Add function rex.count and disallow nil/false as the replacement argument in rex.gsub (issue #10).

10 files changed, 141 insertions, 36 deletions

diff --git a/doc/manual.txt b/doc/manual.txt
index f7d063c..50397b2 100644
--- a/doc/manual.txt
+++ b/doc/manual.txt
@@ -223,25 +223,24 @@ This function searches for all matches of the pattern *patt* in the string
 *subj* and replaces them according to the parameters *repl* and *n* (see details
 below).
 
-  +---------+-----------------------------------+-------------------------+-------------+
-  |Parameter|       Description                 |          Type           |Default Value|
-  +=========+===================================+=========================+=============+
-  |  subj   |subject                            |         string          |     n/a     |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |  patt   |regular expression pattern         |string or userdata       |     n/a     |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |  repl   |substitution source                |string, function, table, |     n/a     |
-  |         |                                   |``false`` or ``nil``     |             |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |   [n]   |maximum number of matches to search| number or function      |   ``nil``   |
-  |         |for, or control function, or nil   |                         |             |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |  [cf]   |compilation flags (bitwise OR)     |         number          |     cf_     |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |  [ef]   |execution flags (bitwise OR)       |         number          |     ef_     |
-  +---------+-----------------------------------+-------------------------+-------------+
-  |[larg...]|library-specific arguments         |                         |             |
-  +---------+-----------------------------------+-------------------------+-------------+
+  +---------+-----------------------------------+--------------------------+-------------+
+  |Parameter|       Description                 |          Type            |Default Value|
+  +=========+===================================+==========================+=============+
+  |  subj   |subject                            |         string           |     n/a     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  patt   |regular expression pattern         |string or userdata        |     n/a     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  repl   |substitution source                |string, function or table |     n/a     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |   [n]   |maximum number of matches to search| number or function       |   ``nil``   |
+  |         |for, or control function, or nil   |                          |             |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  [cf]   |compilation flags (bitwise OR)     |         number           |     cf_     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  [ef]   |execution flags (bitwise OR)       |         number           |     ef_     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |[larg...]|library-specific arguments         |                          |             |
+  +---------+-----------------------------------+--------------------------+-------------+
 
 **Returns:**
   1. The subject string with the substitutions made.
@@ -249,9 +248,9 @@ below).
   3. Number of substitutions made.
 
 **Details:**
-  The parameter *repl* can be either a string, a function, a table,
-  ``false`` or ``nil``. On each match made, it is converted into a
-  value *repl_out* that may be used for the replacement.
+  The parameter *repl* can be either a string, a function or a table.
+  On each match made, it is converted into a value *repl_out* that may be used
+  for the replacement.
 
   *repl_out* is generated differently depending on the type of *repl*:
 
@@ -290,11 +289,6 @@ below).
      same rules as for the return value of *repl* call, described in the above
      paragraph.
 
-  4. If *repl* is ``false`` or ``nil``, no replacement is done. Note
-     that, unusually for Lua, if ``repl`` is absent, it is not taken
-     to be ``nil``. This is to prevent programming errors caused by
-     inadvertently missing out *repl*.
-
   Note: Under some circumstances, the value of *repl_out* may be ignored; see
   below_.
 
@@ -378,6 +372,32 @@ subject.
 
 ------------------------------------------------------------
 
+count
+-----
+
+:funcdef:`rex.count (subj, patt, [cf], [ef], [larg...])`
+
+This function counts matches of the pattern *patt* in the string *subj*.
+
+  +---------+-----------------------------------+--------------------------+-------------+
+  |Parameter|       Description                 |          Type            |Default Value|
+  +=========+===================================+==========================+=============+
+  |  subj   |subject                            |         string           |     n/a     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  patt   |regular expression pattern         |string or userdata        |     n/a     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  [cf]   |compilation flags (bitwise OR)     |         number           |     cf_     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |  [ef]   |execution flags (bitwise OR)       |         number           |     ef_     |
+  +---------+-----------------------------------+--------------------------+-------------+
+  |[larg...]|library-specific arguments         |                          |             |
+  +---------+-----------------------------------+--------------------------+-------------+
+
+**Returns:**
+  1. Number of matches found.
+
+------------------------------------------------------------
+
 flags
 -----
 
