264 lines
13 KiB
Scala
264 lines
13 KiB
Scala
/* Inspired from http://enear.github.io/2016/03/31/parser-combinators/ */
|
|
|
|
import scala.util.parsing.combinator._
|
|
|
|
sealed trait LaTeXToken
|
|
|
|
case class SPACING (str: String) extends LaTeXToken
|
|
case class RAWTEXT (str: String) extends LaTeXToken
|
|
case class COMMAND (str: String) extends LaTeXToken
|
|
case class BEGINENV(prelude: String, str: String) extends LaTeXToken
|
|
case class ENDENV (prelude: String, str: String) extends LaTeXToken
|
|
case object VBACKSLASH extends LaTeXToken /* verbatim backslash */
|
|
case object VSPACE extends LaTeXToken /* verbatim space */
|
|
case object VTILDE extends LaTeXToken /* verbatim tilde */
|
|
case object VCURLYOPEN extends LaTeXToken /* verbatim curly bracket open */
|
|
case object VCURLYCLOSE extends LaTeXToken /* verbatim curly bracket close */
|
|
case object VBRACKETOPEN extends LaTeXToken /* verbatim square bracket open */
|
|
case object VBRACKETCLOSE extends LaTeXToken /* verbatim square bracket close */
|
|
case object CURLYOPEN extends LaTeXToken
|
|
case object CURLYCLOSE extends LaTeXToken
|
|
case object BRACKETOPEN extends LaTeXToken
|
|
case object BRACKETCLOSE extends LaTeXToken
|
|
|
|
trait LaTeXCompilationError
|
|
case class LaTeXLexerError(msg: String) extends LaTeXCompilationError
|
|
|
|
|
|
object LaTeXLexer extends RegexParsers {
|
|
|
|
override def skipWhitespace = false /* No skipping, we want to maintain the layout
|
|
structure of the LaTeX file */
|
|
/* override val whiteSpace = "[ \t\r\f]+".r -- probably not usefule here */
|
|
|
|
def spacing: Parser[SPACING] = {
|
|
"[ \t\r\f_]*".r ^^ { str => SPACING(str) }
|
|
}
|
|
|
|
def raw_text: Parser[RAWTEXT] = {
|
|
"[^\\{\\}\\\\]+".r ^^ { str => RAWTEXT(str) }
|
|
/* should recognize strings without backslash
|
|
and without curly bracket { */
|
|
}
|
|
|
|
def begin0: Parser[String] = {
|
|
"\\\\begin[^\\{]*".r ^^ (_.toString)
|
|
/* grabs whitespace and also env options ... */
|
|
}
|
|
|
|
def end0: Parser[String] = {
|
|
"\\\\end[^\\{]*".r ^^ (_.toString)
|
|
/* grabs whitespace and also env options ... */
|
|
}
|
|
|
|
def arg: Parser[String] = {
|
|
"\\{[^\\}]*\\}".r ^^ (_.toString)
|
|
}
|
|
|
|
def begin_env = begin0 ~ arg ^^ {case begin_txt ~ arg => BEGINENV(begin_txt,arg)}
|
|
def end_env = end0 ~ arg ^^ {case end_txt ~ arg => ENDENV (end_txt,arg)}
|
|
|
|
def command: Parser[COMMAND] = {
|
|
"\\\\[a-zA-Z0-9][a-zA-Z0-9*]*".r ^^ { str => COMMAND(str) }
|
|
}
|
|
|
|
def vbackslash = "\\\\" ^^ (_ => VBACKSLASH )
|
|
def vspace = "\\ " ^^ (_ => VSPACE )
|
|
def vtilde = "\\~" ^^ (_ => VTILDE )
|
|
def vcurlyopen = "\\{" ^^ (_ => VCURLYOPEN )
|
|
def vcurlyclose = "\\}" ^^ (_ => VCURLYCLOSE )
|
|
def vbracketopen = "\\[" ^^ (_ => VBRACKETOPEN )
|
|
def vbracketclose = "\\]" ^^ (_ => VBRACKETCLOSE )
|
|
def curlyopen = "{" ^^ (_ => CURLYOPEN )
|
|
def curlyclose = "}" ^^ (_ => CURLYCLOSE )
|
|
def bracketopen = "[" ^^ (_ => BRACKETOPEN )
|
|
def bracketclose = "]" ^^ (_ => BRACKETCLOSE )
|
|
|
|
def tokens: Parser[List[LaTeXToken]] = {
|
|
phrase(rep1( raw_text |
|
|
vbackslash | vspace | vtilde |
|
|
vcurlyopen | vcurlyclose | vbracketopen | vbracketclose |
