Show simple item record

dc.rights.licenseRestricted to current Rensselaer faculty, staff and students. Access inquiries may be directed to the Rensselaer Libraries.
dc.contributorSchupp, Sibylle
dc.contributorField, John
dc.contributorKrishnamoorthy, M. S.
dc.contributorMilanova, Ana
dc.contributorMusser, David R.
dc.contributor.authorGregor, Douglas
dc.date.accessioned2021-11-03T08:19:21Z
dc.date.available2021-11-03T08:19:21Z
dc.date.created2015-04-10T16:07:11Z
dc.date.issued2004-05
dc.identifier.urihttps://hdl.handle.net/20.500.13015/1341
dc.descriptionMay 2004
dc.descriptionSchool of Science
dc.description.abstractSoftware developers are increasingly reliant on well-tested, reusable software libraries that can rapidly decrease development time by eliminating the need to reimplement standard algorithms and data structures. Generic software libraries in particular, such as the C++ Standard Template Library (STL), provide efficient, reusable solutions that can be customized for nearly any computing environment.
dc.description.abstractThe methodologies and analyses presented are implemented in the STLlint static checker. STLlint is able to detect errors in the use of the Standard Template Library within C++ user programs, including semantic checks on the abstract properties of STL algorithms. Our experiments have shown that STLlint achieves a low false positive rate and is capable of diagnosing common STL usage errors.
dc.description.abstractThis thesis also introduces an organization methodology for extensible specification for generic software libraries based on algorithm concepts. This methodology minimized the effort required to introduce new algorithms, data structures, and semantic properties into generic library specifications.
dc.description.abstractThis thesis introduces a higher-level static analysis that employs executable specifications of software libraries to reduce the complexities of library analysis and improve analysis precision. Moreover, our analysis exploits object-oriented abstractions to improve the precision of analysis for a very important class of iterator movement problems. Other abstractions, such as subroutine abstractions and pointer, are effectively handled with symbolic analysis methods, even in program loops with nontraditional induction variables.
dc.description.abstractGeneric software libraries introduce a large number of abstractions, such as subroutines and classes, which simplify the task of a software developer. Theses abstractions, however, complicate static analyses, causing traditional static analysis techniques to produce poor, imprecise results. In contrast, a higher-level static analysis exploits these same abstractions to produce precise program analysis information at the library's level of abstraction, enabling library-level optimization and verification.
dc.language.isoENG
dc.publisherRensselaer Polytechnic Institute, Troy, NY
dc.relation.ispartofRensselaer Theses and Dissertations Online Collection
dc.subjectComputer science
dc.titleHigh-level static analysis for generic libraries
dc.typeElectronic thesis
dc.typeThesis
dc.digitool.pid174911
dc.digitool.pid174912
dc.digitool.pid174913
dc.rights.holderThis electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
dc.description.degreePhD
dc.relation.departmentDept. of Computer Science


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record