Add some dtl-cpp stuff

.gitignore

@@ -25,3 +25,4 @@ gr
 /src/rtags-esprima.js
 /.ninja_deps
 /compile_commands.json
+/tests/dtlcpptest/dtlcpptest

.gitmodules

@@ -19,3 +19,6 @@
 [submodule "src/estraverse"]
 	path = src/estraverse
 	url = https://github.com/Constellation/estraverse.git
+[submodule "src/dtl-cpp"]
+	path = src/dtl-cpp
+	url = https://code.google.com/p/dtl-cpp/

tests/dtlcpptest/CMakeLists.txt

@@ -0,0 +1,3 @@
+cmake_minimum_required(VERSION 2.8)
+include_directories(${CMAKE_CURRENT_LIST_DIR}/../../src/dtl-cpp/dtl)
+add_executable(dtlcpptest main.cpp)

tests/dtlcpptest/main.cpp

@@ -0,0 +1,98 @@
+#include <string>
+#include <stdio.h>
+#include <string.h>
+#include <vector>
+#include <stdint.h>
+#include "dtl.hpp"
+
+uint32_t translatePoint(uint32_t offset, const dtl::Diff<char, std::string> &diff)
+{
+    const uint32_t old = offset;
+    const dtl::Ses<char> ses = diff.getSes();
+    const std::vector<std::pair<char, dtl::elemInfo> > seq = ses.getSequence();
+    int i = 0;
+    for (std::vector<std::pair<char, dtl::elemInfo> >::const_iterator it = seq.begin(); it != seq.end(); ++it) {
+        if (it->second.type == 1) {
+            if (it->second.afterIdx >= old || !offset) {
+                // printf("0 Breaking at(%c) because %d %lld\n", it->first, offset, it->second.afterIdx);
+                break;
+            }
+            --offset;
+            // printf("Reducing offset to %d for character %lld %lld %c\n", offset,
+            //        it->second.afterIdx, it->second.beforeIdx, it->first);
+        } else if (it->second.type == -1) {
+            if (it->second.beforeIdx >= offset) {
+                // printf("1 Breaking at(%c) because %d %lld\n", it->first, offset, it->second.beforeIdx);
+                break;
+            }
+            ++offset;
+            // printf("Increasing offset to %d for character %lld %lld %c\n", offset,
+            //        it->second.afterIdx, it->second.beforeIdx, it->first);
+        } else if (it->second.afterIdx > old) {
+            // printf("Breaking at %c because of %lld/%d\n", it->first, it->second.afterIdx, offset);
+            break;
+        }
+
+        // printf("%d: %c %d %d %d\n", i++, it->first, it->second.type, it->second
+
+        // printf("%c %lld %lld %d\n", it->first,
+        //        it->second.beforeIdx,
+        //        it->second.afterIdx,
+        //        it->second.type);
+    }
+    // printf("%d => %d\n", old, offset);
+    return offset;
+}
+
+int main(int argc, char **argv)
+{
+    std::string a, b;
+    bool string = false;
+    uint32_t offset = 10;
+    for (int i=1; i<argc; ++i) {
+        if (!strcmp(argv[i], "--string") || !strcmp(argv[i], "-s")) {
+            string = true;
+        } else if (a.empty()) {
+            a = argv[i];
+        } else if (b.empty()) {
+            b = argv[i];
+        } else {
+            offset = atoi(argv[i]);
+        }
+    }
+    if (a.empty() || b.empty()) {
+        fprintf(stderr, "Not enough data\n");
+        return 1;
+    }
+
+    if (!string) {
+        std::string *files[] = { &a, &b, 0 };
+        for (int i=0; files[i]; ++i) {
+            FILE *f = fopen(files[i]->c_str(), "r");
+            if (!f) {
+                fprintf(stderr, "Can't open %s for reading\n", files[i]->c_str());
+                return 1;
+            }
+            files[i]->clear();
+            char buf[1024];
+            while (true) {
+                const int r = fread(buf, sizeof(char), sizeof(buf), f);
+                if (r <= 0)
+                    break;
+                files[i]->append(buf, r);
+            }
+            fclose(f);
+        }
+    }
+
+    dtl::Diff<char, std::string> diff(a, b);
+    diff.compose();
+    diff.composeUnifiedHunks();
+    // diff.printSES(std::cout);
+    dtl::Diff<char, std::string>::printUnifiedFormat(diff.getUniHunks());
+    const uint32_t translated = translatePoint(offset, diff);
+    printf("%d => %d : '%c' => '%c'\n", offset, translated, b.at(offset), a.at(translated));
+
+    return 0;
+}
+