X-Git-Url: https://git.distorted.org.uk/~mdw/sod/blobdiff_plain/a588c77a273681e3cdc85d15fc44f3ddb7da9224..aacae2739986636d9e7921cc0712d506e25c71b8:/doc/parsing.tex diff --git a/doc/parsing.tex b/doc/parsing.tex index 1c4c3cd..8aeb26b 100644 --- a/doc/parsing.tex +++ b/doc/parsing.tex @@ -7,7 +7,7 @@ %%%----- Licensing notice --------------------------------------------------- %%% -%%% This file is part of the Sensble Object Design, an object system for C. +%%% This file is part of the Sensible Object Design, an object system for C. %%% %%% SOD is free software; you can redistribute it and/or modify %%% it under the terms of the GNU General Public License as published by @@ -40,8 +40,164 @@ parser is said to have \emph{failed}, and the result is a list of \emph{indicators}. Finally, the consumed flag is non-nil if the parser consumed any input items. +\begin{describe}{fun}{combine-parser-failures @ @> @} +\end{describe} + +\begin{describe}{fun}{parse-empty \&optional @ @> @} +\end{describe} + +\begin{describe}{fun} + {parse-fail @ \&optional @ @> @} +\end{describe} + %%%-------------------------------------------------------------------------- -\section{File locations} +\section{File locations} \label{sec:parsing.floc} + +\begin{describe}{cls}{file-location} +\end{describe} + +\begin{describe}{fun}{file-location-p @ @> @} +\end{describe} + +\begin{describe}{fun} + {make-file-location @ \&optional @ @ + @> @} +\end{describe} + +\begin{describe*} + {\dhead{fun}{file-location-filename @ @> @} + \dhead{fun}{file-location-line @ @> @} + \dhead{fun}{file-location-column @ @> @}} +\end{describe*} + +\begin{describe}{gf}{file-location @ @> @} + \begin{describe}{meth}{file-location} + {file-location (@ file-location) @> @} + \end{describe} + \begin{describe}{meth}{stream} + {file-location (@ stream) @> @} + \end{describe} + \begin{describe}{meth}{t} + {file-location (@ t) @> @} + \end{describe} +\end{describe} + +\begin{describe}{cls}{condition-with-location (condition) \&key :location} +\end{describe} + +\begin{describe}{meth}{condition-with-location} + {file-location (@ condition-with-location) @> @} +\end{describe} + +\begin{describe*} + {\quad\=\quad\=\kill + \dhead{cls} + {error-with-location (condition-with-location error) \\ \> + \&key :location} + \dhead{cls} + {warning-with-location (condition-with-location warning) \\ \> + \&key :location} + \dhead{cls} + {information-with-location (condition-with-location information) \\ \> + \&key :location} + \dhead{cls} + {enclosing-error-with-location + (enclosing-error-with-location error) \\ \> + \&key :condition :location} + \dhead{cls} + {enclosing-warning-with-location + (enclosing-condition-with-location warning) \\ \> + \&key :condition :location} + \dhead{cls} + {enclosing-information-with-location + (enclosing-condition-with-location information) \\ \> + \&key :condition :location} + \dhead{cls} + {simple-condition-with-location + (condition-with-location simple-condition) \\ \> + \&key :format-control :format-arguments :location} + \dhead{cls} + {simple-error-with-location + (error-with-location simple-error) \\ \> + \&key :format-control :format-arguments :location} + \dhead{cls} + {simple-warning-with-location + (warning-with-location simple-warning) \\ \> + \&key :format-control :format-arguments :location} + \dhead{cls} + {simple-information-with-location + (information-with-location simple-information) \\ \> + \&key :format-control :format-arguments :location}} +\end{describe*} + +\begin{describe}{gf} + {enclosing-condition-with-location-type @ @> @} +\end{describe} + +\begin{describe}{fun} + {make-condition-with-location @ @ + @ \&rest @ + \nlret @} +\end{describe} + +\begin{describe*} + {\dhead{fun}{error-with-location @ @ \&rest @} + \dhead{fun}{cerror-with-location @ @ + @ \&rest @} + \dhead{fun}{cerror*-with-location @ @ \&rest @} + \dhead{fun}{warn-with-location @ @ \&rest @}} +\end{describe*} + +\begin{describe*} + {\dhead{cls}{parser-error (error) \\ \ind + \&key :expected :found \-} + \dhead{gf}{parser-error-expected @ @> @} + \dhead{gf}{parser-error-found @ @> @}} +\end{describe*} + +\begin{describe}{fun} + {report-parser-error @ @ @ @} +\end{describe} + +\begin{describe*} + {\quad\=\kill + \dhead{cls}{base-lexer-error (error-with-location) \&key :location} + \dhead{cls}{simple-lexer-error + (base-lexer-error simple-error-with-location) \\\> + \&key :format-control :format-arguments :location} + \dhead{cls}{base-syntax-error (error-with-location) \&key :location} + \dhead{cls}{simple-syntax-error + (base-syntax-error simple-error-with-location) \\\> + \&key :format-control :format-arguments :location}} +\end{describe*} + +\begin{describe}{mac} + {with-default-error-location (@) @^* @
^* + @> @^*} +\end{describe} + +\begin{describe}{gf}{classify-condition @ @> @} + \begin{describe*} + {\dhead{meth}{error} + {classify-condition (@ error) @> @} + \dhead{meth}{warning} + {classify-condition (@ warning) @> @} + \dhead{meth}{information} + {classify-condition (@ information) + @> @} + \dhead{meth}{base-lexer-error} + {classify-condition (@ base-lexer-error) + @> @} + \dhead{meth}{base-syntax-error} + {classify-condition (@ base-syntax-error) + @> @}} + \end{describe*} +\end{describe} + +\begin{describe}{mac} + {count-and-report-errors () @^* @^* + @> @ @ @} +\end{describe} %%%-------------------------------------------------------------------------- \section{Scanners} \label{sec:parsing.scanner} @@ -93,7 +249,7 @@ depends on the type of scanner. Given a captured place, the scanner can be rewound to the position held in it. Depending on how the scanner works, holding onto a captured place might -consume a lot of memory or case poor performance. For example, if the +consume a lot of memory or cause poor performance. For example, if the scanner is reading from an input stream, having a captured place means that data from that point on must be buffered in case the program needs to rewind the scanner and read that data again. Therefore it's possible to @@ -120,18 +276,20 @@ the scanner and read that data again. Therefore it's possible to \end{describe} \begin{describe}{mac} - {with-scanner-place (@ @) @^* @> @^*} - Capture the @'s current position as a place, evaluate the - @s as an implicit progn with the variable @ bound to the captured - place. When control leaves the @s, the place is released. The return - values are the values of the final @. + {with-scanner-place (@ @) @^* @^* + @> @^*} + Capture the @'s current position as a place, evaluate the @s + as an implicit progn with the variable @ bound to the captured + place. When control leaves the @s, the place is released. The + return values are the values of the final @. \end{describe} \subsection{Scanner file-location protocol} \label{sec:parsing.scanner.floc} -Some scanners participate in the file-location protocol (\xref{sec:floc}). -They implement a method on @|file-location| which collects the necessary -information using scanner-specific functions described here. +Some scanners participate in the file-location protocol +(\xref{sec:parsing.floc}). They implement a method on @|file-location| which +collects the necessary information using scanner-specific functions described +here. \begin{describe}{fun}{scanner-file-location @ @> @} Return a @|file-location| object describing the current position of the @@ -147,32 +305,17 @@ information using scanner-specific functions described here. specialized methods on @|file-location|. \end{describe} -\begin{describe}{gf}{scanner-filename @ @> @} - Return the name of the file the scanner is currently processing, as a - string, or nil if the filename is not known. -\end{describe} +\begin{describe*} + {\dhead{gf}{scanner-filename @ @> @} + \dhead{gf}{scanner-line @ @> @} + \dhead{gf}{scanner-column @ @> @}} + Return the filename, line and column components of the @'s current + position, for use in assembling a @: see the + @|scanner-file-location| function. -\begin{describe}{meth}{scanner-filename (@ t) @> @} - Returns nil. -\end{describe} - -\begin{describe}{gf}{scanner-line @ @> @} - Return the line number of the @'s current position, as an integer, - or nil if the line number is not known. -\end{describe} - -\begin{describe}{meth}{scanner-line (@ t) @> @} - Returns nil. -\end{describe} - -\begin{describe}{gf}{scanner-column @ @> @} - Return the column number of the @'s current position, as an - integer, or nil if the column number is not known. -\end{describe} - -\begin{describe}{meth}{scanner-column (@ t) @> @} - Returns nil. -\end{describe} + There are default methods on all three generic functions which simply + return nil. +\end{describe*} \subsection{Character scanners} \label{sec:parsing.scanner.char} @@ -192,7 +335,7 @@ Character scanners are scanners which read sequences of characters. \begin{describe}{gf}{scanner-unread @ @} Rewind the @ by one step. The @ must be the previous current character, and becomes the current character again. It is an error - if: the @ has reached end-of-file; the @ is never been + if: the @ has reached end-of-file; the @ has never been stepped; or @ was not the previous current character. \end{describe} @@ -286,15 +429,15 @@ file-location protocols. \begin{describe}{gf} {charbuf-scanner-map @ @ \&optional @ - \nlret @ @ @} + \nlret @ @ @} Read characters from the @'s buffers. This is intended to be an efficient and versatile interface for reading characters from a scanner in bulk. The function @ is invoked repeatedly, as if by \begin{prog} - (multiple-value-bind (@ @) \\ \ind\ind - (funcall @ @ @ @) \- \\ + (multiple-value-bind (@ @) \\ \ind\ind + (funcall @ @ @ @) \-\\ \textrm\ldots) \end{prog} The argument @ is a simple string; @ and @ are two @@ -302,7 +445,7 @@ file-location protocols. @ (inclusive) and @ (exclusive) should be processed. If @'s return value @ is nil then @ is ignored: the function has consumed the entire buffer and wishes to read more. If - @ is non-nil, then it must be a fixnum such that $@ \le + @ is non-nil, then @ must be a fixnum such that $@ \le @ \le @$: the function has consumed the buffer as far as @ (exclusive) and has completed successfully. @@ -310,8 +453,8 @@ file-location protocols. fails: the @ function is called with no arguments, and is expected to return two values. If omitted, @ defaults to \begin{prog} - (lambda () \\ \ind - (values nil nil))% + (lambda () \\ \ind + (values nil nil)) \end{prog} The @|charbuf-scanner-map| function returns three values. The first value @@ -356,11 +499,359 @@ file-location protocols. \end{describe} %%%-------------------------------------------------------------------------- -\section{Parsing macros} +\section{Parsing syntax} -%%%-------------------------------------------------------------------------- +\begin{describe}{gf}{expand-parser-spec @ @ @> @} +\end{describe} + +\begin{describe}{gf} + {expand-parser-form @ @ @ @> @} +\end{describe} + +\begin{describe}{gf}{wrap-parser @ @ @> @} +\end{describe} + +\begin{describe}{mac} + {defparse @ (@[[ :context (@ @) @]] + @^*) \\ \ind + @[[ @^* @! @ @]] \\ + @^* + \-\nlret @} +\end{describe} + +\begin{describe}{mac} + {with-parser-context + (@ @{ @ @ @}^*) \\ \ind + @^* \\ + @^* + \-\nlret @^*} +\end{describe} + +\begin{describe}{lmac} + {parse @ @> @ @ @} +\end{describe} + +\begin{describe}{mac} + {parser @ + @[[ @^* @! @ @]] + @ + @> @} +\end{describe} + +\begin{describe}{gf}{parser-at-eof-p @ @> @} +\end{describe} + +\begin{describe}{gf}{parser-step @ @> @} +\end{describe} + +\begin{describe}{mac} + {if-parse (@[[ \=:result @ @! + :expected @ @! \+\\ + :consumedp @ @]]) \-\\ \ind\ind + @ \-\\ + @ \\ + @[@@] + \-\nlret @^*} +\end{describe} + +\begin{describe}{mac} + {when-parse (@[@@]) @ \\ \ind + @^* + \-\nlret @^*} +\end{describe} + +\begin{describe}{mac} + {cond-parse (@[[ \=:result @ @! + :expected @ @! \+\\ + :consumedp @ @]]) \-\\ \ind + @{ (@ @^*) @}^* + \-\nlret @^*} +\end{describe} + +\begin{describe}{parse}{:eof} +\end{describe} + +\begin{describe}{parseform}{lisp @^*} +\end{describe} + +\begin{describe}{parseform}{label @} +\end{describe} + +\begin{describe}{parse}{t} +\end{describe} + +\begin{describe}{parseform}{t @} +\end{describe} + +\begin{describe}{parse}{nil} +\end{describe} + +\begin{describe}{parseform}{nil @} +\end{describe} + +\begin{describe}{parseform}{when @ @} +\end{describe} + +\begin{describe}{parseform} + {seq (@{ @ @! + (@[@@] @) @}^*) \\ \ind + @^*} +\end{describe} + +\begin{describe}{parseform}{and @^*} +\end{describe} + +\begin{describe}{parseform}{or @^*} +\end{describe} + +\begin{describe}{parseform}{? @ @[@@]} +\end{describe} + +\begin{describe}{parseform} + {many (\=@ @ @ \+\\ + @[[ \=:new @ @! :final @ @! \+\\ + :min @ @! :max @ @! \\ + :commitp @ @]]) \-\-\\ \ind + @ @[@@]} +\end{describe} + +\begin{describe}{parseform} + {list (@[[ :min @ @! :max @ @! + :commitp @ @]]) \\ \ind + @ @[@@]} +\end{describe} + +\begin{describe}{parseform} + {skip-many (@[[ :min @ @! :max @ @! + :commitp @ @]]) \\ \ind + @ @[@@]} +\end{describe} + +\begin{describe}{fun}{call-pluggable-parser @ \&rest @} +\end{describe} + +\begin{describe}{parseform}{plug @ @^*} +\end{describe} + +\begin{describe}{fun} + {pluggable-parser-add @ @ @} +\end{describe} + +\begin{describe}{mac} + {define-pluggable-parser @ @ @ + @[[ @^* @! @ @]] + @^*} +\end{describe} + +\begin{describe}{gf}{parser-capture-place @ @> @} +\end{describe} + +\begin{describe}{gf}{parser-restore-place @ @ @> @} +\end{describe} + +\begin{describe}{gf}{parser-release-place @ @ @> @} +\end{describe} + +\begin{describe}{gf} + {parser-places-must-be-released-p @ @> @} +\end{describe} + +\begin{describe}{mac} + {with-parser-place (@ @) + @[[ @^* @! @ @]] + @^*} +\end{describe} + +\begin{describe}{parseform}{peek @} +\end{describe} + +\begin{describe}{parseform}{commit} +\end{describe} + +\begin{describe}{cls}{character-parser-context () \&key} +\end{describe} + +\begin{describe}{gf}{parser-current-char @ @> @} +\end{describe} + +\begin{describe}{parseform} + {if-char (@[@@]) @ @ @} +\end{describe} + +\begin{describe}{parseform}{char @} +\end{describe} + +\begin{describe}{parse}[char]{@} +\end{describe} + +\begin{describe}{parse}[string]{@} +\end{describe} + +\begin{describe}{parse}{:any} +\end{describe} + +\begin{describe}{parseform}{satisfies @} +\end{describe} + +\begin{describe}{parseform}{not @} +\end{describe} + +\begin{describe}{parseform}{filter @} +\end{describe} + +\begin{describe}{parse}{:whitespace} +\end{describe} + +\begin{describe}{cls}{token-parser-context () \&key} +\end{describe} + +\begin{describe}{gf}{parser-token-type @ @> @} +\end{describe} + +\begin{describe}{gf}{parser-token-value @ @> @} +\end{describe} + +\begin{describe}{parseform}{token @ @[@@] @[:peekp @@]} +\end{describe} + +\begin{describe}{parse}[atom]{@} +\end{describe} + +\begin{describe}{cls}{scanner-context () \&key :scanner} +\end{describe} + +\begin{describe}{gf}{parse-scanner @ @> @} +\end{describe} + +\begin{describe}{cls} + {character-scanner-context (scanner-context character-parser-context) + \&key :scanner} +\end{describe} + +\begin{describe}{cls} + {token-scanner-context (scanner-context token-parser-context) + \&key :scanner} +\end{describe} + +\begin{describe}{gf}{push-operator @ @} +\end{describe} + +\begin{describe}{gf}{push-value @ @} +\end{describe} + +\begin{describe}{gf}{apply-operator @ @} +\end{describe} + +\begin{describe}{gf}{operator-push-action @ @} +\end{describe} + +\begin{describe}{parseform} + {expr \=(@[[ :nestedp @ @]]) \+\\ + @ @ + @ @} +\end{describe} + +\begin{describe}{gf}{operator-left-precedence @ @> @} +\end{describe} + +\begin{describe}{gf}{operator-right-precedence @ @> @} +\end{describe} + +\begin{describe}{gf}{operator-associativity @ @> @} +\end{describe} + +\begin{describe}{cls}{prefix-operator () \&key} +\end{describe} + +\begin{describe}{cls}{simple-operator () \&key :name :function} +\end{describe} + +\begin{describe}{cls} + {simple-unary-operator (simple-operator) \&key :name :function} +\end{describe} + +\begin{describe*} + {\quad\=\kill + \dhead{cls}{simple-binary-operator (simple-operator) \\ \> + \&key :name :function + :lprec :rprec :associativity} + \dhead{cls}{simple-postfix-operator (simple-unary-operator) \\ \> + \&key :name :function :lprec :rprec} + \dhead{cls}{simple-prefix-operator + (prefix-operator simple-unary-operator) \\ \> + \&key :name :function :rprec}} +\end{describe*} + +\begin{describe*} + {\dhead{mac}{preop @ (@ @) + @^* @^* + @> @} + \dhead{mac}{postop @ + (@ @ @[[ :rprec @ @]]) + @^* @^* + \nlret @} + \dhead{mac}{binop @ (@ @ @ @) + @^*@^* + @> @}} +\end{describe*} + +\begin{describe*} + {\dhead{cls}{parenthesis () \&key :tag} + \dhead{cls}{open-parenthesis (parenthesis prefix-operator) \&key :tag} + \dhead{cls}{close-parenthesis (parenthesis) \&key :tag}} +\end{describe*} + +\begin{describe*} + {\dhead{fun}{lparen @ @> @} + \dhead{fun}{rparen @ @> @}} +\end{describe*} + +%%%------------------------------------------------------------------------- \section{Lexical analyser} +\begin{describe}{cls} + {sod-token-scanner (token-scanner) + \&key :filename (:line 1) (:column 0) :char-scanner} +\end{describe} + +\begin{describe}{fun}{define-indicator @ @} +\end{describe} + +\begin{describe*} + {\dhead{cls}{lexer-error (parser-error base-lexer-error) \\ \ind + \&key :expected :found :location \-} + \dhead{cls}{syntax-error (parser-error base-syntax-error) \\ \ind + \&key :expected :found :location \-}} +\end{describe*} + +\begin{describe}{fun} + {syntax-error @ @ \&key :continuep :location} +\end{describe} + +\begin{describe}{fun} + {lexer-error @ @ \&key :location} +\end{describe} + +\begin{describe}{parseform} + {skip-until (@[[ :keep-end @ @]]) @^*} +\end{describe} + +\begin{describe}{parseform} + {error (@[[ :ignore-unconsumed @ @! + :force-process @ @]]) \\ \ind\ind + @ @ \-\\ + @^* \\ + @^*} +\end{describe} + +\begin{describe}{parseform}{must @ @[@@]} +\end{describe} + +\begin{describe}{fun} + {scan-comment @ + @> @ @ @} +\end{describe} + %%%----- That's all, folks -------------------------------------------------- %%% Local variables: