Implementing "includes" when parsing in Attoparsec - parsing

I am writing a DSL for fun. I decided to use attoparsec because I was familiar with it.
I want to implement parsing of includes with relative filenames like this:
include /some/dir/file.ext
or URLs:
include http://blah.com/my/file.ext
So when I'm parsing I expect to read the referenced resource and parse the entire thing, appending its contents to the "outer" parsing state.
The problem is that although the parsing of these statements is easy, I can't run IO (as I understand it) within my Attoparsec parsers.
How do I use Attoparsec to achieve this? Do I chop the initial input up using some string filtering and then parse each "block" into parse and feed accordingly? Essentially a two-pass parse approach?

Attoparsec is pure (Data.Attoparsec.Internal.Types.Parser is not a transformer and doesn’t include IO) so you’re right that you can’t expand includes from within a parser directly.
Splitting the parser into two passes seems like the right approach: one pass acts like the C preprocessor, accepting a file with include statements interleaved with other stuff. The “other stuff” only needs to be basically lexically valid, not your full parser—just like the C preprocessor only cares about tokens and matching parentheses, not matching other brackets or anything semantic. You then replace the includes, producing a fully expanded file that you can give to your existing parser.
If an included file must be syntactically “standalone” in some sense†, then you can parse a whole file first, interleaved with includes, then replace them. For instance:
-- Whatever items you’re parsing.
data Item
-- A reference to an included path.
data Include = Include FilePath
parse :: Parser [Either Include Item]
-- Substitute includes; also calls ‘parse’
-- recursively until no includes remain.
substituteIncludes :: [Either Include Item] -> IO [Item]
† Say, if you’re just using attoparsec for lexing tokens that can’t cross file boundaries anyway, or you’re doing full parsing but want to disallow an include file that contains e.g. unmatched brackets.
The other option is to embed IO in your parser directly by using a different parsing library such as megaparsec, which provides a ParsecT transformer that you can wrap around IO to do IO directly in your parser. I would probably do this for a prototype, but it seems tidier to separate the concerns of parsing and expansion as much as possible.

Related

Is Pug context free?

I was thinking to make a Pug parser but besides the indents are well-known to be context-sensitive (that can be trivially hacked with a lexer feedback loop to make it almost context-free which is adopted by Python), what otherwise makes it not context-free?
XML tags are definitely not context-free, that each starting tag needs to match an end tag, but Pug does not have such restriction, that makes me wonder if we could just parse each starting identifier as a production for a tag root.
The main thing that Pug seems to be missing, at least from a casual scan of its website, is a formal description of its syntax. Or even an informal description. Perhaps I wasn't looking in right places.
Still, based on the examples, it doesn't look awful. There will be some challenges; in particular, it does not have a uniform tokenisation context, so the scanner is going to be complicated, not just because of the indentation issue. (I got the impression from the section on whitespace that the indentation rule is much stricter than Python's, but I didn't find a specification of what it is exactly. It appeared to me that leading whitespace after the two-character indent is significant whitespace. But that doesn't complicate things much; it might even simplify the task.)
What will prove interesting is handling embedded JavaScript. You will at least need to tokenise the embedded JS, and the corner cases in the JS spec make it non-trivial to tokenise without parsing. Anyway, just tokenising isn't sufficient to know where the embedded code terminates. (For the lexical challenge, consider the correct identification of regular expression literals. /= might be the start of a regex or it might be a divide-and-assign operator; how a subsequent { is tokenised will depend on that decision.) Template strings present another challenge (recursive embedding). However, JavaScript parsers do exist, so you might be able to leverage one.
In other words, recognising tag nesting is not going to be the most challenging part of your project. Once you've identified that a given token is a tag, the nesting part is trivial (and context-free) because it is precisely defined by the indentation, so a DEDENT token will terminate the tag.
However, it is worth noting that tag parsing is not particularly challenging for XML (or XML-like HTML variants). If you adopt the XML rule that close tags cannot be omitted (except for self-closing tags), then the tagname in a close tag does not influence the parse of a correct input. (If the tagname in the close tag does not match the close tag in the corresponding open tag, then the input is invalid. But the correspondence between open and close tags doesn't change.) Even if you adopt the HTML-5 rule that close tags cannot be omitted except in the case of a finite list of special-case tagnames, then you could theoretically do the parse with a CFG. (However, the various error recovery rules in HTML-5 are far from context free, so that would only work for input which did not require rematching of close tags.)
Ira Baxter makes precisely this point in the cross-linked post he references in a comment: you can often implement context-sensitive aspects of a language by ignoring them during the parse and detecting them in a subsequent analysis, or even in a semantic predicate during the parse. Correct matching of open- and close tagnames would fall into this category, as would the "declare-before-use" rule in languages where the declaration of an identifier does not influence the parse. (Not true of C or C++, but true in many other languages.)
Even if these aspects cannot be ignored -- as with C typedefs, for example -- the simplest solution might be to use an ambiguous CFG and a parsing technology which produces all possible parses. After the parse forest is generated, you could walk the alternatives and reject the ones which are inconsistent. (In the case of C, that would include an alternative parse in which a name was typedef'd and then used in a context where a typename is not valid.)

