# -*- coding: utf-8; -*- """Find and expand dialects, i.e. whole-module source and AST transformers.""" __all__ = ["Dialect", "DialectExpander"] import ast import functools import re import sys from .colorizer import setcolor, colorize, ColorScheme from .coreutils import ismacroimport, get_macros from .unparser import unparse_with_fallbacks from .utils import format_macrofunction class Dialect: """Base class for dialects. `expander`: the `DialectExpander` instance. The expander provides this automatically. Stored as `self.expander`. During dialect expansion, the source location info of the dialect-import statement that invoked this dialect-import is available as `self.lineno` and `self.col_offset`. You can pass those to `mcpyrate.splicing.splice_dialect` to automatically mark the lines from your dialect template as coming from that dialect-import in the user source code. """ def __init__(self, expander): self.expander = expander self.lineno = None self.col_offset = None def transform_source(self, text): """Override this to add a whole-module source transformer to your dialect. If not overridden, the default is to return `NotImplemented`, which tells the expander this dialect does not provide a source transformer. Rarely needed. Because we don't (yet) have a generic, extensible tokenizer for "Python-plus" with extended surface syntax, not to mention that none of the available Python dev tools support any such, this is currently essentially a per-module hook to plug in a transpiler that compiles source code from some other programming language into macro-enabled Python. The dialect system autodetects the text encoding the same way Python itself does. That is, it reads the magic comment at the top of the source file (such as `# -*- coding: utf-8; -*-`), and assumes `utf-8` if not present. So your source transformer gets its input as `str` (**not** `bytes`). The input is the full source text of the module, as a string (`str`). Output should be the transformed source text, as a string (`str`). To put it all together, this allows implementing things like:: # -*- coding: utf-8; -*- '''See https://en.wikipedia.org/wiki/Brainfuck#Examples''' from mylibrary import dialects, Brainfuck ++++++[>++++++++++++<-]>. >++++++++++[>++++++++++<-]>+. +++++++..+++.>++++[>+++++++++++<-]>. <+++[>----<-]>.<<<<<+++[>+++++<-]>. >>.+++.------.--------.>>+. while having that source code in a file ending in `.py`, executable by `macropython`. Implementing the actual BF->Python transpiler is left as an exercise to the reader. Maybe compare how Matthew Butterick did this in Racket: https://beautifulracket.com/bf/intro.html """ return NotImplemented def transform_ast(self, tree): """Override this to add a whole-module AST transformer to your dialect. Dialect AST transformers run before the macro expander. If not overridden, the default is to return `NotImplemented`, which tells the expander this dialect does not provide an AST transformer. This is useful to define custom dialects that use Python's surface syntax, but with different semantics. Another use case is to plug in an AST optimizer. Input is the full AST of the module (in standard Python AST format), but with the dialect-import for this dialect already transformed away, into an absolute module import for the module defining the dialect. Output should be the transformed AST. To easily splice `tree.body` into your template, see the utility `mcpyrate.splicing.splice_dialect` (it automatically handles future-imports, macro-imports, dialect-imports, the magic `__all__`, and the module docstring). As an example, see the `dialects` module in `unpythonic` for example dialects. https://github.com/Technologicat/unpythonic To give a flavor, *Lispython* is essentially Python with TCO, and implicit `return` in tail position:: # -*- coding: utf-8; -*- '''Lispython example.''' from unpythonic.dialects import dialects, Lispython def fact(n): def f(k, acc): if k == 1: return acc f(k - 1, k * acc) f(n, acc=1) assert fact(4) == 24 fact(5000) # no crash """ return NotImplemented def postprocess_ast(self, tree): """Like `transform_ast`, but runs after the macro expander.""" return NotImplemented _message_header = colorize("**StepExpansion: ", ColorScheme.HEADING1) class StepExpansion(Dialect): # actually part of public API of mcpyrate.debug, for discoverability """[dialect] Show each step of expansion while dialect-expanding the module. Usage:: from mcpyrate.debug import dialects, StepExpansion When the dialect expander invokes the source transformer of this dialect, it causes the expander to enter debug mode from that point on. It will show the source code (or unparsed AST, as appropriate) after each transformer. So, to see the whole chain, place the import for this dialect first. This dialect has no other effects. """ def transform_source(self, text): # We pass through the input (instead of returning `NotImplemented`) to # consider this as having taken a step, thus triggering the debug mode # output printer. (If this was the first dialect applied, our output is # actually the original input; but there's no way to know to show it # before this dialect has run.) self._enable_debugmode() return text # This exists only so that dialect-enabled AST compiles can use `StepExpansion`, too. def transform_ast(self, tree): # If the debug mode was already enabled, we're ok - behave as if this method didn't exist # (to suppress the debug printout of a step that did nothing). if self.expander.debugmode: return NotImplemented self._enable_debugmode() return tree def _enable_debugmode(self): self.expander.debugmode = True c, CS = setcolor, ColorScheme msg = f"{c(CS.SOURCEFILENAME)}{self.expander.filename} {c(CS.HEADING1)}enabled {c(CS.ATTENTION)}DialectExpander debug mode {c(CS.HEADING1)}while taking step {self.expander._step + 1}.