SERVER-6773 Aggregation now supports the split expression.

1 files changed, 66 insertions, 0 deletions

diff --git a/src/mongo/db/pipeline/expression.cpp b/src/mongo/db/pipeline/expression.cpp
index 465c98aa450..f462052d856 100644
--- a/src/mongo/db/pipeline/expression.cpp
+++ b/src/mongo/db/pipeline/expression.cpp
@@ -3069,6 +3069,72 @@ const char* ExpressionSize::getOpName() const {
     return "$size";
 }
 
+/* ----------------------- ExpressionSplit --------------------------- */
+
+namespace {
+
+bool stringHasTokenAtIndex(size_t index, const std::string& input, const std::string& token) {
+    if (token.size() + index > input.size()) {
+        return false;
+    }
+
+    return input.compare(index, token.size(), token) == 0;
+}
+
+}  // namespace
+
+Value ExpressionSplit::evaluateInternal(Variables* vars) const {
+    Value inputArg = vpOperand[0]->evaluateInternal(vars);
+    Value separatorArg = vpOperand[1]->evaluateInternal(vars);
+
+    if (inputArg.nullish() || separatorArg.nullish()) {
+        return Value(BSONNULL);
+    }
+
+    uassert(40085,
+            str::stream() << "$split requires an expression that evaluates to a string as a first "
+                             "argument, found: " << typeName(inputArg.getType()),
+            inputArg.getType() == BSONType::String);
+    uassert(40086,
+            str::stream() << "$split requires an expression that evaluates to a string as a second "
+                             "argument, found: " << typeName(separatorArg.getType()),
+            separatorArg.getType() == BSONType::String);
+
+    std::string input = inputArg.getString();
+    std::string separator = separatorArg.getString();
+
+    uassert(40087, "$split requires a non-empty separator", !separator.empty());
+
+    std::vector<Value> output;
+
+    // Keep track of the index at which the current output string began.
+    size_t splitStartIndex = 0;
+
+    // Iterate through 'input' and check to see if 'separator' matches at any point.
+    for (size_t i = 0; i < input.size();) {
+        if (stringHasTokenAtIndex(i, input, separator)) {
+            // We matched; add the current string to our output and jump ahead.
+            StringData splitString(input.c_str() + splitStartIndex, i - splitStartIndex);
+            output.push_back(Value(splitString));
+            i += separator.size();
+            splitStartIndex = i;
+        } else {
+            // We did not match, continue to the next character.
+            ++i;
+        }
+    }
+
+    StringData splitString(input.c_str() + splitStartIndex, input.size() - splitStartIndex);
+    output.push_back(Value(splitString));
+
+    return Value(output);
+}
+
+REGISTER_EXPRESSION(split, ExpressionSplit::parse);
+const char* ExpressionSplit::getOpName() const {
+    return "$split";
+}
+
 /* ----------------------- ExpressionSqrt ---------------------------- */
 
 Value ExpressionSqrt::evaluateNumericArg(const Value& numericArg) const {