Citadel Sourcecode Documentation

Libical was originaly started by Softwarestudio.org; It got into use within a range of Projects; after the upstream lost interest to continue maintaining libical, a couple of big projects started maintaining their bugfixes in forks of libical. After some more or less successfull tries to find a new maintainer, Art managed to get in touch with the original author so we could revive the original http://freeassociation.sf.net project. Next effort was to merge in bugfixes from the various branches, and “get the band back together”. Currently libical is maintained in a cooperative effort of the downstream projects and others.

libcitadel started out as a set of utilities used in several citadel components. Meanwhile it contains some components that might be of use for others:

StrBuf goes new ways to handle strings in C more efficiently. Main problems with general C-String tackled are:

• getting their length is expensive (strlen)
• managing their memory is complicated:
  • most functions don't know about the buffer size, or if, can't readjust it on themselves to beir their results
  • if you manage memory to strict, you have to alloc, memcpy & free a lot
  • alloc / free doesn't scale well in multithreadded environemnts if invoked to frequent (at least with current glibc implementations), so doing less of them is a good thing.
• while strcat might look like a smart idea, its not. StrBuf commands with an Append in their name start their work at the end of a current string, like strcat does. Flush it first, if you want it to be an empty string.
• StrBuf hides its content to you, think of them as being 'private' in a C++ object.

StrBuf manages its memory by itself; it will double its size until everything fits in. StrBuf has a default buffer size, which you can set at the start of your application. There are also tools to evaluate such values (counting reallocs, slags pace on delete..) For not wasting memory nor doing to frequent reallocs, you can hint StrBuf for instances that are going to be extraordinary big by using NewStrBufPlain(NULL, <size>);

StrBuf also has a bunch of string-Transcoding functions which will escape / transcode / de-escape contents for commonly used patterns, like Quoted-printable encoding, URL-encoding, XML/HTML encoding, gzip compression.

StrBufs are assumed to have ASCII or UTF-8 contents in many Places; many other places don't care about the content, even pure binary data like images are no problem. Its up to you to treat it appropriate, or use for example the QP decoder with its internal iconv wrapper to get utf8.

For more details see the current StrBuf Doxygen documentation.