"Batteries" for Parsec in Haskell

I am new to Haskell, and I have been trying to write a JSON parser using Parsec as an exercise. This has mostly been going well, I am able to parse lists and objects with relatively little code which is also readable (great!). However, for JSON I also need to parse primitives like
Integers (possibly signed)
Floats (possibly using scientific notation such as "3.4e-8")
Strings with e.g. escaped quotes
I was hoping to find ready to use parsers for things like these as part of Parsec. The closest I get is the Parsec.Tokens module (defines integer and friends), but those parsers require a "language definition" that seems way beyond what I should have to make to parse something as simple as JSON -- it appears to be designed for programming languages.
So my questions are:
Are the functions in Parsec.Token the right way to go here? If so, how to make a suitable language definition?
Are "primitive" parsers for integers etc defined somewhere else? Maybe in another package?
Am I supposed to write these kinds of low-level parsers myself? I can see myself reusing them frequently... (obscure scientific data formats etc.)
I have noticed that a question on this site says Megaparsec has these primitives included [1], but I suppose these cannot be used with parsec.
Related questions:
How do I get Parsec to let me call `read` :: Int?
How to parse an Integer with parsec
Are the functions in Parsec.Token the right way to go here?
Yes, they are. If you don't care about the minutiae specified by a language definition (i.e. you don't plan to use the parsers which depend on them, such as identifier or reserved), just use emptyDef as a default:
import Text.Parsec
import qualified Text.Parsec.Token as P
import Text.Parsec.Language (emptyDef)
lexer = P.makeTokenParser emptyDef
integer = P.integer lexer
As you noted, this feels unnecesarily clunky for your use case. It is worth mentioning that megaparsec (cf. Alec's suggestion) provides a corresponding integer parser without the ceremony. (The flip side is that megaparsec doesn't try to bake in support for e.g. reserved words, but that isn't difficult to implement in the cases you actually need it.)

How to write parser which handles import statements?