|
|
curlyopen | curlyclose | bracketopen | bracketclose |
|
|
begin_env | end_env | command))
|
|
}
|
|
|
|
def printTokens(tokens: List[LaTeXToken]) : Unit = {
|
|
tokens.headOption match {
|
|
|
|
case Some(SPACING(spaces)) => {println(spaces); printTokens(tokens.tail)}
|
|
|
|
case Some(RAWTEXT(txt)) => {println(txt); printTokens(tokens.tail)}
|
|
|
|
case Some(COMMAND(txt)) => {println(txt); printTokens(tokens.tail)}
|
|
|
|
case Some(BEGINENV(pre,txt)) => {println(pre + txt); printTokens(tokens.tail)}
|
|
|
|
case Some(ENDENV(pre,txt)) => {println(pre + txt); printTokens(tokens.tail)}
|
|
|
|
case Some(VBACKSLASH) => {println("\\\\"); printTokens(tokens.tail)}
|
|
case Some(VSPACE) => {println("\\ "); printTokens(tokens.tail)}
|
|
case Some(VTILDE) => {println("\\~"); printTokens(tokens.tail)}
|
|
case Some(VCURLYOPEN) => {println("\\{"); printTokens(tokens.tail)}
|
|
case Some(VCURLYCLOSE) => {println("\\}"); printTokens(tokens.tail)}
|
|
case Some(VBRACKETOPEN) => {println("\\["); printTokens(tokens.tail)}
|
|
case Some(VBRACKETCLOSE) => {println("\\]"); printTokens(tokens.tail)}
|
|
case Some(CURLYOPEN) => {println("{"); printTokens(tokens.tail)}
|
|
case Some(CURLYCLOSE) => {println("}"); printTokens(tokens.tail)}
|
|
case Some(BRACKETOPEN) => {println("["); printTokens(tokens.tail)}
|
|
case Some(BRACKETCLOSE) => {println("]"); printTokens(tokens.tail)}
|
|
|
|
case Some(token) => {println("+++ INTERNAL ERROR +++");
|
|
printTokens(tokens.tail)}
|
|
|
|
case None => {println("\n\n")}
|
|
}
|
|
}
|
|
|
|
def apply(code: String): Either[LaTeXLexerError, List[LaTeXToken]] = {
|
|
parse(tokens, code) match {
|
|
case NoSuccess(msg, next) => Left(LaTeXLexerError(msg))
|
|
case Success(result, next) => Right(result)
|
|
}
|
|
}
|
|
}
|
|
|
|
/* pre: head(S) <> begin */
|
|
/* post: res = (A,B) where S == A ++ B and
|
|
where begin@A is a begin-end
|
|
balanced list if existing,
|
|
otherwise B = Nil */
|
|
def parse_group0 (begin:LaTeXToken,end:LaTeXToken, level:Int)
|
|
(S:List[LaTeXToken])
|
|
: (List[LaTeXToken],List[LaTeXToken]) = {
|
|
S match {case Nil => (Nil,Nil)
|
|
case h1 :: rest if h1 == begin
|
|
=> (parse_group0 (begin,end, level+1)(rest) match {
|
|
case (s1, rest2) => (h1 :: s1, rest2)
|
|
}
|
|
)
|
|
case h1 :: rest if h1 == end
|
|
=> if(level == 0) { (List(h1), rest) }
|
|
else parse_group0 (begin,end, level-1)(rest) match {
|
|
case (s1, rest2) => (h1 :: s1, rest2)
|
|
}
|
|
case h1 :: rest
|
|
=> (parse_group0 (begin,end,level)(rest) match {
|
|
case (s1, rest2) => (h1 :: s1, rest2)
|
|
}
|
|
)
|
|
}
|
|
}
|
|
|
|
def parse_group (begin:LaTeXToken,end:LaTeXToken)(S:List[LaTeXToken])
|
|
: (List[LaTeXToken],List[LaTeXToken]) =
|
|
parse_group0 (begin,end,0)(S) match {
|
|
case (s1, s2) => (s1.reverse, s2)
|
|
}
|
|
|
|
def purifier (S:List[LaTeXToken]) : List[LaTeXToken] =
|
|
S match {
|
|
case RAWTEXT(s1)::CURLYOPEN::COMMAND("\\isacharminus")::CURLYCLOSE ::rest
|
|
=> purifier(RAWTEXT(s1 + "-") :: rest)
|
|
case RAWTEXT(s1)::CURLYOPEN::COMMAND("\\isacharunderscore")::CURLYCLOSE::rest
|
|
=> purifier(RAWTEXT(s1 + "_") :: rest)
|
|
case COMMAND("\\isacharminus")::CURLYCLOSE::RAWTEXT(s1)::CURLYOPEN::rest
|
|
=> purifier(RAWTEXT("-" + s1) :: rest)
