Object-Oriented Legacy System Trace-based Logic Testing

When reengineering legacy systems, it is crucial to assess if the legacy behavior has been preserved or how it changed due to the reengineering effort. Ideally if a legacy system is covered by tests, running the tests on the new version can identify potential differences or discrepancies. However, writing tests for an unknown and large system is difficult due to the lack of internal knowledge. It is especially difficult to bring the system to an appropriate state. Our solution is based on the acknowledgment that one of the few trustable piece of information available when approaching a legacy system is the running system itself. Our approach reifies the execution traces and uses logic programming to express tests on them. Thereby it eliminates the need to programatically bring the system in a particular state, and handles the test-writer a high-level abstraction mechanism to query the trace. The resulting system, called TESTLOG, was used on several real-world case studies to validate our claims.

Isotopes Trace Biogeochemistry and Sources of Cu and Zn in an intertidal soil

River floodplain soils are sinks and potential sources for toxic trace metals like Cu and Zn. We hypothesize that stable Cu and Zn isotope ratios reflect both the mobilization and the sources of metals. We determined the soil properties, the concentrations and partitioning of Cu and Zn, and variations in δ65Cu and δ66Zn values in a core obtained from an Aquic Udifluvent developed on a freshwater intertidal mudflat of the River Elbe, Germany. The core was sampled at 2 cm intervals to a depth of 34 cm, which corresponds to approximately 9 yr of sedimentation. Elevated concentrations of Cu (up to 320 μg g−1) and Zn (up to 2080 μg g−1) indicated anthropogenic pollution. At the time of sampling the redox conditions changed from oxic (Eh 200 to 400 mV, above 22 cm deep) to strongly anoxic conditions (-100 to -200 mV, below 22 cm deep). The δ65Cu values varied systematically with depth (from -0.02 to 0.16‰) and were correlated with the Fe, C, and N concentrations. Although pre-depositional variations cannot be ruled out, the systematic variation with depth suggests post-sedimentation fractionation of δ65Cu in response to seasonally variable organic matter deposition and redox conditions. In contrast, the δ66ZnIRMM values were uniform (from -0.07 to 0.01‰) throughout the core, indicating that the Zn isotopes did not significantly fractionate after deposition and that the Zn sources were homogeneous throughout the sedimentation.