Root Zanli
Home
Console
Upload
information
Create File
Create Folder
About
Tools
:
/
home
/
pokigamesmy
/
www
/
static
/
img
/
ALFA_DATA
/
alfasymlink
/
root
/
usr
/
share
/
doc
/
re2c
/
manual
/
eof
/
Filename :
eof.rst_
back
Copy
One of the main problems for the lexer is to know when to stop. There are a few terminating conditions: - the lexer may match some rule (including default rule ``*``) and come to a final state - the lexer may fail to match any rule and come to a default state - the lexer may reach the end of input The first two conditions terminate the lexer in a "natural" way: it comes to a state with no outgoing transitions, and the matching automatically stops. The third condition, end of input, is different: it may happen in any state, and the lexer should be able to handle it. Checking for the end of input interrupts the normal lexer workflow and adds conditional branches to the generated program, therefore it is necessary to minimize the number of such checks. re2c supports a few different methods for end of input handling. Which one to use depends on the complexity of regular expressions, the need for buffering, performance considerations and other factors. Here is a list of all methods: - **Sentinel character.** This method eliminates the need for the end of input checks altogether. It is simple and efficient, but limited to the case when there is a natural "sentinel" character that can never occur in valid input. This character may still occur in invalid input, but it is not allowed by the regular expressions, except perhaps as the last character of a rule. The sentinel character is appended at the end of input and serves as a stop signal: when the lexer reads it, it must be either the end of input, or a syntax error. In both cases the lexer stops. This method is used if ``YYFILL`` is disabled with ``re2c:yyfill:enable = 0;`` and ``re2c:eof`` has the default value -1. | - **Sentinel character with bounds checks.** This method is generic: it allows to handle any input without restrictions on the regular expressions. The idea is to reduce the number of end of input checks by performing them only on certain characters. Similar to the "sentinel character" method, one of the characters is chosen as a "sentinel" and appended at the end of input. However, there is no restriction on where the sentinel character may occur (in fact, any character can be chosen for a sentinel). When the lexer reads this character, it additionally performs a bounds check. If the current position is within bounds, the lexer will resume matching and handle the sentinel character as a regular one. Otherwise it will try to get more input with ``YYFILL`` (unless ``YYFILL`` is disabled). If more input is available, the lexer will rematch the last character and continue as if the sentinel never occurred. Otherwise it is the real end of input, and the lexer will stop. This method is used if ``re2c:eof`` has non-negative value (it should be set to the ordinal of the sentinel character). ``YYFILL`` must be either defined or disabled with ``re2c:yyfill:enable = 0;``. | - **Bounds checks with padding.** This method is the default one. It is generic, and it is usually faster than the "sentinel character with bounds checks" method, but also more complex to use. The idea is to partition the underlying finite-state automaton into strongly connected components (SCCs), and generate only one bounds check per SCC, but make it check for multiple characters at once (enough to cover the longest non-looping path in the SCC). This way the checks are less frequent, which makes the lexer run much faster. If a check shows that there is not enough input, the lexer will invoke ``YYFILL``, which may either supply enough input or else it should not return (in the latter case the lexer will stop). This approach has a problem with matching short lexemes at the end of input, because the multi-character check requires enough characters to cover the longest possible lexeme. To fix this problem, it is necessary to append a few fake characters at the end of input. The padding should not form a valid lexeme suffix to avoid fooling the lexer into matching it as part of the input. The minimum sufficient length of padding is ``YYMAXFILL`` and it is autogenerated by re2c with ``/*!max:re2c*/``. This method is used if ``re2c:yyfill:enable`` has the default nonzero value, and ``re2c:eof`` has the default value -1. ``YYFILL`` must be defined. | - **Custom methods with generic API.** Generic API allows to override basic operations like reading a character, which makes it possible to include the end of input checks as part of them. Such methods are error-prone and should be used with caution, only if other methods cannot be used. These methods are used if generic API is enabled with ``--input custom`` or ``re2c:flags:input = custom;`` and default bounds checks are disabled with ``re2c:yyfill:enable = 0;``. Note that the use of generic API does not imply the use of custom methods, it merely allows it. The following subsections contain an example of each method.