|
|
case CURLYOPEN::COMMAND("\\isacharunderscore")::CURLYCLOSE::RAWTEXT(s1)::rest
|
|
=> purifier(RAWTEXT("_" + s1) :: rest)
|
|
case RAWTEXT(s1)::RAWTEXT(s2)::rest
|
|
=> purifier(RAWTEXT(s1 + s2) :: rest)
|
|
case a :: rest => a :: purifier(rest)
|
|
case Nil => Nil
|
|
}
|
|
|
|
def transducer (S:List[LaTeXToken]) : List[LaTeXToken] =
|
|
S match {
|
|
case COMMAND("\\isacommand")::CURLYOPEN::RAWTEXT(cmd)::CURLYOPEN
|
|
::COMMAND("\\isacharasterisk")::CURLYCLOSE::CURLYCLOSE
|
|
::COMMAND("\\isamarkupfalse")::RAWTEXT("%\n"):: CURLYOPEN::rest
|
|
=> COMMAND("\\"+cmd+"*")::CURLYOPEN::CURLYCLOSE::rest
|
|
case h :: rest => h :: transducer(rest)
|
|
case Nil => Nil
|
|
}
|
|
|
|
/* A more abstract approach, that is based on a two-level parsing approach */
|
|
|
|
object LaTeXParser extends Parsers {
|
|
override type Elem = LaTeXToken
|
|
|
|
class LaTeXTokenReader(tokens: Seq[LaTeXToken]) extends Reader[LaTeXToken] {
|
|
override def first: LaTeXToken = tokens.head
|
|
override def atEnd: Boolean = tokens.isEmpty
|
|
override def pos: Position = NoPosition
|
|
override def rest: Reader[LaTeXToken] = new LaTeXTokenReader(tokens.tail)
|
|
}
|
|
|
|
}
|
|
|
|
/* Unit Testing Zone */
|
|
|
|
>>> LaTeXLexer("\\\\") => Right(List(VBACKSLASH)) // ok
|
|
>>> LaTeXLexer("qsdqsd{dfgdfg") => Right(List(RAWTEXT(qsdqsd), CURLYOPEN, RAWTEXT(dfgdfg))) // ok
|
|
>>> LaTeXLexer("qsdqsd}dfgdfg") => = Right(List(RAWTEXT(qsdqsd), CURLYCLOSE, RAWTEXT(dfgdfg))) // ok
|
|
>>> LaTeXLexer("qsdqsd\tdfgdfg\n\n \n") => ... // ok
|
|
>>> LaTeXLexer("qsdqsd\\\\dfgdfg") => Right(List(RAWTEXT(qsdqsd), VBACKSLASH, RAWTEXT(dfgdfg))
|
|
>>> LaTeXLexer("qsdqsd\\dfgdfg*{sdfsdf}")=> Right(List(RAWTEXT(qsdqsd), COMMAND(\dfgdfg*), CURLYOPEN,
|
|
RAWTEXT(sdfsdf), CURLYCLOSE)) // ok
|
|
>>> LaTeXLexer("qsdqsd\\begin [sdgfsdf] {ghjghj}")
|
|
=> Right(List(RAWTEXT(qsdqsd), BEGINENV(\begin [sdgfsdf] ,{ghjghj})))
|
|
// ok
|
|
/* ... but note the following: */
|
|
>>> LaTeXLexer("qsdqsd\\begin [sdgfsdf] ")
|
|
=> Right(List(RAWTEXT(qsdqsd), COMMAND(\begin), RAWTEXT( [sdgfsdf] )))
|
|
|
|
/* Integration Testing Zone */
|
|
|
|
def sample = "\\isacommand{subsubsection{\\isacharasterisk}}\\isamarkupfalse%\n{\\isacharbrackleft}{\\isachardoublequoteopen}Encoder{\\isacharminus}state{\\isacharminus}diagrams{\\isachardoublequoteclose}{\\isacharbrackright}\\ {\\isacharverbatimopen}\\ Encoder\\ State\\ Diagrams\\ {\\isacharverbatimclose}%"
|
|
|
|
>>> LaTeXLexer(sample) =>
|
|
Right(List(COMMAND(\isacommand), CURLYOPEN, RAWTEXT(subsubsection), CURLYOPEN, COMMAND(\isacharasterisk), CURLYCLOSE, CURLYCLOSE, COMMAND(\isamarkupfalse), RAWTEXT(%
|
|
), CURLYOPEN, COMMAND(\isacharbrackleft), CURLYCLOSE, CURLYOPEN, COMMAND(\isachardoublequoteopen), CURLYCLOSE, RAWTEXT(Encoder), CURLYOPEN, COMMAND(\isacharminus), CURLYCLOSE, RAWTEXT(state), CURLYOPEN, COMMAND(\isacharminus), CURLYCLOSE, RAWTEXT(diagrams), CURLYOPEN, COMMAND(\isachardoublequoteclose), CURLYCLOSE, CURLYOPEN, COMMAND(\isacharbrackright), CURLYCLOSE, VSPACE, CURLYOPEN, COMMAND(\isacharverbatimopen), CURLYCLOSE, VSPACE, RAWTEXT(Encoder), VSPACE, RAWTEXT(State), VSPACE, RAWTEXT(Diagrams), VSPACE, CURLYOPEN, COMMAND(\isacharverbatimclose)...
