Following settings can be made on the command line:
For seeking in character streams a temporary buffer is required. By means of this macro its size can be specified.
The buffer limit code is by default 0x0. If it is intended that this character code appears inside patterns, then it need to be reset by this setting.
Buffers in quex keep a certain fallback region when loading new content into the buffer. This prevents the lexer from reloading backwards if the input pointer is navigated below the current reload position. Note, that this has only effect in case of manual navigation of the input pointer, or in scanners with pre-conditions.
The size of a buffer on which the quex engine does its analysis. The larger the buffer, the less reloads are required. But at a certain buffer size (usually about 32Kb) an increase in buffer size does not show measurable performance increase.
The buffer size must be greater or equal the largest conceivable lexeme that the lexer can match.
When using IBM’s ICU library for character conversion (command line option --icu), the intermediate buffer size is determined by this setting.
Defines the size of the mode stack. It defines the number of ‘sub mode calls’ that can be done via the .push_mode and .pop_mode member functions. See also section Mode Transitions.
Defines the initial number of elements of the token queue–if it is enable (see also command line option --no-token-queue)
Defines the size of the safety border. In other words, this corresponds to the maximum number of tokens to be sent caused by a pattern match ‘-1’. If each pattern sends only one token, it can be set safely to zero.
When using a converter (--icu or --iconv) this is the size of the buffer where the original stream data is read into.
When printing the text content of the default token type, it can be converted to UTF8. The conversion requires a temporary buffer whose size is defined by means of this macro.