{c()}" print(_message_header + msg, file=sys.stderr) # -------------------------------------------------------------------------------- _dialectimport = re.compile(r"^from\s+([.0-9a-zA-z_]+)\s+import dialects,\s+([^(\\]+)\s*$", flags=re.MULTILINE) class DialectExpander: """The dialect expander. Due to modularity requirements introduced by `mcpyrate`'s support for multi-phase compilation (see the module `mcpyrate.multiphase`), this class is a bit technical to use. See `mcpyrate.compiler`. Roughly, for a single-phase compile:: dexpander = DialectExpander(filename=...) text = dexpander.transform_source(text) ... # parse `text` into an AST `tree` here tree, dialect_instances = dexpander.transform_ast(tree) ... # macro-expand `tree` here tree = dexpander.postprocess_ast(tree, dialect_instances) """ def __init__(self, filename): """`filename`: full path to `.py` file being expanded, for module name resolution and error messages.""" self.filename = filename self.debugmode = False # to enable, `from mcpyrate.debug import dialects, StepExpansion` self._step = 0 self._seen = set() def transform_source(self, text): """Apply all whole-module source transformers. Return value is the transformed text. """ new_text, _ = self._transform(text, kind="source", find_dialectimport=self.find_dialectimport_source, transform="transform_source", format_for_display=lambda text: text) return new_text def transform_ast(self, tree): """Apply all whole-module AST transformers. Return value is `transformed_tree, dialect_instances`. `dialect_instances` is a `list` of the `Dialect` instances that ran, in the order in which they ran. That list can be passed to `postprocess_ast` to run their AST postprocessors. """ formatter = functools.partial(unparse_with_fallbacks, debug=True, color=True) return self._transform(tree, kind="AST", find_dialectimport=self.find_dialectimport_ast, transform="transform_ast", format_for_display=formatter) def _transform(self, content, *, kind, find_dialectimport, transform, format_for_display): c, CS = setcolor, ColorScheme if self.debugmode: plural = "s" if self._step != 1 else "" msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}before dialect {c(CS.TRANSFORMERKIND)}{kind} {c(CS.HEADING1)}transformers ({self._step} step{plural} total):{c()}\n" print(_message_header + msg, file=sys.stderr) print(format_for_display(content), file=sys.stderr) # We collect and return the dialect object instances so that both # `transform_ast` and `postprocess_ast` can use the same instances. The # dialect-imports vanish at `transform_ast`, so the information about # which dialects ran (and hence should have their postprocessors run) # must be preserved separately from `tree` itself. # # We could store this data in `self`, but keeping nontrivial mutable # state is so last decade. dialect_instances = [] while True: theimport = find_dialectimport(content) if theimport: module_absname, bindings, lineno, col_offset = theimport else: # no more dialects break for dialectname, cls in bindings.items(): if not (isinstance(cls, type) and issubclass(cls, Dialect)): raise TypeError(f"{self.filename}: {module_absname}.{dialectname} is not a `Dialect`, got {repr(cls)}") try: dialect = cls(expander=self) except Exception as err: raise ImportError(f"Unexpected exception while instantiating dialect `{module_absname}.{dialectname}`") from err # make the dialect-import source location info available to the transformers dialect.lineno = lineno dialect.col_offset = col_offset try: transformer_method = getattr(dialect, transform) except AttributeError as err: raise ImportError(f"Dialect `{module_absname}.{dialectname}` missing required transformer method `{transform}`") from err try: result = transformer_method(content) except Exception as err: raise ImportError(f"Unexpected exception in dialect transformer `{module_absname}.{dialectname}.{transform}`") from err # We should run the corresponding `postprocess_ast` even if the # dialect doesn't use `transform_ast`. dialect_instances.append(dialect) if result is NotImplemented: continue # no step taken; proceed to next binding if not result: raise ImportError(f"Dialect transformer `{module_absname}.{dialectname}.{transform}` returned an empty result.") content = result self._step += 1 if self.debugmode: msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}after {c(CS.DIALECTTRANSFORMERNAME)}{module_absname}.{dialectname}.{transform} {c(CS.HEADING1)}(step {self._step}):{c()}\n" print(_message_header + msg, file=sys.stderr) print(format_for_display(content), file=sys.stderr) if self.debugmode: plural = "s" if self._step != 1 else "" msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}completed all dialect {c(CS.TRANSFORMERKIND)}{kind} {c(CS.HEADING1)}transforms ({self._step} step{plural} total).