I'm using lex & yacc to write a VHDL parser. VHDL has some languages features which make it context sensitive in a manner similar to C. For example, typedef-like constructs which impact whether the parser should tokenize something as an IDENTIFIER vs. TYPEDEF_NAME.
The difficulty comes in when you need to build a symbol table based on another file which is referenced by "use" statements (similar to "import" in Java or Python).
library ieee;
use ieee.std_logic_1164.all;
-- code which uses something defined in ieee.std_logic_1164 package
In C, this is fairly straight-forward because the preprocessor has already combined all of the header files into a single translation unit which can be scanned from top to bottom. But 'use' statements in VHDL are not preprocessor commands.
So, somehow, as I'm parsing the file, I have to recognize when I see a use statement and then go off and parse the relevant file, and then continue parsing the original file with that symbol table.
Is there an elegant way to do this with lex/yacc? I know there is yyrestart but I'm not sure if that's going down the right track.
If you are using flex, then it is pretty easy.
The basic mechanism (including two functioning code samples) is described in the "Multiple Input Buffers" chapter of the flex manual. You can also take a glance at this question on SO.
The parser (yacc/bison) reduction which recognizes the use construction can include the code which calls yy_push_buffer. In the example code, the end of the included file is recognized by the scanner (lex/flex), which simply pops the buffer stack.
Depending on the formal rules of file inclusion, you might want the parser to know that the included file has finished, in order to avoid having syntactic constructs which start in the included file and continue in the includer. (C allows this, even though it is almost always an error; I don't know anything about VHDL, but there are definitely languages which do not allow it.) One possibility is to recursively call the parser in order to parse the included file, which will require a re-entrant ("pure") parser. In that case, the scanner should return an end-of-included-file token when it hits the end of the included file, because your included file grammar production will need to be terminated with such a token.
You may need to worry about the possibility that the parser has already requested the next input token. Most LALR(1) grammars do not depend on the lookahead token for semi-colon terminated statements, and bison usually doesn't request a lookahead token in a context in which it doesn't need it. But that behaviour is not guaranteed by all Posix-compatible yacc implementations and you might be using one which doesn't.
In that case, you would have to preserve the lookahead token so that you can reread it after the included file has been parsed. That would most conveniently be done by stashing the lookahead token somewhere the scanner can see it, and having the scanner return that token (if set) when it sees the end of the included file. In a bison action, you can find the lookahead token in yychar and its semantic value and location (if locations are enabled) are in yylval and yylloc. If bison has not read the lookahead token, the value of yychar will be YYEMPTY, and the simplest possible bison implementation would assert(yychar == YYEMPTY) when it is about to push the input buffer. If the assert fails, you'll need to implement a more sophisticated strategy.

Multiple parsers calling each other?

I am working on a complicated system that uses a number of XML schemas and associated parsers. One of the schemas is used to hold general data that are accessed by all of the other schemas. I would like to maintain this division in the (flex and bison) parsers. So, if I parse the main XML file and get to, say, the tag <matrix>, I would like to call a <matrix> parser as a subroutine, return its content to the calling program and continue parsing there after the </matrix> tag. I have been looking around the net, but I have not found anything useful. Is it even possible to do this?
It seems easiest to maintain the common pieces in a separate file and to split the individual parser components into two more files: Part 1 has the Prologue and the individual grammar rules, part 2 has the epilogue. Then the three files can be concatenated (in a Makefile) before calling the parser:
parser.y: parser.part1 common.inc parser.part2
cat parser.part1 common.inc parser.part2 >parser.y
Your approach is wrong. You shouldn't need a special parser for each distinctive tag. You should parse all tags regardless of their properties and link them to a tree. Afterwards you can validate the tree to ensure a correct consistency of nested tags. If the markup language you're talking about is really that special, then you could create a parser that takes rules describing each tag. In this case parsing and checking are done at the same time, most HTML parsers are implemented like this.

Parsing binary data

I got interested in parser generators. But I don't have the theoretical background. I just read a few things on the internet.
Currently I'm trying to do something with ANTLR
So my questions:
I have a special format of my dataframes:
The first byte of a frame is a tag that describes the nature of the data
The second byte contains the length (number of bytes) of the data itself
Then follows the data itself
The data can contain dataframes itself, and dataframes can be listed one after the other
I hope my description is clear. My questions:
Can I create such a parser with ANTLR that reads the lengs of the frame and then knows when the frame ends?
In ANTLR can I load the different tags I use from a generated file?
Thank you!
I'm not 100% sure about this, but:
Parser generators like antlr require a grammar that is at least context-free
using length-fields in your data makes your grammar not context free (context-sensitive i think)
It is the latter point i'm not sure about - maybe you want to research some more on that.
You probably have to write a packet "parser" yourself (which then has to be a parser for your context-sensitive packet grammar)
Alternatively, you could drop the length field, and use something like s-expressions, JSON or xml; these would be parseable by something generated with antlr.
I think you will be better off to create a hand written binary parser instead of using ANTLR because ANTLR is primarily intended to read and make sense of a text file and not binary data. The lexer part is focused on tokenizing text so trying to make it read binary data instead would be an uphill battle.
It sounds as if your structure would need some kind of recursive way of reading the data although it could be done easier just having a tree structure and then fill it as you read your file.

Resources