From 96010ea9ae2d814e2322813028553a70ef6fb820 Mon Sep 17 00:00:00 2001
From: Josh Holtrop <jholtrop+git@gmail.com>
Date: Sat, 15 May 2010 00:43:55 -0400
Subject: [PATCH] lexer working with pcre, not building any objects for
 handling tokens yet

---
 .gitignore           |  1 +
 tests/build/Makefile |  2 +-
 tests/build/itest.I  | 11 ++++++++---
 tests/build/main.cc  | 16 ++++++++++++++++
 tmpl/parser.cc       | 12 +++++++++++-
 5 files changed, 37 insertions(+), 5 deletions(-)
 create mode 100644 tests/build/main.cc

diff --git a/.gitignore b/.gitignore
index 5f8b7e8..bcf881b 100644
--- a/.gitignore
+++ b/.gitignore
@@ -6,3 +6,4 @@ tags
 tmpl.*
 tests/*/itest.cc
 tests/*/itest.h
+tests/*/test
diff --git a/tests/build/Makefile b/tests/build/Makefile
index 078fd11..158a360 100644
--- a/tests/build/Makefile
+++ b/tests/build/Makefile
@@ -6,7 +6,7 @@ LDFLAGS  := -lpcre
 
 all: $(TARGET)
 
-$(TARGET):
+$(TARGET): $(shell which imbecile)
 	imbecile $(I_SOURCE).I
 	$(CXX) -o $@ *.cc $(LDFLAGS)
 
diff --git a/tests/build/itest.I b/tests/build/itest.I
index 9c99b75..986921e 100644
--- a/tests/build/itest.I
+++ b/tests/build/itest.I
@@ -1,9 +1,14 @@
 
 [tokens]
 
-ASSIGN     :=
-DASSIGN    :==
-IDENTIFIER [a-zA-Z_][a-zA-Z_0-9]*
+AND         and
+OR          or
+NOT         not
+LPAREN      \(
+RPAREN      \)
+WS          \s+
+EQUALS      =
+IDENTIFIER  [a-zA-Z_][a-zA-Z_0-9]*
 
 [rules]
 
diff --git a/tests/build/main.cc b/tests/build/main.cc
new file mode 100644
index 0000000..924c9f4
--- /dev/null
+++ b/tests/build/main.cc
@@ -0,0 +1,16 @@
+
+#include <sstream>
+#include <string>
+
+#include "itest.h"
+
+using namespace std;
+
+int main(int argc, char * argv[])
+{
+    Parser p;
+    stringstream t(string(
+                "hi there (one and two and three and four) or (two = nine)"
+                ));
+    p.parse(t);
+}
diff --git a/tmpl/parser.cc b/tmpl/parser.cc
index e03a327..3e15416 100644
--- a/tmpl/parser.cc
+++ b/tmpl/parser.cc
@@ -88,7 +88,7 @@ bool I_CLASSNAME::parse(istream & i)
     while (buff_pos < buff_size)
     {
         int longest_match_length = 0;
-        int longest_match_index;
+        int longest_match_index = -1;
         int longest_match_ovector[ovector_size];
         for (int i = 0; i < sizeof(tokens)/sizeof(tokens[0]); i++)
         {
@@ -108,6 +108,16 @@ bool I_CLASSNAME::parse(istream & i)
                 }
             }
         }
+        if (longest_match_index >= 0)
+        {
+            cout << "Matched a " << tokens[longest_match_index].name << endl;
+            buff_pos += longest_match_length;
+        }
+        else
+        {
+            /* no pattern matched the input at the current position */
+            return false;
+        }
     }
 }