diff --git a/src/algo.h b/src/algo.h
index e5ac525..e4a98e1 100644
--- a/src/algo.h
+++ b/src/algo.h
@@ -173,10 +173,8 @@ static void checkarg_gsub (lua_State *L, TArgComp *argC, TArgExec *argE) {
   lua_tostring (L, 3);    /* converts number (if any) to string */
   argE->reptype = lua_type (L, 3);
   if (argE->reptype != LUA_TSTRING && argE->reptype != LUA_TTABLE &&
-      argE->reptype != LUA_TFUNCTION && argE->reptype != LUA_TNIL &&
-      (argE->reptype != LUA_TBOOLEAN ||
-       (argE->reptype == LUA_TBOOLEAN && lua_toboolean (L, 3)))) {
-    luaL_typerror (L, 3, "string, table, function, false or nil");
+      argE->reptype != LUA_TFUNCTION) {
+    luaL_typerror (L, 3, "string, table or function");
   }
   argE->funcpos = 3;
   argE->funcpos2 = 4;
@@ -187,6 +185,16 @@ static void checkarg_gsub (lua_State *L, TArgComp *argC, TArgExec *argE) {
 }
 
 
+/* function count (s, patt, [cf], [ef], [larg...]) */
+static void checkarg_count (lua_State *L, TArgComp *argC, TArgExec *argE) {
+  check_subject (L, 1, argE);
+  check_pattern (L, 2, argC);
+  argC->cflags = ALG_GETCFLAGS (L, 3);
+  argE->eflags = luaL_optint (L, 4, ALG_EFLAGS_DFLT);
+  ALG_GETCARGS (L, 5, argC);
+}
+
+
 /* function find  (s, patt, [st], [cf], [ef], [larg...]) */
 /* function match (s, patt, [st], [cf], [ef], [larg...]) */
 static void checkarg_find_func (lua_State *L, TArgComp *argC, TArgExec *argE) {
@@ -337,10 +345,6 @@ static int algf_gsub (lua_State *L) {
       }
     }
     /*----------------------------------------------------------------*/
-    else if (argE.reptype == LUA_TNIL || argE.reptype == LUA_TBOOLEAN) {
-      buffer_addlstring (pBuf, argE.text + from, to - from);
-    }
-    /*----------------------------------------------------------------*/
     if (argE.reptype == LUA_TTABLE || argE.reptype == LUA_TFUNCTION) {
       if (lua_tostring (L, -1)) {
         buffer_addvalue (pBuf, -1);
@@ -425,6 +429,68 @@ static int algf_gsub (lua_State *L) {
 }
 