{c()}" print(_message_header + msg, file=sys.stderr) return content, dialect_instances def postprocess_ast(self, tree, dialect_instances): """Apply AST postprocessors of dialect objects in `dialect_instances`. Return value is the postprocessed tree. """ format_for_display = functools.partial(unparse_with_fallbacks, debug=True, color=True) c, CS = setcolor, ColorScheme if self.debugmode: plural = "s" if self._step != 1 else "" msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}before dialect {c(CS.TRANSFORMERKIND)}AST postprocessors {c(CS.HEADING1)}({self._step} step{plural} total):{c()}\n" print(_message_header + msg, file=sys.stderr) print(format_for_display(tree), file=sys.stderr) content = tree for dialect in dialect_instances: try: transformer_method = dialect.postprocess_ast except AttributeError as err: raise ImportError(f"Dialect `{format_macrofunction(dialect)}` missing required transformer method `postprocess_ast`") from err try: result = transformer_method(content) except Exception as err: raise ImportError(f"Unexpected exception in dialect transformer `{format_macrofunction(dialect)}.postprocess_ast`") from err if result is NotImplemented: continue # no step taken; proceed to next dialect if not result: raise ImportError(f"Dialect transformer `{format_macrofunction(dialect)}.postprocess_ast` returned an empty result.") content = result self._step += 1 if self.debugmode: msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}after {c(CS.DIALECTTRANSFORMERNAME)}{format_macrofunction(dialect)}.postprocess_ast {c(CS.HEADING1)}(step {self._step}):{c()}\n" print(_message_header + msg, file=sys.stderr) print(format_for_display(content), file=sys.stderr) if self.debugmode: plural = "s" if self._step != 1 else "" msg = f"{c(CS.SOURCEFILENAME)}{self.filename} {c(CS.HEADING1)}completed all dialect {c(CS.TRANSFORMERKIND)}AST postprocessors {c(CS.HEADING1)}({self._step} step{plural} total).{c()}" print(_message_header + msg, file=sys.stderr) return content def find_dialectimport_source(self, text): """Find the first dialect-import statement by scanning source code `text`. As a side effect, import the dialect definition module. A dialect-import is a statement of the form:: from ... import dialects, ... To keep the search simple, the dialect-import **must**: - Be on a single line; not use parentheses or a line continuation. - Start at the first column on the line where it appears. When this runs, the input is just text. It is not parseable by `ast.parse`, because a dialect that has a source transformer may introduce new surface syntax. Similarly, it's not tokenizable by `tokenize`, because a dialect may customize what constitutes a token. So we can only rely on the literal text "from ... import dialects, ...", similarly to how Racket heavily constrains the format of its `#lang` line. Return value is the tuple `(module_absname, bindings, lineno, col_offset)`: - `module_absname` is the absolute module name referred to by the import - `bindings` is a dict `{dialectname: class, ...}`, with all bindings collected from that one dialect-import statement. Each binding is a dialect, so usually there is just one. - `lineno` is the line number of the import statement, determined by counting the lines of `text`. - `col_offset` is the corresponding column offset. Currently not extracted; is always set to 0. The return value refers to the first not-yet-seen dialect-import (according to the private cache `self._seen`). Note that this does not transform away the dialect-imports, because the expander still needs to see them in the AST transformation step. If there are no more dialect-imports that have not been seen already, the return value is `None`. """ matches = _dialectimport.finditer(text) try: while True: match = next(matches) statement = match.group(0).strip() if statement not in self._seen: # apply each unique dialect-import once self._seen.add(statement) lineno = 1 + text[0:match.start()].count("\n") # https://stackoverflow.com/a/48647994 col_offset = 0 # TODO: extract the correct column offset break except StopIteration: return None dummy_module = ast.parse(statement, filename=self.filename, mode="exec") dialectimport = dummy_module.body[0] module_absname, bindings = get_macros(dialectimport, filename=self.filename, reload=False, allow_asname=False) return module_absname, bindings, lineno, col_offset def find_dialectimport_ast(self, tree): """Find the first dialect-import statement by scanning the AST `tree`. Transform the dialect-import into `import ...`, where `...` is the absolute module name the dialects are being imported from. As a side effect, import the dialect definition module. Primarily meant to be called with `tree` the AST of a module that uses dialects, but works with any `tree` that has a `body` attribute, where that `body` is a `list` of statement AST nodes. A dialect-import is a statement of the form:: from ... import dialects, ... Return value is the tuple `(module_absname, bindings, lineno)`, where: - `module_absname` is the absolute module name referred to by the import - `bindings` is a dict `{dialectname: class, ...}`, with all bindings collected from that one dialect-import statement. Each binding is a dialect, so usually there is just one. - `lineno` is the line number from the import statement node, or `None` if the statement had no `lineno` attribute. - `col_offset` is the corresponding column offset. It is also taken from the same import statement node. The return value refers to the first dialect-import that has not yet been transformed away. If there are no more dialect-imports, the return value is `None`. """ for index, statement in enumerate(tree.body): if ismacroimport(statement, magicname="dialects"): break else: return None module_absname, bindings = get_macros(statement, filename=self.filename, reload=False, allow_asname=False) # Remove all names to prevent dialects being used as regular run-time objects. # Always use an absolute import, for the unhygienic expose API guarantee. thealias = ast.copy_location(ast.alias(name=module_absname, asname=None), statement) tree.body[index] = ast.copy_location(ast.Import(names=[thealias]), statement) # Get source location info lineno = statement.lineno if hasattr(statement, "lineno") else None col_offset = statement.col_offset if hasattr(statement, "col_offset") else None return module_absname, bindings, lineno, col_offset