Various tech of the week

This was a fun week with varied tech. I worked on several projects, bouncing from one to the other as time allowed (and as questions were answered).

C++: I found and fixed one problem, which lead to a number of changes to the output. Some were obvious and expected, others were not. I traced the execution to identify the path from our change to the modified output. Visual Studio is quite good at debugging and examining values, yet I had to supplement it with custom-written logging routines. (The amount of data was larger than I could hold in my head, and some values are hidden from the debugger.)

C#: This was refactoring code to break classes into separate but corresponding 'mutable' and 'immutable' classes. The mutable classes become builder classes (such as ClassNameBuilder) and the immutable classes retained the original name. The division into two classes helps organize the code, and also reduces class size. A win all around, I think.

PDF: Format PDF files and add bookmarks. Perhaps not programming -- I'm not parsing or generating the raw PostScript or PDF files -- yet it was a fun diversion.

Perl and HTML: This was a fun, short task. We had to extract text from an HTML file, adjust the text, and then format the text back into HTML (albeit a different set of HTML). Perl's HTML::Parser class did the heavy lifting of extracting the text, and then it was simple to adjust and re-HTML-ize the text.

Ruby and GraphViz: I was back at the "web site links vizualizer" this week, enhancing the program and fixing some problems. It now grabs the web site files, extracts the links, and builds a GraphViz script for rendering into an image. (GraphViz supports multiple image formats, and we can pick whichever we desire.) Ruby's 'openuri' and 'hpricot' packages helped here, handling the parsing of HTML. (There are lots of HTML parsers available; perhaps I will never have to parse HTML again!)