 
+static int algf_count (lua_State *L) {
+  TUserdata *ud;
+  TArgComp argC;
+  TArgExec argE;
+  int n_match = 0, st = 0, retry;
+  /*------------------------------------------------------------------*/
+  checkarg_count (L, &argC, &argE);
+  if (argC.ud) {
+    ud = (TUserdata*) argC.ud;
+    lua_pushvalue (L, 2);
+  }
+  else compile_regex (L, &argC, &ud);
+  /*------------------------------------------------------------------*/
+  SET_RETRY (retry, 0);
+  while (st <= (int)argE.textlen) {
+    int to, res;
+    res = GSUB_EXEC (ud, &argE, st, retry);
+    if (ALG_NOMATCH (res)) {
+#ifdef ALG_USERETRY
+      if (retry) {
+        if (st < (int)argE.textlen) {  /* advance by 1 char (not replaced) */
+          st += ALG_CHARSIZE;
+          retry = 0;
+          continue;
+        }
+      }
+#endif
+      break;
+    }
+    else if (!ALG_ISMATCH (res)) {
+      return generate_error (L, ud, res);
+    }
+    ++n_match;
+    to = ALG_BASE(st) + ALG_SUBEND(ud,0);
+#ifdef ALG_PULL
+    {
+      int from = ALG_BASE(st) + ALG_SUBBEG(ud,0);
+      if (st < from)
+        st = from;
+    }
+#endif
+    /*----------------------------------------------------------------*/
+    if (st < to) {
+      st = to;
+      SET_RETRY (retry, 0);
+    }
+    else if (st < (int)argE.textlen) {
+#ifdef ALG_USERETRY
+      retry = 1;
+#else
+      /* advance by 1 char (not replaced) */
+      st += ALG_CHARSIZE;
+#endif
+    }
+    else break;
+  }
+  /*------------------------------------------------------------------*/
+  lua_pushinteger (L, n_match);
+  return 1;
+}
+
+
 static int finish_generic_find (lua_State *L, TUserdata *ud, TArgExec *argE,
   int method, int res)
 {
diff --git a/src/gnu/lgnu.c b/src/gnu/lgnu.c
index 54b73ec..a49560e 100644
--- a/src/gnu/lgnu.c
+++ b/src/gnu/lgnu.c
@@ -299,6 +299,7 @@ static const luaL_Reg r_functions[] = {
   { "find",       algf_find },
   { "gmatch",     algf_gmatch },
   { "gsub",       algf_gsub },
+  { "count",      algf_count },
   { "split",      algf_split },
   { "new",        algf_new },
   { "flags",      Gnu_get_flags },
diff --git a/src/oniguruma/lonig.c b/src/oniguruma/lonig.c
index d622445..fa43afb 100644
--- a/src/oniguruma/lonig.c
+++ b/src/oniguruma/lonig.c
@@ -329,6 +329,7 @@ static const luaL_Reg r_functions[] = {
   { "find",             algf_find },
   { "gmatch",           algf_gmatch },
   { "gsub",             algf_gsub },
+  { "count",            algf_count },
   { "split",            algf_split },
   { "new",              algf_new },
   { "flags",            LOnig_get_flags },
diff --git a/src/pcre/lpcre.c b/src/pcre/lpcre.c
index 87d1a56..7550ecf 100644
--- a/src/pcre/lpcre.c
+++ b/src/pcre/lpcre.c
@@ -392,6 +392,7 @@ static const luaL_Reg r_functions[] = {
   { "find",        algf_find },
   { "gmatch",      algf_gmatch },
   { "gsub",        algf_gsub },
+  { "count",       algf_count },
   { "split",       algf_split },
   { "new",         algf_new },
   { "flags",       Lpcre_get_flags },
diff --git a/src/posix/lposix.c b/src/posix/lposix.c
index 52908d8..988a3aa 100644
--- a/src/posix/lposix.c
+++ b/src/posix/lposix.c
@@ -263,6 +263,7 @@ static const luaL_Reg r_functions[] = {
   { "find",       algf_find },
   { "gmatch",     algf_gmatch },
   { "gsub",       algf_gsub },
+  { "count",      algf_count },
   { "split",      algf_split },
   { "new",        algf_new },
   { "flags",      Posix_get_flags },
diff --git a/src/tre/ltre.c b/src/tre/ltre.c
index d734434..6c9ed7c 100644
--- a/src/tre/ltre.c
+++ b/src/tre/ltre.c
@@ -331,6 +331,7 @@ static const luaL_Reg r_functions[] = {
   { "find",          algf_find },
   { "gmatch",        algf_gmatch },
   { "gsub",          algf_gsub },
+  { "count",         algf_count },
   { "match",         algf_match },
   { "split",         algf_split },
   { "config",        Ltre_config },
diff --git a/src/tre/ltre_w.c b/src/tre/ltre_w.c
index c335346..0a6e370 100644
--- a/src/tre/ltre_w.c
+++ b/src/tre/ltre_w.c
@@ -212,6 +212,7 @@ static const luaL_Reg r_functions[] = {
   { "wfind",         algf_find },
   { "wgmatch",       algf_gmatch },
   { "wgsub",         algf_gsub },
+  { "wcount",        algf_count },
   { "wmatch",        algf_match },
   { "wsplit",        algf_split },
   { NULL, NULL }
diff --git a/test/common_sets.lua b/test/common_sets.lua
index 8489030..b8c0c7b 100644
--- a/test/common_sets.lua
+++ b/test/common_sets.lua
@@ -37,6 +37,18 @@ local function set_f_gmatch (lib, flg)
   }
 end
 
+local function set_f_count (lib, flg)
+  return {
+    Name = "Function count",
+    Func = lib.count,
+  --{  subj             patt         results }
+    { {"ab",            lib.new"."}, { 2 } },
+    { {("abcd"):rep(3), "(.)b.(d)"}, { 3 } },
+    { {"abcd",          ".*" },      { 2 } },
+    { {"abc",           "^." },      { 1 } },
+  }
+end
+
 local function set_f_split (lib, flg)
   -- split (s, p, [cf], [ef])
   local function test_split (subj, patt)
@@ -308,6 +320,7 @@ return function (libname)
     set_m_tfind     (lib),
     set_m_find      (lib),
     set_m_match     (lib),
+    set_f_count     (lib),
     set_f_gsub1     (lib),
     set_f_gsub2     (lib),
     set_f_gsub3     (lib),
diff --git a/windows/mingw/rex_onig.mak b/windows/mingw/rex_onig.mak
index deaeb40..5ce7b24 100644
--- a/windows/mingw/rex_onig.mak
+++ b/windows/mingw/rex_onig.mak
@@ -10,7 +10,7 @@ MYINCS   = -I$(REGEXINC)
 MYLIBS   = -lonig -Wl,--enable-auto-import
 OBJ      = lonig.o lonig_f.o common.o
 PROJDIR  = oniguruma
-TESTNAME = onig
+TESTNAME = oniguruma
 
 include _mingw.mak