|
|
|
|
|
|
should become:
|
|
|
|
def sample' = "\\isasubsubsection*{Encoder-state-diagrams}{\\ Encoder\\ State\\ Diagrams\\ }"
|
|
|
|
which could in the LaTeX be set to:
|
|
|
|
def sample'' = "\\isasubsubsection{\\ Encoder\\ State\\ Diagrams\\}\label{sec:Encoder-state-diagrams}"
|
|
|
|
def L = List(CURLYOPEN, COMMAND("\\isachardoublequoteopen"), CURLYCLOSE, RAWTEXT("Encoder")
|
|
, CURLYOPEN, COMMAND("\\isacharminus"), CURLYCLOSE, RAWTEXT("state"), CURLYOPEN, COMMAND("\\isacharminus"), CURLYCLOSE, RAWTEXT("diagrams"), CURLYOPEN, COMMAND("\\isachardoublequoteclose"), CURLYCLOSE, CURLYOPEN, COMMAND("\\isacharbrackright"))
|
|
|
|
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
|
|
|
|
def sample3 = "\\isacommand{text{\\isacharasterisk}}\\isamarkupfalse%{\\isacharbrackleft}wheel{\\isacharunderscore}ass{\\isacharcolon}{\\isacharcolon}exported{\\isacharunderscore}constraint{\\isacharbrackright}\\ {\\isacharverbatimopen}\\ The\\ number\\ of\\ teeth\\ per\\ wheelturn\\ is\\ assumed\\ to\\ be\\isanewline\\ positive{\\isachardot}{\\isacharverbatimclose}"
|
|
|
|
LaTeXLexer(sample3) =>
|
|
Right(List(COMMAND(\isacommand), CURLYOPEN, RAWTEXT(text), CURLYOPEN, COMMAND(\isacharasterisk), CURLYCLOSE, CURLYCLOSE, COMMAND(\isamarkupfalse), RAWTEXT(%), CURLYOPEN, COMMAND(\isacharbrackleft), CURLYCLOSE, RAWTEXT(wheel), CURLYOPEN, COMMAND(\isacharunderscore), CURLYCLOSE, RAWTEXT(ass), CURLYOPEN, COMMAND(\isacharcolon), CURLYCLOSE, CURLYOPEN, COMMAND(\isacharcolon), CURLYCLOSE, RAWTEXT(exported), CURLYOPEN, COMMAND(\isacharunderscore), CURLYCLOSE, RAWTEXT(constraint), CURLYOPEN, COMMAND(\isacharbrackright), CURLYCLOSE, VSPACE, CURLYOPEN, COMMAND(\isacharverbatimopen), CURLYCLOSE, VSPACE, RAWTEXT(The), VSPACE, RAWTEXT(number), VSPACE, RAWTEXT(of), VSPACE, RAWTEXT(teeth), VSPACE, RAWTEXT(per), VSPACE, RAWTEXT...
|
|
|
|
\begin{isamarkuptext*}[wheel_ass::exported_constraint]%
|
|
\\ The\\ number\\ of\\ teeth\\ per\\ wheelturn\\ is\\ assumed\\ to\\ be\\isanewline\\ positive
|
|
\end{isamarkuptext*}\isamarkuptrue
|
|
|
|
or even:
|
|
|
|
\begin{isamarkuptext*}[label=wheel_ass, label_type=exported_constraint, attributes={}]%
|
|
\\ The\\ number\\ of\\ teeth\\ per\\ wheelturn\\ is\\ assumed\\ to\\ be\\isanewline\\ positive
|
|
\end{isamarkuptext*}\isamarkuptrue
|
|
|
|
|
|
|
|
/* Rudimentary Topevel Code */
|
|
object TestLaTeXLexer extends LaTeXLexer {
|
|
|
|
def main(args: Array[String]) = {
|
|
parse(tokens, "johnny 121") match {
|
|
case Success(matched,_) => println(matched)
|
|
case Failure(msg,_) => println("FAILURE: " + msg)
|
|
case Error(msg,_) => println("ERROR: " + msg)
|
|
}
|
|
}
|
|
}
|