fredcpp is a third-party C++ library to access Federal Reserve Economic Data (FRED®) database via the web-API as published and maintained by Federal Reserve Bank of St. Louis.

FRED® represents an extensive collection of economic data series, compiled and regularly updated from a large number of data sources. It is one of the world's most comprehensive online economic databases. Please visit the official St. Louis Fed FRED® site for more information about FRED® and details on how to obtain FRED API access. There you can also find excellent FRED API documentation which references the supported queries and parameters.

DISCLAIMER: fredcpp (FRED API Client for C++) is not affiliated in any way with St. Louis Fed and as such is not officially maintained or otherwise supported by St. Louis Fed.

fredcpp is a free software; the use, copy, and distribution rights are granted under the terms of the @ref fredcpplicense | MIT License.

The main objective behind fredcpp is to shield C++ users from the details of constructing and executing web-requests to FRED API, then parsing the results; at the same time to provide an expressive way to specify FRED data-queries in C++ context.

fredcpp main features:

• access to full set of available FRED entities (series, release, category, source) and related sub-queries
• uniform and expressive public C++ interface
• built-in HTTP request execution, XML response parsing, and central logging facilities
• limited external dependencies (effectively only for HTTP handling)
• stand-alone ready with the supplied implementation for binding with cURL-library
• flexible integration with external HTTP stack, XML parser, and logging framework alternatives

Additionally, the following concerns are kept in focus:

• portability (relies mainly on STL support)
• ease of build configuration (CMake-based)
• ease of use (example programs supplied)
• extensibility (includes suites of unit and acceptance tests with Google's gtest)
• customization (source-code available to changes per virtues of free software)

Refer to @ref fredcppusage | fredcpp Usage page for more details.

Updates and details about the current version listed in @ref fredcppchangelog | fredcpp ChangeLog.

A shortcut to get started with fredcpp, assuming you are familiar with CMake, have a C++ compiler, cURL (libcurl) library and headers installed.

HTTPS-NOTE: Starting 08/18/2015, FRED API requires HTTPS access for added security. fredcpp has been updated to generate HTTPS requests. However, fredcpp users need to make sure that HTTP executor used with fredcpp has been properly configured to handle HTTPS requests. CA Certificate used by FRED API must be available for verification in the user's environment.

To configure CurlHttpClient supplied with fredcpp, download the CA Certificate bundle cacert.pem from cURL CA Extract page to the fredcpp root -- this is needed to run acceptance tests. For use once installed, set the evironment variable CURL_CA_BUNDLE to point to the downloaded file. Alternatively, call CurlHttpClient::withCACertFile(path) method to configure this explicitly from the user code, by default the file is expected in the running process' work directory.

NOTE: On Windows generate CMake files with "NMake Makefiles" for MSVC NMAKE.

From fredcpp source directory:

echo "$FRED_API_KEY" > api.key

mkdir build
cd build
cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=RelWithDebInfo ..
make clean all
make test
ctest -V

cd examples
./example1
cat example.txt

Successful?

make install

Use fredcpp library in your projects:

#include <fredcpp/fredcpp.h>

#include <fredcpp/external/CurlHttpClient.h>
#include <fredcpp/external/PugiXmlParser.h>
#include <fredcpp/external/SimpleLogger.h>

......

  fredcpp::Api api;

  api.withLogger( fredcpp::external::SimpleLogger::getInstance() )
     .withExecutor( fredcpp::external::CurlHttpClient::getInstance() )
     .withParser( fredcpp::external::PugiXmlParser::getInstance() )
     ;


  // set your FRED API key

  std::string yourApiKey;

.......

  api.withKey( yourApiKey );


  // request the recent 10 observations for series "CBIC1"

  fredcpp::Api response;

  api.get( ApiRequestBuilder::SeriesObservations("CBIC1")
                              .withSort("desc")
                              .withLimit("10")
          , response );

  if ( !response.good() ) {
    FREDCPP_LOG_ERROR( response.error );
    exit( response.error.status );

  } else {
    FRECPP_LOG_INFO( response );
  }

......

Compile and link against the installed fredcpp library (assuming installation with default /usr/local prefix):

g++ example.cpp -lfredcpp -lcurl

More docs:

make doc

See more examples in examples source sub-directory. See the official FRED API documentation for details on available requests and parameters.

CMake is needed to build and install fredcpp library. CMake supports a variety of platforms and C++ compilers and automatically selects comprehensive compiling options; however, it also allows manual configuration if needed. Please download and install an up-to-date CMake release for your platform if required.

1. Unpack the fredcpp source-code into a directory and create a build directory inside it (for CMake out-of-source build):

   mkdir build
   

2. Change to the build directory and configure CMake build-parameters (use CMake GUI tools, e.g. cmake-gui OR specify these on CMake command-line):

   • Generator type - by default CMake chooses platform-specific generators; for consistency we prefer makefiles:

     -G "Unix Makefiles"
      OR
     -G "NMake Makefiles"
     

   • FRED API key file location (FREDCPP_API_KEY_FILE) - needed here for acceptance testing. By default key is expected in api.key file in the fredcpp source directory

   • Build type (CMAKE_BUILD_TYPE) - Debug by CMake default, but for optimized performance should be set to RelWithDebInfo or Release

   • Installation prefix (CMAKE_INSTALL_PREFIX) - path to directory under which to install the resulting fredcpp library and headers. CMake provides platform-specific defaults.

3. Generate build files - once the parameters are set from CMake GUI, do Configure then Generate to produce the make files/projects:

   Example Windows command line to both configure and generate makefiles (for NMAKE) at once:

   cmake -G "NMake Makefiles" -DCMAKE_BUILD_TYPE=RelWithDebInfo ..
   

4. Build all fredcpp artifacts - static library, test-runners, and example programs will be built in respective sub-directories under the build directory.

   With "NMake Makefiles" on Windows, NMAKE should be used for make command:

   make clean all
   

5. Run unit and acceptance tests - expect all tests passed:

   NOTE: a valid FRED API key is needed for acceptance tests.

   make test

   To see a more detailed test report, especially when any of the tests has failed:

   ctest -V
   

6. Optionally, examine and run the supplied examples - source code and executable files are in examples sub-directory of source or build directory respectively.

7. Optionally, generate documentation files (needs Doxygen installed) - the resulting files are in doc sub-directory under build:

   make doc
   

8. Install fredcpp library and headers:

   NOTE: By default CMake uses platform-specific install locations (e.g. /usr/local on Linux, %ProgramFiles% on Windows etc.). Define CMAKE_INSTALL_PREFIX CMake variable to override the default install location. E.g. to install under a directory defined by $FREDCPP_HOME environment variable:

   cmake -DCMAKE_INSTALL_PREFIX="$FREDCPP_HOME" .. make install

9. Compile and link your program with fredcpp referring to the installed include and library locations.

We have already been using fredcpp in several production applications and consider it stable for our needs; a couple of new features are planned (additional HTTP stack bindings). However, with the public release made widely available, it could generate more feedback - perhaps, feature requests or improvements, and not too many bug-reports, hopefully.

A public issue tracker may be eventually set up to address possible user-feedback. Meanwhile, please direct your feedback to fredcpp GitHub project page.

Of course, Fork us on GitHub! as it goes to welcome your contribution.