|
|
writer.axdx
Srbová, Kateřina ; Fincato, André (referee) ; Fišerová, Barbora (advisor)
Write a sentence, read it again, select the part that should be in italics, mark it with the mouse or by moving your fingers on the touch pad or keyboard (oopsies, one character off), repeat the marking (this time correctly, ugh), find the right button or use the appropriate keyboard shortcut, write another sentence, paragraph, chapter, essay, book... What happens if the internet crashes, autosaving fails, the version of the software is not up to date? Most of the commonly used text writing tools offer a lot of formatting options, thus directing attention away from the writing itself. This thesis explores their features, (inter)compatibility and user experience. Its goal is to create a specification of the writer.axdx tool which should be used as a text editor for projects within the community-software.axdx (an ecosystem of tools and methods serving communities). The goal of this tool is to bring the focus back to the writing itself, in a format that is not constrained by proprietary software and does not break when opened in another editor. To provide the ability to work with text according to one's own needs, which are otherwise defined by the needs or methods of the so-called power user. To have everything that is needed for writing and nothing that is not.
|
|
|
automatic recognition of encoding and language
Hron, Michal ; Pinkas, Otakar (advisor) ; Pavlíčková, Jarmila (referee)
Processing simple or complex texts (MIME type - application) often requires automatic recognition of encoding and language. Some types of files or pages contain an internal information about the encoding method. There might be some conflicts, however, eg. between the HTTP header and the meta tag. Sometimes it may be useful to verify the accuracy of the file encoding even when the encoding is known. In case that the identification of encoding is not available it is necessary to use a method of automatic recognition of encoding and language. One such method is an n-grams method. It has been used many times to categorize texts in many programs and in various programming languages. Based on tests results, it seems that the automatic recognition of the Czech language and other Slavic languages is less successful than the recognition of Western languages. Determining the reasons and searching for better solutions is therefore beneficial even nowadays. The length of the input text and the use of various languages in one document are important parameters. This thesis does not take text consisting of sentences in several different languages into consideration. In addition to a basic analysis of the topic, the thesis also includes a software solution to particular problems in a form of independent programs or plug-ins.
|
guest ::
