7mM-Up\033[m   \033[7mM-Down\033[m  scroll ½ pg");
 #endif
-    Screen::put(11, 0, "");
-    Screen::put(12, 0, "\033[7mHome\033[m \033[7mEnd\033[m begin/end of line");
+    Screen::put(11, 0, "\033[7mHome\033[m   \033[7mEnd\033[m     of pattern");
+    Screen::put(12, 0, "");
     Screen::put(13, 0, "\033[7m^K\033[m delete after cursor");
     Screen::put(14, 0, "\033[7m^L\033[m refresh screen");
     Screen::put(15, 0, "\033[7m^Q\033[m quick exit and output");
     Screen::put(16, 0, "\033[7m^R\033[m or \033[7mF4\033[m restore bookmark");
     Screen::put(17, 0, "\033[7m^S\033[m scroll to next file/dir");
-    Screen::put(18, 0, "\033[7m^T\033[m toggle colors on/off");
+    Screen::put(18, 0, "\033[7m^T\033[m toggle colors");
     Screen::put(19, 0, "\033[7m^U\033[m delete before cursor");
     Screen::put(20, 0, "\033[7m^V\033[m verbatim character");
     Screen::put(21, 0, "\033[7m^W\033[m scroll back one file/dir");
@@ -502,9 +502,9 @@ void Query::redraw()
     if (!message_)
     {
 #ifdef WITH_MACOS_META_KEY
-      Screen::put(0, 0, "\033[7mF1\033[m help and options:        \033[7m^\033[m=\033[7mCtrl\033[m  \033[7mS-\033[m=\033[7mShift\033[m  \033[7mM-\033[m=\033[7mAlt\033[m/\033[7mOption\033[m or use \033[7m^O\033[m+key");
+      Screen::put(0, 0, "     HELP AND OPTIONS       \033[7m^\033[m=Ctrl  \033[7mS-\033[m=Shift  \033[7mM-\033[m=Alt/Option or \033[7m^O\033[m+key");
 #else
-      Screen::put(0, 0, "\033[7mF1\033[m help and options:        \033[7m^\033[m=\033[7mCtrl\033[m  \033[7mS-\033[m=\033[7mShift\033[m  \033[7mM-\033[m=\033[7mAlt\033[m or use \033[7m^O\033[m+key");
+      Screen::put(0, 0, "     HELP AND OPTIONS       \033[7m^\033[m=Ctrl  \033[7mS-\033[m=Shift  \033[7mM-\033[m=Alt or \033[7m^O\033[m+key");
 #endif
     }
 
@@ -750,6 +750,8 @@ void Query::query_ui()
   eof_        = true;
   buflen_     = 0;
 
+  Screen::clear();
+
   // if -e PATTERN specified, collect patterns on the line to edit
   if (!flag_regexp.empty())
   {
@@ -787,8 +789,11 @@ void Query::query_ui()
 
     move(len_);
   }
-
-  Screen::clear();
+  else
+  {
+    // display a brief help message
+    message("press CTRL-Z or F1 for help");
+  }
 
   set_prompt();
 

diff --git a/src/query.hpp b/src/query.hpp
@@ -63,7 +63,7 @@
 
 // the max time that a message (to confirm a command) is shown at the query line, in steps of 100ms
 #ifndef QUERY_MESSAGE_DELAY
-#define QUERY_MESSAGE_DELAY 12
+#define QUERY_MESSAGE_DELAY 15
 #endif
 
 class Query {

diff --git a/src/ugrep.cpp b/src/ugrep.cpp
@@ -7601,9 +7601,12 @@ void ugrep()
   // reflex::Matcher options
   std::string matcher_options;
 
-  // -Y: permit empty pattern matches
+  // -Y: permit empty pattern matches and match closing ) when no opening (
   if (flag_empty)
+  {
+    convert_flags |= reflex::convert_flag::closing;
     matcher_options.push_back('N');
+  }
 
   // -w: match whole words, i.e. make \< and \> match only left side and right side, respectively
   if (flag_word_regexp)

diff --git a/src/ugrep.hpp b/src/ugrep.hpp
@@ -38,7 +38,7 @@
 #define UGREP_HPP
 
 // ugrep version
-#define UGREP_VERSION "3.12.6"
+#define UGREP_VERSION "3.12.7"
 
 // disable mmap because mmap is almost always slower than the file reading speed improvements since 3.0.0
 #define WITH_NO_MMAP