Sun Microsystems, Inc.

Wayne C Gramlich

Soren J Tirfing

John Robert Rose

Terrence C Miller

Eduardo Pelegri-Llopart

Achut Reddy

Shyam Desirazu

Thomas Preisler

The present invention provides an improved method and apparatus for generating executable computer code for an application program written in C++ source code. In typical prior art systems, application program source code that has not itself been modified must still generally be recompiled in the event that object-oriented class definitions used by the application program and contained in separate header files have been modified. The methods and apparatus of the present invention reduce the need for such recompilation, by using procedural interfaces to implement object-oriented interfaces at the compiled code level. Thus, in accordance with the present invention, compiled header file code is generated that includes accessors for accessing object instances of the class definitions, each of the accessors being a procedure operative to access the object instances of the corresponding class definition. Compiled application program code is generated which replaces object references with procedure calls to appropriate accessors. In this way, the application program and the header file class definitions remain relatively independent of each other with respect to implementation details, even at the compiled code level.

Method and apparatus for generating executable code from object-oriented

A system and method for providing annotation overlays from diverse sources of commentary for World-Wide Web documents is disclosed. Sources of commentary contribute annotation overlays regarding particular documents on the World-Wide Web. The annotation overlays from a particular source are stored on one or more overlay servers, which are connected to the Web. A user of a Web browser opens an annotation proxy server between the Web browser and the Web servers that intercepts all documents retrieved by the Web browser and merges with the retrieved documents commentary from sources designated by the user of the Web browser that refer to the requested documents. Multiple annotation overlay proxies can be serially connected. The annotation proxy can perform the merge operation by first creating a local annotation directory of annotation overlays from sources designated by the user then, when the user requests a document, merging with the requested document information only from the annotation directory. Alternatively, the annotation proxy can perform the merge operation on the fly by pulling the annotation overlays directly from the Web servers without the use of a local annotation directory.

System for annotation overlay proxy configured to retrieve associated

This Continuation-In-Part describes a part of this run-time debugger operation which is designated the 'Fix-and-Continue' invention. This invention permits a user to begin a debugging session wherein if an error in the code is encountered, the user can edit the corresponding source code to correct the error and then execute a 'Fix and Continue' command all without leaving the debugging session. The Fix and Continue code calls the compiler to recompile the source code file with the edited text in it, receives the resulting recompiled object code file from the compiler, uses the dynamic linker to link the recompiled object code into the target application program process, patches the previous version of this same object code fie to refer to the newly recompiled code, resets any required variables and registers, resets the program counter to the line of code being executed when the error was discovered. The debugger then continues in the debug session thereby saving the time it would ordinarily take to quit the debug session, relink and reload the target program and start the debug session once again.

Method and apparatus for a fast debugger fix and continue operation

The invention provides a method and apparatus for dynamic patching for run-time error checking. With the present invention, a program to be error checked is read into memory where a debugging module resides. Such in-memory copy of the program is scanned for load objects. The load objects comprise libraries used by the program being error checked, as well as the main routines of such program. Next, a list of patch sites in the load objects is created and these patch sites corresponds to address locations in the load objects to be patched. Address space is then allocated for sections of the patch area where calls to real checking code resides. Finally, patches are written out to the in-memory copy of the program to be error checked. The original instruction in a patch site is replaced with branch to patch area instruction and the original instruction displaced is stored in the patch site's corresponding section of the patch area. The call to real checking code in this section of the patch area calls an error checking code. After the error checking is complete, the process transfers from the patch area to the next instruction in the segment of load object that was patched. In addition, the present invention is able to load new load objects. Such new load objects are scanned and the aforementioned patching process is repeated for patching the new load objects.

Method and apparatus for run-time error checking using dynamic patching

In the method and apparatus of the present invention a file to be added to the database is given a unique name that is dependent upon the contents of the file such that, when the contents of the source file changes, the name of the database component file to be added to the database also changes. Conversely, if two files of the same name have the same information contained therein, the same file name will be generated and the duplication of information in the database is prevented by providing a simple test that checks for the existence of the name of the database file before the generation and addition of the new file to the database. If the file name exists in the database, information is already contained in the database and the file is not generated and added to the database information. Preferably the name of the file is generated by computing a hash value from the contents of the file concatenating the hash value to the name of the source file. Because the source file name is used in conjunction with the hash value to construct the database file name, the hash value does not have to be unique for all files but only for those source files having the same name.

Method and apparatus for the naming of database component files to avoid

Each database component file contains information regarding the text contained in the source file which enables the user to quickly determine the frequency of occurrence of the specified text and the location of the specified text in the source file. For each textual word (referred to herein as a 'symbol') an entry in the database file is provided containing symbol information. The symbol information comprises the symbol name, symbol type and line number in the source file where the symbol is located. Line indentification information is also provided which contains the line numbers in the source file and corresponding line lengths and hash values which are computed from the contents of the lines of the text in the source file. Before a line of text identified in a query is displayed to the user indicating a result of the query, the line identification information provides the means to verify that the line of text identified in the symbol information is the same line of text contained in the source file. The hash value and line length corresponding to the line text is compared to a line length and hash value computed for the text retrieved from the source file. If the computed line length and hash value does not match the line length and hash value contained in the line identification information, the text does not match the database reference because the source file has been changed subsequent to the generation of the database. Thus, the problems that arise because the source file version does not match the version of the database component file is prevented.

Method and apparatus for searching database component files to retrieve

A database system for text files is employed which comprises a compiler which receives as input a description of the source file types or languages and search filter and outputs a tag set definition file and filter file to be used by the browsing mechanism. The tags set definition file is input to the database builder mechanism and is used by the database builder to select and translate the information from the input source file to a database component file. Thus, by providing a description of the source file type, any type of source file in any language can be input to the database system of the present invention to generate a database component file which may be subsequently reviewed and searched by the same browsing mechanism.

User extensible, language sensitive database system

A file system locking mechanism is provided which prevents the errors which arise when race conditions occur in a multitasking system. Atomic operators are utilized to perform certain operations for the creation and renaming of files and directories in the file system wherein if a race condition occurs, the atomic operations fail and the file creation process is halted to a predetermined state whereby the errors which occur during race conditions are prevented.

Growing community of inventors

Sunnyvale, CA, United States of America

Wayne C Gramlich